journal of applied economics and business studies, volume. 2, issue 1 (2018) 19-28 https://doi.org/10.34260/jaebs.212 19 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x impact of shocks in domestic aggregate demand for natural gas on domestic aggregate energy consumption in china manzoor ahmad1, shehzad khan2, kiran alim3 1 phd scholar, department of industrial economics, nanjing university, nanjing, china 2 institute of business studies and leadership, abdul wali khan university mardan, garden campus 3 faculty of management, universiti teknologi malaysia (utm), skudai, malaysia abstract this paper examines the possible asymmetric transmissions from domestic aggregate demand for natural gas to domestic aggregate energy consumption in china using time series data from 1970 to 2016. the nonlinear autoregressive distributed lags (nardl) model is employed to check the possibility of long-term asymmetric nexus among variables. the empirical findings confirm the existence of symmetric cointegration between the domestic aggregate demand for natural gas and domestic aggregate energy consumption. the results also indicate that positive shocks in domestic aggregate energy consumption lead an increase in domestic aggregate energy consumption in both short-run and long-run. while the nardl dynamic multiplier graph suggests that the positive component of domestic aggregate demand for natural gas has deep impact on domestic aggregate energy consumption. keywords energy consumption, demand of natural gas; natural gas production; china jel classification e20; q11; q41 copyright © 2018 saebr all rights reserved 1. introduction natural gas (ng) is a low-carbon, clean, high-efficient energy and is a part of three mainstays of the global energy mix together with coal and oil. in 2016, ng constituted 24.13% of the world’s primary energy consumption mix, and this ratio is expected to increase to 25.17% in 2035 (zhang, ji, & fan, 2018). in addition to this, ng is the rapidly mounting primary energy source and is anticipated to be the fuel of choice for many developed and under developing economies (lim & yoo (2012). these days, the use of gas is rising due to many reasons, including fuel diversification, price, energy security, overall market growth, market deregulation and price (yazici & demirbas, 2001). generally, the most important use of ng is not only generating electricity but also used in heating. additionally, natural gas has grown in importance and therefore is one of the most important sources of clean energy for heating in domestic uses during the last several decades (boran, 2015). besides, in the literature of economics total energy use has drawn open access manzoor ahmad, shehzad khan & kiran alim 20 an immense attention than other commodities due to its economic, social, overlapping generation, political and multi-dimensions, and environmental issues (bilgili, 2014). china is affluent in natural resources. as stated by the ministry of land and resources (mlr), china’s geological conservative ng reserve potential volumes to sixty-two trillion cubic meters, and china’s mineable reserve conceivable quantities are estimated to be thirty-two trillion cubic meters (dong et al. (2017). as china has recognized the significance of uncontaminated energy resources and advances the utilization of the natural gas, the output of china’s ng is gradually raising (bp statistical review of world energy, 2015). an expansion of ng in china presents a confusing picture to worried stakeholders comprises of industrial participants, academics as well as policymakers. the growing pressure to enhance air quality appears to be compelling china to aggressively take up ng as a cleaner source of energy (li, yin, & wang, 2018). the domestic demand for ng has been staggering too. the use of ng has increased rapidly over the period from 2007 to 2014, expanding from 70.52 to 186.89 billion cubic meters (bcm), about an annual percentage change of 15 percent. however, in 2015 the domestic aggregate demand for ng 193.2 bcm, an increase of merely 5.7% (natural gas industry, 2016). the growth of china’s domestic aggregate demand for ng was quite slow (qian, duan, & wang, 2013). after 2000, china built many long-distance gas pipelines as well as branch pipelines, stimulating the fast development of the ng industry (see figure 1). the domestic aggregate demand for ng expanded from 25.3 bcm to 185.5 bcm, with a mean annual growth rate of 15.3 percent. figure 1: china’s domestic aggregate demand for natural gas 1985-2014 (source: dong et al. (2017) journal of applied economics and business studies, volume. 2, issue 1 (2018) 19-28 https://doi.org/10.34260/jaebs.212 21 to meet the domestic demand, china started to import ng during 2006, after which the total import quantity started to swiftly grow as well, and the dependence on other country’s ng has gradually expanded (zhang, 2014). in 2014, the ng import in china reached to 58.4 billion bcm, and the degree of foreign dependence was 27.5 percent (see figure 2). figure 2: china’s natural gas import situation (source: dong et al. (2017) due to the aforementioned major development in production, extraction, import of ng and an increase in domestic aggregate demand for ng, its impact on domestic aggregate energy consumption of china cannot be ignored. due to the variations in demand for energy consumption, it is expected that the relationship between domestic aggregate demand for ng and domestic aggregate energy consumption will be non-linear. thus, to capture the asymmetry arises due to shocks in domestic aggregate demand for ng, i use the non-linear ardl to explore the long and short run relationships. the rest of this article is arranged as follows. section 2 examines the data source and econometric methodology. section 3 evaluates the estimated results and interpret the computed findings. finally, the paper is wind up in part 4. 2. research methods this study uses secondary data on domestic aggregate energy consumption (daec) in kg oil equivalent, gross domestic product (gdp) in current us dollar, domestic aggregate demand for natural gas (dadng), domestic aggregate natural gas production (dangp) in billion cubic meters. the data series on gdp and daec have been gathered from world bank database. while the data series on dadng and dangp have been compiled from (bp statistical review of world energy (2017). the data used in this study is from 1970-2016. all the data are transformed into a logarithm. manzoor ahmad, shehzad khan & kiran alim 22 2.1 empirical model the general relationship between domestic demand for ng, gross domestic product, and aggregate domestic energy consumption can be represented in the following linear regression model: 𝐷𝐴𝐸𝐶𝑡 = 𝜍0 + 𝜍1𝐺𝐷𝑃𝑡 + 𝜍2𝐷𝐴𝐷𝑁𝐺𝑡 + 𝜍3𝐷𝐴𝑁𝐺𝑃𝑡 + 𝜇𝑡 (1) whereas 𝐷𝐴𝐸𝐶𝑡 represents domestic aggregate energy consumption in time period𝑡, 𝐺𝐷𝑃𝑡 indicates gross domestic product in time period 𝑡 , 𝐷𝐴𝐷𝑁𝐺𝑡 signifies domestic aggregate demand for natural gas in time period 𝑡 , 𝐷𝐴𝑁𝐺𝑃 shows domestic aggregate natural gas production, 𝜍0 is constant, 𝜍1, 𝜍2 and 𝜍3 are respective coefficients. while 𝜇𝑡 is error term. to employ the ardl method, based on pesaran et al. (2001) ardl framework, the following equation is employed. ∆𝐷𝐴𝐸𝐶𝑡 = 𝜒0 + ∑ 𝜒1 𝑝 𝑘=0 ∆𝐷𝐴𝐸𝐶𝑡−𝑘 + ∑ 𝜒2 𝑝 𝑘=0 ∆𝐺𝐷𝑃𝑡−𝑘 + ∑ 𝜒3 𝑝 𝑘=0 ∆𝐷𝐴𝐷𝑁𝐺𝑡−𝑘 + ∑ 𝜒4 𝑝 𝑘=1 ∆𝐷𝐴𝑁𝐺𝑃𝑡−𝑘 + 𝜗0𝐷𝐴𝐸𝐶𝑡−1 + 𝜗1𝐺𝐷𝑃𝑡−1 + 𝜗2𝐷𝐴𝐷𝑁𝐺𝑡−1 + 𝜗3𝐷𝐴𝑁𝐺𝑃𝑡−1 + θ𝑡 (2) whereas ∆𝐷𝐴𝐸𝐶𝑡 indicates the first difference of domestic aggregate energy consumption in time period 𝑡 , ∆𝐺𝐷𝑃𝑡−𝑘 shows the first difference of gross domestic product in time period 𝑡 and lagged value 𝑘, ∆𝐷𝐴𝐷𝑁𝐺𝑡−𝑘 signifies the first difference of domestic aggregate demand for natural gas in time period 𝑡 and with lagged value 𝑘 , ∆𝐷𝐴𝑁𝐺𝑃𝑡−𝑘 is the first difference of domestic aggregate natural gas production with lagged value 𝑡 − 𝑘 . whereas 𝜒1 → 𝜒4 are short run coefficients, 𝜗0 → 𝜗3 are long run coefficients and θ𝑡 is error correction term. in equation (2), the possible existence of longterm co-integrating association among sample variables is found out by testing the joint significance of lagged levels employing the standard f-test. specifically, the null hypothesis of 𝜗0 = 𝜗1 = 𝜗2 = 0 or no long run association among sample variables is tested. the typical critical values of f-statistics cannot be utilized because it does not follow the standard normal f-distribution. to deal with this problem, the pesaran et al. (2001) developed two different critical values that take into consideration the integrating characteristics of variables. the null hypothesis is rejected only if the estimated value of f-statistic is more than the upper (lower) critical values. if the sample variables provide an evidence of co-integration, then equation (2) is estimated to produce both short-term and long-term dynamics responses of china’s domestic aggregate energy consumption to journal of applied economics and business studies, volume. 2, issue 1 (2018) 19-28 https://doi.org/10.34260/jaebs.212 23 changes in gross domestic product, domestic aggregate demand for natural gas and domestic natural gas production. the previous two equations are based on the assumption that domestic aggregate demand for natural gas has symmetric impacts on domestic aggregate energy consumption. as our primary objective is to test whether or not aggregate domestic demand for natural gas has symmetric or asymmetric affect on domestic aggregate energy consumption. thus, to capture the asymmetric effects in the model, following shin et al. (2013) methodology, domestic aggregate demand for natural gas is decomposed into two new variables where one variable captures merely positive shocks in domestic aggregate demand for natural gas (𝐷𝐴𝐷𝑁𝐺𝑡 +) and the other variable merely captures the negative shocks in domestic aggregate demand for natural gas (𝐷𝐴𝐷𝑁𝐺𝑡 −) as follows: 𝐷𝐴𝐷𝑁𝐺𝑡 + = ∑ ∆𝐷𝐴𝐷𝑁𝐺𝑗 + 𝑡 𝑗=1 = ∑ 𝑚𝑎𝑥 𝑡 𝑗=1 (∆𝐷𝐴𝐷𝑁𝐺𝑗 +, 0) (3) 𝐷𝐴𝐷𝑁𝐺𝑡 − = ∑ ∆𝐷𝐴𝐷𝑁𝐺𝑗 − 𝑡 𝑗=1 = ∑ 𝑚𝑖𝑛 𝑡 𝑗=1 (∆𝐷𝐴𝐷𝑁𝐺𝑗 −, 0) (4) to get the full version of nonlinear ardl model, i substitute 𝐷𝐴𝐷𝑁𝐺𝑡 with 𝐷𝐴𝐷𝑁𝐺𝑡 + and 𝐷𝐴𝐷𝑁𝐺𝑡 − in equation (2): ∆𝐷𝐴𝐸𝐶𝑡 = �̀�0 + ∑ �̀�1 𝑝 𝑘=0 ∆𝐷𝐴𝐸𝐶𝑡−𝑘 + ∑ �̀�2 𝑝 𝑘=0 ∆𝐺𝐷𝑃𝑡−𝑘 + ∑ �̀�3 𝑝 𝑘=0 ∆𝐷𝐴𝐷𝑁𝐺𝑡 + + ∑ �̀�4 𝑝 𝑘=0 ∆𝐷𝐴𝐷𝑁𝐺𝑡 + + ∑ �̀�5 𝑝 𝑘=1 ∆𝐷𝐴𝑁𝐺𝑃𝑡−𝑘 + 𝜗0𝐷𝐴𝐸𝐶𝑡−1 + 𝜗1𝐺𝐷𝑃𝑡−1 + 𝜗2𝐷𝐴𝐷𝑁𝐺𝑡−1 + + 𝜗3𝐷𝐴𝐷𝑁𝐺𝑡−1 − + 𝜗4𝐷𝐴𝑁𝐺𝑃𝑡−1 + θ𝑡 (5) after estimating equation (5), the wald test is carried out to identify asymmetry effects of domestic aggregate demand for natural gas on china’s domestic aggregate energy consumption as follows: (1) short-run asymmetrical effects is recognized if the null hypothesis of − �̀�3 �̀�1 ⁄ = − �̀�4 �̀�1 ⁄ or, no short run symmetrical affects is rejected; and (2) long-term symmetrical affects is found out if the null hypothesis of − 𝜗2 𝜗0 ⁄ = − 𝜗3 𝜗0 ⁄ , or no long-term asymmetrical affects is rejected. manzoor ahmad, shehzad khan & kiran alim 24 3. empirical results and discussion to identify the order of integration among the main variables, augmented dickeyfuller (adf) and phillips-perron (pp) test have been employed. the findings of the unit root tests are depicted in table 1. the outcome validates that all the main variable series are non-stationary at the level but become stationary at first difference. in addition to this, none of the variables is integrated of order two. so, we can carry out nardl in our study. table 1: adf and pp unit root test variables level first difference adf pp adf pp daec 0.643031 1.338733 -3.565457** -3.568900** gdp 1.658236 1.649734 5.275870*** 5.344616*** dadng 0.984485 -0.424803 2.981044** -3.872180** dangp 0.386355 1.149038. -2.788603** -2.724029** note: *** and ** represent 1% and 5% significance level, respectively the estimated short-term and long-term findings are reported in table 2. as our central theme of this paper is to identify the asymmetrical effects, thus first we focus on estimated positive and negative components of dadng. the short-term outcomes reveal that both coefficients of positive and negative components of dadng are positive. for instance, the computed elasticities of dadng increases (decreases) with respect to daec are 0.24 (0.05), pointing that a one percent increase (decrease) in dadng is expected to increase (decrease) daec by 0.24% (0.05%). from table 2, it is clear that the coefficient of the positive component of dadng is highly statistically significant. on the other hand, the coefficient of the negative component of dadng is statistically insignificant. based on the different statistical significance and magnitude of the computed elasticities, dadng changes appear to have an asymmetric effect on china daec in the short run. in contrast, the result of wald test (see table 3) divulges that the null hypothesis of no asymmetrical effects, in short, is accepted, providing a strong evidence of short-term symmetric effects of dsdng changes. the long-term outcomes also reveal that the positive component as a well negative component of dadng are positive. for an instant, the elasticity of dadng rises (decreases) with respect to daec is 0.47 (0.11), inferring that one percent rise (decrease) in dsdnd is anticipated to increase (decrease) daec by 0.47% (0.10%). from table 2, it is clear that the coefficient of the positive component of dadng is statistical significance at one percent level. on the other side, the long-term coefficient of the negative component of dadng is statistically insignificant. based on the different statistical significance and magnitude of the computed elasticities, dadng changes seem to have journal of applied economics and business studies, volume. 2, issue 1 (2018) 19-28 https://doi.org/10.34260/jaebs.212 25 an asymmetric effect on china daec in the long-run. however, the outcome of wald test (see table 3) indicates that the null hypothesis of no asymmetrical effects in long-run is accepted, providing a strong support of long-term symmetric effects of dadng changes. table 2: results of short-run and long-run nonlinear ardl equation (2, 0, 1, 0) variable short run coefficient std. error t-statistic c 0.6596664*** 0.212307 3.107116 d(daec(-1)) 0.494879*** 0.128730 3.844327 𝐷𝐴𝐷𝑁𝐺 + 0.235091*** 0.062274 3.775113 𝐷𝐴𝐷𝑁𝐺 − 0.052579 0.091964 0.571732 gdp(-1) 0.059466** 0.021805 2.727163 gp(-1) -0.094435 0.060277 -1.566679 d(gdp) 0.191420*** 0.045569 4.200699 d(dangp) 0.011761 0.105387 0.111600 ect(-1) -0.500258*** 0.091981 -5.438691 variable long run coefficient std. error t-statistic 𝐷𝐴𝐷𝑁𝐺 + 0.469940*** 0.088145 5.331415 𝐷𝐴𝐷𝑁𝐺 − 0.105104 0.182473 0.575997 gdp 0.118870*** 0.035743 3.325670 dangp -0.188772 0.115850 -1.629458 c 1.328648*** 0.342256 3.852816 nardl bound test f − statistic (hyp1) = 4.281223 [lb = 2.2, ub = 3.09 at 10%] [lb = 2.56, ub = 3.49 at 5%] [lb = 2.88, ub = 3.87 at 2.5%] note: *** and ** represent 1% and 5% significance level, respectively. the short-run coefficient of gdp is not only positive but also significant at one percent level. accordingly, one percent increase in gdp leads an increase daec by 0.19% in the short run. while the short-run coefficient of dngp is positive and statistically insignificant. the long-run coefficient of gdp is positive as well as statistically significant at one percent level. this means that gdp increases china’s domestic aggregate energy consumption in the long run; for instance, when gdp increases by one percent, daec increase by about 0.12% in the long-run. while the long-run coefficient of dangp is statistically insignificant. table 3: testing hypothesis of an asymmetrical effect null hypothesis alternative hypothesis f-statistic decision − 𝜗2 𝜗0 ⁄ = − 𝜗3 𝜗0 ⁄ − 𝜗2 𝜗0 ⁄ ≠ − 𝜗3 𝜗0 ⁄ 0.884 long run symmetry − �̀�3 �̀�1 ⁄ = − �̀�4 �̀�1 ⁄ − �̀�3 �̀�1 ⁄ ≠ − �̀�4 �̀�1 ⁄ 0.659 short run symmetry manzoor ahmad, shehzad khan & kiran alim 26 -1.2 -0.8 -0.4 0.0 0.4 0.8 1.2 domestic aggregate demand for natural gas +1% domestic aggregate demand for natural gas -1% difference figure 3: dynamic multiplier graph it is north worthy to mention that the short-run and long-run estimated coefficients in equation (5) are important. we test for cointegration among the variables and we compare the computed f-statistic with the upper critical values with the tabulated values of pesaran et al. (2001). when we carry out this test, the obtained f-statistic is 4.28 and 5% (2.5%) upper critical value is 3.49 (3.87). hence, we strongly reject the hypothesis of no cointegration among variables. another method of conforming cointegration among variables is that the equilibrium occurs when the error correction coefficient is not only negative but also statistically significant (shin, baek, & heo, 2018). as a matter of fact, we obtained the negative coefficient of ect (-1) (-0.50), and this is statistically significant at one percent levels. hence, there is strong evidence supporting long-run association among sample variables. in order to evaluate the asymmetric adjustment of negative and positive shocks in the prevailing long-run equilibrium with respect to the newfangled long run equilibrium a nardl’s dynamic multiplier graph has been plotted (see figure 3). in summary, the result of dynamic multiplier graph infers that negative dadng shock has less or no impact than the positive daec shocks. journal of applied economics and business studies, volume. 2, issue 1 (2018) 19-28 https://doi.org/10.34260/jaebs.212 27 table 4 validation test test test-statistic probability arch 1.662586 0.1973 lm test 3.99755 0.1374 ramsay reset test 0.003244 0.9549 ultimately, to validate our model selection, we carry out different diagnostic tests for the residuals of our data (see table 4). the lagrange multiplier (lm) statistic of residual serial correlation is lm = 3.99. hence, there is no indication of the serial correlation in our model. the arch statistic of residual serial correlation is arch = 1.66, which confirmed that our model is free of the heteroscedasticity problem. the ramsey’s reset statistic is lm = 0.003, supporting a correctly specified optimum model. 4. conclusion the current explored the linkages between gross domestic demand (gdp), domestic aggregate demand for natural gas (dadng), domestic aggregate natural gas production (dangp) and domestic aggregate energy consumption (daec) by utilizing the methodology of non-linear ardl over the period of 1970-2016 for china. the adf and pp unit root tests are employed to check the variables for an order of integrations. the results of both unit root tests indicated that all the variables are i(1). the nardl bound test showed that there exists a long-run association among sample variables. while wald test confirmed that there exist both short-run and long-run symmetric linkage between dadng and daec. the findings also inferred that gdp is an important factor of daec in both long-run and short-run. references bilgili, f. (2014). long run elasticities of demand for natural gas: oecd panel data evidence. energy sources, part b: economics, planning, and policy, 9(4), 334341. boran, f. e. (2015). forecasting natural gas consumption in turkey using grey prediction. energy sources, part b: economics, planning, and policy, 10(2), 208213. dong, x., pi, g., ma, z., & dong, c. (2017). the reform of the natural gas industry in the pr of china. renewable and sustainable energy reviews, 73, 582-593. dudley, b. (2015). bp statistical review of world energy 2016. london, uk. global, b. p. (2017). bp statistical review of world energy june 2017. relatório. disponível em: http://www. bp. com/en/global/corporate/energyeconomics/statistical-review-of-world-energy. html. manzoor ahmad, shehzad khan & kiran alim 28 kang, z. (2014). natural gas supply-demand situation and prospect in china. natural gas industry b, 1(1), 103-112. li, b., yin, h., & wang, f. (2018). will china’s “dash for gas” halt in the future?. resources, conservation and recycling, 134, 303-312. lim, h. j., & yoo, s. h. (2012). natural gas consumption and economic growth in korea: a causality analysis. energy sources, part b: economics, planning, and policy, 7(2), 169-176. natural gas industry, 2016. (n.d.). natural gas operational performances. (accessed 02.04.18). pesaran, m. h., shin, y., & smith, r. j. (2001). bounds testing approaches to the analysis of level relationships. journal of applied econometrics, 16(3), 289-326. qian, x., duan, z., & wang, h. (2013). development status and prospects of china's natural gas industry. econ anal china pet chem ind, 11:22–26. shin, c., baek, j., & heo, e. (2018). do oil price changes have symmetric or asymmetric effects on korea’s demand for imported crude oil?. energy sources, part b: economics, planning, and policy, 13(1), 6-12. shin, y., yu, b., & greenwood-nimmo, m. (2014). modelling asymmetric cointegration and dynamic multipliers in a nonlinear ardl framework. in festschrift in honor of peter schmidt (pp. 281-314). springer, new york, ny. yazici, n., & demirbaş, a. (2001). turkey's natural gas necessity and consumption. energy sources, 23(9), 801-808. zhang, y., ji, q., & fan, y. (2018). the price and income elasticity of china's natural gas demand: a multi-sectoral perspective. energy policy, 113, 332-341. journal of applied economics and business studies, volume 1, issue 1 (2017) 45-52 https://doi.org/10.34260/jaebs.115 45 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online): 2663-693x a causality analysis between trade and economic growth of pakistan syyeda farhana shah1, saleem khan2*1& abdur rauf3 1 ms scholar, islamia college peshawar 2 assistant professor in economics, department of economics, abdul wali khan university mardan, garden campus 3 assistant professor, department of economics, abbottabad university of science and technology (aust) abstract the present study core objective is to find the linkages between pakistan’s exportsimports and gross domestic product (gdp), using data over the period 1981-2016. unit root in each series is checked with the help of augmented dickey fullers' (adf) test. the engle grange approach is utilized to determine the long run relationship among variables of the study. moreover, causality among the selected variables is tested by the methodology of vector error correction model (vecm). in results, we found that all selected variables are first difference stationary and causality run from gdp to exports and imports of the country. in both cases, the causality is positive and significant. no reverse causation was observed from exports-imports to gdp. finally, the results demonstrate that the causal connection among gdp and imports is stronger than the gdp and exports. keywords imports, exports, economic growth, causality jel classification f14; f43 copyright © 2017 saebr all rights reserved 1. introduction economic growth can be viewed as a shift in its production possibility frontier or an expansion in the country’s total output or production. many factors affect economic growth that also includes country's imports and exports. the exports growth has generally been associated with the economic growth. exports represent higher demand for the country's products and to meet more demand there is an increase in prices and production of that particular commodity. export-led growth (elg) hypothesis states that an increase in exports enhances economic growth. exports directly contribute to the gross domestic product (gdp). exports allow a country to benefit from the large global markets or economies of scale (helpman & krugman, 1985). rodrick (1988) stated that exports lead firm to large-scale production, technological change and an increasing rate of capital formation. 1*corresponding author: drsaleem.eco@awkum.edu.pk syyeda farhana shah, saleem khan & abdur rauf 46 when domestic output increases more than domestic demand, then the surpluses are exported to various markets. this further increases the demand, boosting the output and thus there is a direct relationship between economic growth and export growth (sharma & dhakal, 1994). ramos (2001) examined the export-import and gdp growth relationship for portugal economy. he found bidirectional relationship between gdp and imports as well as gdp and exports while no link between export and import was observed. generally, literature emphasizes country’s exports are the major source of economic growth. exports play an important role in society’s welfare by improving people’ living standard through increase in per capita income. the advocates of economic growth view that the rate of economic growth can be accelerated with the help of exports. the empirical research has produced bi-directional results such as economic growth can be accelerated with increase in exports but the exports can also be encouraged with economic growth. consequently, both the relationships that export leading to promote economic growth and growth leading to promote exports are possible. thus, it is important to estimate the relationship between the exports and economic growth of a country. sharma et al. (1991) employed granger causality approach and identified positive relationship between economic growth and exports. dodaro (1993) employing a country-wise time series data and using causality approach found weak support for the existence of elg. he also found a weak support for the alternative augmented growth-led-export hypothesis too but somewhat stronger than the elg. hatemi (2002) on the other hand reported a bi-directional causality between economic growth and exports for the japan. since the previous research has produced conflicting results such as positive, negative, unidirectional or bidirectional relationship while studying the exports and gdp growth relationship. therefore, this research is undertaken with a view to revisit the country’s data and explore the direction of relationship between pakistan’s exports and economic growth. this research paper uses engle-granger and vector error correction model (vecm) methods where the aim is to test for the existence of co-integration and causality relationships among the exports-imports and gross domestic product or economic growth of pakistan. previously, in case of pakistan using the time series approaches, a bidirectional relationship has been found between economic growth and exports (shirazi and manap, 2004, and iqbal et al., 2012). this research not only uses more recent larger data set but also investigates the imports and economic growth nexus as well. 2. literature review exports and imports play a pivotal role in getting a high economic growth and economic development of a country. examination of the impact of these two variables on economic growth has been studied both empirically and theoretically albeit with conflicting results. the mixed results from the previous research motivated this research. some of the previous research is reviewed here with the aim to identify the gaps in the literature, get guidance in the development of both the theoretical and empirical models and finally compare the results. the relevant literature is outlined here as follows: journal of applied economics and business studies, volume 1, issue 1 (2017) 45-52 https://doi.org/10.34260/jaebs.115 47 in literature, it has been found that either unidirectional or bidirectional gdp growth and exports nexus existed. gibba and molnar (2016) explored the economic growth and export relationship for gambia using data for the period 1980-2010. they found that the export causes economic growth to grow. oxley (1993) utilized granger-causality test over portuguese data, 1983-1995, and failed to accept the hypothesis of export-led growth while favoured reverse causality that income growth causes export growth. ram (1985) studied the export led growth relatively for large number of developing countries over the period of 1960-1977. using granger causality test, found that exports performance did not seem important for economic growth, however, and the exports impact on economic growth was small in the low income developing countries as compared to the middle income developing countries. furthermore, some studies found causality from growth rate to exports such as shihab et al. (2014) and findly (1984). alkhateeb et al. (2016) analyzed the relationship between exports and economic growth for saudi arabia over the period 1980-2013 by utilizing co-integration approach and causality test where the study found bidirectional causality between gdp and exports. hussain and afaf (2014) used both the causality test and co-integration approach to estimate the impact of exports and imports on saudi arabia economic growth over the period of 1990-2011. they found a statistically significant unidirectional causality running from exports to gdp growth but imports to gdp growth was not found to be statistically significant. ajmi et al. (2015) tested the causal link between export and economic growth by applying the methodology of linear and non-linear granger causality tests. this study used annual data for south africa from 1911-2011. from linear granger causality test, ajmi et al. (2015) did not find significant evidence of causality between export and gdp for south africa. however, with non-linear causality test found unidirectional causality from gross domestic output to exports. asafu-adjaye et al. (1999) also did not find causality relationship in the case of india. therefore, the studies of the ajmi et al. (2015) and asafuadjaye et al. (1999) did not confirm the evidence of export-led growth hypothesis. yuhong et al. (2010) found direct relationship between the economic growth and imports for china also between growth and imports. achchuthan (2013) also concluded that as imports increases, economic growth also increases. the studies of haseeb et al. (2014), irwan et al. (2015), and khan et al. (2016) tested causality between exports and economic growth relationships. the estimated results of all the above studies concluded that exports and growth are significantly positive correlated. yuksel and zengin (2016) analysed the export, import and growth causality relationships for six developing countries (i.e. argentina, brazil, china, malaysia, mexico and turkey) using data for the period of 1961-2014. using time series econometrics techniques to measure for causality and co-integration, it was found in this study that in case of argentina, increase in exports lead to an increase in economic growth. however, in the case of china and turkey, imports caused exports to grow in the country as well. the imports economic growth relationship varied for the countries under study. syyeda farhana shah, saleem khan & abdur rauf 48 due to the mixed results indicated in the studies reviewed above, we aim to investigate the casual relationships among economic growth, imports and exports. this is achieved by estimating the co-integration and causality between imports and gdp growth, exports and gdp growth for pakistan economy over the period 1981-2016. 3. research methods 3.1.data in this research we have employed annual data to test for possible relationship as well as to found the direction of causality between exports and economic growth of pakistan. the data period is from 1981 to 2016. the data were collected from two major databases of international financial statistics (ifs) and pakistan economic survey. all variables of the study are in nominal term and in millions of rupees. moreover, these variables have been converted into natural logarithm. 3.2. empirical model earlier empirical formulations tested the causality by having only the gdp growth and exports in the model (emely, 1968). other studies examined this relationship in the neoclassical framework (balassa, 1978; fosu, 1990). however, guntukula (2018) examined the relationship among exports, imports and growth for different nations. we used the following specifications of the economic growth, exports and imports: 𝐺𝐷𝑃 = 𝑓(𝑋, 𝑀) (1) the log-linear econometric form of equation (1) can be represented as follows: 𝑙𝑛𝑅𝐺𝐷𝑃𝑡 = 𝑎 + 𝛾𝑙𝑛𝑅𝑋𝑡 + 𝛿𝑙𝑛𝑅𝑀𝑡 + 𝑡 (2) where rgdp stand for real gross domestic product, rx and rm stand for real exports and real imports respectively. the last term of equation 𝑡 is the white noise term. the parameters γ and δ represents the magnitude of change in dependent variable due to changes in the independent variables. the ln stands for natural logarithm which means that we get changes in percentage terms we get elasticities. 3.3. econometric analysis 3.3.1. engle-granger approach of co-integration this study utilizes the engle-granger (eg) procedure of co-integration for long run analysis. engle granger analysis of co-integration requires that all the variables of model are stationary at same level of integration. the stationarity can be examined with the help of augmented dickey filler (adf) test. 3.3.2. the vector error correction model (vecm) vector of error correction model (vecm) is used to test the causal relationships among variables. the vecm is helpful for measuring the causality when variables are stationary at first order rather than level (engle-granger, 1987). the vecm model based on equation 2 can be written as follows: journal of applied economics and business studies, volume 1, issue 1 (2017) 45-52 https://doi.org/10.34260/jaebs.115 49 ∆𝑙𝑛𝑅𝐺𝐷𝑃𝑡 = 𝛼0 + ∑ 𝛼𝑖 ∆𝑙𝑛𝑅𝐺𝐷𝑃𝑡−𝑖 𝑚 𝑖=1 + ∑ 𝛾𝑖 ∆𝑙𝑛𝑅𝑋𝑡−𝑖 𝑛 𝑖=1 + ∑ 𝛿𝑖 ∆𝑙𝑛𝑅𝑀𝑡−𝑖 𝑝 𝑖=1 + 𝜎𝐸𝐶𝑡−1 + 𝑡 (3) where δ in equation 3 is the difference operator; the subscripts m, n, and p over summation are the numbers of lags. the ec represents the error correction term and can be obtained from the regression model of co-integration. the coefficient on the ec term needs to be negative and significant. it shows the correction of the disequilibrium adjustment towards the equilibrium. the rate at which this disequilibrium can be adjusted/corrected depends on the sign, magnitude and statistical significance of the error correction term. 4. results and discussions use of time series data techniques require that the data is tested for stationarity and accordingly use the time series techniques as estimating the ordinary least squares (ols) regressions on non-stationary data may produce spurious results. therefore, prior to using co-integration, first we need to test for unit roots. for this we use augmented dickey fuller (adf) test. after that, we do co-integration analysis. the vecm will be used to test for the nature of causal relationships. as stated, adf test is used to check the stationarity of variables. the results of the adf test are provided in table 1 as follows: table 1: adf test results of unit root with constant and trend variable at level at 1st difference t-test critical values at 5 % t-test critical values at 5 % rgdp -2.332 (0.406) -3.548 -4.414* (0.006) -3.548 rx -0.571 (0.974) -3.548 -7.457* (0.000) -3.548 rm -2.201 (0.474) -3.548 -6.371* (0.000) -3.548 the values in parenthesis are probability values. * indicates 1 percent level of significance. in table 1 we find that none of the data series is stationary at level since p-values in parenthesis are larger than the standard 5% and 1% level of significance and the absolute values of test-statistics (t-test) are smaller than its critical values. therefore, fail to reject the null hypothesis and conclude that all the data series have unit roots at level. however, when the first difference of the data is taken, we find that all the data become stationary. knowing that all the data is stationary at first difference, we can estimate the co-integration relationships. we first estimate the model given in equation 2 using the ordinary least squares method, so as to get the estimated residuals which shall be utilized to test for the cointegration. unit syyeda farhana shah, saleem khan & abdur rauf 50 root tests of adf and phillips-perron (pp) are used to test for stationarity. consequently, table 2 reports the adf test and pp test results of an error term as following. table 2: unit test results of residuals equation of regression adf test at level phillips perron test at level gdp = f(x, m) t-test p-value t-test p-value -1.836 0.063 1.950 0.050 from above results, it can be seen that the residuals are stationary at level utilizing both the unit roots. we can conclude that variables in model are co-integrated or have long run relationship. finally, the vecm is used for testing causality among the variables. in order to run the vecm model to determine the causalities, we first specify the lag length using lag length criteria. the optimum lag length is calculated as one as indicated by esteric (*) in table 3 below on the basis aic and sc. table 3: results for lag order selection lag fpf aic sc hq 1 1.32e-07* -7.331098* -6.922959* -7.193772* 2 1.40e-07 -7.279104 -6.462827 -7.004452 3 1.95e-07 -6.981379 -5.756964 -6.569401 *show lag order specification based on different criteria such as aic, sc and hq etc, after calculating the lag length, the vecm results are obtained and given in the table 4. table 4, indicates the direction of causality. the p-value is less than 0.05 when causality direction is from gdp to exports and gdp to imports; it implies an existence of the casual relationship. however, no converse causal relationship either from exports or imports to gdp is observed. therefore, there is unidirectional causality indicating that gdp stimulates exports and imports. table 4: results of vecm for exports, imports and growth causality direction lag length chi-sq p-value results exports →gdp 1 2.998 0.223 no causality relationship imports →gdp 1 2.500 0.286 no causality relationship gdp →exports 1 6.791 0.033 causality relationship gdp →imports 1 8.006 0.018 causality relationship 5. conclusion the main aim of this research was to explore the long and short run relationship between the economic growth, exports and imports. owing to the mixed results obtained in previous research regarding the causality between the gdp growth and imports and gdp growth and exports, we undertook this study for pakistan using the time series data from 1981-2016. first, we employed adf test to check for the unit root and then applied engle-grange approach for co-integration and the vecm for exploring the causal relationships. the journal of applied economics and business studies, volume 1, issue 1 (2017) 45-52 https://doi.org/10.34260/jaebs.115 51 engle-granger approach is performed to identify co-integration among exports, imports and economic growth. the engle granger result revealed that variables exhibited long-run relationship or had co-integration. the vecm of causality analysis was used to test the causality. we found a unidirectional causality. the estimation results lead to a confirmation of causality from gdp to exports and gdp to imports but not vice versa. the causal effect from gdp to exports as well as imports is positive and statistically significant. this implies that an increase in gdp leads to an increase in exports and imports in context of the pakistan economy over the period of analysis. further, we can conclude that contrary to the general expectations there is no evidence of elg hypothesis for the pakistan. the reason for this could be that pakistan's exports are not very diversified and not a major contributor to the gdp. however, causality direction from exports to gdp growth is positive but not statistically significant. it is suggested that efforts may be made to increase the volume of gdp by mainting a healthy growth rate over a long period of time to increase the exports. references achchuthan, s. (2013). export, import, and economic growth: evidence from sri lanka. journal of economics and sustainable development, 4(9), 147-55. ajmi, a. n., aye, g. c., balcilar, m., & gupta, r. (2015). causality between exports and economic growth in south africa: evidence from linear and nonlinear tests. the journal of developing areas, 49(2), 163-181. alkhateeb, t. t. y., mahmood, h., & sultan, z. a. (2016). the relationship between exports and economic growth in saudi arabia. asian social science, 12(4), 117-124. asafu-adjaye, j. & chakraborty, d. (1999). export-led growth and import compression: further time series evidence from ldcs. australian economic papers, 38: 164-75. balassa, b. (1978). exports and economic growth: further evidence. journal of development economics, 5(2), 181-89. dodaro, s. (1993). exports and growth: a reconsideration of causality. journal of developing areas, 27(2), 227-44. engle, r. f., & granger, c. w. (1987). co-integration and error correction: representation, estimation, and testing. econometrica: journal of the econometric society, 251-276. findlay, r. (1984). growth and development in trade models. handbook of international economics, (1), 185-236. fosu, a. k. (1990). exports and economic growth: the african case. world development, 18(6), 831-35. gibba, a., & molnar, j. (2012). a study on exports as a determinant of economic growth in the gambia. some studies of economics changes, 237-244. haseeb, m., n.h. hartani, n.a.a. bakar, azam m. & s. hassan, (2014). exports, foreign direct investment and economic growth: empirical evidence from malaysia (19712013). american journal of applied sciences, 11: 1010-1015. syyeda farhana shah, saleem khan & abdur rauf 52 hatemi, j. a. (2002). export performance and economic growth nexus in japan: a bootstrap approach. japan and the world economy, 14 (1), 25-33. helpman, e. & krugman, p. (1985). market structure and foreign trade, cambridge, ma: mit press. hussain m (2014). export and gdp in pakistan: evidence from causality and cointegration analysis. international journal of management cases, 16 (1), 37-49. hussain, m., & afaf, s. (2014). relationship between exports, imports, and economic growth in saudi arabia: 1990-2011. evidence from cointegration and granger causality analysis. journal of emerging trends in economics and management sciences, 5(3), 364-370. irwan s., haseeb, m., azam m. & islam, r., (2015). foreign direct investment, financial development, international trade and energy consumption: panel data evidence from selected asean countries. international journal of energy economics and policy, 5(3), 841-850. iqbal, a., hameed, i. & devi, k. (2012). relationship between exports and economic growth of pakistan. european journal of social sciences, 32(3), 453-460. khan, s., azam, m., & chandra, e., (2016). import demand, income elasticity and growth rate in pakistan: the impact of trade liberalization. foreign trade review, 51(3),1-12. oxley, l. (1993). cointegration, causality and export-led growth in portugal, 1865-1985. economics letters, 43(2),163-166. ram, r. (1985). exports and economic growth: some additional evidence. economic development and cultural change, 33(2), 415-425. ramos, f. (2001). exports, imports and economic growth in portugal: evidence from causality and cointegration analysis. economic modelling, 18(4), 613-623. rodrik, d. (1988). closing the technology gap: does trade liberalization really help? cambridge nber working paper no 2654. sharma, s.c., & dhakal, d., (1994). causal analysis between exports and economic growth in developing countries. applied economics, 26: 1145-1157. sharma, s., norris, m., & cheung, d., (1991). exports and economic growth in industrialized countries. applied economics, 23: 697-708. shihab, r. a., soufan, t., & abdul-khaliq, s. (2014). the causal relationship between exports and economic growth in jordan. international journal of business and social science, 5(3), 302-308. yuhong li, zhongwen c. & changjian s. (2010). research on the relationship between foreign trade and the gdp growth of east china-empirical analysis based on causality. modern economy, 1: 118-124. yuksel, s. & zengin, s. (2016). causality relationship between import, export and growth rate in developing countries. international journal of commerce and finance, 2(1), 147156. journal of applied economics and business studies, volume. 2, issue 1 (2018) 29-36 https://doi.org/10.34260/jaebs.213 29 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x households discrimination in school enrolment in pakistan: does gender matter? ali muhammad1*, zahoor ul haq2 & imad khan31 1 msc graduate, department of economics, khyber pakhtunkhwa, kp, pakistan 2 professor, department of economics, pakhtunkhwa economic policy research institute, abdul wali khan university mardan, kp, pakistan 3 lecturer, department of economics, university of swat, kp, pakistan abstract this study uses pakistan social and living measurement survey 2016 to study gender discrimination in school enrollment across the four provinces of pakistan using bi-variate analysis. results show that there is highly significant difference between male and female education in rural areas (x^2=4940.50 and p<0.05). analysis indicate that gender disparity in enrollment is significantly higher in low income households (x^2=115.468 and p<0.05). the study also showed that as compared to male, fewer female are enrolled in both public and private sectors. hence, socio-economic factors play important role in making decision about children enrollment in different types of school. the study recommends that government to take appropriate steps to reduce gender discrimination in school enrollment by offering subsidy on female education in the country. keywords gender discrimination; enrolment; socio-economic factors; enrolment; gender jel classification j10; j18; j19 copyright © 2018 saebr all rights reserved 1. introduction knowledge, skills and improvements in personality, attitudes, talents and behaviours are the outcome of education. education is considered as a critical factor while evaluating a society’s development (undp, 1990). in today’s modern era, human capital is considered the best national resource for a country. on one hand, educated people can access better opportunities for livelihoods while at the same time their creative work can benefit the entire nation (tripathi et al., 2014). education is one of the basic and essential rights of a person and its access should not be discriminated on gender, race, language or religion. the right to educate is common in all the religions and cultures. unfortunately, in south asia, access to education is discriminated based on gender, financial status, marital 1*corresponding author: safi72325@gmail.com open access ali muhammad, zahoor ul haq & imad khan 30 status and other socio-economic factors and pakistan is no exception (faizi & butt, 2017). such discrimination negatively affects the development process of a country. genders shall not be discriminated in the provision of opportunities to get their real potential unleashed. many other studies have also highlighted the significance of male and female education in economic development of a country (afzal et al., 2013). all these research studies emphasized and documented the significance of education without any disparity based on gender, region or nation for socio-economic growth and development that remains slow in asia. as compared to other countries, literacy rate is not only very low but highly unequal among males and females in pakistan. according to economic survey of pakistan, literacy rate has declined from 60 percent to 58 percent with literacy rate of 70 percent in males and 48 percent in females. pakistan has been ranked the second worst country in the world for gender inequality. according, to the world economic forum (wef), pakistan ranks 143 out of 144 countries in the gender inequality index, way behind bangladesh and india which rank 72nd and 87th respectively in 2016. this gape is the focus of this study in the case of education. irfan (2016) states that in pakistan sisters as compared to their brothers are less likely to attend primary education. in some cases, where children are sent to school, it is commonly seen that parents decide to invest in their sons’ education rather than their daughters. this may reflect the fact that upon marriage, daughters may no longer contribute to family income and are therefore not seen as worth investing. consequently, girls’ otherwise intelligent and capable are deprived of education. this also limits their role in the development of the society. studies that have investigated gender discrimination in education also indicated that such discrimination limits countries’ potential to develop and prosper. several factors could affect households’ discrimination of gender in school enrolment. these factors may include education level of the household head, poverty status of household, age, gender of household head and a host of other factors. these factors enhance our understanding of the links between a social phenomenon and gender description in school enrolment. this study is an attempt to analyse the socio-economic factors that lead to the gender gap in the enrolment. moreover, objective of this study is to know about the status of children distribution as per their socio-economic status and to highlight the factors that lead to gender gap in school enrolment across the provinces of pakistan. the findings of the study might be helpful for parents, researchers, policy makers and the government. results can be used in the development of policies for reducing gender discrimination in school enrolment among school going children. the findings of the study are also helpful for recognizing the worth and contribution of female education for economic development of the country. journal of applied economics and business studies, volume. 2, issue 1 (2018) 29-36 https://doi.org/10.34260/jaebs.213 31 2. literature review zarar et al. (2017) examined the causes and effects of gender discrimination against women in quetta by using primary data. the study concluded that literacy is directly linked with the willingness of the parents/guardians towards education attainment. the study further stated that majority of the girls do not give any monetary incentives as they get married in the early age while boys are required to have more education as compared to girls and the reason is the earnings. the study also highlighted certain other religious factors which hinder female education in the society. another study conducted by luqman et al. (2017) analysed the factors contributing to gender disparity in education in rural areas of the punjab province in pakistan. the main data were collected from three hundred household heads and from their spouses through specific individual interviews to find the actual difference in view of the male and female respondents. the collected data was analysed using spss and applied t-test method to find out the difference in responses of male and female participants. in order to determine the difference in intensity of factors in three districts, f-test was applied. the analysed results confirmed that there is highly significant difference between age of male family heads and their spouses (wives) (χ=50.121 and p <0.05). cross tabulation suggested that educational status of wives was low as compared to their husbands. analyses showed that a few socio-economic factors are the existing gender discrepancy with reference to education in the study areas. the researcher determined low income, high educational expenditures, low educational level of parents (especially the mother), and security concerns of parents regarding sexual harassment as the major factors impeding female education. the study recommended that the government at the national level should take serious steps to creating conducive environment to increase female enrolment rate in rural localities. another study held in punjab province by afzal et al. (2013) analysed the gender disparity in schools. they showed that parents differently treat male and female education. this disparity is high in rural as compared to urban areas. shayan (2015) also studied disparity in females and males education access. he showed that women participation in education is less than men and it further decreases in the higher education. the fear of terrorist attack and kidnaping by taliban has also reduced women participation in certain areas. moreover, the study identified that the most basic problem of women is the very paternal structure of the society and discrimination originate from extreme religious beliefs and traditions. shaukat et al. (2014) analyse the discrimination practices in the higher institutions of pakistan focusing on the variables such as decision making, professional development, utilization of resources, academic affairs and job satisfaction. the study used the t-test and anova test for analysing the data. the study concluded that gender differences is only in decision making and academic affair and less discrimination is observed in higher ali muhammad, zahoor ul haq & imad khan 32 positions than in lower positions. qureshi (2012) examined the gender differences in school enrolment and return to education in pakistan by using primary data. the study suggested that there is under investment in female education and return to education’s are much higher for female than male. it is found that private rate of return to time spent in school is higher than labour market for a female but return that goes to parents are lower for female than sons because parents are dependent on son at old age support. the study also concluded that mother roles are more significant and more impassive than father in term of magnitude of all level of female education. khan et al. (2013) states that female feel more secure after getting education and association between female education in urban areas is higher than rural areas while white et al. (2016) believes in supportive attitude of parents can help in female education. other studies such as mian et al. (2016), ara and malik (2012), cooray and potrafke (2010) and chaudhry and rahman (2009) investigated the impact of gender inequality in education and reached to conclusion the gender discrimination affects the education attainment. the next section discusses the data and analytical tool used in the analysis. 3. research methods the principal objective of the study is to investigate gender discrimination in school enrolment in pakistan. the study used pslm data for the year 2016. frequencies and cross tabulation are used to analyze the data. this report contains the data collected from 24,238 household based on 1605 urban & rural primary sampling units (psus). the period of field enumeration of hies as part of hiics 2015-16 was from september 2015 to june 2016. the variables used in the analysis are gender, age, income of the household, institution and region. table 1 presents the variables and their definitions. table 1: definition of the variables used in the analysis variable definition gender male/female age age of the respondent in years income of household annual income of households in rupees institution public, private and others (deeni madras etc.) region urban and rural the study mainly uses cross-tabulations and chi-square for analysis of the data. it helps to establish association between dependent variable, gender discrimination and the independent variables chi square test is also used to test the hypotheses. 𝑋2 = ∑ ∑(0𝑖𝑗 − 𝑒𝑖𝑗)2 𝑘 𝑖=1 / 𝑒𝑖𝑗 𝑗 𝑖=1 (1) where, 𝑋2 represents chi-square for two categorical variables, 0𝑖𝑗 represents the observed frequencies in the cross-classified category at ith row and jth column. and 𝑒𝑖𝑗 journal of applied economics and business studies, volume. 2, issue 1 (2018) 29-36 https://doi.org/10.34260/jaebs.213 33 represent the expected frequency for the same category, assuming no association between variables under investigation. the resulting frequency is distributed as chi-square with relevant degree of freedom. the degree of freedom is calculated as df = (r-1) (c-1) where df is equal to degree of freedom, r represents the number of rows and c represents number of columns. 4. results and discussion several steps were taken to select the right observations before analyzing the data. first, households with school going children were selected in pakistan social and living measurement (pslm) survey. second, households with both male and female school children were selected. third, individual child’s profile was obtained from the pslm data. fourth and finally, household characteristics were placed against each selected school going child in the database. this gets us to 56,218 out of 157,775 observations. crosstabulations were generated against different socio-economic characteristics of the households to understand the phenomenon of household gender discrimination in school enrolment. table 2 shows the number of schools going children according to their age. about 35.6 percent of the children are in the age bracket of more than five and less than 18 years. the rest do not fall in this category. the importance of school education could not be more over emphasized as it forms the base of the pyramid of education. primary education also makes the basis of any human capital development and hence it is very important for the economic growth of a country from the perspective of new growth theory. glewwe and jacoby, (1995) highlight the importance of age of a child for starting a school. it is natural that if a child starts education at a lower age then he/she also completes it quickly, which increases returns to schooling as he/she can work for more years. typically, school enrolment can be increased by increasing the number of institutions as well as increasing the proportion of gdp spent on education. poverty, income of households, education level of parents, family size is some of the other factors that affect enrolment at school level. the difference between school going and those who do not attend schools is statistically significant. table 2: distribution of respondent according to school age ages frequency percent valid percent cumulative percent not school age 101557 64.4 % 64.4 % 64.4 % school age>=5 and <=18 56218 35.6 % 35.6 % 100.0 % total 157775 100.0 100.0 100.0 % source: own estimation using hies 2015-16 data. next it is important to understand that proportion of children attending schools in rural and urban areas. less than fifty percent (48 percent) of the children attend schools in rural areas while the 71.6 percent attend in urban areas (table 3). one way of increasing enrolment in schools could be by increasing the numbers of schools and government ali muhammad, zahoor ul haq & imad khan 34 spending. it could also decrease the imbalance of school prevalence in urban and rural areas. the access to higher education for female gets even worse due to many socioeconomic factors including early marriages, social pressure and financial constraints are the major restraints as reported by different research studies in different regions/provinces of pakistan (khan et al., 2013). table 3: distribution of respondents ever attended any educational institution province yes no all rural urban rural urban kp 6981 15475 7250 8240 37946 12.80 % 15.00 % 13.30 % 8.00 % 24.10 % punjab 11721 33900 8569 8866 63056 21.40 % 32.90 % 15.70 % 8.60 % 40.00 % sindh 5769 16571 9207 6256 37803 10.50 % 16.10 % 16.80 % 16.80 % 24.00 % baluchistan 1762 7820 3428 5960 18970 3.20 % 7.60 % 6.30 % 5.80 % 12.00 % overall 26233 73766 28454 29322 157775 48.00 % 71.60 % 52.00 % 28.40 % 100.00 % chi square=4940.50 p<0.05 source: own estimation using hies 2015-16 data. next it is important to understand that how much gap exist between male and female according to institution. table 4 shows that 6278 (30.1) male while 5200 (24.9%) female are studying in public sector schools. in case of private sector, 6502 (21.6%) male while 3635(17.4%) female are studying in these institutions. the others institution like deeni madras etc. accommodate 663 (3.2%) male and 601(2.9%) female. these results also show that enrolment in institution is statistically significantly affected by gender. in the private institutions, male account for 55.3 of the enrolments while female contributes 44.7. hence, households prefer male to be educated in private schools. table 4: distribution of respondents according their institution gender institutions totals public private others male 6278 30.1 % 4502 21.6 % 663 3.2 % 11443 54.8 % female 5200 24.9 % 3635 17.4 % 601 2.9 % 9436 45.2 % all 11478 55.0 % 8137 39.0 % 1264 6.1 % 20879 100.0 % chi square = 3.776 p>0.0 table 5 shows the distribution of educational institution according to income groups. the lowest income group was of the household with annual income of less than rs. 50000 and the highest group was with earnings of more than rs.500000 per year. afzal et al. (2013) showed that income is an important determinant of education between male and female. income not only determines the disparity between male and female education. the table shows that the number of households opting for private institutions as income level journal of applied economics and business studies, volume. 2, issue 1 (2018) 29-36 https://doi.org/10.34260/jaebs.213 35 increases and this relationship is statistically significant. table 4.5 shows that for households of income of less than rs. 50,000 per year, households with male and female tend to send their children to public institutions while a large proportion of households with male only send their children to private institutions. as income increases this phenomenon becomes more noticeable. table 5: distribution of respondent according to their income institutions total public private others less than 50000 count 1792 1690 201 3683 % of total 3.9 % 3.7 % 0.4 % 8.0 % 50001-150000 count 2978 847 324 4149 % of total 6.5 % 1.8 % 0.7 % 9.0 % 150001-300000 count 9259 4997 1018 15274 % of total 20.1 % 10.9 % 2.2 % 33.2 % 300001-500000 count 5925 5092 636 11653 % of total 12.9 % 11.1 % 1.4 % 25.3 % above 500000 count 4217 6482 517 11216 % of total 9.2 % 14.1 % 1.1 % 24.4 % total count 24171 19108 2696 45975 % of total 52.6 % 41.6 % 5.9 % 100.0 % 5. conclusion the study concludes that a highly significant age difference exists between male and female education in rural areas. similar trend is found in educational enrolment of male and female. educational enrolment of female is low as compared to male children. analysis showed that several socio-economic factors are responsible for existing gender disparity with reference to education. gender disparity is found to be the maximum for lowest income group. it is high in baluchistan, followed by kp, sindh and punjab. the findings suggest that the root cause of the gender discrimination are the socio-economic factors that directly as well as indirectly affect the education attainment. as pakistan is a developing country and majority of the population live in rural areas living standards of the population is not up to the mark. income plays a great role in the attainment of education and other facilities. furthermore, gender discrimination specifically with women can be reduced by giving more opportunities them in the shape of women quotas and subsidized which will eventually lead to the reduction in gender gap enrolment. awareness campaigns about the importance of women education in the society, increasing the educational funds, and scholarships, could be other steps to enhance female enrolment and improve the facilities and infrastructure of the institutions. references afzal, m., butt, a. r., akbar, r. a., & roshi, s. (2013). gender disparity in pakistan: a case of middle and secondary education in punjab. research and reflections in education, 7(2): 113 -124. ali muhammad, zahoor ul haq & imad khan 36 ara, n., & malik, k. d. (2012). gender discrimination in educationa barrier in development of female education at higher secondary level. interdisciplinary journal of contemporary research in business, 4(5): 330-339 cooray, a. v., & potrafke, n. (2010). absence of democracy and gender inequality in education. 6th australasian development economics workshop, pp. 1-30. chaudhry, i. s., & rahman, s. u. (2009). the impact of gender inequality in education on rural poverty. european journal of economics, finance and administrative sciences, 15: 174188. faizi, w. u.-n., & butt, m. n. (2017). higher education for women in peshawar: barriers and issues. al-idah, p# 72. glewwe, p and jacoby, h. (1995). an economic analysis of delayed primary school enrolment and childhood malnutrition in a low-income country. review of economics and statistics, 77(1). irfan. (2016). gender discrimination in pakistaan. retrieved 2011, from www.pakistantoday.com.pk luqman, m., babar, s., shiewi, x., & wen, y. (2017). factors contributing to gender disparity in education in. development policy, research and practice, 1: 1-14. mian, r., rehman, m., & rehman, c. a. (2006). gender discrimination at work place: a case study on education sector of pakistan. international journal of management sciences and business research, 5(12): 104-113. qureshi, m. g. (2012). the gender differences in school enrolment and returns to education in pakistan. pide working papers, 219-256. shayan, z. (2015). gender inequality in education in afghanistan: access and barriers. open journal of philosophy, 5(5), 277. shaukat, s., siddiquah, a., & pell, a. w. (2014). gender discrimination in higher education in pakistan: a survey of university faculty. eurasian journal of educational research, 56(56). tripathi, s. (2014). importance of education. retrieved from importantindia: https://www.importantindia.com/10509/importance-of-education/ white, g., ruther, m., & kahn, j. (2016). educational inequality in india: an analysis of gender differences in reading and mathematics. in ihds working paper 2016-2. khan, h., ali, a., khan, p. r., & zia, d. a. (2013). social constraints to femlale higher education in pakhtoon society. european journal of business and social sciences, 2(4): 25-31. zarar, r., moula bukhsh, m., & khaskheli, w. (2017). causes and consequences of gender discrimination against women in quetta city. arts and social sciences journal, 08(03). undp. (1990). human development report. new york: for the united nations development program (undp). journal of applied economics and business studies, volume 1, issue 1 (2017) 1-10 https://doi.org/10.34260/jaebs.111 1 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online): 2663-693x do gross domestic product changes have asymmetric effect on india’s energy use? manzoor ahmad1*, zia ullah khan2 & shehzad khan3 1 phd scholar, department of industrial economics, nanjing university, nanjing, china 2 school of business administration, department of industrial economics, southwestern university of finance and economics, china 3 institute of business studies and leadership, abdul wali khan university mardan, khyber pakhtunkhwa pakistan abstract the existing literature on the linkage between gross domestic product (gdp) and energy use in both industrialized and developing economies usually assumes that the impacts of gross domestic product changes are symmetric. in this study, we utilized nonlinear autoregressive distributed lag (nardl) model and test whether or not the effect of variations in the gross domestic product on energy use is symmetric or asymmetric from the context of india. using time series data over 1971-2014, the findings depict that the change in the gross domestic product has a symmetric effect on energy use both in short-run and the long-run. our conclusions infer that there is no asymmetrical association between gdp and energy use, leading to support the symmetric impact of gdp on energy use. keywords gdp, energy use, nardl, india, asymmetric jel classification e10; n75;o13;q43 copyright © 2017 saebr all rights reserved 1. introduction along with globalization, energy becomes the most important inputs in the process of economic development. consequently, not only the demand for energy has increased but also a dependency on energy has rapidly risen in both developed and developing countries (ghali & el-sakka, 2004). although the demand for energy in third world economies is even now considerably lower than the global standards, there has been an upturn equivalent to industrial growth and levels of income. moreover, the elasticity coefficient computed to indicate the association between economic growth and energy use both in strong and weak economies assume values approximately equal to one, which infers that a one percent upsurge in economic growth leads increase energy use by 100 percent in developing economies (kapusuzoglu & karan, 2012). besides, the elasticity coefficient estimated between the demand for energy use and gdp is relatively less than one for developing economies. the difference between developed and developing economies from the 1*corresponding author: manzoorahmad94@163.com manzoor ahmad, zia ullah khan & shehzad khan 2 perspective of the association between the demand for energy (energy use) and economic growth generally stems from the continuously escalating need for energy in emerging economies (dorsman et al., 2012). india as the world second most populous country, while consuming only six percent of the globe’s primary energy and three-quarters of energy use are met by fossil fuels. since 2000, the demand for primary energy use has nearly increased by two times and also the potential for further fast growth is relatively high. in addition to this, the targeted policy interventions and economic growth in india have lifted more than millions out of tremendous poverty, however, per capita energy use is still close to one-third of the world average and at least 240 million masses have no access to use of electricity. in this situation, even with expanding concentration on subsidy reform and energy efficiency, there are reasons to anticipate continued rapid expansion in energy demand (iea, 2015). thus, india being a third-largest big economy; is growing swiftly and policies are in a position to continue with the country’s growing and modernization of its manufacturing. in this sense, the increase in india’s gross domestic product is a prime driver of energy trends (bakirtas & akpolat, 2018). the primary energy use of india between 2010 and 2016 are depicted in figure 1. data for the figure is obtained from statista, 2018. figure 1: primary energy use in india between 2010 and 2016, by fuel in million metric tons of oil equivalent. in 2016, the demand for renewable energy, hydro, nuclear energy, coal, natural gas, and oil were 16.5, 29.1, 8.8, 411.9, 45.1 and 212.7 million metric tons, respectively. contrariwise, figure 2 shows the position of india’s energy use with respect to total top 5 emerging economies (brics)’ aggregate energy use. 212.7 45.1 411.9 8.8 29.1 16.5 0 100 200 300 400 500 oil natural gas coal nuclear energy hydro renewable energy consumption in million metric tons of oil-equivalent p r im a r y e n e r g y c o n su m p ti o 2016 2015 2013 2012 journal of applied economics and business studies, volume 1, issue 1 (2017) 1-10 https://doi.org/10.34260/jaebs.111 3 figure 2: energy use share of each brics member nations with respect to the total energy use of brics two approaches that explain the reasons behind the existence of the nexus between economic growth and energy uses are the neoclassical and the ecological approaches. the neoclassical approach primarily considers the structure of the economy as a closed system. goods are produced by employing labor, capital and are exchanged between sellers and buyers. in this course, the higher gross domestic product is intended to be attained by raising the human capital and labor inputs. moreover, this approach postulates that rise in the capital, labor quality and technological advancement will also contribute to attaining economic growth. the neoclassical growth model involves three typical models. the first model sort out changes in technology, the second with natural resources and the last model combine the first two models (barro, 1998; stern & cleveland, 2004; ockwell, 2008). on the other hand, the ecological approach regard energy as an underlying determinant that permitting economic production (dorsman et al., 2012). there exists a disagreement on the association between energy use and economic growth in the existing literature. some of this literature include (longxing et al., 2011; chen et al., 2012; mulali et al., 2013; muhammad et al., 2015; irwan et al., 2015; faisal et al., 2015; bennouna & hebil, 2016; bah & azam, 2017; shahbaz et al., 2017; muhammad et al., 2018; mahalingam & orman, 2018; cai et al., 2018). it is noteworthy to mention that very few studies have concentrated on investigating the asymmetrical impact of gdp on energy use in india. therefore, this is the first study that attempts to inspect the asymmetry arises due to change in gross domestic product and also dissect its effect on energy use from the perspective of india. the remaining paper is divided into three sections. part two establish a nonlinear ardl model. part 3 elaborates the estimated findings, while part 4 ends up with conclusions. manzoor ahmad, zia ullah khan & shehzad khan 4 2. research methods following the methodology of (shin et al., 2013) the relationships between positive and negative components energy use and gdp are represented by the following long-run regression: eut = ϑ0 + ϑ1 +gdpt + + ϑ2 −gdpt − + εt (1) where eu is energy use integrating of order one, gdp represents gross domestic product integrating of order one, ϑ = ϑ0, ϑ1 +, ϑ2 −) is a vector of long-run unknown parameters. it is noted that ϑ1 + represent coefficients of the positive component of gdp and ϑ2 − indicated the negative component of gdp. while gdpt = gdp0 + gdpt + + gdpt − .where gdpt + and gdpt −are partial sum process of positive and negative variation in gdpt follow as; gdpt + = ∑ ∆gdpj + = t j=1 ∑ max (∆gdpj, 0) , gdpt − t j=1 = ∑ ∆gdpj − = t j=1 ∑ min (∆gdpj, 0) t j=1 (2) the equation 2 is a simple modelling to inspect asymmetrical behaviour among variables included in the model. this modelling was first employed by (schorderet, 2001) from the perspective of the nonlinear nexus between unemployment and output. following (shin et al., 2013), equation 1 can be fitted in an ardl setup under the context of (pesaran et al., 2001) as: ∆eut = ς0 + ς1eut−i + ς2 +gdpt−i + + ς3 −eut−i − + ∑ θi p i=1 ∆eut−i + ∑(ϖi +∆ m i=0 gdpt−i + + ϖi −∆gdpt−i − ) + ϵt (3) where 𝑝 and m are lag orders. ∑ ϖi +m i=0 estimates the short-run possible response of gdp increases on the energy use emissions while ∑ ϖi −m i=0 measures the short run impact of gdp reduction on co2 emissions. hence, in this setup, along with asymmetric long-term association, the asymmetric short-run impact of variations in gdp on energy use is also captured. the error correction model (ecm) of the equation 3 is depicted as: ∆ect = ∑ αi p i=1 ∆eut−i + ∑(ηi +∆gdpt−i + m i=1 + ηi −∆gdpt−i − ) + υiectt−i + ρt (4) where αi , depicts short-run coefficient and ηi + , ηi − indicate short-run adjustment symmetry. while υiindicates the coefficient of error correction term. journal of applied economics and business studies, volume 1, issue 1 (2017) 1-10 https://doi.org/10.34260/jaebs.111 5 we follow the following steps and procedures in order to estimate nardl model. in step 1, we test each series for an order of integration with the help of augmented dickey fuller (adf) and phillips-perron (pp) unit root tests (dickey & fuller, 1979; phillips & perron , 1988; elliott, et al., 1996). in the second step, we estimate equation 3, using the standard ols procedure. in step 3, the bound testing technique is carrying out to test the existence of a long-run association between variables (shin, et al., 2013; pesaran, et al., 2001). this technique is based on the wald f test having the null hypothesis of h0 (hyp1): ς1 = ς2 + = ς3 − = 0 and the null hypothesish0: ς1 ≠ ς2 + ≠ ς3 − ≠ 0. in the fourth step, the possible existence of relationship between gdp and energy use in the long-run and short-run asymmetries in established. furthermore, we also estimated the symmetric increasing “dynamic multiplier effects” of one percent difference in gdpt−i + and gdpt−i − respectively as: κb + = ∑ ∂eut_j ∂gdpt−1 + b j=0 , κb − = ∑ ∂eut_j ∂gdpt−1 − b j=0 , b = 1,2,3 … … .. (5) it is noted that as b → ∞, κb + → ϑ1 + and κb − → ϑ1 −. in the last step, the following two hypotheses are tested against alternative hypothesis: h0: (hyp2): − ς2 + ς1 ⁄ = − ς3 − ς1 ⁄ (no long run asymmetrical relationship) h1: − ς2 + ς1 ⁄ ≠ − ς3 − ς1 ⁄ ≠ 0 (long run asymmetrical relationship) h0: (hyp3): − ϖi + θi ⁄ = − ϖi − θi ⁄ (no short run asymmetrical relationship) h1: − ϖi + θi ⁄ ≠ − ϖi − θi ⁄ (short run asymmetrical relationship) the results of these hypotheses and empirical model are presented in the next section. 3. results and discussion this study used secondary data on energy use in kg oil equivalent and gross domestic product in current us dollar. all the data series have been compiled from the wdi (2018), world bank database. the data range used in this study is 1971-2014. as an important condition of time series data, adf and pp unit root tests are utilized to check the unit root in variables. the estimated results in table 1 infer that both energy use and gdp are nonstationary at levels but become stationary only at first difference under the 0.001 level of significance. moreover, the estimated outcomes of pp validated that all the designated variables can be labelled under the i(1) process. manzoor ahmad, zia ullah khan & shehzad khan 6 table 1: adf and pp unit root test level first difference variables adf pp adf pp eu 3.749478 3.619111 -4.814941** -5.042372** gdp 0.032307 -0.006617 -5.954680** -5.973082** * depicts significance at 5% the main difference between linear ardl and nonlinear ardl is that the nonlinear ardl capture asymmetries arise from the positive and negative shocks of macroeconomic variables. on the other hands, linear ardl do not account asymmetrical relationship among variables. thus, we start our analysis from the positive and negative components of gdp. figure 3 depicted the positive and negative shocks of gdp and its effect on energy use. our main objective is to compute whether or not the effect of a variation in gdp on india’s energy use is asymmetric or symmetric. table 2 presents the result of estimated nonlinear adrl coefficients of short-run as well as the long-run. the reported result of shortrun coefficients is positive for gdp increases as well as for gdp decreases. as such, the projected elasticities of gdp increase (decrease) relating to energy use is 0.033 (0.0334), infers that a 1% growth (reduction) in gdp is expected to rise (decline) energy use by 0.033% (0.0334%). it is further inferred that the estimated rise in coefficient of gdp is highly significant, nevertheless, the coefficient of gdp decreases is statistically insignificant. keeping in view, the variations in the significance level and the directions of reported elasticities, the difference in gdp indicates towards an asymmetric effect in the short-run of indian energy uses. 0.0 0.4 0.8 1.2 1.6 2.0 1975 1980 1985 1990 1995 2000 2005 2010 positive shocks of gdp -.12 -.10 -.08 -.06 -.04 -.02 .00 1975 1980 1985 1990 1995 2000 2005 2010 negative shocks of gdp figure 3: positive & negative components of gdp though, the reported result of wald test indicates that the null hypothesis of symmetry should be accepted. in contrast, the result of long run divulge that the two estimates are positive. for example, the expansions/reductions in the elasticity of gdp relating to energy use is 0.375 (0.379), recommending that increase/decrease of one percent change in gdp is likely to increase (decrease) energy use by 0.375% (0.379%). journal of applied economics and business studies, volume 1, issue 1 (2017) 1-10 https://doi.org/10.34260/jaebs.111 7 table 2: results of short-run and long-run nardl equation variable short run coefficient std. error t-statistic c 0.210570* 0.122684 1.716355 𝐿𝐸𝑈(−1) -0.088130* 0.051533 -1.710170 𝐺𝐷𝑃+ 0.032951*** 0.011565 2.849271 𝐺𝐷𝑃− 0.033371 0.035813 0.3572 ect(-1) -0.088130 0.007487 -11.77051 variable long run coefficient std. error t-statistic 𝐺𝐷𝑃+ 0.373890*** 0.103710 3.605138 𝐺𝐷𝑃− 0.378662 0.528090 0.717039 c 2.389308*** 0.021840 4.598309 nardl bound test f − statistic (hyp1) = 32.16222 [lb = 2.63, ub = 3.35 at 10%] [lb = 3.1, ub = 3.87 at 5%] [lb = 3.55, ub = 4.38 at 1%] *** and * represent 1% and 10% significance level, respectively. the estimations provided in equation (3) are statistically significant, which indicates a “co-integration” relationship between the explained variables in the model. thus, it is important to conduct a “co-integration” for the association between gdp and energy use. the most common test for this is to evaluate and compare the upper critical value to the estimated f-statistic value as suggested by the pesaran, shin, and smith (2011). when we follow this procedure, the assessed f-statistic is 32.16 and the one percent and five percent upper critical value is 3.55 (4.38). henceforth, our study strongly accepts that there is a long-run cointegration association between variables. table 3: testing hypothesis of asymmetrical effect null hypothesis f-statistic probability long run symmetries (hyp2) 1.0156 0.2455 short run symmetries (hyp3) 1.9761 0.4912 *** and * indicate that the null hypothesis is accepted at 1% and 10% level of significant besides, the previous estimation results we got in equation 3 do not suffer the spurious regression issue. another form of validating long-run relationship is that the equilibrium takes when the error-correction coefficient should not only negative but also significant. in reality, we get the negative coefficient of ect (-0.088), and this is also statistically significant at 1% levels. thus, our estimated results provide an evidence of supporting cointegration. manzoor ahmad, zia ullah khan & shehzad khan 8 table 4: results of short and long-run ardl equations variable short run coefficient std. error t-statistic c -0.145726*** 0.030242 -4.818702 𝐿𝐸𝑈(−1) -0.088478** 0.039997 -2.212100 gdp 0.032999*** 0.010578 3.119425 ect(-1) -0.088478*** 0.007517 -11.77049 variable long run coefficient std. error t-statistic gdp 0.372958*** 0.057021 6.540666 c -1.647040** 0.619384 -2.659140 nardl bound test f − statistic = 43.98235 [lb = 3.02, ub = 3.51 at 10%] [lb = 3.62, ub = 4.16 at 5%] [lb = 4.18, ub = 4.79 at 1%] *** and ** represent 1% and 5% significance level after validating the symmetric relationship between gdp and energy use, we estimate a linear ardl model. the estimated outcomes of linear ardl are reported in table 4. the findings show that there exist short-run and long-run positive relationships between gdp and energy use. ultimately, to validate and confirm our estimated model, we carry out several diagnostic tests for the residuals of our data. table 5 signifies the estimated values of different diagnostic tests. all the results of diagnostic tests divulge that our model is free of any statistical problem. table 5: validation tests test test-statistic probability white test 2.891395 0.7167 lm test 0.860972 0.6502 ramsay reset test 0.495172 0.4858 cusum stable 0.05 cusum square stable 0.05 4. conclusion changes in the gross domestic product are expected to have asymmetric effects on energy use. however, the previous literature on the nexus between gdp and energy used not directly test the asymmetry hypothesis in their works of research. thus, the current study is conducted to test whether or not the effect of changes in the gross domestic product on energy use is asymmetric from the perspective of india, currently one of the top emerging economies among developing countries and third largest end user of energy in all over the world. the findings of nonlinear ardl validate that variations in gdp have an asymmetric effect on energy use in both short-run and the long-run. in other words, the asymmetry of changes in gdp is not observed in both short-run and long-run. journal of applied economics and business studies, volume 1, issue 1 (2017) 1-10 https://doi.org/10.34260/jaebs.111 9 references bah, m. m., & azam, m., (2017). "investigating the relationship between electricity consumption and economic growth: evidence from south africa." renewable and sustainable energy reviews, 80 (2017) 531–537 bakirtas, t., & akpolat, a. g. (2018). the relationship between energy use, urbanization, and economic growth in new emerging-market countries. energy , 147, 110-121. barro, r. j. (1998). "government spending in a simple model of endogenous growth." nberworking paper, no. 2588. washington, dc: nber. bennouna, a., & hebil, c. e. (2016). "energy needs for morocco 2030, as obtained from gdp-energy and gdp-energy intensity correlations." energy policy , 88: 45-55. cai, y., sam, c. y., & chang, t. (2018). "nexus between clean energy use, economic growth and co2 emissions." journal of cleaner production, 182: 1001-1011. chen, p. y., chen, s. t., & chen, c. c. (2012). "energy use and economic growth—new evidence from meta analysis." energy policy , 44: 245-255. dickey, d. a. & fuller, w. a. (1979) distribution of the estimators for autoregressive time series with a unit root. journal of american statistical association, 74, 4273 dorsman, a., simpson, j. l., & westerman, w. (2012). "energy economics and financial markets." springer science & business media. elliott, g., (1996). efficient tests for an autoregressive unit root. econometrica, 64(4), 813-36 faisal, tursoy, t., & resatog, n. g. (2015). "energy use, electricity, and gdp causality; the case of russia, 1990-2011. 3rd global conference on business, economics, management and tourism,26-28 november 2015, rome, italy. procedia economics and finance. ghali, k. h., & el-sakka, m. i. (2004). "energy use and output in canada: a multivariate cointegration analysis." energy economics, 26, 225–238. irwan s., haseeb, m., azam, m. & islam, r., (2015). "foreign direct investment, financial development, international trade and energy consumption: panel data evidence from selected asean countries." international journal of energy economics and policy 5(3), 841-850. iea. (2015). mexico energy outlook, world energy outlook special report. oecd/iea. kapusuzoglu, a., & karan, m. b. (2012). the drivers of energy use in developing countries. in energy economics and financial markets (pp. 49-69). springer science & business media. longxing, j., meizhen, l., xiuqing, l., & zhen, z. (2011). "energy use per gdp in various regions of china and its mode. iaceed2010, (pp. 2335–2339). mahalingam, b., & orman, w. h. (2018). "gdp and energy use: a panel analysis of the us." applied economics , 213: 208-218. muhammad, a., qayyum, k. bakhtayar, b. & emirullah, c (2015). "the causal relationship between energy consumption and economic growth in the asean-5 countries." renewable & sustainable energy reviews, 47: 732–745 muhammad, a., khan, a.,q., & ihan, o., (2018). "the effects of energy on investment, human health, environment and economic growth: fresh empirical evidence from china." environmental science and pollution research doi : 10.1007/s11356-01904497-4 manzoor ahmad, zia ullah khan & shehzad khan 10 mulali, u. a., fereidouni, h. g., lee, j. y., & sab, c. n. (2013). "examining the bidirectional long run relationship between renewable energy use and gdp growth." renewable and sustainable energy reviews , 209-222. ockwell, d. (2008). "energy and economic growth: grounding our understanding in physical reality." energy policy , 36, 4600–4604. pesaran, m. h., shin, y., & smith, r. j. (2001). "bounds testing approaches to the analysis of level relationships." journal of applied econometrics, 16: 289–326. phillips, p. c. & perron, p. (1988). "testing for a unit root in time series regression. " biometrika, 75, 335-46. schorderet, y. (2001). discussion paper 2001-13; revisiting okun’s law:an hysteretic perspective. university of califonia . shahbaz, m., hoang, t. h., mahalik, m. k., & roubaud, d. (2017). "energy use, financial development and economic growth in india: new evidence from a nonlinear and asymmetric analysis." energy economics, 63: 199-212. shin, y., yu, b., & greenwoo, m. (2013). modelling asymmetric cointegration and dynamic multipliers in a nonlinear ardl framework. new york: springer. stern, d. i., & cleveland, c. j. (2004). "energy and economic growth." rensselaer working papers in economics, no. 0410). troy, ny: rensselaer polytechnic institute . world development indicators (wdi) (2018), the world bank journal of applied economics and business studies, volume 1, issue 1 (2017) 53-64 https://doi.org/10.34260/jaebs.116 53 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online): 2663-693x the impact of military spending on economic growth of pakistan sajjad1, tariq2 & muhammad tariq31* 1 assistant professor of economics at higher education department, khyber pakhtunkhwa. 2 phd scholar, department of economics, abdul wali khan university, mardan. 3 assistant professor, department of economics, abdul wali khan university, mardan abstract a sound national defence is extremely essential for a country’s sovereignty. the geostrategic position of pakistan and its deterrence policy against neighbouring india have generally been the reasons for stringent military financing. defence spending affects all sectors of the economy directly or indirectly. this study aims to investigate the influence of government military expenditures on the economic growth of pakistan over the period 1987-2016. augmented dickey-fuller test has been used for checking the unit root in the data. autoregressive distributed lag (ardl) bounds testing approach to cointegration has been applied to analyze the relationship between military spending and economic growth. the findings indicate that military expenditure has a positive impact on pakistan's economic growth in the long-run, however it has negative effect on economic growth in the short-run. keywords military spending, economic growth, ardl, pakistan jel classification h 55 copyright © 2017 saebr all rights reserved 1. introduction a sound national defence is extremely essential for a country’s sovereignty. all countries around the world defend their boundaries and control internal riots by keeping forces such as navy, air force, and the army. all citizens get benefits from the defence services of forces. likewise, these citizens also bear the entire expenses of the troops which are called military expenditures. most studies have shown that military spending is positively linked to the economic growth of less developed countries (augier et al., 2015). benoit (1973) used cross-sectional data for 44 less-developed countries and found a positive and significant association between military spending and economic growth. 1*corresponding author: tariq_noor@awkum.edu.pk sajjad, tariq & muhammad tariq 54 the proponents of keynesian economics argue that when aggregate demand is insufficient in relation to potential supply, the increase demand for military spending positively affects labour, employment, and capital stock. so efficient utilization leads to higher profit, which stimulates investment and as result growth rate increases (khilji and mahmood, 1997; benoit, 1973; and fiani et al, 1984). secondly, there is an opportunity cost of military expenditures and could also have implications for balance of payments (bop). thirdly, positive effect of military spending on growth is also anticipated in the sense that in many developing countries like pakistan military forces are engaged in the developmental projects such as constructions of schools, roads, and hospitals. they are also engaged in the provision of health care, educational facilities, and other technical trainings not only to military but also to general public. the nature of military or defence expenditures varies from region to region and country to country. the most common indicators to measure the military expenditures are the number of armed forces, the percentage of military expenditure of the country's total gross domestic product, and expenditure per military person. the stockholm international peace research institute reported recently that about 1686 billion us dollars were spent on global defence expenditure in 2016, more than 0.4 percent of real defence spending in 2015. the report also shows that most countries of the world are continually increasing their defence expenditures. similarly, most of the asian countries also increased their military spending. in terms of military expenditures, the countries of asia and oceania were the second largest regions during 2017. moreover, the governments of u.s, china, saudi arabia, and russia are spending more on their defences as compared to other nations of the world. india has also increased their military expenditure during the last decades. indian authorities have extended their spending on military by 5.5%, reaching the total amount to $ 63.9 billion in 2017 (sipri, 2017; wef, 2017). pakistan allocates and spends a huge amount on defence expenditures every year. the total outlay of budget 2017-2018 was estimated at rs. 5,103.8 billion in which defence affairs and services expenditures were estimated to be rs. 920.2 billion. the estimated increase in the defence budget is 7 percent for the financial year 2017-2018. pakistan ranks at the 23rd in world military spending and at 17th in the list of countries holding weapons. the geo-strategic position of pakistan and its least deterrence policy against neighbouring india have generally been the reason for stringent military financing. there is an immense interest in pakistan's internal security in allocating funds to the military to curb terrorism and extremism in the country. it has been discussed on various forums around the world that pakistan and its neighbouring country india have engaged in a race in acquiring weapons for the last seven decades. this race has been caused due to various security, territorial disputes, economic journal of applied economics and business studies, volume 1, issue 1 (2017) 53-64 https://doi.org/10.34260/jaebs.116 55 and political factors. therefore, this study investigates the impact of military spending on economic growth of pakistan. 2. literature review federal government budget spending on defence is one of the major concerns as this spending competes with the budget allocation in other social economic factors. the defence expenditure directly or indirectly affects all other sectors of the economy. researchers have carried out several studies to verify the impact of military expenditure on the growth of the economy. different studies show different effects of military spending on the country’s growth. some relevant literature studies show that military expenditure has a negative association with economic growth in a country, whereas most literature studies show a positive association between economic growth in a country and military expenditure. review of some of the studies is given below. khilji and mahmood (1997) studied the link between pakistan's defence spending and its economic growth between 1972 and 1995. they used different single equation models for the estimation which are widely used in defence spending literature. the study used the causality test known as granger causality for the analysis and found bi-directional feedback between the variables. the study concluded from the estimated results that in pakistan the relationship between the growth of an economy and its military spending is negative. sezgin (2000) used an error correction mechanism (ecm) for studying the impact of turkey and greece military spending on their countries ' economic growth. they investigated turkish and greece defence burden growth related to the economic growth of both countries. the study found that greece spent 5.74% of the gdp on defence spending, while turkey spent 4.42% of the gdp on defence spending. the study concluded that in both countries the effects of spending on the military demonstrate a positive association with short-term and long-term growth. baker (2007) used the global insight model for studying the impact of high military spending and expenditure of the iraq war on the u.s economy. the projections showed that the annual increase in the military spending of the u.s is 1 percent of the gdp. the estimated results showed that the u.s increasing and higher military spending on the war in iraq drain precious resources from the productive economy. the results further indicated that the higher spending on military leads to higher deficits in trade, slower economic growth, and less investment. hou (2009) used the panel and cross-sectional data for examining the nexus between indian economic growth and its expenditure on defence. the study examined numerous facets of the efforts of the indian defence. he used richardson action-reaction model for sajjad, tariq & muhammad tariq 56 studying pakistan and indian race for arms. the results indicated and supported that there exists continuing arms race between the two neighbouring countries. in addition, the estimated result indicated threat from pakistan, political status, and income as the determinant of military expenditure of india. the study concluded that indian’s defence expenditures and growth of the economy have a negative relationship. furthermore, d’agostino et al. (2010) studied military spending impact on economic growth. they used fader-ram model for estimation modifying solow and endogenous growth model using 28 countries data. the study concluded that the fader-ram model has so many weaknesses. they further indicated that both solow and endogenous growth models show that spending on military and economic growth has a negative relationship. brauoveanu (2010) used techniques of quintile analysis, cluster analysis and granger causality for analysing the association between romania’s military expenditure and growth expansion. the study examined the direction, presence, and intensity of the romanian defence spending and their growth. he concluded that due to high proportional spending on equipment and operations in the country has negatively affected economic growth in the country. ali and dimitraki (2014) used the regime-switching technique for studying the effects of chinese military expenditure on growth expansion. they used a very interesting technique of markov switching specification of two states using data from 1953 to 2010. they found that the growth of the economy and changes in military spending is dependent on the state. furthermore, they found a negative link between military spending and growth in the midst of slower growth in the chinese economy, while positive during faster economic growth. the study suggested that the increase in military spending may be harmful to growth during a slowdown growth. apanisile and okunlola (2014) studied nigerian defence spending and its effects on the growth in the economy using ardl bound testing approach. they found a negative association between economic growth and military spending in the short term, while the association was positive in the long run. the study further found that capital and labour, which is the most important determinants of economic growth in a country have a significant and positive effect in the short and long run. they concluded and suggested that the government may reduce the expenditure on the military because of the reason that military expenditure contributing nothing to the nigeria’s economy in short run. the nigerian government may concentrate on human capital development because human capital is the most important determinant of output. khalid et al. (2015) studied defence expenditure effect on u.s economic growth between 1970 and 2011, using autoregressive distributed lag (ardl) model. they investigated the externality effect on the economy of the united states of america. they focused on journal of applied economics and business studies, volume 1, issue 1 (2017) 53-64 https://doi.org/10.34260/jaebs.116 57 whether the military expenditure promotes technological progress, develop infrastructure and human capital or impeding a country’s economic growth? they found and concluded that economic growth in the us is negatively linked to national defence expenditure. augier et al. (2015) examined defence spending effect on china’s economic growth for the period 1952-2012. the study used data recently published on government expenditure, economic activity and defence spending using feder-ram and augmented solow model for estimation. they found that the feder-ram model explained china's economic growth poorly. they concluded that the augmented solow model shows that military expenditure and economic growth in china have positive links. dash et al. (2016) empirically studied the growth and defence expenditure link in bric bloc of economies for 1993-2014 using panel cointegration and causality for probing the link. the study mainly emphasized to know about the advantages of increasing military spending in that region as well as to know strategic and pre and post-war situation. the study concluded that military spending in the bric bloc economies has a positive impact on the growth in the economy. 3. research methods feder (1983) presented a model for examining military spending and economic growth relationship, which was later extended by biswas and ram in 1986. feder (1983) model was consisting of defence and civilian production function in a two-sector economy. equations (1) and (2) shows both of these models as follows. 𝐷 = 𝐷(𝐿𝑑 𝐾𝑑) (1) 𝐶 = 𝐶(𝐿𝑐 𝐾𝑐𝐷) (2) where, the terms 𝐿𝑑 and 𝐾𝑑 shows the capital and labour shares to the defence sector and 𝐿𝑐 and 𝐾𝑐 shows civilian sector labour and capital shares. as indicated by biswas and ram (1986) in equation (2) “d” shows the externality effect to the civilian sector causing by the defence sector, which is the relative productivity factor differential for labour and capital in the two sectors mentioned above. the aggregate supply of labour and capital is shown in equations (3) and (4). 𝐿 = 𝐿𝑑 𝐿𝑐 (3) 𝐾 = 𝐾𝑑 𝐾𝑐 (4) 𝑌 = 𝑀 + 𝐺 (5) by taking the total differential of the equation (5) and dividing it by y (income), it gives the following equation (6). 𝑑𝑌 𝑌 = 𝑑𝐶 𝑑𝐿 𝑑𝐿 𝑌 + 𝑑𝐺 𝑑𝐾 𝑑𝑘 𝑌 + 𝑑𝐶 𝑑𝐷 𝑑𝐷 𝑌 (6) now by multiplying the first and last terms of equation (6) with𝐿 𝐿⁄ and 𝑀 𝑀⁄ we will get the following equation (7). it indicates that how economic growth depends on weighted sajjad, tariq & muhammad tariq 58 shares of labour growth as well as capital growth and defence in output. the partial derivatives, f are then found as estimated coefficients. �̃� = 𝐾𝑙 𝐿 𝐿 𝑌 + 𝐹𝑘 𝑑𝑘 𝑌 + 𝐹𝑑 𝐷 𝐷 𝑌 (7) hence, the empirical model designed for this study is a modified form of the equation (7) shows in equation (8) as follows. 𝑙𝑛𝑌 = 𝑎1 + 𝑎2𝑙𝑛𝑙 + 𝑎3𝑙𝑛𝑘 + 𝑎4𝑙𝑛𝐷 + 𝑎5𝑙𝑛𝑅 + 𝑒 (8) whereas 𝑌 is the real gdp, 𝑙 shows labour force, 𝑘 show capital stock, d shows spending on military, and 𝑅 represents real rate of interest and ε indicates the error term. as at aggregate equilibrium in an economy total supply is equal to total demand. the modified shape of the model derived from equation (8) is therefore as follows. 𝑙𝑛 𝑌(𝑡) = 𝑎1 + 𝑎2𝑙𝑛𝐺𝐸(𝑡) + 𝑎3𝑙𝑛𝑀𝐸(𝑡) + 𝑎4𝑅(𝑡) + (𝑡) (9) in equation (9) 𝑙𝑛 𝑌(𝑡) shows the log of gross domestic product, 𝑙𝑛𝐺𝐸(𝑡) indicates the log of non-military expenditures, 𝑙𝑛𝑀𝐸(𝑡) is the log of military expenditures, 𝑅(𝑡) represents the real interest rate in annual percentage form and (𝑡) is the error term. whereas, 𝑎1 is the intercept. and a1, a2 and a3 are the respective coefficients in the model. in order to examine the shortand longterm relationship between military expenditure and pakistan’s economic growth, the auto regressive distributive lag (ardl) bound testing approach for co-integration is as follows: ∆𝐺𝐷𝑃(𝑡) = 𝑎0 + ∑ 𝑎1 𝑝 𝑖=1 ∆𝐺𝐷𝑃(𝑡 − 𝑖) + ∑ 𝑎2∆𝐺𝐸(𝑡 − 𝑖) 𝑝 𝑖=0 + ∑ 𝑎3∆𝑀𝐸(𝑡 − 𝑖) 𝑝 𝑖=0 + ∑ 𝑎3∆𝑅(𝑡 − 𝑖) 𝑝 𝑖=0 + 𝛿1𝐺𝐷𝑃(𝑡 − 1) + 𝛿2𝐺𝐸(𝑡 − 1) + 𝛿3𝑀𝐸(𝑡 − 1) + 𝛿4𝑅(𝑡 − 1) + (𝑡) (10) the ardl bound test for cointegration, which is based on the f-statistics or waldstatistics is suggested by pesaran et al. (2001). the null hypothesis shows the absence of co-integration amongst the variables, while the alternate hypothesis shows the cointegration between the variables. the acceptance or rejection of the null hypothesis is determined by the calculated value of the fstatistics and the critical value of the bound test, using the given degrees of freedom (narayan, 2005). if the estimated value of f-statistics is greater than the critical upper value, it means that the variables have a long-term relationship. however, if the calculated value of the f-statistic is lower than the lower limit, the variables do not have a long-term relationship. the long-term relationship between the variables is uncertain if the value of fstatistics is between the lower and the upper limits (hassan & kalim, 2012). 4. data this section provides details of the data, results, and discussion of the study. section 4.1 shows details of the data. this section provides information about the variables, data, and journal of applied economics and business studies, volume 1, issue 1 (2017) 53-64 https://doi.org/10.34260/jaebs.116 59 data collection sources. after that, section 4.2 presents information about the analysis of the study. the details are as follows. for the analysis, the current study used annual time series data for the period 1987 to 2016.the details of the data are given as follows in table 1. table 1: description of variables, time series annual data (1987-2016) variables definition symbols sources gross domestic product gross domestic product in pakistani millions rupees gdp world development indicators (2017) military expenditures military expenditure in pakistani millions rupees me sipri (2017) non-military government expenditures non-military government expenditures in pakistani millions rupees ge pakistan economic survey real interest rate real interest rate in annual percentage r state bank of pakistan 5. results and discussion this study used augmented dickey fuller (adf) test for checking variables stationarity (dickey and fuller, 1979, 1981). the adf test results are presented in table 2 as follows: table 2: results of the stationarity test table 2 indicates adf-test results of the stationarity for the case of pakistan. the test shows that the variable gdp is non-stationary which becomes stationary at i (1). the general government expenditure variable is also become stationary at the 1st difference at the 10 percent significance level, while military expenditure variable is also stationary at first difference while variable real rate of interest is stationary at level. as stationary results indicate that the variables for estimation are stationary at the level and at the first difference, the study can therefore confidently apply the ardl model for this analysis (khalid et al., 2015). table 3 shows the ardl bound test for co-integration. the result that is estimated illustrates that the f-calculated value is 5.345. the relevant upper and lower bounds at the level of 10% are 2.676 and 3.586 respectively. the results indicate that the calculated value variables level z(t)* critical values 1% 5% 10% p-values lgdpt at 1 st difference -5.369 -3.689 -2.971 -2.625 0.000 lget at 1 st difference -2.826 -3.689 -2.971 -2.625 0.067 lmet at 1 st difference -4.997 -2.650 -1.653 -1.609 0.000 rt at level -3.151 -3.711 -2.981 -2.629 0.035 sajjad, tariq & muhammad tariq 60 of the f-statistics is greater than the critical value of the upper bound, which suggests that co-integration exists. table 3: ardl bound test results f-statistics critical value at 1% critical value at 5% critical value at 10% i(0) i(1) i(0) i(1) i(0) i(1) 5.345 4.614 5.966 3.272 4.306 2.676 3.586 n=30, k=3 table 4 illustrates a parsimonious long-run ardl bound test for cointegration. the estimated results show a long-term positive and significant relationship between gdp(t) and ge(t), which means that the increase in general government expenditure has a positive effect on the gross domestic product in the country. so an expansion of 1% in ge(t) prompts increment per capita gdp(t) by 0.856%. similarly, r(t) also having a negative relationship with gdp(t), and khalid et al. (2015) found similar results. me(t), on the other hand, has a positive relationship with gdp(t) in the long term and is statistically significant at 1 percent. the long-term positive and significant association of military expenditure with economic growth is not so ironic because the military runs a large number of schools, hospitals, etc. in pakistan, which definitely have a positive effect on pakistan's growth. the evaluated effects of the analysis are in accordance with (sezgin, 2000; alptekin and levine, 2012; augier et al., 2015; mirza et al., 2015; dash et al., 2016). similarly, table 5 below shows the relationship of the variable in the short run. the estimated short-term coefficient illustrates that all the estimated variables are significant statistically at 1 percent level of significance. we found that military expenditure in the country has a negative effect on economic growth. table 4: estimated long run coefficients dependent variable: lgdpt variables coefficient t-statistics p-value constant 6.533* 7.228 0.000 lget 0.856* 22.866 0.000 lmet 0.443* 8.631 0.000 rt -0.022** -2.283 0.045 diagnostic test test statistics jarque-bera (normality) 1.1318 [.568] not applicable lm test (1) correlation 3.085 [.213] f(2, 8)= .5386 [.603] heteroscedasticity 2.802[.487] f(15,10)= .8413[.630] cusum test stable stable cusumq test stable stable journal of applied economics and business studies, volume 1, issue 1 (2017) 53-64 https://doi.org/10.34260/jaebs.116 61 table 5: short run results dependent variable: dlgdpt variables coefficient t-ratio p-value dlget-1 -0.619* -3.659 0.004 dlmet-1 -0.274** -4.575 0.013 dlmet-2 -0.163** -2.985 0.013 drt-1 -0.008 -0.508 0.622 drt-2 0.032* 3.387 0.006 ectt-1 -0.864* -6.116 0.000 *, ** and *** represents coefficient significant at 1, 5 and 10% percent, respectively. we also examine the stability of coefficients in the short and long run, following pesaran et al., (2001) which is cited in bahmani-oskooee (2001), the study applied the tests for checking the stability. the objective of stability test is to evaluate regression coefficient whether equation of the model is stable or not overtime. the test of stability is only apposite for annual time series data, particularly in the event that we are vague about a basic change in the data or uncertain about structural change in the data. so, for checking the stability of the coefficients cusum and cusumq statistics are plotted against boundaries which are known as critical boundaries. according to khalid et al., (2015) and bahmani-oskooee (2001), if the plotted statistics fall inside the critical boundaries, then we cannot reject the null hypothesis and conclude that the coefficients are stable. figures 1 and 2 show the plot of the two residuals cusum and cusumq that fall within the 5 percent basic limits. as clear from given graphs, both the graphs of cusum and cusumq are stable and confirms the long run stability of the coefficients. -15 -10 -5 0 5 10 15 94 96 98 00 02 04 06 08 10 12 14 16 cusum 5% significance figure 1: cumulative sum of recursive residuals sajjad, tariq & muhammad tariq 62 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 94 96 98 00 02 04 06 08 10 12 14 16 cusum of squares 5% significance figure 2: cumulative sum of squares of recursive residuals 6. conclusion in this study, we investigated the short and long-term relationship between pakistan's military spending and economic growth. time series data from 1987 to 2016 were used for the analysis. augmented dickey fuller test and autoregressive distributed lag (ardl) bounds testing approach used for data analysis. the findings indicate that military expenditure has a positive impact on pakistan's economic growth in the long-run, however it has negative effect on economic growth in the short-run. this study concluded that military spending is an important positive factor in pakistan's economic growth in the longrun. references alptekin, a., & levine, p. (2012). military expenditure and economic growth: a metaanalysis. european journal of political economy, 28, 636–650. ali, f. m., & dimitraki, o. (2014). military spending and economic growth in china: a regime-switching analysis. applied economics, 1-39. apansile, o. t., & okunlola, o. c. (2014). an empirical analysis of effects of military spending on economic growth in nigeria: a bound testing approach to co-integration 1989-2013. journal of public administration, finance and la, 6, 117-130. augier, m., mcnab, r., guo, j., & karber, p. (2015). defence spending and economic growth: evidence from china. journal of defence and peace economics, 1-26. bahmani-oskooee, m. (2001). real and nominal effective exchange rates of middle eastern countries and their trade performance. applied economics, 33, 103-111. baker, d. (2007). the economic impact of the iraq war and higher military spending. centre for economic and policy research 1611 connecticut avenue, nw, suite 400 washington, d.c. 20009. journal of applied economics and business studies, volume 1, issue 1 (2017) 53-64 https://doi.org/10.34260/jaebs.116 63 benoit, e. (1973). defence and economic growth in developing countries. boston ma: health. lexington books. biswas, r., & ram, r. (1986). military expenditure and economic growth in ldc: an augmented model and further evidence. economic development and cultural change, 34(2), 361-37. brauoveanu, l. o. (2010). the impact of defence expenditure on economic growth. romanian journal of economic forecasting, 13(4), 148-167 · d’agostino, g., dunne, j.p., & pieroni, l. (2010) assessing the effects of military expenditures on economic growth. forthcoming chapter in the oxford handbook of the economics of peace and conflict, edited by stegios skaperdas and michelle garfinkel. oxford: oxford university press. dash, d. p., bal, d. p., & sahoo, m. (2016). nexus between defence expenditure and economic growth in bric economies: an empirical investigation. theoretical and applied economics, 1, 89-102. dickey, d., & fuller, w. (1979). distribution of the estimators for autoregressive time series with a unit root. journal of the american statistical association, 74(366), 427-431. dickey, d., & fuller, w. (1981). likelihood ratio statistics for autoregressive time series with a unit root. econometrica, 49(4), 1057-1072. feder, g. (1983). on exports and economic growth. journal of development, 12(1-2), 5973. fiani, r., annez, p., & taylor, l. (1984). defence spending, economic structure, and growth: evidence among countries and over time. economic development and cultural change, 32, 487-98. hassan, m. s., & r. kalim (2012). the triangular causality among education, health and economic growth: a time series analysis of pakistan. world applied sciences journal, 18(2) 196-207. hou, n. (2009). arms race, military expenditure and economic growth in india. department of economics business school the university of birmingham. khalid, m. a., & abdul razaq, m. a. (2015). the impact of military spending on economic growth: evidence from the us economy. research journal of finance and accounting, 6(7), 183-190. khilji, n, m., & mahmood, a. (1997). military expenditure and economic growth in pakistan. pakistan development review, 36(4), 791-808. mirza, m. n., jaspal, z., & malik, a. i. (2015). military spending and economic growth in pakistan. margalla papers, 151-184. https://www.researchgate.net/journal/1582-6163_romanian_journal_of_economic_forecasting sajjad, tariq & muhammad tariq 64 narayan, p. (2005). the saving and investment nexus for china: evidence from cointegration tests. applied economics, 37(17), 1979–1990. pesaran, m. h., shin, y., & smith, r. j. (2001). bounds testing approaches to the analysis of level relationship. journal of applied econometrics, 16(3), 289–326. sezgin, s. (2000). defence expenditure and economic growth in turkey and greece: a cointegration analysis. muğlaüniversitesi sbe dergisi güz 2000 cilt: 1sayı: 1, 1-12. sipri yearbook 2017: armaments, disarmament and international security. oxford: oxford university press. world economic forum (2017). https://www.weforum.org/agenda/2018/05/india-worldsbiggest-defence-military-spender/ journal of applied economics and business studies, volume 1, issue 1 (2017) 35-44 https://doi.org/10.34260/jaebs.114 35 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online): 2663-693x an economic analysis of mobile banking in pakistan muhammad ali chohan1*, suresh ramakrishnan2 & shamaila butt1 1 phd scholar, azman hashim international business school, universiti teknologi malaysia, johor bahru, malaysia 2 senior lecturer, azman hashim international business school, universiti teknologi malaysia, johor bahru, malaysia abstract around the globe, various initiatives use the mobile phone to provide financial services to those without access to traditional banks. the current trend is that banks and telecommunication companies together are bringing the unbanked people under the financial umbrella through mobile banking. the aim of the current study is to find out the impact of mobile banking on various economic indicators of pakistan. for this purpose, the current research was conducted in two phases; in the first phase, primary data were collected through a well-structured interview schedule from the users of mobile banking. in the second phase, primary data were collected through structured interviews from the mobile industry financial experts. this study found that mobile banking brings economic benefits for individuals, companies and countries. these economic benefits lead to bigger economic effects such as an increase in gdp growth, more foreign direct investment, remittances, increase in government’s tax base, creates new businesses, employment generation and transparency in transactions. it is concluded that access to finances through mobile banking leads to the country’s gdp and the entrepreneurial activities. keywords mobile banking, gdp growth, fdi, remittances & domestic payments, government tax base jel classification g28; g29 copyright © 2017 saebr all rights reserved 1. introduction in recent years, pakistani financial institutions have gained momentum in terms of activity and effectiveness. the current trend is that banks and telecommunication companies together are bringing the unbanked people under the financial umbrella through mobile banking in pakistan. people, businesses and the country receive economic benefits from mobile banking. through mobile banking, most of population living in villages and remote areas having cell phones has access to the financial services. there mobile financial services offer many economic benefits to individuals, companies and to the country which translates into wider economic effects such as increased gdp growth, expanded entrepreneurship, foreign direct investment (fdi), high government tax base, growing domestic payments and remittances, increased transparency by limiting corruption. the mobile banking helps in improving savings in terms of time and cost 1 *corresponding author: ali.chohan@hotmail.com muhammad ali chohan, suresh ramakrishnan & shamaila butt 36 because of reduction in transportation cost. more active use of mobile banking brings subsidiary effects and improvement in family reserves, improvement in earnings, and reduction in financial failures. due to this rapid diffusion of mobile banking in pakistan, there may be opportunities for financial developments that could boost the country’s gdp in future. barnhart & barnhart (2000) described banking in terms of business keeping, lending, exchanging and issuing money. it can also be expressed as the business of bankers. at present, banks perform vital role in the development of financial markets. it offers customers an alternative financial services access compared to traditional banks. it helps introducing new products on media channels and find new competitors. this type of banking uses information technology (it) too. in the face of increasing competition among banks, it is important for the banks that they offer services to customers at lower costs. by using it resources efficiently, banks can lower costs and thus gain comparative advantage. also, using services through it tools, banks can distinguish its self from other banks that do not offer such services. anyasi & otubu (2009) found that banking industry has made enormous progress by using the banking guidelines and advanced information technology. automation in banking sector has gained importance due to a combination of regulatory rules and competition in the banking sector. the use of information technology in the banking sector improves communication, connectivity and new business development. mobile banking is commonly termed as m-banking and sms banking. it can be used to perform tasks as cheque clearing, and to make transactions. at present, mobile banking offer services via sms or mobile internet. clients download specified application software on their mobile devices to use the financial services. mobile banking is although extensively used throughout the world, but its objective and structure vary from state to state. information technology applications can be used to maintain and provide account history along with ministatements, information about term deposits, loan statements, credit and debit card statements, mutual funds/ equity statement of secondary markets, information about insurance policy, pension funds. transfer of funds nationally and internationally, micro-finance facility, bill payments can also be made through mobile banking. information about investment in various portfolios and real-time stock quotes can also be obtained through mobile banking. support services for obtaining credit, mortgage and insurance can also be requested through mobile banking services. request for cheque books and credit & debit cards can be made through mobile banking too. mobile banking applications can be used to locate the automated services machines (atms). furthermore, companies about the services can be recorded through the mobile bank’s applications. most people in pakistan have access to mobile phones and they have bank accounts too, however there is a lack of coordination between the banks and telecommunication companies. many people have mobile phones but don’t have bank accounts and vice versa. mobile banking can fill this gap by bringing these unbanked individuals under the financial umbrella. people, businesses and overall the country can gain many economic benefits by using the mobile banking. this study attempts to find out the economic journal of applied economics and business studies, volume 1, issue 1 (2017) 35-44 https://doi.org/10.34260/jaebs.114 37 benefits that pakistan can derive from using the mobile banking. analysis of economic impacts of mobile financial services warrants inclusion of both direct and indirect benefits. therefore, this research will focus on the benefits that can accrue to individuals, companies and overall country. 2. literature review banks play an important role in today's age. banks compete in introducing new products and services to get an edge in the increasingly competitive industry. provisions of services such that customers are facilitated are of utmost importance. banking industry is going through many transformations such as offering their customers services over the handheld mobile devises. such services are not constrained by the geographic locations, legal or other barriers and therefore if it has facilitated the customers on hand. the growth and coverage of wireless telecommunications has provided a platform for mobile banking services. hibberd (2007) informed there were three billion mobile phones users worldwide and it has been keep growing at a fast pace. banks have recognised the potential of phone devices to reach out to customers through advertisements. . furthermore, keen and mackintosh (2001) described that innovation in technology played an important role in the selection of mobile value-added services. he further predicted that services offered through mobile devices may further grow over time with new innovations in mobile banking. waverman et. al, (2005) found that many studies have been conducted in different countries worldwide over the use of mobile phones for banking. their study found that an increasing use of mobile devices has a positive effect on the gdp. mckinsey and company (2006) studied the indirect impact of wireless services in india, china and philippines and found that the indirect effect were at least three times more than the direct impacts. it was also described that the total wireless economic impact comprised an immediate effect on mobile operators, a subsidiary effect on telecom industry and the surplus is enjoyed by end users. it was predicted that the use of mobile services would increase by 10% in china. it was further found that it contributes $10 billion to the indian economy. vodafone (2005) reported that an increase of 10 handsets per 100 people increases the gdp by 0.6%. deloitte and touche (2007) found that a 10% increase in the use of mobile services, would lead to 1.2% gdp increase. the users of m-banking save time and money as they are not required to physically visit the bank branches. mobile users find it easy to track the transactions and thus help in better management of funds. the user of m-banking can access his bank account from a remote location. advancement and innovation of mobile technology enhanced the potential to better use the financial services available. there have major cost reductions through branchless banking and using third party networks (can be post offices and small retailers) as a financial service provider. utilization of cell phones can also facilitate provision of micro-finances. remittances play an important role in the development of a country and it has been observed that many people use mobile banking for receiving and sending remittances. transferring funds through conventional banks, western union, money gram has shown to be costly and therefore mobile banking can also play a huge role here too. mobile banking muhammad ali chohan, suresh ramakrishnan & shamaila butt 38 also facilitates the transfer of funds in small amounts into remote areas in case of emergencies. reamonn and williams (2005) found that the foreign direct investment is likely to be higher in countries where cell phones use, and coverage is more widespread. although recently there has been more scrutiny of funds through mobile banking due to the concern about funds being transferred to wrong hands. however, m-banking/m-payments has still been growing. use of m-banking services also lead to subsidiary effects, such as improvement in domestic savings, revenues generations. it changes the dynamics of family savings and distribution (reijswoud, 2007). generally speaking, mobile banking compliments the traditional banking. infodev (2006) reported that mobile banking has reduced the dependence on traditional banking system. the boston consulting group (bcg) in coordination with the telenor group studied the potential economic and social impacts of mobile banking for pakistan, bangladesh, india, serbia and malaysia. it has been found that in these countries; around 72% of population do not use financial services through mobile devices although the number of people having cell phones was similar to that of developed countries. therefore, they suggested that these people having cell phones be provided financial services through their wireless devices, so that they can come under financial umbrella. the report studied the possible economic and social impacts of each developing country under the study and also looked at regulation requited to provide such services. the bcg reported that if a country can minimize the ‘unbanked’ by 5 percent to 20 percent through mobile banking till 2020, it will boost gdp up to 5 percent. pakistan can increase its gdp by 3 percent by 2020 if the mobile financial services are provided to people and this will generate more jobs, new businesses opportunities and also result in increased government’s tax revenues. in this report, bcg showed mobile banking was an important tool for social and economic development of all the rising nations. therefore, governments and other regulatory bodies need to provide a favourable environment for the mobile banking system to flourish in future. it has been documented that a country's social and economic development depends on the access of inexpensive mobile banking services to the ‘unbanked’ people in a country. according to the bcg report, only 15 percent of people were banked while remaining 85 percent of adults were unbanked due to access issues. even among those 15 percent, only 5 percent have full access to the services of banks while 10 percent have only accessed to the basic banking services i.e. current accounts or saving accounts. as mobile financial services are growing in pakistan, this could result in better financial developments. telecommunication companies and banks are offering different mobile banking and mobile commerce services in pakistan. some of the important partners in offering the financial service through mobile devices are mobilink, telenor, ufone, warid, zong, united bank limited (ubl), mcb bank limited, standard chartered bank, bank alfalah, faysal bank, bank al-habib, dubai islamic bank and many others. 3. research methods primary data were collected through a structured interview schedule which asked the questions relating to the economic impact of mobile banking in pakistan. similarly, structured interviews were held also with the experts. information and data were extracted journal of applied economics and business studies, volume 1, issue 1 (2017) 35-44 https://doi.org/10.34260/jaebs.114 39 the from the interview schedules. the interviews were conducted in different financial institutions and multinational companies dealing in mobile financial services. the interviews were conducted with the experts of state bank of pakistan as well. the interviews were also held with senior professors of economics of different universities. the sample covers responses from 27 individuals. individuals includes 6 experts of commercial banks, 4 cfo’s of telecom companies which were dealing with mobile banking, six prominent economist of pakistan, 6 experts from sbp monetary policy division, and 5 senior professors of different universities. 4. results and discussion the following questions were frequently asked from different individuals. the feedback given by them was as follows: why there is a need of mobile banking or branchless banking in pakistan? a total of 18 experts stated that globally there is a fundamental transformation in all the sectors of the banking industry has been happening such as new products and services are introduced on continuous basis. furthermore, many people (especially adults) have mobile phones but they are unbanked in pakistan due to various reasons. therefore, most of banks and telecommunication companies realised that they can utilise the mobile networks and devices to provide basic value-added services to their customers through a single click on their cell phones. the remaining 09 experts answered that it was a way to decrease the customer's load on traditional bank branches and to facilitate the customers at home. they opined that branchless mobile banking is both cost and time effective. why banks are investing in mobile banking rather than traditional branch banking? twenty (20) experts responded that although most of the banks were providing their products and services in the urban areas of the pakistan mainly cities and developed towns through branch banking, but the banking sector also realized that most of the people still lived in rural areas which were unbanked. therefore, banks are keenly interested to bring those unbanked in their financial umbrella through mobile banking services or branchless banking. so, the banks and telecommunication companies are now investing in mobile banking rather than branch banking. the remaining 7 experts said that it was cost effective as establishing and running a physical bank branch is expensive. it is an effective way to manage the ones resources both for banks and consumer. it is cheaper and easier to manage than the conventional branches. what is the economic impact of mobile banking on gdp of pakistan? according to all experts, there is a positive impact of the mobile banking on the gdp expansion in developing countries like pakistan. the gdp growth may be high due to rapidly increasing number of mobile users. banks can generate more profits by bringing more customers under the mobile banking networks thus leading to more tax payments. mobile banking involves better quality phones which would be bought by consumers for utilizing this service thus generating revenues. muhammad ali chohan, suresh ramakrishnan & shamaila butt 40 what is the economic impact of mobile banking on fdi of pakistan? all experts commented that there was a positive impact of mobile banking on the foreign direct investment in the pakistan. the potentially large market for mobile banking may attract foreign investments in banking sectors and mobile devices and mobile networks. many international banks and telecommunication companies are already working in pakistan and they are providing mobile financial services to their customers. the foreign direct investment in this sector may also help boosting the economic growth. what is the economic impact of mobile banking on savings of a common man? according to 22 experts, mobile financial services bring unbanked people under the financial umbrella. this type of banking also saves the time and money. most people in pakistan have no access to the banks because of unavailability of branches in villages and remote areas. the mobile banking service fills this gap by providing financial services on their cell phones. the remaining 5 experts answered that mobile banking service had also indirect impacts i.e. bigger family savings, higher incomes and availability of funds on urgent basis if needed. the cost of travel and time can be utilized for earning extra income. all the information relating to the funds are available online and therefore funds can be managed more efficiently. what is the economic impact of mobile banking on remittances and domestic payments in pakistan? twenty-one (21) experts described that remittances play an important role in boosting the economic growth of a country. in pakistan, mobile banking service helps people to send and receive funds locally or internationally within a short time. mobile banking service offers a quick transfer of funds more cheaply, quickly and easily. on the other hand, the traditional funds transfer providers such as money gram and western union are charging high fees. the high cost of traditional transfer companies makes the funds transfer almost prohibitive and some time people use the illegal transfers such as hawala and other money laundering methods. therefore, mobile banking service offers a cheap and convenient method of transferring funds for low income people. the mobile banking services increase the remittances. the remaining 6 experts opined that both remittances and domestic payments would increase due to the wide spread availability of mobile banking services. how mobile banking benefits the government to increase their tax revenues? all the experts shared opinion that the government of pakistan can also benefit from mobile banking service. the government's revenue increases in tax revenues from banks. taxes on use of network services, taxes on sale of new mobiles, taxes on businesses focusing on mobile banking applications, tax collection from software and hardware companies. therefore, mobile banking service use results in increasing the tax revenues of the government. journal of applied economics and business studies, volume 1, issue 1 (2017) 35-44 https://doi.org/10.34260/jaebs.114 41 how mobile banking helps in increasing the transparency in pakistan? according to 19 experts, bribery was the most common evil in all the sectors both in private and government sectors. by increasing the documentation, and traceability of the transactions, more transparency can be maintained. banking by phone reduces the risk of fraud. you will receive a text message urgently when you perform any account activity. this includes deposits, withdrawals, transfers, as soon as any amount deducted or deposited into your account. the remaining 8 experts answered that mobile financial services helped increased the transparency in all the sectors by removing the middlemen and minimizing cash flow. furthermore, illegal activities such as hawala and other means of money laundering can be curbed. how can mobile banking reduce the economic inequality in pakistan? according to all experts, today most people in pakistan are unbanked because of the inaccessibility of traditional bank branches in the villages and remote areas. so, these ‘unbanked people’ cannot benefit from financial services such as opening and maintaining current accounts, saving accounts, local money transfers and remittance, paying utility bills, insurance etc. therefore, mobile banking service will help this financial service exclusive group of people to come under the financial umbrella and enjoys these services with their cell phones. when unbanked people have more excess to financial services, it ultimately reduces the economic inequality in the country in future by empowering them. what is the satisfaction level of mobile banking customers in pakistan? according to 20 experts, as the mobile banking was in infancy stage in pakistan such as telenor introduced “easy paisa” in october 2009 and ubl introduced “ubl omni” in april 2010. also, the number of mobile banking users has been tremendously increasing every year. this indicates the increasing level of satisfaction and trust on the mobile financial system in pakistan. thus, accessibility, affordability, safety, time saving is the key factors which increase the level of satisfaction of mobile banking users and their potential uses. the remaining 7 experts answered that the satisfaction level from the mobile bank can range from good to fair depending on the bank and the services being provided by that bank. the major advantage is that one does not have to go to the traditional banks, thus avoiding the roads congestions, and line up at the banks. discussion based on the above results, it can be concluded that there has been a major transformation in the banking sector all over the world due to mobile banking. new companies, new products and services are emerging due to mobile banking. technology plays an important role because mobile banking transformed the traditional processes of banks. hibberd (2007) and keen & mackintosh (2001) studies support the results of this study on transformation in banking sector and role of technology. due to emergence of mobile banking in developed countries, it brings some economic benefits to individuals and companies, which translate into bigger economic effects in a country. in the developing countries, mobile banking is growing too but with a little pace due to many muhammad ali chohan, suresh ramakrishnan & shamaila butt 42 issues like technology hurdles, unawareness, lack of education, lack of rules and regulations for implementing mobile financial services. one of the foremost reasons of the need of mobile banking in developing countries like pakistan is the number of people which are not under the financial umbrella and have no access to the traditional branch banking because of inaccessibility, time cost, travelling cost, and low literacy rate. in pakistan, mobile banking is flourishing gradually. this study showed that some banks and telecommunications companies individually and collectively doing investments in mobile banking system rather than traditional branch banking because of the huge potential users of mobile financial services throughout pakistan. many banks and telecommunication companies are providing mobile financial services to the people on their cell phones. as mobile banking in pakistan is in initial stage, therefore most of the companies are providing basic financial services such as monitoring accounts 24\7, balance inquiry, generating mini statements, paying utility bills, fund transfers, and sms alerts. most of the already “banked” people are using mobile banking services in pakistan while ‘unbanked’ living in villages or remote areas have showing little interest in mobile banking due to different reasons. with the passage of time, mobile banking will deepen its roots across in pakistan and individuals and companies will get the economic benefits such as increase in local reserves, availability of finances from banks, time saving, cost saving etc. these economic benefits turn in the bigger economic effects such as growth in gdp, entrepreneurship, jobs, increase in foreign direct investments (fdi), increase in remittances and domestic payments, increase in government tax revenues, increase transparency by curbing corruption (waverman et. al, 2005, reamonn & williams, 2005, and reijswoud, 2007). reaching out to unbanked people and providing them the financial inclusion through mobile banking service, there will be an increased growth in pakistan’s gdp. the bcg report also supports the results of this study that the availability of credit to the entrepreneurs will result in more jobs thus reducing the unemployment in our country especially in rural areas. therefore, mobile financial services will play a very important role in the sustainable development of pakistan in the future. 5. conclusion there is a major revolution in the banking industry. due to the mobile banking services, new products and services, new companies and new channels are being introduced every day. technology plays an important role in this major revolution in the banking industry and overcoming the legal, physical and industrial hurdles. estimated number of mobile users worldwide has increased to 3 billion dues to the advancements in wireless telecommunication services. banks, advertisement agencies and cell phone operators see the larger number of people using the wireless devices as tremendous opportunity for mobile banking. providing mobile value-added services lead to adoption of mobile banking services. due to rapid mobile diffusion and use of mobile financial services, there has been growth in economic activities. this research found that there is a great potential for mobile banking in pakistan and it has grown rapidly for the last few years. in pakistan, total number of mobile users are more than 120 million and the number of mobile banking accounts are 1.4 million. in pakistan, many banks and mobile networks have introduced the mobile banking services such as telenor and united bank journal of applied economics and business studies, volume 1, issue 1 (2017) 35-44 https://doi.org/10.34260/jaebs.114 43 limited bank. this growth in mobile banking in pakistan have been at a very fast pace and shows that there is a great potential of mobile banking in our country. through the mobile banking and mobile commerce, most of the people which were unbanked will be empowered and will come under financial umbrella. the state bank of pakistan has placed regulation on mobile banking services to ensure the security, reliability and competence of the mobile banking system to protect the concern of end users. the sbp also introduced “branchless banking regulations” in 2008 to help the new entrants in the market. this study showed that there are economic benefits from mobile banking to individuals and companies such as increase in domestic capital development, time and cost savings. the availability of credit to the entrepreneurs may improve the economic activities and help to create more jobs especially in the rural areas. it is important to make the general public more motivated towards understanding the use of mobile banking services. by using the mobile banking, they can get the information about their bank balance, their payments and funds transfer through their mobile devices at remote locations. more entrepreneurs and businesses may engage in cell phone businesses as there will be more demand for the wireless devices sets. furthermore, mobile banking can increase the entrepreneurial activities especially in rural areas on smaller scales. mobile banking can help in direct debit and multi payments authorization, related account access and multi user transaction authorizations. mobile banking offers an edge over internet banking. mobile phones can easily be connected from remote areas through 3g/4g networks and wifi networks which is sometime not possible through internet or internet is not available in many areas in pakistan. furthermore, mobile transactions made by mobile phones can be saved easily. transfer of money can be made instantly from one account in to the other account of the same bank by using mobile banking. the mobile banking has a very important role in the development of banking industry. as a part of banking industry, the mobile banking has also a big role in the economic growth of our country. banking industry in pakistan must be encouraged for competition that will help to improve future economic growth. it is concluded that mobile banking is cost effective and progressive. in this research we have consolidated the opinions of experts about the increase of mobile banking that will ultimately help the future researchers to conduct the detailed study on mobile banking. references anyasi, f. i., & otubu, p. a. (2009). mobile phone technology in banking system: its economic effect. research journal of information technology, 1(1), 1-5. barnhart, c. l., & barnhart, r. k. (eds.). (2000). the world book dictionary (vol. 2). published exclusively for world book-childcraft international. boston consulting group. (2011). financial inclusion: from obligation to opportunity. https://www.bcg.com/en-in/perspectives/28890 deloitte & touche. (2007). global mobile tax review 2006-2007, lowering mobile taxes boosts usage and fuels growth. https://www.gsma.com/mobilefordevelopment/wpcontent/uploads/2007/10/global-mobile-tax-review-2006-2007.pdf. muhammad ali chohan, suresh ramakrishnan & shamaila butt 44 hibberd, m. (2007). mobile communication international, pp. 40-44. infodev. (2006). micro-payment systems and their application to mobile networks, http://infodev.org/files/3014_file_infodev.report_m_commerce_january.2006.pdf . keen, p. g., mackintosh, r., & foreword by-heikkonen, m. (2001). the freedom economy: gaining the mcommerce edge in the era of the wireless internet. mcgrawhill professional. lydon, r., & williams, m. (2005). communications networks and foreign direct investment in developing countries. communications & strategies, (58), 43. mckinsey & company. (2006). wireless unbound, the surprising economic value and untapped potential of the mobile phone. december 2006 © copyright 2006 mckinsey & company, pp. 6-7. von reijswoud. (2007). v mobile banking an african perspective. http://www.regulateonline.org/content/view/948/63. waverman, l., meschi, m., & fuss, m. (2005). the impact of telecoms on economic growth in developing countries. the vodafone policy paper series, 2(03), 10-24 journal of applied economics and business studies, volume 1, issue 1 (2017) 11-20 https://doi.org/10.34260/jaebs.112 9 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online): 2663-693x effect of transformational leadership on employees’ job satisfaction in public sectors universities of northern areas of khyber pakhtunkhwa muhammad khalil ur rehman1, ghayyur qadir2, saqib shahzad3*1 1 ms scholar, institute of business studies and leadership, abdul wali khan university, mardan, pakistan 2 lecturer, institute of business studies and leadership, abdul wali khan university, mardan, pakistan 3 demonstrator, institute of business studies and leadership, abdul wali khan university, mardan, pakistan abstract the aim of this study is to examine the impact of transformational leadership on employees’ job satisfaction. the study was conducted in the public sectors universities of northern areas of khyber pakhtunkhwa (kp), pakistan. a sample size of 100 employees was selected using convenient sampling techniques. the data were collected from grade 16 and above employees’. a total of 130 questionnaires were distributed among these employees out of whom 100 filled questionnaires. thus, the survey participation was 76.92%. these 100 properly filled questionnaires were used for data analysis. data were analyzed through descriptive statistics, correlation analysis, and regression analysis. the result of the study shows that there is a statistically significant positive relationship between the job satisfaction and the intellectual stimulation of the transformational leadership. therefore, steps needs to be taken to convince and prepare leaders in universities that focuses on this aspect of the transformational leadership. a larger sample size with the inclusion of other universities is recommended to further validate or otherwise the results of this study. keywords transformational leadership, inspirational motivation, job satisfaction. jel classification m12; m52 copyright © 2017 saebr all rights reserved 1. introduction organizations are a social system in which the most valuable assets for the efficiency and effectiveness are the availability of skilled human resources. human resources play an important role in achieving organizational goals and objectives (mosadragh, 2003). in the more competitive business climate due to globalization, organization are doing their try best to achieve competitive advantages regarding the competency of human resources. selection of employees from different regions, cultures and backgrounds is a hard work 1*correspondence author: saqibshahzad@awkum.edu.pk muhammad khalil ur rehman, ghayyur qadir & saqib shahzad 10 and depends on the quality of the organization's leadership which they have (albion & gagliardi, 2007). the success of organizations depends upon the involvement, efforts and commitment of their labour or workforce. the major factors for organization's effectiveness are their manager’s leadership style. organizations are usually more concerned about the development and further improvement of their leadership. leadership is an activity which is mostly relevant with the people, it is the process of attracting people to achieve the goals of organizations (skansi, 2000). the awareness that a leader encourages its subordinates to achieve and reach their potential. the transformational leadership also provide the resources which are required for devolvement and achieving the goals (ansari & arasto, 2008). in addition, transformational leadership is an advisor to his subordinates for training, mentoring, encouragement and individual development. the transformational leadership serves as role a model and provide and foster climate of good work for his subordinate employees (hassan et. al, 2013). the transformational leadership has been divided into five categories which namely idealized influence (attributed), idealized influence (behaviour), inspirational motivation, individual consideration and intellectual stimulation (bass & avolio, 1998). the socialized personality of a leader and whether he or she is perceived as being confident and commitment of high order usually has an idealized influence (attributed). similarly, compelling actions of the leader is based on the value, beliefs or ideals and that can be termed as an idealized influence (behaviour). moreover, the perception of the leaders to his or her followers through mentoring, encouragement and maintaining followers to self-actualization is termed as individual consideration. furthermore, when a leader inspires and appeals to his or her followers can be referred to as inspirational motivation. finally, the extent to which a leader engages in a behaviour that causes its followers is called intellectual stimulation. leadership style plays a vital role for the employee’s job satisfaction. the behaviour and attitude of employees towards their job and organization, their success in the organization is generally called employee's job satisfaction (mohammad & hossein 2006). similarly, employee’s happiness leads to more productive and profitability for organization (sarri & judge, 2004). it has been argued that transformational leadership has an impact on employee’s job satisfaction. the aim of the study is to find out the impact transformational leadership on employee’s job satisfaction in the public sectors universities of the northern areas of kp, pakistan. this study will examine the impact of idealized influence, individualized consideration, inspirational motivation, and intellectual stimulation on employees’ job satisfaction. journal of applied economics and business studies, volume 1, issue 1 (2017) 11-20 https://doi.org/10.34260/jaebs.112 11 2. literature review review of the literature below presents a systematic overview definition of the leadership, its various types and how do they have impact on job satisfaction of the employees. the literature review is organized as follows: 2.1 leadership the systematic study of leadership began in early 1930s with the study of the “great man” (house & aditya, 1997). house & aditya, (1997) described the various traits and quality of a leadership which can distinguish leaders from non-leaders. they found that only a few traits and quality of the leader were positively related to the leader appearance and leadership effectiveness. the word “leadership” in an organization generally means the approaches adopted by the leaders in the daily activities with the workers. it included many dimensions like standards, norms, items or issues observed in job conditions and effect on worker emotional performance, personality and behaviour (lok & crawford, 2004). leadership is one of the most important topics that has been studied to examine its effcet on firms performance and jobs staifactions all over the world (kuchler, 2008). moreover, leadership is the process of convincing people of the accomplishment of the desired goals or outcomes (de jong & hartog, 2008). in addition, researchers stated that leadership performs a key role in determining the success or failure of an organization (lok & crawford, 2009). gill (2010) argued that leaders help to motivate, stimulate, encourage and recognize their followers of the accomplishments and thus further efficiency is achieved. furthermore, leaders guide his or her subordinates to do work as a team in order to complete the goals and objectives of the organizations in a timely and efficient manner (weihrich, cannice, & koontz, 2010). it has been further demonstrated that even when the employees do not agree with their boss, they will still accept the influence of the boss because of their power or position they hold in the organization (cialdini & griskevicius, 2010). effective leaders take his or her subordinates into confidence and groom them as an important part of the whole team work which affect the outcomes (bennis & nanus, 1985). similarly, leadership not only provide vision and guidelines to the follower’s leaders but also provide support, cooperation, coordination and guiding the team efforts towards achieving the common purpose and goals (morgan, 2012). recently, the term transformational leadership has been popularized which has been explained along with its types in the following sections. 2.2 transformational leadership the concept of transformational leadership was initially introduced in the study of (burns, 1978). burns (1978) focused initially on the value of the contemporary type of muhammad khalil ur rehman, ghayyur qadir & saqib shahzad 12 leadership and then related it to more people specific as world continuously and rapidly observed changes. according to him a leader is transformational when he/she supports, cooperates and encourages his or her subordinates to raise their moral and motivational levels, perceptions, beliefs and perception to the goals and objective of the organization (burns, 1978). furthermore, burns leadership theory explained that the leader is one who motivates his subordinates to do more work than originally expected of him/her (bass, 1985). transformational leadership has four components which are an idealized influence, individualized consideration, intellectual stimulation and inspirational motivation (avolio et al., 1995). idealized influence is also known as charisma. in this type of leadership styles, leaders act as a strong role model for his or her followers due to their exceptional skills and high principled, ethical and moral standards (avolio et al., 1995). moreover, idealized influence has a very strong positive relationship with the efforts of the subordinates at work (bycio, hackett & allen, 1995). in addition, idealized influence leaders are trained to take risks while maintaining ethics in their conduct and actions (bass & riggio, 2006). under the individualized types of leaderships, leaders pay attention to each and every individual needs and to get best performance out of each individual (bass & avolio, 1990). furthermore, researchers also demonstrated that in this type of leadership roles, leaders provide customized socio-emotional support to his or her subordinates to promoting and guide to get best outcomes from them as per their abilities. intellectual stimulation type of leadership is related with the activities of the leaders to challenge creativity and innovation, the leaders motivates and encourages his or her followers to find new ways of doing things, provide opportunities to learn and explore solutions to difficult problems (hater & bass, 1988). moreover, researcher stated that intellectual stimulation is the type of leadership where the leaders strongly encourages the followers to explore and go beyond their potential levels of their abilities (bycio, hackett & allen, 1995). in inspirational motivation type of leadership, leaders motivate and attract his or her followers to achieve the goals and make himself a motivation as a high achiever (bass, 1985). in addition, in this type of leadership, the leader inspires his or her subordinates by communicating his vision (shibru & darshan, 2011). 2.3 employee’s job satisfaction and leadership style job satisfaction has been discussed in several studies. researchers have also discussed different types and determinants of job satisfaction. it involved different types of satisfactions relevant to pay, supervision, promotion, working conditions, organizational journal of applied economics and business studies, volume 1, issue 1 (2017) 11-20 https://doi.org/10.34260/jaebs.112 13 practice and co-worker relationships (misener, et al., 1996). job stress is a combination of feelings or effective responses which are associated with the circumstances of performing jobs. job satisfaction is how different human being feel about different dimensions of their jobs (spector, 1997). moreover, job satisfaction is extent of positive effective orientation towards job (gurney, mueller & price, 1997). in addition, job satisfaction is the limit up to which an employee likes his or her job (stamps, 1997). the evaluation of three different items of a human being which are, his own duty, the situation or an environment in which he or she is doing a job and the contents of the job he or she really performs (mosadeghrad, 2003). furthermore, job satisfaction is present inside in the psyche of a person where an employee will express the level of his or her job satisfaction (stone & petterson, 2005). in addition, when the workers are satisfied from their job, then they have positive and pleasant attitude towards his or her job. on the other hand, if the employees are dissatisfied from his or her job so they have negative and unpleasant situation towards his or her job (armstrong, 2001). job satisfaction is an attitude which is related with overall life attitude and service quality (iles, preece & chuai, 2010). 2.4 transformational leadership and employee’s job satisfaction the leadership research examines the significance of leadership in building differences in employee’s job satisfaction. in 1950s and 1960 several studies explored manager use of his or her leadership behaviours to increase the level of employee’s job satisfaction (northouse, 2018). the leader's quality to cooperate and motivate his or her subordinates has great impact on employee’s job satisfaction and self-confidence (chen and spector 1991; brockner 1988; de cremer 2003). leader who is cooperative and supportive to his or her subordinates enjoys high level of employee’s job satisfaction than those leaders who are unsupportive and uncooperative with subordinates. leadership style is an important indicator of employee’s job satisfaction (al-ababneh, 2013). the behaviour of a leader is positively related with employee’s job satisfaction. therefore, a leader needs to follow an effective leadership behaviour because of its significant impact on employee’s job satisfaction (yousef, 2000). it has been shown that the followers have more interest in transformational leadership rather than transactional leadership (druskat, & kayes, 2000). transformational leadership and organizational effectiveness have significant relationship in the south africa cricket administration (ristow et al., 1999). some researchers also found that work withdrawal and transformational leadership has negative relationship (gill et al., 2010). muhammad khalil ur rehman, ghayyur qadir & saqib shahzad 14 3. research methods the conceptual model for this study is shown in figure 1. it has been demonstrated in the graph that the transformational leadership style that is composed of various types of the leadership styles have an impact on the employees jobs satisfactions. all the various leadership types are assumed to have significant and style relationship with the job satisfaction. figure 1: leadership style and job satisfaction this study was conducted in the public sectors universities of the northern areas of kp, pakistan. the employees from the university of malakand, university of swat, shaheed benazir bhutto university sheringal upper dir, abdul wali khan university sub-campus timergara and abdul wali khan university bacha khan campus chitral cmprised the samle for this study. transformational leadership questionnaires were adopted from the study of avolio, bass & jung (1995) while job satisfaction questionnaires were taken from the study of edward & rothbard (1999) and weiss, et al. (1967). the data were collected from grade 16 and above employees. a self-administered questionnaire was used to collect the primary data. originally, 130 questionnaires were distributed among these employees out of which 100 filled questionnaires were received forming percentage of 76.92%. these properly filled questionnaires were used for data analysis. data were analyzed through spss version 16. descriptive statistics correlation analysis, and regression analysis were used for analysis to find the relationship between transformational leadership and employees’ job satisfaction. 4. results and discussion in this section, we discuss the descriptive statistics, correlation analysis and finally the multiple regression is used to looks at factors determining the employees job satisfaction. we first begin with the descriptive statistics which are given as follows: transformational leadership idealized influence individualized consideration inspirational motivation intellectual stimulation employees’ job satisfaction journal of applied economics and business studies, volume 1, issue 1 (2017) 11-20 https://doi.org/10.34260/jaebs.112 15 4.1 descriptive statistics this section discusses the basic characteristics of our respondents in the sample such as age, gender, qualification and experience. further correlation analysis has been performed among the variables. table 1 presents the age of the respondents. there were 11 of the respondents having age 21-25 years. while, 14 of the respondents having age 26-30 years and they comprised 14 % in the sample. almost 50% of the employees belonged to the age group from 31-40 years. only about 27% of the employees had an age of greater than 4o years. this indicates that major of the employees were quite young had an age of less than 40 years. this could be due to the fact that most of the universities that we collected the data were established recently and therefore had younger staff compared to more established older universities. table 1: age of the respondents years frequency percent valid percent cumulative percent 21-25 11 11.0 11.0 11.0 26-30 14 14.0 14.0 25.0 31-35 21 21.0 21.0 46.0 36-40 27 27.0 27.0 73.0 41-45 7 7.0 7.0 80.0 46-50 7 7.0 7.0 87.0 51-55 5 5.0 5.0 92.0 56-60 3 3.0 3.0 95.0 above 60 5 5.0 5.0 100.0 total 100 100.0 100.0 source: survey data table 2 presents the qualification of the respondents in our sample. the table shows that about 65% has the graduate and post graduate qualification. about 28% had a qualification of ms/phd education. again this indicates that a smaller portion had a phd staff (9%). again it can be stated that in these new universities and with small campus, they are going through developmental stages and therefore shall take time to develop a more qualified staff and faculty at their universities. table 2: qualification of the respondents qualification frequency percent valid percent cumulative percent intermediate 7 7.0 7.0 7.0 graduate 22 22.0 22.0 29.0 post graduate 43 43.0 43.0 72.0 msm phil 19 19.0 19.0 91.0 phd 9 9.0 9.0 100.0 total 100 100.0 100.0 source: survey data table 3 exhibits gender wise distribution of our respondents. it shows that about 87 of the employees were male while only 13 of the employees’ were female. some of these muhammad khalil ur rehman, ghayyur qadir & saqib shahzad 16 universities are located in a very conservative regions of the pakistan. the education level in these regions is generally low and women generally don't work or work in a very restricted fewer organizations. similarly, these universities are located in much remote areas and therefore also struggles to attract female staff or faculty from big centres such as peshawar, islamabad or lahore areas. table 3: gender of the respondents gender frequency percent valid percent cumulative percent male 87 87.0 87.0 87.0 female 13 13.0 13.0 100.0 total 100 100.0 100.0 source: survey data table 4 shows the experience of the respondents in our sample. it can be that most of the universities in or sample had a staff with only a few years experience. about 97% staff has the experience less than 10 years. this is again natural as the universities in our sample are newer and they had the staff recently recruited so had less experience staff and faculty members. table 4: experience of the respondents experience frequency percent valid percent cumulative percent less than 1 years 19 19.0 19.0 19.0 1-2 year 25 25.0 25.0 44.0 3-5 years 35 35.0 35.0 79.0 6-10 years 18 18.0 18.0 97.0 over 10 years 3 3.0 3.0 100.0 total 100 100.0 100.0 source: survey data the given table 5 shows the correlation between job satisfaction and the various types of transformational leadership styles. it can be seen that the intellectual simulation has a positive and significant relationship with the job satisfaction. while the others leadership styles had a negative relationship with job satisfaction. these results need to be taken with care as they are not controlling for other variables. table 5: correlation analysis between job satisfaction and leadership styles idealized influence individualized consideration inspirational motivation intellectual stimulation job satisfaction idealized influence 1 individualized consideration .466** 1 inspirational motivation -0.049** 0.225** 1 intellectual stimulation -0.134** -0.116** .187** 1 job satisfaction -0.031** -0.262** -0.084** .301** 1 ** correlation is significant at the 0.01 level (2-tailled). ** correlation is significant at the 0.05 level (2-tailled). source: survey data journal of applied economics and business studies, volume 1, issue 1 (2017) 11-20 https://doi.org/10.34260/jaebs.112 17 4.2 econometrics estimation table 6 shows the econometric estimation of the effect of the transformational leadership styles on the employee's job satisfaction for those employed at various universities located in the northern areas of khyber pakhtunkhwa, pakistan. the coefficient value for idealized influence is 0.135 which shows that one unit increases in idealized influence causes increase in job satisfaction by 0.135 units. however, t value is 1.233, which shows that idealized influence does not have a statistically significant impact on job satisfaction. similarly, individualized consideration is negatively related to the job satisfaction and it is also significant. inspirational motivation is negatively related to the job satisfaction as well although it is not statistically significant and so is the case with the inspirational motivation as well. intellectual stimulation is the only varaible that has the psotive and signicant realtionship with the job satisfaction. on average a one unit increase in intellectual stimulation increases the job satisfcation by 0.226 units. the above results although may look odd but in the case of universities it is intellectual stimulation that probably matters the most for the employees to be satisfied with their job. universities are meant to be creating new knowledge and engage in intellectual work. also, the production of the intellectual work and outcomes and goals therein achieved will motivate the staff and faculty and therefore more job satisfaction. therefore, this not unexpected that the most that matters in universities is intellectual stimulation that leads to more job satisfaction. the model shows a coefficient determination of 0.16 which is not uncommon in cross sectional or survey data. the value of f is 4.655 and statistically significant indicating the all the variables in the model jointly determines the job satisfaction. table 6: effect of various factors on university's employees job satisfaction variables unstandardized coefficients standardized coefficients t sig. b std. error beta constant 2.514 .401 6.271 .000 idealized influence .135 .109 .133 1.233 .221 individualized consideration -.309 .126 -.273 -2.455 .016 inspirational motivation -.082 .113 -.072 -.723 .471 intellectual stimulation .226 .073 .300 3.096 .003 r2 0.16 f 4.655 source: survey data muhammad khalil ur rehman, ghayyur qadir & saqib shahzad 18 5. conclusion this part of the study summaries the overall study. the aim of the study was to find the impact of transformational leadership on employees’ job satisfaction in the public sectors universities of the northern area of the khyber pakhtunkhwa, pakistan. the result of the study shows that there is a statistically significant positive relationship between the job satisfaction and the intellectual stimulation of the transformational leadership. others dimension of the transformational leadership such idealized influence, individualized consideration, inspirational motivations were either negatively related or/and insignificant. this research shows that the most important for the universities was the intellectual stimulation. therefore, steps needs to be taken to convince and prepare leaders in universities that focuses on this aspect of the transformational leadership. further, this study was conducted in a few newer northern universities in kp, pakistan with a small sample size and the results may not generalize to other areas. therefore, a larger sample size with the inclusion of other universities may yield different results or validate these results of this study. there is a replication needed of this study in a wide areas and bigger sample size. references al-ababneh, m. (2013). leadership style of managers in five-star hotels and its relationship with employees job satisfaction. vol. 3, no.2, pp. 94. albion, m. j., & gagliardi, r. e. (2007). a study of transformational leadership, organisational change and job satisfaction. in proceedings of the 7th industrial & organisational psychology conference and 1st asia pacific congress on work and organisational psychology: better work. better organisations. better world (pp. 1-5). australian psychological society. ansari renani, q., & arastoo, i. (2008). emotional intelligence and transformational managers in iranian insurance companies. journal of the insurance industry, 23, 91-121. armstrong, m. (2001). a handbook of management techniques: the best-selling guide to modern management methods. kogan page publishers. avolio, b. j., bass, b. m., & jung, d. i. (1995). mlq multifactor leadership questionnaire: technical report. redwood city, ca: mindgarden. bass, b. m. (1985). leadership and performance beyond expectation. new york: free press. bass, b.m., & avolio, b.j. (1998). you can drag a horse to water but you can’t make it rink. journal of leadership studies, 5, 1-17 bass, b. m., & avolio, b. j. (1990). transformational leadership development: manual for the multifactor leadership questionnaire. consulting psychologists press. bass, b. m., & riggio, r. e. (2006). transformational leadership. psychology press. bennis, w., & nanus, b. (1985). the strategies for taking charge. leaders, new york: harper. row. brockner, j. (1988). the effects of work layoffs on survivors: research, theory, and practice. research in organizational behaviour, 10, 213-255. burns, j. m. g. (1978). leadership. new york: harper & row. journal of applied economics and business studies, volume 1, issue 1 (2017) 11-20 https://doi.org/10.34260/jaebs.112 19 bycio, p., hackett, r. d., & allen, j. s. (1995). further assessments of bass's (1985) conceptualization of transactional and transformational leadership. journal of applied psychology, 80(4), 468. chen, p. y., & spector, p. e. (1991). negative affectivity as the underlying cause of correlations between stressors and strains. journal of applied psychology, 76(3), 398. cialdini, r. b., & griskevicius, v. (2010). social influence. in r. f. baumeister & e. j. finkel (eds.), advanced social psychology: the state of the science (pp. 385-417). new york, ny, us: oxford university press. de cremer, d. (2003). why inconsistent leadership is regarded as procedurally unfair: the importance of social self‐esteem concerns. european journal of social psychology, 33(4), 535-550. de jong, j. p., & den hartog, d. n. (2008). innovative work behaviour: measurement and validation. eim business and policy research, 8(1), 1-27. druskat, v. u., & kayes, d. c. (2000). learning versus performance in short-term project teams. small group research, 31(3), 328-353. edwards, j. r., & rothbard, n. p. (1999). work and family stress and well-being: an examination of person-environment fit in the work and family domains. organizational behavior and human decision processes, 77(2), 85-129. gill, r. (2010). leadership in public services and the private sector: a comparison and the implications for handling crises and emergencies. international journal of leadership in public services, 5(4), 20-26. gill, a., fitzgerald, s., bhutani, s., mand, h., & sharma, s. (2010). the relationship between transformational leadership and employee desire for empowerment. international journal of contemporary hospitality management, 22(2), 263-273. gurney, c. a., mueller, c. w., & price, j. l. (1997). job satisfaction and organizational attachment of nurses holding doctoral degrees. nursing research, 46(3), 163-171. hassan, s., mahsud, r., yukl, g., & prussia, g. e. (2013). ethical and empowering leadership and leader effectiveness. journal of managerial psychology, 28(2), 133-146. hater, j. j., & bass, b. m. (1988). superiors' evaluations and subordinates' perceptions of transformational and transactional leadership. journal of applied psychology, 73(4), 695. house, r. j., & aditya, r. n. (1997). the social scientific study of leadership: quo vadis?. journal of management, 23(3), 409-473. iles, p., preece, d., & chuai, x. (2010). talent management as a management fashion in hrd: towards a research agenda. human resource development international, 13(2), 125-145. kuchler, w. j. (2008). perceived leadership behaviour and subordinates' job satisfaction in midwestern ncaa division iii athletic departments. the sport journal, 11(2). lok, p., & crawford, j. (2004). the effect of organisational culture and leadership style on job satisfaction and organisational commitment: a cross-national comparison. journal of management development, 23(4), 321-338. misener, t. r., haddock, k. s., gleaton, j. u., & ajamieh, a. r. a. (1996). toward an international measure of job satisfaction. nursing research, 45(2), 87-91. mohammad mosadegh rad, a., & hossein yarmohammadian, m. (2006). a study of relationship between managers' leadership style and employees' job satisfaction. leadership in health services, 19(2), 11-28. morgan, t. l. (2012). an examination of project managers' leadership contributions to project success using critical success factors (doctoral dissertation, capella university). muhammad khalil ur rehman, ghayyur qadir & saqib shahzad 20 mosadragh, a. m. (2003). the role of participative management (suggestion system) in hospital effectiveness and efficiency. research in medical sciences, 8(3), 85-89. mosadeghrad, a.m. (2003). principles of health care administration .tehran: dibagran tehran northouse, p. g. (2018). leadership: theory and practice. sage publications. ristow, a. m., amos, t. l., & staude, g. e. (1999). transformational leadership and organisational effectiveness in the administration of cricket in south africa. south african journal of business management, 30(1), 1-5. saari, l. m., & judge, t. a. (2004). employee attitudes and job satisfaction. human resource management: published in cooperation with the school of business administration, the university of michigan and in alliance with the society of human resources management, 43(4), 395-407. shibru, b., & darshan, g. m. (2011). transformational leadership and its relationship with subordinate satisfaction with the leader (the case of leather industry in ethiopia). interdisciplinary journal of contemporary research in business, 3(5), 686-697. skansi, d. (2017). relation of managerial efficiency and leadership styles–empirical study in hrvatska elektroprivreda dd. management: journal of contemporary management issues, 5(2), 51-67. spector, p. e. (1997). job satisfaction: application, assessment, causes, and consequences (vol. 3). sage publications. stamps p. l. (1997). nurses and work satisfaction: an index for measurement. chicago, il: health administration press. stone, a.g, & petterson, k. (2005). the history of leadership focus, servant leadership roundtable. virgina beach va, : regent university available at : http://www.regent.edu/acad/sls/servantleadershiproundtable weihrich, h., cannice, m. v. & koontz harold (2010). management: a global and entrepreneurial perspective, 13. weiss, d. j., dawis, r. v., england, g. w., & lofquist, l. h. (1967). manual for the minnesota satisfaction questionnaire. work adjustment project. industrial relations center, university of minnesota, 1977. yousef, d. a. (2000). organizational commitment: a mediator of the relationships of leadership behavior with job satisfaction and performance in a non-western country. journal of managerial psychology, 15(1), 6-24. journal of applied economics and business studies, volume. 4, issue 4 (2020) 1-14 https://doi.org/10.34260/jaebs.441 1 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x wage disparities between high and low wage cities with and without the cost of living within punjab and sindh: an application of oaxaca-blinder using pslm with hies farrukh mahmood*1, shumaila hashim2, uzma iram3 & muhammad zubair chishti4 1pakistan institute of development economics (pide), quaid-e-azam university, islamabad, pakistan 2applied economics research centre (aerc), university of karachi, karachi, pakistan. 3assistant professor at applied economics research centre (aerc), university of karachi, karachi, pakistan. 4quaid-e-azam university, islamabad, pakistan. abstract wage disparities research hardly incorporate for the cost of living differences due to data restriction, while the wage disparity issue is the crucial area of economist interest. the study aims to examine the wage disparities between high and low wage cities for punjab and sindh province of pakistan with and without the cost of living, deploying the data of pakistan social and living standards measurement survey (pslm) with household integrated economic survey (hies) for 2005, 2007, 2010, and 2013. applying the oaxaca-blinder estimation method, the findings infer that wage dispersion is high without the cost of living model for both provinces (punjab and sindh) as compared to with cost of the living model. moreover, the results reveal that the wage dispersion is greater in punjab province than sindh province. for policymakers, our study suggests that the cost of living is an essential component of the wage dispersion in pakistan’s cities; it should be considered while formulating for wage policy. keywords wage disparity, cost of living, micro data, oaxacablinder. jel classification i21;i25; i26; o11, q10 1. introduction in the labor market, workers have wage differences based on their skills, and hence they have a different marginal product of labor, which is often known as wage disparity. * farrukh.mahmood@jinnah.edu mailto:farrukh.mahmood@jinnah.edu farrukh mahmood, shumaila hashim, uzma iram & muhammad zubair chishti 2 it is, therefore, one of the crucial phenomena of the labor market, due to which it has always remained one of the significant areas of research for the labor economist. labor economists have found consistent wage dispersion within and between countries (beeson et al., (1989), fiss et al., (2012)). wage disparity determines distress social and moral values among individuals, and also the level of living standards among workers. the reasons for wage disparity are quite controversial and complicated. therefore, economists presented different reasons to explain wage disparity phenomenon, like; amenities, education, gender, (greenwood et al., 1991, ashraf and ashraf, 1993, partrideg et al., 2010, partridge et al., 2010, seyfried and college 2015). several studies are available on this topic, which confirmed the existence of wage disparity (nazli, (2004), jaffry et al., (2006), sabir and aftab, (2007), blau and khan 2009, partrideg et al., (2010), shah, (2010), ali and akthar (2013)). jaffry et al. (2006) confirmed the existence of wage disparity among workers across different sectors of the same skills. consequently, the wage disparity exists across provinces, across cities, and also within cities. education and gender are also a source of wage disparity (awan et al. 2007, sabir and aftab 2007, ali and akthar 2013). subsequently, individuals who have completed their education from private institutions have higher wages as compared to public institutions (awan et al. 2007, and asadullah, 2009). moreover, the believer of human capital theory, emphasize a higher level of education and work experiences, to get a higher level of wages as compared to others (becker, 1962). while some labor economist gives attention to the law of one price, which contradicts to the real-world application (mortensen, 2003), subsequently, it influences inequality in living standards. to overcome this inequality, the worker put pressure to adjust the wage accordingly (pasha & pasha, 2002). the procedure of the government is quite different, i.e., packages of wages by employers in working in multiple cities, so leading to improved equalisation of the real wages across places. this decision of the government is entirely independent of the cost of living (col); instead, government wage remuneration packages are adapted based on some political factors (dumond et al. 1999). the procedure of implementing the policy of minimum wage is quite unrealistic because the col is varying across the economy. therefore, it is fair to say that revision of minimum wage policy should be altered accordingly and set on the bases of a standard tool, i.e., col. secondly, it also quite hard to study each individual in an economy on the support of skills, abilities, and efficiency and set his wage accordingly. therefore, by using the col as a standard tool in each region for the wage disparity (combes et at. 2008). lastly, if a city has a journal of applied economics and business studies, volume. 4, issue 4 (2020) 1-14 https://doi.org/10.34260/jaebs.441 3 particular level of amenities, the col will be according to higher (dumond et al., 1999). so, the decision made on the standard of col will help to resolve all these issues. literature has been undertaken on wage disparity and col in the developed economies (roback 1988; dumond et al. 1999; azzoni et al., 2002; winters 2009; kim et al. 2009; pereira and galego, 2014; and mchenry and mcinerney, 2014, handbury, j. 2019). col (walden, 1997), nominal wage (gerking and weirick, 1983; dumond et al., 1999; kim et al., 2009; adamchik and hyclak, 2016; manduca, 2019), housing price (winters, 2009; desmond, m., & wilmers, n. 2019), labor market characteristics (melo, 2015; azzoni et al., 2002; fontes et al., 2010; garcia et al., 2002; combes et al., 2008; adamchik and hyclak, 2016; beeson & eberts, 1989; lee, n., & clarke, s. 2019), amenities (beeson & eberts, 1989; azzoni et al. 2002; kurre, 2003; yankow, 2006; atuesta & paredes, 2012; lundh 2012), and the marginal product of labor (mortensen, 2003); subsequently, intercity wage disparity. empirically, the weighted least square method and roback-blanchflower-oswald model is used by kim et al., (2009) and deller (2009). however, the oaxaca-blinder decomposition estimation technique is used by herrera et al. (2015). nazli (2004) estimated the effect of labor market characteristics on an individual’s wage. the empirical results imply that wage disparity exists in pakistan. subsequently, education-experience has a positive and statistically significant effect on wages. the impact of the experience is weaker than education. furthermore, these wage disparities were also established across the province, gender, and area. jaffry et al. (2006) concluded that the intra and inter-industry wage disparity exist, and it also extends to different sectors. sabir and aftab (2007) explored the gender-wise wage disparity in pakistan by using the labour force survey. this study employed pool and oaxacablinder decomposition methodology, and results highlighted that the change in unobserved characteristics play an essential role in shrinking the gender-wise wage disparity. the literature on wage disparity has analyzed in detail, but the inter-city wage disparity is not explored for the labour market of pakistan. it is a well-known hypothesis that there is wage disparity across pakistan, which is mostly because of the job skills, quality, and other attributes of labor. but this hypothesis is not explored empirically. furthermore, wage disparity is not analyzed at the micro-level (across cities) in the case of pakistan. subsequently, the present study also observes the consequence of the col on the wage disparity to analyze the issue of real wage disparity. in literature, the col is used as a measure of real wage disparity, but it does not reflect the accurate picture of the col since this method has faced numerous theoretical, practical difficulties farrukh mahmood, shumaila hashim, uzma iram & muhammad zubair chishti 4 (dumond et al. (1999). therefore, this study not only contributes towards the real picture of comparison based on the col but also examing the wage disparity across the cities of pakistan. this study has two primary objectives, which will be modified after estimating the wage disparities between high wage city (hwc) and low wage city (lwc) with and without the col for the province of punjab and sindh. the first objective of the study is to explore the difference between hwc and lwc within the province, i.e., punjab and sindh. consequently, the second objective of the study is to compare the difference between hwc and lwc across the provinces. the results based on the goal mentioned above will give a clue to minimize the disparity across cities and provinces, respectively. oaxaca-blinder (1973) estimation technique is used to examine wage disparity between high wage city (hwc) and low wage city (lwc) with and without the col within punjab and sindh. with and without the col comparison of average wage disparity between hwc and lwc encourage to explain the patterns of inter-city wage disparity. total wage disparity is a combination of explained and unexplained components in hwc and lwc equation, respectively. therefore, inter-city wage disparity is decomposed by with and without the col. the first section has introduced the topic. the second section is based on the literature review. section three contains a discussion on data and methodology. section four describes the empirical results based on data and methodology. lastly, section five concludes the study. 2. data and methodology: 2.1 the data the sample of the present study is based on the workers’ data of 11 large cities in pakistan.1 workers who are living in urban cities for the period of 2005, 2007, 2010, and 2013. microeconomic data of these large cities are taken from pakistan social and living standards measurement (pslm) survey with household integrated economic survey (hies) for the period of 2005, 2007, 2010 and 2013. this data is collected by the pakistan bureau of statistics (pbs). in the present study, the sample of workers is restricted between the age of 10 to 60 years.2 there are 3670 workers in the slab of 10 to 60 years in the year 2005; however, it was 3985, 3759, and 3964 in 2007, 2010, and 2013, respectively. 1 it includes eight from punjab (rawalpindi, sargodha, faisalabad, gujranwala, sialkot, lahore, multan, and bahawalpur), three from sindh (sukkur, hyderabad, and karachi) 2 described as in labour force survey lfs journal of applied economics and business studies, volume. 4, issue 4 (2020) 1-14 https://doi.org/10.34260/jaebs.441 5 data description of thevariables variable definition monthly wage monthly wage receive by paid workers with the age from 10 to 60 years as described in lfs col cost of living index is constructed by monthly expenditure of household (food and non food expenditure).3 experience square of age of worker education continuous variable of years of schooling occupation dummy variable, if worker is white collar job41 and blue collar job5 0 marital status dummy variable, if married =1, otherwise(widow, divorced and single) =0 source: author’s illustration note 2.2 methodology: economist proposed that difference in productivity leads to the wage disparity. several empirical studies have been undertaken to test the hypothesis of wage disparity statistically. the first decomposition method was employed by kitagawa (1955). later on, blinder-oaxaca (1973) enhances this model by incorporating the effect of the unobserved part. this new version of the mean difference model is mostly used in the labour market, such as wage disparity. moreover, this model is an essential tool for estimating the wage disparity. suppose 𝜔, is the wage for the individual “𝑖” which is our outcome variable of interest. assume there are two groups, i.e., high wage group and low wage group6 which is represented by 𝜔ℎ and 𝜔𝑙 , respectively. furthermore, consider 𝜔𝑖 is a vector of explained determinants, 𝑥𝑖 , according to a regression model: 𝜔𝑖 = { 𝛽𝑙 𝑥𝑖 + 𝜀𝑖 𝑙 𝑖𝑓 𝑙𝑜𝑤 𝑤𝑎𝑔𝑒 𝛽ℎ 𝑥𝑖 + 𝜀𝑖 ℎ 𝑖𝑓 ℎ𝑖𝑔ℎ 𝑤𝑎𝑔𝑒 1 where β is the vector of parameters, including intercepts. it is assumed that the value of the 𝛽ℎ is higher than the 𝛽𝑙. also, consider that the mean of the high wage individual is higher than the low wage individuals. the result is that the low wage have a lower mean value of 𝜔𝑖 than do high wage individuals. the gap between these two groups can be represented as follow; 3 the items consists of food, clothing, housing, fuel and lighting, transport, education, footwear, tobacco, misalliance expenses and others. 4 chief executives, senior officers, legislators, managers, professors and associates professors in all fields 5 clerks, services and sales workers, craft and related trade workers, plant and machine operators, and assembles 6 yearly city wise average is computed by province then found high wage and low wage city in both provinces, made pool data by using it farrukh mahmood, shumaila hashim, uzma iram & muhammad zubair chishti 6 𝜔ℎ − 𝜔𝑙 = 𝛽ℎ 𝑥ℎ − 𝛽𝑙 𝑥𝑙 2 where 𝑥ℎ and 𝑥𝑙 are the vector of explanatory variables evaluated at the means for the high and low wage, respectively. in our empirical model, col,7 experience, education, marital status, and occupation are included in the explanatory variables set. however, for simplicity consider two explanatory variables, then eq. 2 can be modified as follow; 𝜔ℎ − 𝜔𝑙 = (𝛽0 ℎ − 𝛽0 𝑙 ) + (𝛽1 ℎ 𝑥1 ℎ − 𝛽1 𝑙 𝑥1 𝑙 ) + (𝛽2 ℎ 𝑥2 ℎ − 𝛽2 𝑙 𝑥2 𝑙 ) 𝜔ℎ − 𝜔𝑙 = 𝐺0 + 𝐺1 + 𝐺2 3 so that the gap in 𝜔 between the high and the low wage can be thought of as being due in part to (i) differences in the intercepts (𝐺0), (ii) differences in 𝑥1 and 𝛽1 (𝐺1), and (iii) differences in 𝑥2 and 𝛽2 (𝐺2). the parameter of equation 2 can be estimated by substituting the mean of a sample of the 𝑥’𝑠. this process can be advance further if we calculate differently between specific groups of the explained and unexplained component. the mean difference between the two outcomes can be written in either of two ways: 𝜔ℎ − 𝜔𝑙 = ∆𝑥𝛽𝑙 + ∆𝛽𝑥ℎ 4 where ∆𝑥 = 𝑥ℎ − 𝑥𝑙 and ∆𝛽 = 𝛽ℎ − 𝛽𝑙, or as 𝜔ℎ − 𝜔𝑙 = ∆𝑥𝛽ℎ + ∆𝛽𝑥𝑙 5 these decompositions are equally valid. the first difference of the 𝑥’𝑠 is the weighted by the value of the estimated parameter of the high wage group, and the differences in the estimated parameter are weighted by the 𝑥’𝑠 of the low wage group, vice versa. in both ways, it is dividing the gap in outcomes between the low and the high wage groups. equation 4 and 5 are the special cases, while the more general decomposition is as follows: 𝜔ℎ − 𝜔𝑙 = ∆𝑥𝛽𝑙 + ∆𝛽𝑥𝑙 + ∆𝑥∆𝛽 𝜔ℎ − 𝜔𝑙 = 𝐸 + 𝐶 + 𝐶𝐸 6 the gap in mean outcomes can be divided into the endowments (𝐸), the gap in coefficients (𝐶), and a gap arising from the interaction of endowments and coefficients (𝐶𝐸). equations 4 and 5 are special cases in which 𝜔ℎ − 𝜔𝑙 = ∆𝑥𝛽𝑙 + ∆𝛽𝑥ℎ = 𝐸 + (𝐶𝐸 + 𝐶) 4′ 7 cost of living is construct through principal component analysis (pca) econometric methodology by utilizeing 37 group of 288 commodities. journal of applied economics and business studies, volume. 4, issue 4 (2020) 1-14 https://doi.org/10.34260/jaebs.441 7 𝜔ℎ − 𝜔𝑙 = ∆𝑥𝛽ℎ + ∆𝛽𝑥𝑙 = (𝐸 + 𝐶𝐸) + 𝐶 5′ hence, the first decomposition represents the unexplained part, while the second represents the explained part. oaxaca-blinder’s is a particular situation of alternative decomposition: 𝜔ℎ − 𝜔𝑙 = ∆𝑥[𝐷𝛽ℎ + (𝐼 − 𝐷)𝛽𝑙 ] + ∆𝛽[(𝐼 − 𝐷)𝛽ℎ + 𝐷𝛽𝑙 ] where 𝐼 is the identity matrix and 𝐷 a matrix of weights, in the simple case, where 𝑥 is a scalar rather than a vector, 𝐼 is equal to one, and 𝐷 is a weight. in this case, 𝐷 = 0 in the first decomposition, equation 4, and 𝐷 = 1 in the second, equation 5. 3. empirical results the present study used oaxaca-blinder (1973) econometric method for the estimation of the wage disparity in the province of sindh and punjab, between hwcand lwc, subsequently with col variable and without it. the objective of this research is to observe why the disparity of wage increases without the col for lwc and hwc workers. literature is available of wage disparity in-depth, but the wage disparity of hwc and lwc is not explored for the pakistan labour market. results of this study are explained as follow; the empirical result of the oaxaca-blinder estimation technique for wage disparity is represented in table 1 by using the wage model. wage disparity is analyzed with cost of living (w-col) and without cost of living (wo-col) for the province of punjab. furthermore, wage disparity is also investigated across hwc and lwc. in the wocol model, all the parameters are statistically significant; 9.22 is the monthly average wage for hwc, and the monthly average wage coefficient is 8.63 for lwc, yielding a disparity of wage between hwc and lwc is about 0.59 unit. this difference can further be classified into three groups. the first part of the group is the endowment term. this term represents the average increase in wages of the city that have low wages if it had similar characteristics as hwc. the second element of the group is the measures of difference in lwc when exploiting the coefficients of high wages to the lwc features. the difference in coefficient and endowment effect is measured by the third component, which is called the simultaneous effect. the second column results of table 1 represent that rs.10165 is a geometric average of hwc, whereas that of the lwc is rs.5612; the calculated difference between these geometric averages is about 44% wo-col model. conversely, in the w-col model, for hwc, the gross monthly wage is 9.30, and 8.99 is the average gross monthly wage for lwc, yielding 0.31 a wage disparity. the farrukh mahmood, shumaila hashim, uzma iram & muhammad zubair chishti 8 last column implies that there is approximately about 26% difference between the hwc and lwc of punjab. the central hypothesis of the present study is that the cost of living should be considered before setting the wage. this hypothesis is consistent with the empirical result, i.e., wage disparity w-col and wo-col, it can easily be inferred that the average monthly wage is higher in the case of w-col as compared to wo-col; therefore col should be considered while before the announcement of wage. because col is not only varying across different cities but also the hypothesis of competitive wage is rejected. the average difference has decreased in the case of the w-col model as compared to the wo-col. lastly, the number of observations are different across these two models because col is constructed by using principle component analysis of 37 group of consumption commodities of food and non-food items.8 table 1: threefold decomposition of high-wage and low-wage disparity with and wo-col in punjab province wo-col exp(b) w-col exp(b) hwc 9.22*** 10165 9.30*** 11016 lwc 8.63*** 5612 8.99*** 8066 difference 0.59*** 1.81 44% 0.31*** 1.36 26% endowment 0.25*** 1.28 35% 0.10** 1.11 33% coefficient 0.35*** 1.42 39% 0.17*** 1.19 36% interaction -0.01 0.98 26% 0.02 1.02 31% observation 1246 883 hwc 7249 615 lwc 522 268 the effect of the endowment is 1.28, which implies that the wage of lwc’s workers would increase by 28 % if they had similar attributes as hwc under the model of wo-col. however, endowment amounts are 1.11 under the model of w-col, which climaxes that if lwc had identical characteristics as hwc at that time, 11% more would be received by them. roughly endowment effect is 17% higher in wo-col model, and 1.42 is the coefficient of the component in the wo-col model and 1.19 in 8 the variable of cost of living is constructed by principle component analysis and household data of food and non-food expenditure have been employed for this purpose. the items consists of food, clothing, housing, fuel and lighting, transport, education, footwear, tobacco, misalliance expenses and others. the detail list of each items is available in hise. 9 for each year there is 1 high wage city and 1 low wage city, in each povience. the worker if high wage and low wage are mearge four years across punjab and sindh, that is why this number is seems quite low, similarly for other observations. journal of applied economics and business studies, volume. 4, issue 4 (2020) 1-14 https://doi.org/10.34260/jaebs.441 9 the w-col model, which implies that 42 % amount in wo-col model and 19% in wcol model, rise in lwc. the simultaneous effect is negative in the wo-col model, while in the w-col model, it is positive and insignificant in both models. empirically it is found that wage disparity is more prominent in hwc and lwc for the wo-col as compared to the w-col model. on averages, the high-wage, and low-wage disparities 44% in wo-col while the 26% is w-col model. the results of this study highlighted that the difference in high and low-wage is more significant by 18% in wocol. in the wo-col model, wage disparities are overestimated. in table 2 for cities of high and low-wage, w-col and wo-col models, twofold decomposition has been reported for the province of punjab. to calculate a two-fold wage decomposition model, a pooled model is used for both samples of workers considered as a reference coefficient. the conclusion of the threefold and twofold model is identical: namely, that discrimination of component or unexplained accounts greater than the interwage city gap. the results of the twofold wage decomposition model are analogous to the threefold model that accounts for more than the inter-wage city gap. table 2: twofold decomposition of higher and lower city wage disparity with and wocol in punjab province wo-col exp(b) w-col exp(b) hwc 9.22*** 10165 9.30*** 11016 lwc 8.63*** 5612 8.99*** 8066 difference 0.59*** 1.81 44% 0.31*** 1.36 26% explained 0.24*** 1.27 47% 0.12*** 1.13 48% unexplained 0.34*** 1.41 53% 0.18*** 1. 20 52% observation 1246 883 hwc 724 615 lwc 522 268 the prediction of the mean wage output decomposition model, wcol, and wocol have shown in table 3 with their differences for sindh province by employing pool data of hwc and lwc. in table 3, for hwc and lwc, two models of w-col and wcol are applied separately in table 3. the average gross monthly wage prediction in the wo-col model is 9.08 for hwc, while for lwc, 8.77 is the average total monthly wage prediction; therefore, 0.31 is a disparity of wage. moreover, for hwc, 9.14 is the gross monthly wage in the w-col model, and 9.01 is the average total monthly wage prediction for lwc, having a 0.13 wage farrukh mahmood, shumaila hashim, uzma iram & muhammad zubair chishti 10 disparity. results are summarized that the wo-col model has a higher gross monthly wage as compared to the w-col model, which implies that wage disparity is low in the w-col model. this wage disparity has three components, which are endowment, coefficient, and interaction parts. the second column of table 3 shows that rs.8852 is the log monthly wage of raw geometric average for hwc, and rs.6458 is for lwc. the amount is 27% difference in the w-col model, while rs.9343 is the raw geometric average of the monthly wage of in hwc in the w-col model, and rs.8166 is for lwc which 12% amount to a difference. the effect of the endowment is 1.05, which highlights that the wages would rise by 5% in the lwc if they had similar characteristics as hwc in the wo-col model. in contrast, in the w-col model, 1.11 is the amount of endowment, which shows that if lwc had the matching characteristics as hwc, they would obtain 11% additional. the effect of the endowment is almost 17% higher, and the coefficient of component is 1.32 in the wo-col model, and it's 1.17 in the w-col model which means that 32 % amount in wo-col model and 17% in wcol model, increase in lwc. moreover, the effect of interaction reflects differences of endowments and coefficient, which are called as the simultaneous effect, and results of this research show that in the wo-col model, simultaneous impact is negative but turn into positive in wcol model but insignificant. in the wo-col model, wage dispersion is higher than the w-col model, which implies that wage dispersion in hwc and lwc in wo-col is 27% compare to w-col model which is 12% in province of sindh. so, these results reflect that wage disparity is higher by 15% in wo-col. the wo-col model has not captured a real picture of difference. hence in wo-col model, wage disparity is overestimated table 1.3: threefold decomposition of high and low wage disparity with and wocol in sindh province. wo-col exp(b) w-col exp(b) hwc 9.08*** 8852 9.14*** 9343 lwc 8.77*** 6458 9.01*** 8166 difference 0.31*** 1.37 27% 0.13*** 1.14 12% endowment 0.05** 1.05 31% -0.03 0.96 31% coefficient 0.28*** 1.32 40% 0.16*** 1.17 37% interaction -0.02*** 0.97 29% 0.00 1.00 32% journal of applied economics and business studies, volume. 4, issue 4 (2020) 1-14 https://doi.org/10.34260/jaebs.441 11 observation 4082 3119 hwc 3191 2576 lwc 891 543 in table 4 represents the twofold decomposition of w-col and wo-col for high and lwc in sindh province. the conclusion is the equivalent of the twofold and threefold model of wage decomposition: (unexplained or component) discrimination accounts more than the inter-wage city gap. table 4: twofold decomposition of higher and lower city wage disparity with and wocol sindh province wo-col exp(b) w-col exp(b) hwc 9.08*** 8852 9.14*** 9343 lwc 8.77*** 6458 9.01*** 8166 difference 0.31*** 1.37 27% 0.13*** 1.14 12% explained 0.03* 1.03 43% -0.03 0.96 45% unexplained 0.27*** 1.32 57% 0.16*** 1. 17 55% observation 4082 3119 hwc 3191 2576 lwc 891 543 4. conclusion the present study used the oaxaca-blinder (1973) econometric model for the estimation of the wage disparity between high and low wage cities, subsequently with and without the col within punjab and sindh province. the objective of the study is to examine why this wage dispersion rises between with and without the col model for lwc and hwc workers of similar characteristics. enormous literature is available on the topic of wage disparity; however the wage disparity between cities is not examined for the labour market of pakistan. the empirical results imply that the hypothesis of competitive wage is completed rejected within and across the provinces of pakistan, i.e., punjab and sindh. subsequently, this wage disparity remains consistent with and without col. furthermore, empirical results also established that connecting the col variable in wage disparity equation leads to reduce wage disparities between high and lwc. therefore the col should be considered for the wage-fixing across the cities. farrukh mahmood, shumaila hashim, uzma iram & muhammad zubair chishti 12 references adamchik, v. a., & hyclak, t. j. (2016). regional wage disparity in poland. allied social science associations, san francisco ca. ahmed, s., & gulzar, a. (2008). inter-city variation in prices. pakistan development review, winter, 877-892. asadullah, m. n. (2009). returns to private and public education in bangladesh and pakistan: a comparative analysis. journal of asian economics, 20(1), 77-86. ashenfelter, o., & oaxaca, r. (1987). the economics of discrimination: economists enter the courtroom. the american economic review, 77(2), 321-325. atuesta, l. h., & paredes araya, d. (2012). a spatial cost of living index for colombia using a microeconomic approach and censored data. applied economics letters, vol 19, issue 18, 1799-1805. awan, m. s., & hussain, z. (2007). returns to education and gender disparity in wages in pakistan. azzoni, c. r., & servo, l. m. (2002). education, cost of living and regional wage inequality in brazil. papers in regional science,81(2), 157-175. becker, g. s. (1962). investment in human capital: a theoretical analysis. journal of political economy, 70(5, part 2), 9-49. becker, gary s. "1971. the economics of discrimination." (1957): 1976. beeson, p. e., & eberts, r. (1989). identifying productivity and amenity effects in interurban wage disparity. the review of economics and statistics, vol. 71, issue 3, 443-52. blinder, a. s. (1973). wage discrimination: reduced form and structural estimates. journal of human resources, 436-455. combes, p. p., duranton, g., & gobillon, l. (2008). spatial wage disparities: sorting matters!. journal of urban economics, volume 63, issue 2, 723-742. deller, s. (2009). wages, rent, unemployment and amenities. journal of regional analysis and policy, volume 39, issue 2, 141-154. desmond, m., & wilmers, n. (2019). do the poor pay more for housing? exploitation, profit, and risk in rental markets. american journal of sociology, 124(4), 10901124. dumond, j. m., hirsch, b. t., & macpherson, d. a. (1999). wage disparity across labour markets and workers: does cost of living matter? economic inquiry, volume 34, issue 4, 577-598. fiss pc, kennedy mt, davis gf. (2012). how golden parachutes unfolded: diffusion and variation of a controversial practice. organ. sci. 23:1077–99 journal of applied economics and business studies, volume. 4, issue 4 (2020) 1-14 https://doi.org/10.34260/jaebs.441 13 fontes, g. g., simoes, r. f., & hermeto camilo de oliveira, a. m. (2010). urban attributes and wage disparities in brazil: a multilevel hierarchical model. regional studies, volume 44, issue 5, 595-607. gerking, s. d., & weirick, w. n. (1983). compensating differences and interregional wage disparity. the review of economics and statistics, aug vol. 65, issue. 3, 483-487. greenwood, j. m., hunt, l. g., rickman, s. d., & treyz, i. g. (1991). migration, regional equilibrium, and the estimation of compensating disparity. the american economic review, 1382-1390. handbury, j. (2019). are poor cities cheap for everyone? non-homotheticity and the cost of living across us cities (no. w26574). national bureau of economic research. herrera-idárraga, p., lópez-bazo, e., & motellón, e. (2015). regional wage gaps, education, and informality in an emerging country, the case of colombia. spatial economic analysis, working paper 1/33. jaffry, s., ghulam, y., shah, v., & hasan, l. (2006). inter-industry wage disparity in pakistan. the pakistan development review, winter vol. 45, issue 4, 925-946. kim, y. k. & sax, l. j. (2009). student-faculty interaction in research universities: differences by student gender, race, social class, and first-generation status. research in higher education, 50(5), 437-459. kitagawa, e. m. (1955). components of a difference between two rates. journal of the american statistical association, 50(272), 1168-1194. kurre, j. a. (2003). is the cost of living less in rural areas?. international regional science review, 26(1), 86-116. lee, n., & clarke, s. (2019). do low-skilled workers gain from high-tech employment growth? high-technology multipliers, employment and wages in britain. research policy, 48(9), 103803. lundh, c. (2012). wage forms, costs of living and the urban-rural wage gap: southern sweden, 1881–1930. scandinavian economic history review,vol. 60 issue 2, 123-145. manduca, r. a. (2019). the contribution of national income inequality to regional economic divergence. social forces, 98(2), 622-648. mortensen, d.t. (2003) wage dispersion, why are similar worker paid differently?, mit press. nazli, h. (2004). the effect of education, experience and occupation on earnings: evidence from pakistan. lahore journal of economics, vol. 9, 1-30. farrukh mahmood, shumaila hashim, uzma iram & muhammad zubair chishti 14 oaxaca, r. (1973). male-female wage disparity in urban labour markets. international economic review, vol. 14, no. 3, 693-709. partridge, m. d. (2010). the duelling models: neg vs amenity migration in explaining us engines of growth. regional science, vol 89, issue 3, 513-536. pasha, h. a., & pasha, a. g. (2002). cost of living index by the city of pakistan. social policy and development center, pakistan.(research report 43). pereira, j., & galego, a. (2014). inter-regional wage disparity in portugal: an analysis across the wage distribution. regional studies, vol 48 issue 9, 1529-1546. roback, j. (1988). wages, rents, and amenities: differences among workers and regions. economic inquiry, vol 26, issue 1, 23-41. sabir, m., & aftab, z. (2007). dynamism in the gender wage gap: evidence from pakistan. the. the pakistan development review, vol. 46, issue 4, 865-882. seyfried, w., & college, r. (2015). examining the determinants of earnings disparity across major metropolitan areas. ,. the journal of applied business and economics, vol 17 issue 2. stiglitz, j. e., ocampo, j. a., spiegel, s., efrench-davis, r., & nayyar, d. (1997). stability with growth: macroeconomics, liberalization, and development. new york: oxford university press. walden, m. l. (1997). how much income variation "really" exists within a state? the review of regional studies, vol 27, issue 3, 237-250. winters, j. v. (2009) wages and prices: are workers fully compensated for cost of living differences? regional science and urban economics, vol 39, issue 5, 632-643. yankow, j. j. (2006). why do cities pay more? an empirical examination of some competing theories of the urban wage premium. , 60 (2),. journal of urban economics, vol 60 issue 2, 139-161. journal of applied economics and business studies, volume 1, issue 1 (2017) 21-34 https://doi.org/10.34260/jaebs.113 21 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online): 2663-693x an assessment of underground economy and tax evasion in pakistan minhaj-ud-din1*, javed iqbal2 & zia-ur-rahman3 1assistant professor, department of management sciences, abdul wali khan university mardan, timergara campus 2assistant professor, department of economics, abdul wali khan university mardan, garden campus 3 phd scholar, department of economics, abdul wali khan university mardan, garden campus abstract underground economy has serious implications for economic performance and public policy of a country. the purpose of this paper is to estimate the size of underground economy and tax evasion in pakistan for the period 1973-2016. this study uses monetary approach for estimation of size of the underground economy and tax evasion in pakistan. the results indicated that increase in taxes, intensity of regulation, and inflation were the driving force of underground economy. the estimates show that the size of underground economy in pakistan showed an increasing trend from 1974 onward and attained its maximum value in 1998. thereafter, its size exhibited decreasing trend with small fluctuations. interestingly, the impact of taxation reforms introduced in 1997 was not considerable. results indicated that the tax burden is the driving force for the existence of underground economy which need to be appropriately set and enforced. this may discourage people from indulging in underground economies. the results from this study can be used for effective policy formulations with respect to underground economy. keywords underground economy, tax evasion, tax burden jel classification e26; h26 copyright © 2017 saebr all rights reserved 1. introduction the underground economy may be defined as those activities which have no record in the official statistics and no taxes have been paid on them. these activities consist of monetary as well as non-monetary transactions. underground economy exists not only in developing countries but also in developed countries. many measures are taken such as prosecution, education and growth in different sectors of the formal economy to discourage people' and businesses' participation in the underground economy. gathering reliable information about the nature and size of underground economy is almost impossible (spiro, 1993). many studies have been undertaken to estimate the size of the underground economy, its impact and determinants in various countries. the existence of underground economy and unreliable estimates about its size, makes the data on national accounts such as unemployment rate, inflation rate, and gdp growth rate questionable. 1 *correspondence author :minhaj_din@awkum.edu.pk minhaj-ud-din, javed iqbal & zia-ur-rahman 22 ignoring the underground part of the economy in policy formulations at macro level could have strong implications for the direction of the economy. underground economy can have extremely negative impacts on cultural, social and economic development of a country, especially for pakistan which is already facing severe revenue shortfall with high fiscal deficit. it not only affects the equitable distribution of economic resources but also decreases the efficiency. since the underground and undocumented part of the economy is out of general tax net, it tends to increase tax burden on formal sector of the economy (iqbal, qureshi, & mahmood,1998). similarly, increase in the size of underground economy is a lost revenue for governments. since, underground economy is not counted in the official estimates of the national incomes, it could also have strong effect on inflation as well. there is a lack of public trust on governmental institutions, and quality of goods and services produced in the underground economy is compromised. moreover, underground economy may distort the labour market due to lack of enforcement of the labour laws. monetary transactions remain out of the tax net and national accounts either due to negligence of tax collecting authorities or due to its small size or illegality of its nature. these transactions include income from unregistered employment, assets formation from unreported work (like agriculture, housing, hunting, furs like garments and decoration, transaction in property, fishing, medical services, enterprises, restaurants and hotels, catering, transportation, etc.), working “off the books” or “moon lighting” (second jobbing) for cash, sale of information (publication of books, production of posters, video recording, etc.), informal trade (i.e. working without permit), selling home-grown/homemade products, financial instruments (like trade in bonds and stocks), covert rentals (like rent a car, house, shop, etc.), acceptance of tips, currency transaction (money changer), illegal trade in drugs (alcohol and tobacco), theft, prostitution, bribery, smuggling, gambling, corruption, begging and kidnapping (often called mafia), etc. given the size of the underground economy in many countries, it is pertinent to discuss the factors that could affect the underground economy (tanzi, 1980; thomas, 1999; yasmin, bushra, & rauf, 2003; khalid, 2002; kemal, 2007). the volume of shadow economy is associated with many economic and non-economic factors. rise in tax burden and social transfer, intensity of regulations in the labour market like reduction in working hours, early retirement, decline in tax moral, and restriction on second jobbing are some of the major economic factors which affect its size and extent. the non-economic factor like unwillingness to show the accurate income and tacit cooperation with dishonest officials are some other reasons which play vital role in the expansion of its size (zaman, & goschin, 2015; khan, & khalil, 2017). underground economy has very strong negative impacts on social, cultural and economic conditions of a country. it challenges the writ of the government by violating the established rules and regulations (e.g. no tax payment, no work permits and licenses, provision of illegal unlawful products etc.) and becomes a huge obstacle for the government to achieve the determined budgetary targets. its effect can also be seen from the perspective of economic policy making, where part of the labour force (unemployed in actual statistics) actually work and earn in the underground part of economy which leads journal of applied economics and business studies, volume 1, issue 1 (2017) 21-34 https://doi.org/10.34260/jaebs.113 23 to ineffectiveness of the macroeconomic policies (ahmad, & ahmad, 1995). one the other hand, increase in its size results in further loses in tax revenue, which not only puts additional pressure on revenue generating authorities to increase taxes in the formal sector of the economy, but also increases incentives to hide taxes and escape into the shadow economy (aslam, 1998). likewise, the unregistered firms take benefit of not paying taxes which increases the cost of production to the registered firms which pay high taxes. despite of these negative effects, the underground activities also have some positive effects. the lower cost structure in the informal sector leads to provision of wider opportunities of employment in the unregistered labour market. due to low entry cost and no permit acquisition by the informal sector's firms, their costs are lower making the unregistered firms charge lower prices than the registered ones. furthermore, such underground activities provides better competitive environment (i.e. lower prices with high sale volume) for their sustainability and growth in the long run (tanzi, 1999). keeping in view the unreliability about the extent and size of the underground economy and its detrimental effect on the formal part of economy, it is very important to estimate the size of the underground economy as well what determines the size of such economy. therefore, this study is an effort towards this end. 2. literature review this section discusses the previous literature on estimating the size of the underground economy and its effect in various parts of the world. the relevant literature is summarised as follows: many researchers used the monetary approach to estimate the size of underground economy. shabsigh (1995) used ratio of currency in circulation (cc) to demand deposits as a dependent variable and real interest rate, real per capita income, banking services and tax revenue from imports as explanatory variables. he reported that the size of the underground economy was about 20.74 % of gdp for the period from1975 -1990. similarly, ahmad and ahmad (1995) used the ratio of cc to m2, and ratio of cc plus bearer bonds to m2 as dependent variables while interest rate on time deposits, ratio of total tax revenue to gdp, and a dummy for the period 1960-71 to capture the impact of currency holdings were used as explanatory variables. they reported that underground economy declined to 35.09% in 1990 from 51.96% in 1960. ogunc and yilmaz (2000) estimated the size of underground economy in turkey by applying an indirect monetary approach using data from 1971-1999. they found that that the share of underground economy went from 13.9 % of gdp in 1971 to 20.5 % of gdp in 1999. similarly, schneider and enste (2000) used the monetary approach for different european countries in mid 1990s. they found that the underground economy as a percent of gnp for greece and italy was 27-30, for belgium, spain, and portugal was 20-24, for denmark, norway and sweden was 18-23, for germany, france, ireland, great britain, and netherlands was 13-16 and for united states, japan, switzerland, & austria was 8-10. iqbal, qureshi, & mahmood (1998) also used monetary approach to estimate the size of minhaj-ud-din, javed iqbal & zia-ur-rahman 24 the pakistan's underground economy over the period 1973-1996. they regressed the ratio of cc to m2 on the banking services, growth rate of gdp, international trade tax, interest rate on time deposits, a dummy for structural adjustment program of 1988 and lagged dependent variable. they found an upward trend as the size of the underground economy was 20.2 % of gdp in 1973 to 51.3 % of gdp in 1996. aslam (1998), khalid (2002) and kemal (2003) used ratio of cc plus foreign currency accounts to total money supply in their models. however, in explanatory variables they were quite different from one another. in aslam (1998) the explanatory variables were total tax revenues, interest rate on time deposits and a dummy for foreign currency accounts introduced in 1991. instead of using the same dummy, khalid (2002) used structural adjustment program as a dummy variable in his analysis. he also used the banking services and lagged dependent variable in his explanatory part of the model. instead of using real interest rate and the same dummy of khalid (2002) model, kemal (2003) used the dummy of aslam (1998). aslam (1998) estimates showed an upward trend from 29 % of gdp in 1960 to 43.9 % of gdp in 1990, stagnant between 1990 and 1996 at 43.8 %, and then declined to 35.5 % in 1998. khalid (2002) estimates also showed an upward trend from 13.45 % of gdp in 1976 to 28.51 % of gdp in 1998. in kemal (2003), the estimates went on increasing from 20.27 % of gdp in 1973 to 25.51 % of gdp in 1991, and then with a rapid increase it reached to 54.52 % of gdp in 1998 and then declined to 37.25 % of gdp in 2003. finally, kemal (2007) found an upward trend from 16.3 % of gdp in 1974 to 31.4 % of gdp in 2005, with highest 38.7 % value of gdp in 1998. yasmin, bushra & rauf (2004) reported upward trend from rs.12 billion to rs.1085 billion during the period under analysis (1974-2002). a possible explanation of such enormous increase in underground economy could be the absence of tax reforms over a period under study. qazi and hussain (2006) used two models by adding tax reform dummy to the model used previously by ahmad & ahmad (1995). the estimates of the 1st model reported downward trend from 51.6 % of gdp in 1960 to 20.3 % of gdp in 2003. the estimates of the 2nd model also reported a downward trend. they noted that taxation reforms played a significant role in shrinking the size of underground economy. they found a positive relationship between black economy and corporate and personal tax rates. according to them, when these rates were at its peak (60%) in 1980s the black economy was also at its peak (51.6%), and with gradual decrease in its rate the size of black economy also went on decreasing from 56 % during 1980-86, to 28 % in 1993. arby et. al. (2010) estimated the size of underground economy by using the monetary approach and applying an autoregressive-distributed lag (ardl) model by adding education as an additional factor affecting the size of shadow economy. they found that the size of the underground economy was 30 percent. similarly, about 20 percent of the overall economic transactions were taking place in the informal sector of the economy. blackburn et. al. (2012) found factors which play a role in motivating individuals and firms to conceal their true wealth to avoid taxes. in their analysis, they found that the presence of financial market imperfection, the amount of wealth disclosed by an individual and the level of financial development are the key factors which determine the degree of journal of applied economics and business studies, volume 1, issue 1 (2017) 21-34 https://doi.org/10.34260/jaebs.113 25 involvement in tax evasion and engagement in underground economy. similarly, capasso & jappelli (2013) provided a theoretical and an empirical model to study the relationship between financial development and the size of underground economy in italy. they found that local financial development (reduction in the cost of external finance) can reduce tax evasion and the size of underground economy. mughal and schneider (2018) also used the monetary approach for finding the extent of underground economy in pakistan for a period of 1973-2015 and employing the ardl and granger causality methods. they found a significantly positive relationship between the official sector and shadow economy in the long run. the size of underground economy was recorded 25% on average for the period under analysis. kireenko & nevzorova (2015) studied the effect of shadow economy on the quality of life. they used a sample of 150 countries and divided them into 5 groups based on the size of their underground economies for the period of 1999-2007. they found that the quality of life (measured by life expectance at birth and the number of children in school) had a positive association with the size of underground economy. furthermore, zaman & goschin (2015) developed an index for shadow economy in which they included three indicators: shadow economy measured in euro per inhabitants, shadow economy as percent of gdp and shadow economy of romania as percentage of the total eu-28 shadow economies for the period of 1999-2012. their results showed that the underground and formal economies were co-integrated. ferrer-i-carbonell and gerxhani (2016) estimated a relationship between tax evasion and individual wellbeing in fourteen central and eastern european countries in 2013 and 2014 by focusing on the role of institutions and social capital. they found a negative association between tax evasion and individuals’ life satisfaction. on the other hand, khan and khalil (2017) incorporated some real factors of economy like employment level, political stability, tax to gdp ratio and cost of working to estimate the size of underground economy using the data for a period of 1972-2010. for estimating the size of informal sector, they used hp-prescott filter method for obtaining the potential gdp and actual gdp series through feasible generalized least squares (fgls), and found that 71 percent of the pakistan's economy was informal. 3. research methods various approaches for estimating the size of underground economies have been discussed in the previous literature. these approaches are discussed in the following subsections. 3.1 direct approaches exclusive survey technique surveys the suppliers and consumers engaged in underground activities or services. the basic advantage of this method is that it provides information about the underground economy directly from those who are engaged in it. however, the unwillingness of respondents to cooperate presents a huge obstacle. auditing of tax return technique scrutinizes the tax payers’ files by the tax collecting minhaj-ud-din, javed iqbal & zia-ur-rahman 26 authority in depth. this approach provides information in detail about self-employed persons who have better opportunities of concealment. however, the estimates based on this technique imitate only that portion of income discovered by the relevant authority, which is likely to be a fraction of the hidden ones. the tax audits have proved that this method can’t reveal all tax evasion, and that they are limited to taxable activities only. 3.2 indirect approaches these are mostly macro-economic approaches or indicator approaches which contain information about the shadow economy. income and expenditure approach (generally referred to as the fiscal approach) states that if exact estimates of the expenditures of a nation do exist, then the difference between the expenditure and income of that nation can be used as an indicator of the presence of black economy. however, this approach is criticized on the ground that the gap between expenditure and income may increase due to other reasons, like: illness, unemployment, retirement and over-report of consumption expenditures or under-report of income level, etc. labour market approach measures the total number of hours available to informal sector in terms of monetary units by multiplying it with average productivity of the workers in shadow economy. on other hand, the decrease in availability of labour force for the official economy is linked with increase in underground economy. main problem with this approach is that we can’t measure the average productivity of workers and the number of hours spent by the workers in shadow economy accurately. the identification of second jobbing (which is often the case) is another problem with this approach. that’s why economists consider this approach as a weak indicator for the measurement of black economy. 3.3 monetary approach monetary approach developed by cagan (1958) was used to study the relation between currency ratio and tax pressure for the united state over the period 1919-1955. it was found that a number of factors like interest rate, weighted average tax rate, real per capita income, and income tax affect the currency ratio significantly. direct and indirect tax burden were both included in this model. tanzi's (1983) took the cagan (1958) approach and developed it further. he took all conventional factors in his analysis for the united states over the period 1929-1980. the model he used for his analysis is given as under: 𝑙𝑛⁡(𝐶/𝑀2)𝑡 =⁡𝛽0 +⁡𝛽1 𝑙𝑛(1 + 𝑇𝑊)𝑡 + 𝛽2𝑙𝑛(𝑊𝑆/𝑌)𝑡+𝛽3𝑙𝑛𝑅𝑡 +⁡𝛽4𝑙𝑛(y/n)𝑡 +⁡ξt (1) where β0 > 0, β1 > 0, β2 > 0, β3 < 0, β4 > 0, ln represents natural logarithms, and c/m2 is the ratio of currency in circulation to broad money supply, tw is the weighted average tax rate, ws/y is the ratio of wages and salaries to national income, r is the interest payment on saving deposits, y/n is the ratio of national income to population and finally ξ is the error term. 3.4 physical input (electricity) demand approach this is the latest developed discrepancy approach. this approach focuses on the consumption of electricity which is required to produce the national income of a country. journal of applied economics and business studies, volume 1, issue 1 (2017) 21-34 https://doi.org/10.34260/jaebs.113 27 in this method, the excessive use of electricity is then attributed to black economy. however, many of the underground activities do not use much or even no electricity (e.g. personal services), and that they often use other energy sources like gas, oil, coal, etc. for such types of activities. another problem with this approach is that it does not control for the variations in technological progress across the country which may bring a huge change in quantitative and qualitative aspects of the product. 3.5 empirical model for this study this study will use the monetary approach with slight modifications in the models. four different versions of the models given below shall be used on the data for the period from 1973 to 2016. the significance of using four models in the current study is to check the sensitivity and robustness of the explanatory variables to different dependent variables. these models are as under: [([(cc + dd + fca)/m2]⁡𝑡 = 𝛽0 +⁡𝛽1𝑇𝑌𝑡−1 +⁡𝛽2𝐵𝑆 +⁡𝛽3𝑌𝑔 +⁡𝛽4𝐼𝑁𝐹 + 𝜀𝑡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡(2) with β0, β1, β4 ˃ 0 and β2, β3 ˂ 0 [(cc + dd + fca)/m2]⁡𝑡 = 𝛽0 +⁡𝛽1𝑇𝑌𝑡−1 +⁡𝛽2𝐵𝑆 +⁡𝛽3𝐼𝑁𝐹 +⁡𝛽4[(cc + dd + fca)/m2]⁡𝑡−1 + 𝜀𝑡 (3) with β0, β1, β4 ˃ 0 and β2, β3 ˂ 0 [(cc + dd + fca)/m2]⁡𝑡 = 𝛽0 +⁡𝛽1𝑇𝑌𝑡−1 +⁡𝛽2𝐵𝑆 +⁡𝛽3𝐼𝑁𝐹 +⁡𝛽4𝐷97 + 𝜀𝑡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡ (4) with β0, β1, β3 ˃ 0 and β2, β4 ˂ 0 [(cc + dd + fca)/m2]⁡𝑡 = 𝛽0 +⁡𝛽1𝑇𝑌𝑡−1 +⁡𝛽2𝐵𝑆 +⁡𝛽3𝑌𝑔 +⁡𝛽4𝐼𝑁𝐹 + 𝛽5[(cc + dd + fca)/m2]⁡𝑡−1 +⁡𝜀𝑡 (5) with β0, β1, β4, β5 ˃ 0 and β2, β3 ˂ 0 where cc is the in currency in circulation, dd is the demand deposits, fca is the foreign currency accounts, m2 is the combination of m1, time deposits and other deposits, ty is the tax-to-gdp ratio, bs is the banking services defined as the ratio of total amount of banks deposits to total number of bank accounts, yg is the growth rate of real gdp rate in percentage per annum, inf is the inflation rate in percentage per annum, d97 is the dummy variable to capture the impact of taxation reform on underground economy(year 1997 was selected due to reason that the excise tax reform was on its peak in 1997, even though it was started in 1990) and finally ξ is the error term. 3.6 estimation of underground economy the size of the underground economy shall be estimated as follows: for each year, the liquidity holding with tax variable [{(cc+dd+fca)/m2}t ] and without tax variable [{(cc+dd+fca)/m2} wt ] was calculated for each equation through regression. the difference between [{(cc+dd+fca)/m2}t] and [{(cc+dd +fca)/m2} wt] gave us an indication that how much liquid money holding is tax-induced. in other words, it will indicate the extent to which higher level of taxes (direct and indirect) induce people to hold larger amounts of liquid money like cash, demand deposits and foreign currency accounts. the level of increased demand for liquid money is presumed to minhaj-ud-din, javed iqbal & zia-ur-rahman 28 indicate the magnitude of tax evasion, which is known by illegal money (im). mathematically; ⁡⁡𝐼𝑀 =⁡[{ 𝐶𝐶 + 𝐷𝐷 + 𝐹𝐶𝐴 𝑀2 } 𝑡 − { 𝐶𝐶 + 𝐷𝐷 + 𝐹𝐶𝐴 𝑀2 } 𝑤𝑡 ] ∗ 𝑀2⁡⁡(6) following tanzi (1983), the difference between m1 (i.e. cc+dd) and illegal money (im) was then declared to be legal money (lm) calculated as lm = m1 – im. the division of gdp on legal money will then give us the velocity of legal money. 𝑉1𝑚 =⁡[ 𝐺𝐷𝑃 𝐿𝑀 ]⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡(7) by assuming that vim = vlm, the mathematical expression for the calculation of underground economy will become as follow: underground economy (ue) = im * vim finally, the multiplication of underground economy with tax-to-gdp ratio will give us the tax evasion (te) for that country which is given as follows: 𝑇𝑎𝑥⁡𝑒𝑣𝑎𝑠𝑖𝑜𝑛⁡(𝑇𝐸) = 𝑈𝐸⁡[ 𝑇𝑎𝑥𝑒𝑠 𝐺𝐷𝑃 ]⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡(8) 4. results and discussion following the monetary or liquidity demand approach, simple ordinary last squares (ols) method is applied to models given in equations 2-5 and the results are presented in table 1. these models use the same dependant variable while different independent variables have been included to estimate their effects. in table 1, the results from the various specifications of the underground economy models are presented. results from model 1 shows that the coefficient of lagged tax-togdp ratio (tyt-1) is positive and statistically significant at 1 percent, which gives an indication of high liquid money holding if the tax rate in previous year is high. this finding seems to confirm the main hypothesis that as the tax rate increases, people engage themselves in underground activities and prefer more liquid money holdings. the coefficients of banking services (bs) and annual growth rate of gdp (yg) are negative and significant at 1 & 5 percent, respectively. it implies better banking services and economic growth can decrease the demand for liquid money holdings. coefficient of inflation rate (inf) is positive and statistically significant at 1 percent, which implies that higher inflation will motivate people to hold more liquid money to sustain the same standard of living. the explanatory variables are significantly explaining the dependent variables because r2 is 0.74 and f-statistic is also significant. the value of dw-test confirms no problem of autocorrelation. in model 2, inflation rate variable is replaced with lagged dependent variable (dept-1). results suggest that the coefficient of the lagged tax-to-gdp ratio (tyt-1) is once again positive and statistically significant at 1 percent. the coefficient of the bs is negative and lagged dependent variable (dept-1) and real growth rate is positive, and journal of applied economics and business studies, volume 1, issue 1 (2017) 21-34 https://doi.org/10.34260/jaebs.113 29 statistically significant at 1 percent, 1 percent and 10 percent respectively. values of r2, fstatistic and dw-test give the same results as discussed earlier. in model 3, due to insignificance of the lagged dependent variable (dept-1) at standard 1% and 5%, it was replaced with dummy variable (d97) just to see the impact of taxation reform to see if it had any impact on the underground economy and expected to have a negative sign. the coefficient of the tax reform dummy (d97) is negative, as expected, but turn out to be insignificant. the inclusion of this dummy was made on the ground that though statistically it is insignificant but still it has a negative relation with currency holding and the tax reforms of the 1997, which have not given any structural change to underground economy by decreasing its magnitude. the coefficient of tyt-1 and inf are positive and statistically significant at 1 percent, whereas the coefficient bs is negative but significant at 1 percent. results regarding the values of r2, f-statistic and dw-test are similar to that of previous models. in this model 4, the lagged tax-to-gdp variable (tyt-1) is replaced with the prevailing tax-to-gdp ratio variable (ty). also, the insignificant variables d97 was also replaced with a single lagged dependent variable (dept-1) and real gdp growth rate (yg). results of the estimated model show that tax-to-gdp (ty) variable is insignificant, though has positive relation with liquid money holding. this may indicate that majority of the people engaged in underground economy are backward looking and that they make their decisions regarding involvement in such type of activities while learning from the past as the lagged tax-to-gdp ratio in the initial three models was statistically significant. the coefficients of the bs and gdp growth rate (ty) are negative and statistically significant at 5 percent. the coefficient of the inf is positive and statistically significant at 1 percent, as was in the first three models. the coefficient of the lagged dependant variable (dept-1 ) is also positive and statistically significant at 10 percent. table 1: effect of various factors on the liquid money holdings in pakistan (1973-2016) model 1 model 2 model 3 model 4 constant 0.560*** (0.042) 0.465*** (0.077) 0.598*** (0.020) 0.450*** (0.081) lag of tax-to-gdp ratio 1.73*** (0.611) 0.935*** (0.337) 1.285*** (0.464) ----- tax-to-gdp ratio --- ------1.152 (0.744) banking services (total amount of banks deposits/ total number of bank accounts -487.79*** (157.516) 1641.507*** (462.959) -2109.062 *** (429.686) -402.126** (164.391) growth rate of real gdp (%) per annum -0.005** (0.002) 0.012*** (0.002) -----0.005** (0.002) inflation rate (%) per annum 0.013*** (0.001) ---0.013*** (0.001) 0.011*** (0.001) lag of the dependent variable ------0.206* (0.115) -----0.218* (0.122) dummy variable capturing impact of taxation reform on underground economy ---------0.006471 (0.024515) minhaj-ud-din, javed iqbal & zia-ur-rahman 30 (d97) r2 0.74 0.76 0.74 0.75 adj r2 0.71 0.73 0.70 0.71 s.e. of regression 0.03 0.02 0.02 0.02 sum squared resid 0.02 0.02 0.02 0.02 log likelihood 79.24 80.91 79.18 80.22 durbin-watson 1.92 2.53 2.10 2.34 f-statistic 21.71 24.63 21.61 18.08 note: * , **, *** represent the level of significance at 10%, 5% and 1 % respectively, standard errors are in given parentheses. the estimated results of the underground economy and tax evasion for all these models are reported in table 2. similarly, in figure 1& 2, the trend of the underground economy and tax evasion for various model specifications have been shown as well. results of these models are different in values from each other but they are quite similar in trends. the derived estimates have shown an upward trend from 1974 to 1998 and then have shown a downward trend up-to 2016 with a little bit fluctuation in its path. further, these estimates are highly sensitive to the value of tax evasion and velocity of money. the point at which these values were on their highest level, the values of shadow economy were also at their peak2. the significance of using four models in the current study can be viewed from the resulted values of these models. the trends of these estimates show that size of the informal sector and tax evasion change with change in specification of the model. the main significant variables in these models are inflation, banking services and tax-to-gdp ratio (with lag), which can also be seen from the values of tax evasion, where increase in tax has not only increased the tax evasion but has also increased the size of informal economy. inclusion of lagged tax-to-gdp ratio has confirmed the cited preposition that the engaged portion of society in the underground economy is quite backward looking where they learn from the past while making their decision about involvement in such sort of activities. 2 the estimated value of the velocity of legal money in the current study is greater than its estimated value by state bank of pakistan. the reason is the difference in specification of formulas. estimating its value, the state bank has divided gdp by m1, while in current study velocity of legal money is considered to be the ratio of gdp to legal money. since legal money (lm) is less than m1 that’s why velocity of legal money in this study is greater than velocity of money calculated by state bank of pakistan, i.e. vlm > v m1. journal of applied economics and business studies, volume 1, issue 1 (2017) 21-34 https://doi.org/10.34260/jaebs.113 31 table 2: resulted values of velocity of money, underground economy and tax evasion years as percent of gdp model 1 model 2 model 3 model 4 vm ue te vm ue te vm ue te vm ue te 1974 9.9 10.01 0.46 9.5 5.16 0.23 9.7 7.24 0.33 9.9 9.77 0.45 1975 11.7 13.24 0.61 11.1 7.93 0.37 11.3 9.49 0.44 11.4 10.91 0.5 1976 11.4 13.84 0.65 10.9 8.84 0.41 11 9.91 0.47 11.2 11.65 0.55 1977 10.4 13.72 0.66 10 8.87 0.42 10 9.83 0.47 10.2 11.61 0.56 1978 9.8 14.46 0.76 9.4 9.34 0.49 9.5 10.34 0.54 9.7 12.98 0.68 1979 9.1 15.66 0.83 8.7 9.93 0.53 8.8 11.16 0.59 8.9 13.1 0.69 1980 8.8 15.99 1.01 8.3 10.27 0.65 8.4 11.39 0.72 8.8 15.56 0.98 1981 8.3 18.58 1.25 7.7 11.47 0.77 7.9 13.15 0.88 8.1 16.11 1.08 1982 8.5 20.28 1.36 7.9 12.67 0.85 8 14.29 0.96 8.2 16.6 1.11 1983 8.5 21.44 1.37 7.9 13.59 0.87 8 15.06 0.96 8.1 16.88 1.08 1984 9 21.27 1.39 8.4 13.69 0.89 8.5 14.95 0.97 8.7 17.89 1.16 1985 8.7 21.23 1.3 8.1 13.48 0.82 8.2 14.92 0.91 8.3 16.48 1 1986 9 19.93 1.16 8.5 12.93 0.75 8.6 14.05 0.81 8.7 15.79 0.91 1987 8.9 17.89 0.96 8.4 11.67 0.63 8.5 12.68 0.68 8.6 13.96 0.75 1988 8.3 15.84 0.89 7.9 10.46 0.59 8 11.28 0.63 8.1 13.74 0.77 1989 8.7 16.01 0.92 8.3 10.29 0.59 8.4 11.4 0.66 8.6 13.56 0.78 1990 9 16.63 0.92 8.5 10.66 0.59 8.6 11.82 0.66 8.8 13.35 0.74 1991 9.3 17.14 0.89 8.8 11.12 0.57 8.9 12.17 0.63 9 13.39 0.69 1992 9.4 17.75 1.02 8.9 11.62 0.66 9 12.58 0.72 9.3 16.17 0.93 1993 9.9 22.13 1.26 9.2 13.78 0.78 9.4 15.52 0.88 9.6 18.04 1.03 1994 10.1 24.14 1.41 9.4 15.19 0.89 9.5 16.85 0.98 9.8 20.3 1.19 1995 11.1 24.75 1.47 10.3 15.49 0.92 10.5 17.25 1.03 10.8 20.59 1.23 1996 12.1 27.7 1.74 11.1 17.2 1.08 11.3 19.17 1.2 11.7 23.75 1.49 1997 13.3 34.92 2.23 12 21.06 1.35 12.2 23.75 1.52 12.7 28.68 1.79 1998 13.9 38.69 2.35 12.3 23.23 1.41 12.6 26.08 1.58 13 29.48 1.83 1999 10.8 26.6 1.48 10 16.85 0.93 10.1 18.46 1.02 10.3 20.1 1.12 2000 10.8 22.42 1.2 10.1 14.56 0.78 10.2 15.72 0.84 10.4 17.95 0.96 2001 11.6 22.96 1.24 10.8 14.74 0.8 11 16.07 0.87 11.3 19.11 1.03 2002 11.7 23.29 1.12 10.9 14.74 0.71 11 16.29 0.78 11.1 17.27 0.83 2003 10.6 18.74 0.87 10 12.35 0.57 10.1 13.25 0.61 10.2 15.04 0.69 2004 9.4 17.1 0.78 9 11.19 0.51 9 12.14 0.55 9.2 14.08 0.64 2005 8.8 17.01 0.81 8.4 11.03 0.52 8.5 12.08 0.57 8.7 14.67 0.7 2006 8.5 17.59 0.91 8 11.22 0.58 8.1 12.47 0.64 8.3 15.66 0.81 2007 7.8 17.54 0.9 7.4 11.07 0.61 7.4 12.44 0.69 7.7 15.39 0.85 2008 7.5 18.3 0.88 7 11.53 0.58 7.1 12.90 0.65 7.3 16.50 0.81 2009 7.3 17.3 0.86 6.7 11 0.6 6.9 13.1 0.62 7.0 16.10 0.78 2010 7 16.44 0.9 6.5 11.10 0.7 6.5 12.76 0.67 6.9 15.93 0.76 2011 6.8 16.25 0.85 6.3 10.93 0.67 6.1 12.53 0.7 6.5 15.47 0.8 2012 6.5 15.98 0.84 6.1 10.56 0.64 6.3 12.21 0.63 6.6 14.73 0.76 2013 6.3 15.12 0.82 5.9 10.42 0.61 6.1 12.10 0.6 6.4 14.13 0.71 2014 6.2 14.58 0.8 5.8 10.27 0.58 5.9 12.00 0.6 6.1 13.74 0.67 2015 6.1 13.75 0.71 5.7 9.83 0.54 5.7 11.93 0.58 6.1 13.65 0.64 2016 5.9 13.51 0.68 5.6 9.57 0.5 5.6 11.83 0.55 6 13.31 0.61 where velocity of legal money (vm), underground economy (ue), tax evasion (te) minhaj-ud-din, javed iqbal & zia-ur-rahman 32 figure 1: estimates of underground economy (1974-2016) figure 2: estimates of tax evasion (1974-2016) there could be various reasons of decrease in the size of informal sector and tax evasion after reaching to the highest level in 1998. for instance, rise in private investment level, increase in smuggling, taxation reform in, and better policies of the government like increase in the growth rate of gdp and ease in credit facility to the private sector played a vital role in shrinking the size of underground economy. it worthy to mention here that the derived estimates of these models have not only found a single peak point for its maximum value but have also got smaller variation in the derived estimates by following almost similar trends. a possible explanation of the differences in results found in this study in comparison with other studies is the difference may be the specification of the models. the usage of tax-to-gdp ratio (with lag and 0 5 10 15 20 25 30 35 40 45 1970 1980 1990 2000 2010 2020 model 1 model 2 model 3 model 4 0 0.5 1 1.5 2 2.5 3 1970 1980 1990 2000 2010 2020 model 1 model 2 model 3 model 4 journal of applied economics and business studies, volume 1, issue 1 (2017) 21-34 https://doi.org/10.34260/jaebs.113 33 without lag) as an explanatory variable in these models has confirmed the preposition that the engaged portion of the informal economy are quite rational while making their decisions about involvement in the unregistered sector, because the lagged tax-to-gdp ratio variable was found statistically significant in the first three models, whereas without lag this variable was statistically insignificant in the last model only, though it has a positive relation with currency holding. at the end, the derived estimates of this study should not be considered as an accurate measure of the underground economy and tax evasion; rather, these estimates can be used as overall trends for better policy making. 5. conclusion this study is an attempt towards knowing the unknown part of the economy. the monetary approach is used to estimate its size and what determines it. a set of four models were presented for the extended time period of 43 years, from 1973 to 2016. these estimates show that the size of informal sector and tax evasion went on increasing from 1974 and attained the maximum level in 1998. however, in the coming years its magnitude went on decreasing with slight variation in its path. the decrease in underground economy and tax evasion may be caused by increase in private investment, increase in smuggling regulations, taxation reform, and better policies of the government like increase in the growth rate of gdp and ease in credit facility to the private sector. increase in taxes, intensity of regulation, and inflation were found to be the driving force of underground economy and tax evasion, which should be focused and regularized by the policy makers so that people have little incentives to indulge in underground activities. these estimates are different from one another, with small variation, which shows the sensitivity of the dependent variables to the explanatory variables. moreover, the derived estimates of this study should not be taken an exact measure of underground economy and tax evasion. these results should be treated carefully because they are sensitive to the assumptions made, equations specified, and data used. references ahmed, m., & ahmed, q. m. (1995). estimation of the black economy of pakistan through the monetary approach. the pakistan development review, 34(4), 791-807. aslam, s. (1998). the underground economy and tax evasion in pakistan: annual estimates (1960-1998) and the impact of dollarisation of the economy. the pakistan development review, 621-631. arby, m. f., malik, m. j., & hanif, m. n. (2010). the size of informal economy in pakistan. sbp working paper series, research department, state bank of pakistan blackburn, k., bose, n., & capasso, s. (2012). tax evasion, the underground economy and financial development. journal of economic behavior & organization, 83(2), 243-253. cagan, p. (1958). the demand for currency relative to the total money supply. journal of political economy, 66(4), 303-328. capasso, s., & jappelli, t. (2013). financial development and the underground economy. journal of development economics, 101, 167-178. minhaj-ud-din, javed iqbal & zia-ur-rahman 34 ferrer-i-carbonell, a., & gërxhani, k. (2016). tax evasion and well-being: a study of the social and institutional context in central and eastern europe. european journal of political economy, 45, 149-159. iqbal, z., qureshi, s. k., & mahmood, r. (1998). the underground economy and tax evasion in pakistan: a fresh assessment (no. 1998: 158). lahore: pakistan institute of development economics. kemal, m. a. (2010). underground economy and tax evasion in pakistan: a critical evaluation. working papers & research reports, rr-no. kemal, m. a. (2007). fresh assessment of the underground economy and tax evasion in pakistan: causes, consequences, and linkages with the formal economy. khalid, m. (2002). estimation of underground economy, causality and business cycle analysis of pakistan (doctoral dissertation, m. phil thesis, department of economics, quaid-i-azam university, islamabad). khan, a., & khalil, s. (2017). the real size of underground economy: a case of pakistan. pakistan journal of applied economics, 27(1), 89-100. kireenko, a., & nevzorova, e. (2015). impact of shadow economy on quality of life: indicators and model selection. procedia economics and finance, 25, 559-568. mughal, k., & schneider, f. (2018). shadow economy in pakistan: its size and interaction with official economy. mpra paper no. 87087. öğünç, f., & yılmaz, g. (2000). estimating the underground economy in turkey. cbrt research department discussion paper, 15. hussain, m. h., & ahmed, q. m. (2006). estimating the black economy through monetary approach: a case study of pakistan. pide research report no. 65. schneider, f., & enste, d. h. (2000). shadow economies: size, causes, and consequences. journal of economic literature, 38(1), 77-114. shabsigh, m. g. (1995). the underground economy: estimation, and economic and policy implications: the case of pakistan (no. 95-101). international monetary fund. spiro, p. (1993). evidence of a post-gst increase in the underground economy. canadian tax j. revue fiscale canadienne, vol. 41 (2): pp. 247–58. tanzi, v. (1980). the underground economy in the united states: estimates and implications in “banca nazionale del lavoro quarterly review n. 135”. tanzi, v. (1983). the underground economy in the united states: annual estimates, 1930-80. staff papers, 30(2), 283-305. tanzi, v. (1999). uses and abuses of estimates of the underground economy. the economic journal, 109(456), 338-347. thomas, j. (1999). quantifying the black economy: measurement without theory 'yet again? the economic journal, 109(456), f381-f389. yasmin, b., & rauf, h. (2004). measuring the underground economy and its impact on the economy of pakistan. the lahore journal of economics, 9:2, 93–103. zaman, g., & goschin, z. (2015). shadow economy and economic growth in romania. cons and pros. procedia economics and finance, 22, 80-87. journal of applied economics and business studies, volume. 4, issue 4 (2020) 237-250 https://doi.org/10.34260/jaebs.4411 237 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x foreign remittances and income inequality in pakistan: a pooled regression analysis sana suleman1* and ahmed raza cheema2 1 phd scholar, department of economics, university of sargodha 2assistant professor, department of economics, university of sargodha abstract the people from developing countries like pakistan move to developed countries to earn their bread and butter. consequently, such migrants remit a handsome part of their earnings to their dependents living in homeland. foreign remittances have multidimensional impact on the economy of a developing country. the study evaluates the impact of foreign remittances on income inequality in pakistan by estimating the set of fixed effect and random effect models using the pooled data from eight household income and expenditure surveys between 1998/99 and 2015/16. gini coefficient as well as generalized entropy measure is used to estimate income inequality, but the results remain intact. it is observed that foreign remittances have statistically significant favorable impacts on income inequality in pakistan. further, the results are robust and insensitive to control variables (e.g. income and poverty measures, headcount ratio, poverty gap and squared poverty gap). the policy measure is that bureau of emigration and overseas employment (beoe) should be empowered to explore the job opportunities in developed countries. the government should assist the migrants through subsidizing the visa and migration processes to capitalize the foreign remittances. keywords foreign remittances, income inequality, gini coefficient, generalized entropy, pakistan jel classification b17, b27, d6 1. introduction people move from place to place to get better compensation for their services. remittance is income sent back by migrants to their homes. it is proven to be a major financial tool for many household living all over the world. the remittances received at household level improve the living standard of that family. the received income is used in purchasing basic commodities, attaining good health facilities and gaining better education level and use in for further investment (buch et al, 2004; el-sakka and mcnabb, 1999). the remittances are less * sana_economist@yahoo.com sana suleman, and ahmed raza cheema 238 subject to economic and political conditions, so the welfare of the household is more stable at the country like pakistan (qureshi, 2016). the remittances are the important component of bop. balance of payments of the countries of origin improves via inflow of remittances (solimano, 2003; world bank, 2006). the balance of payments of several third world labor exporting countries has been largely reliant on the inflow of foreign remittances from migrants (nishat and bilgrami, 1991; azad, 2005). international remittances have major impact at macroeconomic level when remittances are 4% to 31% of gdp (world bank, 2010). it is the largest source of foreign exchange in developing nations resulting in reduction in external borrowings (ratha, 2012). remittances can cause economic growth to accelerate by increasing aggregate consumption and investment (anyanwu & erhijakpor, 2010) foreign remittances over the decades are growing. flow of remittances increased in all six regions of the world. at world level the overall remittances received in 2010 were 470$ billion, in 2015 596$ billion, and in 2017 633$ billion. the overall estimated flow of remittances at world level in 2018 was $689 billion (world bank, 2018). these were 12.4 percent of total remittances which is $131billion in south asia on the whole (world bank, 2018). in 2019, annual remittance flows to lowand middle-income countries (lmics) were $550 billion which were three times higher than official development assistance oda in flows (world bank, 2020). remittance flows were larger than fdi in 2019 as shown in diagram i in appendix (world bank, 2020). as far as pakistan is concerned, transfers flow to pakistan have shown track record of variations in resources of remittances in last few decades. remittances took first jump in the 1970’s, when during the construction boom millions of pakistani temporary migrants were engaged in in the persian gulf. before it uk was leading with 54% share in the total inflow of remittances. near the end of 1980s, the portion of overseas remittances from uae and saudi arabia came to be extensively high. these streams reduced during the cheap oil turmoil of 1980's and the 1990's when arab economies were dwindling. the gulf war in the early 1990's also had a drastic effect on remittances to pakistan. later, in 1990s usa announced the green card policy, which make it a favorite workplace of labors from pakistan and a huge number of heads moved to usa which significantly contributed to the home remittances. after 9/11 terrorism incident in usa, the instability arose in usa which forced pakistani workers to transfer their investments to homeland and intensive growth in remittances started. in fy 2001-02 transfer of money to pakistan increased more than doubled due to usa inflows. this scale of sharp increase in remittance inflows from all foreign pakistanis is stretched to a decade (hasan, 2012). remittances from the usa raised most, from $73.3 million in 2000 to over $1.7 billion in financial year 2008-09 (kock, 2011). further, remittances to pakistan increased at the time of natural tragedies, too. in october 2005 earthquake which affected northern pakistan and victims got back in the better position financially by the remittances sent back by the pakistanis abroad (suleri and savage, 2006). the foundation of pakistan remittances initiative (pri) in 2009 was an effort to attract more journal of applied economics and business studies, volume. 4, issue 4 (2020) 237-250 https://doi.org/10.34260/jaebs.4411 239 remittances through official channels, due to which remittances grew from under $1 billion in 2000 to $7.8 billion in 2008-09 (government of pakistan, 2010). there has been an intense growth in the stream of remittances to pakistan in recent few years, too (sbp, 2015). remittances to pakistan raised by 25.8 percent in 2011 over the previous year, making pakistan the fifth leading remittance-recipient emergent state (gop, 2011). the remittances showed increasing trend on yoy. these were 13.2 $billion in 2012, 13.9$billion 2013, in 15.8 $billion 2014 (pri, 2015). in 2015 pakistan stood on 5th position in remittances receiving in the asian economies (world bank, 2016). around the globe 9 million pakistanis were working. such pakistanis made over us$ 19.3 billion in remittances in the year 2017(ali, 2020). pakistan received $2.097billion foreign remittances in december 2019 demonstrating a 15.25percent month on month (mom) increment when contrasted with $1.819bn transmitted in nov 2019, and roughly 20pc increment than $1.748bn recorded in december 2018(sbp,2020). these substantial inflows of remittances assisted in dropping current account deficit, reducing poverty and growing foreign exchange reserves. there is almost consensus that remittances reduce poverty, but about income inequality there are conflicting views. milanovic, (1987), stark et al. (1988), adams (1991), ahlburg (1996), handa and king (1997), barham and boucher (1998), rodriguez (1998), lerman and feldman (1998), adger (1999), adams et al. (2008), beyene (2014), devkota (2014), taylor (1992), taylor and wyatt (1996), adams (1989), mishra (2007), acosta et al (2008) showed that the remittances had adverse effects on income inequality. it may be outcome of different reasons. remittances are included in top income quintile group because the migrants are from upper income groups and have adverse effect on income distribution in the country (adams, 1991). the opportunity of migration may not generally be available among the populace. the poor household cannot move abroad easily due to financial issues and complex documentations (morton et al. 2010). remittances may not generally improve the circulation of income, and as opposed to what many accept they may indeed, add to inequality inside the nation at income level (ravanilla & robleza, 2003,). usually remitters come from wealthier families and urban regions. these households can manage the expenses and risks related to migration (rodriguez and horton 1995; rodriguez 1998; taylor, lopez and feldman 2010; devkota, 2014).thus, only rich people has access to migration facilities and this is the reason the gap between rich and poor become widened and income inequality increased among masses. the following studies found that remittances had favorable effects on income inequality (taylor et al. 2005; mckenzie & rapoport, 2007; zhu and luo, 2008; chiwuzulum odozi et al. 2010; zhu and xubei, 2010; pfau and giang, 2011; gubert et al , 2010; bang et al,2016;ahmad, 2017). the studies showed that emigration facility may be available to the lower classes (e.g. poor) and they send back money to their homeland and after some period of time their families become better off. in this way income inequality decreased among different classes. but, beyene (2014) did not find any impact of remittances on inequality. sana suleman, and ahmed raza cheema 240 as far as pakistan is concerned, here also mixed results regarding the relationships between income inequality and foreign remittances were produced by previous studies (e.g. mughal and diawara, 2010; mughal and anwar, 2012, mushtaq et. al., 2017) using the time series data consisting of 30 observations. there are some reservations about the use of time series income inequality data. 1) interpolated values were used more than the actual ones1. 2) non-comparable2 inequality data were used. 3) only one inequality measuregini coefficient was used. our study contributes to the existing literature of pakistan in following ways: firstly we use dis-aggregated data (i.e. province-level) to address mixed existing findings of the impact of the foreign remittances on income inequality for analysis. secondly, our study uses the actual, comparable and larger data set consisting of 64 observations derived from micro datasets from 1998-99 to 2015-16. thirdly, this study uses two inequality measures (gini coefficient and generalized entropy3). remaining part of the paper is as follows: the second section explains the data used and empirical strategies employed. then, section three discusses the results. the final section concludes and suggests some policy implications. 2. data and methodology 2.1 data there are so many household income and expenditure surveys (henceforth hies) available from 1963 to 2015-16 (e.g. 1963/64, 66/67, 68/69, 69/70, 70/71, 71/72, 79, 84/85,85/86, 86/87, 87/88, 90/91, 92/93, 93/94, 96/97, 98/99, 2001/02, 04/05, 05/06, 07/08, 10/11, 11/12, 13/14, 15/16). there are three phases of change in questionnaires/methods of collection of data. first, there was one questionnaire from 1963 to 1987-88 consisting of 11 data sets. second, in 1990 hies questionnaire was revised and used from 1990 to 1996-97 consisting of 4 years. thirdly, in 1998-99 again questionnaires were revised and data collection methods were improved. this questionnaire is being used for onward years. so, the previous studies (e.g. (mughal and diawara, 2010; mughal and anwar, 2012; mushtaq et. al. 2017) used the time series data consisting of actually only 14 data points from 1963 to 2006, 1979 to 2007-08 and 1980-2010, respectively that was not comparable. these studies used more interpolated values than actual values. but, this study uses all of the data sets (e.g. 1998-99, 2001-02, 2005-06, 2007-08, 201011, 2011-12, 2013-14, 2015-16) that are comparable and from where the variable foreign remittance can be measured. between these surveys there is only one hies data set for year 2004-05, where foreign remittances are not available. so the data for that year are not being used. there are four provinces including urban and rural areas in pakistan. so by using one observation for urban and rural provinces each, we get eight observations for one year. by 1 pakistan’s inequality data series are derived from hies over time, but there are breaks between different hies data sets. so a time-series approach requires interpolation for missing years. thus, the previous studies used only actual 14 values and the others (about 16 values for income inequality) were interpolated. 2 the data collection methods and questionnaires have changed over time. see for details data section. 3 this measure satisfies all of the axioms to be fulfilled by an appropriate inequality measure see cheema and sial 2013. journal of applied economics and business studies, volume. 4, issue 4 (2020) 237-250 https://doi.org/10.34260/jaebs.4411 241 pooling the data we get 64 observations for the eight years. such like data have never been used in pakistan. 2.2 methodology this study estimates inequality measures in pakistan before estimating long run relationship between foreign remittances and income inequality. 2.2.1 measuring inequality there are different methods to measure inequality. some measures that are apparently sensible do not behave in a reasonable way. for instance, variance of any distribution is dependent on income scale: e,g; if incomes doubles it quadruples the estimates of inequality. this property of a measure is not appropriate. a measure of inequality is appropriate when it satisfies the set of axioms [see for detail, (cheema and sial ,2013)]. this study estimates the following inequality measures to find the inequality in pakistan: 2.2.1.1 gini coefficient corrado gini a statistician developed gini coefficient measure. it is the area between lorenz curve and line of equality divided by the total area below the line of equality and lorenz curve (todaro, 2002) see figure-2. gini coefficient = a/a+b a: area between line of equality and lorenz curve b: total area below line of equality and lorenz curve mathematically it is written as: the gini coefficient lies between 0 and 1.the value near to zero presents equality of income distribution. as the value approaches to one means the distribution of income becoming more unequal. zero represents perfect equality, whereas one depicts perfect inequality. the first four axioms of measuring inequality, gini coefficient fulfilled these too. 2.2.1.2 generalized entropy measure the general form of generalized entropy is as: ( ) 2 1 1 1 1 n i i y ge n y     =    = −   −      where n: number of individuals’, yi : the individual i’s income, i є (1, 2…n) and 1 i y y n   =      , the average income. ge measures ranges between 0 to infinity. the value zero represents equal income distribution. when ge approaches to higher values, it illustrates inequality of income. the α takes any value, denotes the difference of income at different 2 1 1 1 2 n n i j i j gini y y n y = = = − sana suleman, and ahmed raza cheema 242 levels in the income distribution. ge measure is more sensitive for lower values of α for lower tail distribution and for higher values α for upper tail distribution. usually, α takes the value of 0,1 and 2. where α = 0 represents that the lower tail income has more weights for income differences; α = 1, implies that equal weights are given to the distribution; where as α = 2 applies more weights to upper tail gaps. l’hopital’s rule implies that when ge uses the parameter 0 and 1, then it will become two of theil’s indices (theil, 1967). the mean log deviation and theil index are as given: ( ) 1 0 log n i i y ge n y =  ( ) 1 1 1 log n i i i y y ge n y y= =  2.2.2. relationship between income inequality and foreign remittance in order to find the long run relationship between income inequality and foreign remittances, the study pools the data by taking one observation from urban and rural areas each of four provinces from 1998-99 to 20015-16 making 8 observations per year in pakistan. there are eight data sets. so, we use 64 observations for analysis. the characteristics of rural and urban areas of provinces in pakistan are different. so following the framework used in studies of mueller and sial (1993) and cheema & sial (2012), techniques of panel data are employed. at first the study estimate the fixed effect model whose functional form is given below: ( ) it i it it inequality fr   = + + + 1, 2....., re tan int it it i n fr foreign mit ces ercept slope error term    = = = = = after estimation of fixed effect model, f-test is employed to choose between pooled regression and fixed effect model. the null hypothesis is formulated as: the parameters are zero. the null hypothesis fails to accept at 5% level of significance. so, the fixed effect model is apt choice. random effect model in order to check the role of foreign remittance on income inequality, the random effect model is estimated and that is given as follows: journal of applied economics and business studies, volume. 4, issue 4 (2020) 237-250 https://doi.org/10.34260/jaebs.4411 243 ( )0 1 0 1 1 1 : 0 : 0 it it i it inequality fr v h h      = + + + =  where inequality= gini coefficient, ge (0), ge (1), fr=foreign remittance the null hypothesis as: variance is zero for cross sections. breush-pagan lm test is used to select between pooled regression and random effect model. the null hypothesis fail to accept at 5% level of significance by using chi-square test showing that random effect model is optimal. after this, the hausmen test is used to choose between the fixed effect and random effect models. the null hypothesis formulated under this technique as: individual effects are independent of the model regressors (hausman 1978;park, 2008 ). the hypothesis of hausmen test can be written as 0 1 : ( , ) 0 : ( , ) 0 i i h cov x random effect model h cov x fixed effect model   =  hausman test specifies that random rem is the optimal selection. 3. empirical results the study estimates income inequality and foreign remittances, from these surveys to find the long run relationship between them. the study also calculates per capita income and poverty measures (e.g. headcount ratio (henceforth hc), poverty gap (henceforth pg) and squared poverty gap (henceforth spg) from these surveys to use them as control variables. a set of descriptive statistics is presented in the table 1. table 1: descriptive statistics variables n mean sd min max gini (%) 64 26.478 5.299 18.12 37.6 ge (0) (%) 64 11.860 4.7034 5.3 23 ge(1) (%) 64 11.8671 4.744 5.33 23.92 fr (rs p.m.) 64 1885.78 1433.34 0 6712.19 income (rs p. m.) 64 2724.69 1647.70 813.7888 7144.311 hc (%) 64 20.264 12.35 2.14 57.047 pg (%) 64 3.58 2.92 0.24 14.88 spg (%) 64 0.98 0.97 0.05 5.26 gini=gini coefficient, ge=generalized entropy, fr=foreign remittances, hc=headcount ratio,pg= poverty gap, spg=squared poverty gap, p.m= per month source: authors’ own calculations *zero remittances were calculated in rural areas of sindh from hies 2001-02. so, minimum value of remittance is 0 for this year. to find the relationship between inequality and foreign remittances in the long run, this study estimated the random effect model using the pooled data consisting of 64 observations and the results are provided in the table 2. after applying different statistical tests, this study concludes that our preferred model is random effect. however to lend further credibility to our results this study also estimates the sana suleman, and ahmed raza cheema 244 fixed effect model. the results of both models show that there is negative and statistically significant relationship between income inequality and foreign remittances. when remittances increase, income inequality decreases. our findings are robust and not sensitive to fixed effect and random effect models. the results are consistent with those of taylor et al (2005) in mexico, acosta et al (2007) in latin american countries, ebeke and le goff (2009) in mediterranean basin countries, kimhi (2010)in dominican republic, wouterse (2010) in intra-african, odozi (2010) in nigeria , margolis(2013) in algeria. checking of robustness in order to check the robustness of our empirical analysis, we include some variables like income, and poverty indices4 (e.g. headcount ratio, poverty gap and squared poverty gap). income is an important variable to affect income inequality. many studies focused the relationships between income inequality and economic growth (simon kuznets,1955; paukert,1973; ahluwalia,1976; papamek & kyn,1987; tsakoglou,1988; randolph & lott,1993; jha, 1996; dawson,1997;eusufzi 1997; mubaku, 1997; huang, 2004; deiniger and squire, 1996 & 1998). as far as poverty is concerned, there is positive relationship between income inequality and poverty (cheema & sial, 2012; ali & tahir, 1999; saboor, 2004). the results of the robustness checks are presented in table 3. the results still show a negative and statistically significant relationship between income inequality and foreign remittances in pakistan. so, we can conclude that our findings are robust and not sensitive to the inclusion of these controls as the determinants of income inequality. these results are consistent with those of taylor et al. (2005), mckenzie& rapoport (2007), zhu and luo (2008), chiwuzulum odozi et al. (2010), zhu and xubei (2010), pfau and giang (2011), bang et al. (2016), ahmad (2017). income inequality is negatively related with income, but it not statistically significant. so, income has no effect on income inequality. income inequality is positively related to poverty, but this relationship is statistically insignificant. moreover, the paper also uses other inequality measures ge (i.e. ge (0) and ge (1)) that satisfies all of axioms to become a suitable inequality measure (reyes, 2005; atkinson, 1970; allison, 1978; champernowne, 1974; cowell, 2011; cheema and sial, 2012). again our results remain intact. so, we can conclude that income inequality is negatively and statistically significantly related with foreign remittances. 4. conclusions and policy suggestions the study established the role of foreign remittances on income inequality by estimating the set of fixed effect and random effect models using the pooled data from eight household income and expenditure surveys between 1998/99 and 2015/16. gini coefficient and generalized inequality measures are used to measure income inequality. the results depicted that income inequality is inversely related with foreign remittances in pakistan. the results 4 these poverty indices were estimated applying the same technique as was applied by cheema and sial (2012). journal of applied economics and business studies, volume. 4, issue 4 (2020) 237-250 https://doi.org/10.34260/jaebs.4411 245 are robust and not sensitive to income and poverty measures (e.g. headcount ratio, poverty gap and squared poverty gap). bureau of emigration and overseas employment (beoe) is a centralized agency of the federal government for processing recruitment demands of the pakistani manpower through licensed overseas employment promoters, etc. for the different manpower importing countries in the world especially in the middle east. the government should empower beoe to explore the job opportunities in developed countries and assist the poor migrants through subsidizing the visa and migration processes to capitalize the foreign remittances. references acosta, p.a., calderon, c., fajnzylber, p., lopez, j.h. (2008). what is the impact of international remittances on poverty and inequality in latin america? world development, 36(1),89–114. acharya, c. p., & leon-gonzalez, r. (2012). the impact of remittance on poverty and inequality: a microsimulation study for nepal. national graduate institute for policy studies discussion paper, (11-26). adams jr., r.h. (1989). worker remittances and inequality in rural egypt. economic development and cultural chang. 38(1), 45–71. adams, r. h. (1991). the effects of international remittances on poverty, inequality, and development in rural egypt (vol. 86). intl food policy res inst. adams, r. h. (1992). the effects of migration and remittances on inequality in rural pakistan. the pakistan development review, 31(4), 1189-1206. adams, r.h., page, j. (2003). international migration, remittances and poverty in developing countries, policy research working paper no. 3179, world bank poverty reduction group, washington, dc. adams, j., cuecuecha, a., page, j. (2008). the impact of remittances on poverty and inequality in ghana. world bank policy working research paper 4732.the world bank. adger, w. n. (1999). exploring income inequality in rural, coastal viet nam. the journal of development studies, 35(5), 96-119. ahlburg, d. a. (1996). remittances and the income distribution in tonga. population research and policy review, 15(4), 391-400. ahmad, m. t. a. (2017). rural-urban migration and rural inequality in china. ahluwalia, m. s. (1976). inequality, poverty and development. journal of development economics, 3(4), 307342. ali, s. s., tahir, s., & arif, g. m. (1999). dynamics of growth, poverty, and inequality in pakistan. the pakistan development review, 837-858. ali, s. r. (2020). the pattern, sources, and growth of remittances to pakistan: the kinked exponential approach. journal of research in emerging markets, 2(1), 1-6. allison, p. d. (1978). measures of inequality. american sociological review, 865-880. anyanwu, j. c., & erhijakpor, a. e. o. (2010). do international remittances affect poverty in africa? african development review, 22, 51–91. atkinson, a. b. (1970). on the measurement of inequality. journal of economic theory 2, 244-263 azad, a.k. (2005). migrant workers remittances; a source of finance for micro-enterprise development in bangladesh’ in maimbo s.m. and ratha d (eds), remittance; development impact and future prospects, world bank pp. 103-118. anyanwu, j. c. (2011). international remittances and income inequality in africa. review of economic and business studies, 4(1), 117-48. bang, j. t., mitra, a., &wunnava, p. v. (2016). do remittances improve income inequality? an instrumental variable quantile analysis of the kenyan case. economic modelling, 58, 394-402. barham, b.& boucher, s.(1998). migration, remittances, and inequality: estimating the neteffects of migration on income distribution. journal of development and economics, 55, 307–331. sana suleman, and ahmed raza cheema 246 beyene, b. m. (2014). the effects of international remittances on poverty and inequality in ethiopia. the journal of development studies, 50(10), 1380-139. buch, c. m., kuckulenz, a. le manchec, m.-h., 2002. worker remittances and capital flows. kiel working paper 1130, kiel institute for world economics. chiwuzulum odozi, j., taiwo awoyemi, t., & omonona, b. t. (2010). household poverty and inequality: the implication of migrants’ remittances in nigeria. journal of economic policy reform, 13(2), 191-199. champernowne, d. g. (1974). a comparison of measures of inequality of income distribution. the economic journal, 84(336), 787-816. cheema, a. r., & sial, m. h. (2012). poverty, income inequality, and growth in pakistan: a pooled regression analysis. the lahore journal of economics, 17(2), 137. cheema, a. r., & sial, m. h. (2013). estimating income inequality in pakistan: hies 1992-93 to 200708. interdisciplinary journal of contemporary research in business, 4(9), 886-902. cowell, f. (2011). inequality among the wealthy. lse sticerd research paper no. case150. dawson, p. j. (1997). on testing kuznets' economic growth hypothesis. applied economics letters, 4(7), 409410. daianu, d., 2001. balance of payments financing in romania – the role of remittances, romanian center for economic policies, bucharest. devkota, j. (2014). impact of migrants’ remittances on poverty and inequality in nepal. in forum of international development studies, 44, 36-53. deininger, k., & squire, l. (1996). a new data set measuring income inequality. the world bank economic review, 10(3), 565-591. deininger, k., & squire, l. (1998). new ways of looking at old issues: inequality and growth. journal of development economics, 57(2), 259-287. docquier, f., rapoport, h., 2003. remittances and inequality: a dynamic migration model, iza discussion paper no. 808, institute for the study of labor, bonn. el-sakka, m. i., & mcnabb, r. (1999). the macroeconomic determinants of emigrant remittances. world development, 27(8), 1493-1502. gilani, i., khan, f., and iqbal, m. (1981), labor migration from pakistan to the middle east and its impact on the domestic economy (part i), pakistan institute ofdevelopment economics research report series no. 126 (islamabad, june 1981), 2-3. godoy, r., byron, e., reyes-garcía, v., vadez, v., leonard, w. r., apaza, l., ... & wilkie, d. (2005). income inequality and adult nutritional status: anthropometric evidence from a pre-industrial society in the bolivian amazon. social science & medicine, 61(5), 907-919. handa, s., & king, d. (1997). structural adjustment policies, income distribution and poverty: a review of the jamaican experience. world development, 25(6), 915-930. hasan, s. m. (2012). sbp's home remittance policies and its effects in pakistan: a study of the last ten years (2001-2010). journal of independent studies and research, 10(1), 15-21. hausman, j. a. (1978). specification tests in econometrics. econometrica: journal of the econometric society, 1251-1271. jha, s. k. (1996). the kuznets curve: a reassessment. world development, 24(4), 773-780. kock, u., & sun, y. (2011). remittances in pakistan: why they have gone up and why they are not coming down. the pakistan development review, 189-208. kuznets, s. (1995) .economic growth and income inequality. the american economic review, vol.45, pp. 128. león‐ledesma, m., & piracha, m. (2004). international migration and the role of remittances in eastern europe. international migration, 42(4), 65-83. lerman, j. p.,and feldman, s., (1998). nonfarm activity and rural household income: evidence from philippine microdata. economic development and cultural change, 46(4), 789-806. litchfield, j. a. (1999). inequality: methods and tools. world bank, 4. journal of applied economics and business studies, volume. 4, issue 4 (2020) 237-250 https://doi.org/10.34260/jaebs.4411 247 lowell, b.l., de la garza, r.o., 2000. the developmental role of remittances in us latino communities and in latin american countries. a final project report, inter-american dialogue. mckenzie, d., & rapoport, h. (2007). network effects and the dynamics of migration and inequality: theory and evidence from mexico. journal of development economics, 84(1), 1-24. mishra, p., 2007. emigration and wages in source countries: evidence from mexico. journal of develoment economics. 82, 180–199. milanovic, b. (1987). remittances and income distribution. journal of economic studies, 14(5), 24-37. morton, j., panday, p., & kula, m. (2010). remittances, poverty andeconom ic growth. internat ional journal of art s andsciences, 3(7), 390-99. mueller, w. f., & sial, m. h. (1993). cyclical variation in the profit-concentration relationship. review of industrial organization, 8(3), 277-291. mughal, m., & diawara, b. (2010). impact of remittances on inequality and poverty: macro and micro-evidence from pakistan. ncde paper, 1-32. mughal, m. y., & anwar, a. i. (2012). remittances, inequality and poverty in pakistan: macro and microeconomic evidence (no. 2012-2013_2). mushtaq, m., hayat, m. a., ahmad, s., & nadeem, m. (2017). distributional impact of foreign remittances in pakistan. pakistan business review, 19(2), 320-338. nishat, m. & bilgrami, n. (1991) the impact of migrant workers' remittances on pakistan',pakistan development review, 30 (1), 400-480. pakistan bureau of statistics. (1994–2017). household integrated economic survey (hies). islamabad, pakistan: author. park, h. m. (2008). linear regression models for panel data using sas, stata,limdep, and spss (working paper). bloomington, in: indiana university, center for statistical and mathematical computing. papanek, g. f., & kyn, o. (1987). flattening the kuznets curve: the consequences for income distribution of development strategy, government intervention, income and the rate of growth. the pakistan development review, 26(1), 1-54. paukert, f. (1973). income distribution at different levels of development: a survey of evidence. international labour review, 108, 97–125. pfau, w. d., and giang, l. t. (2011).determinants and impacts of international remittances on household welfare in vietnam. international social science journal.60(197-198), 431-443. portes, l. s. v. (2009). remittances, poverty and inequality. journal of economic development, 34(1), 127. qureshi, j. a. (2016). the pakistan remittance initiative and remittance flows to pakistan (vol. 35). intl food policy res inst. quibria, m. g. (1997). international migration, remittances and income distribution in the source country: a synthesis” bulletin of economic research 49 (1): 29-46. ravanilla, n. m., & robleza, e. j. (2003). the contribution of ofw remittances to income inequality: a decomposition analysis. undergraduate thesis, ba economics, university of the philippines, quezon city. randolph, s. m., & lott, w. f. (1993). can the kuznets effect be relied on to induce equalizing growth?. world development, 21(5), 829-840. rodriguez, e. r., & horton, s. (1995). international return migration and remittances in the philippines (no. horton-95-01). rodriguez, e. r. (1998). international migration and income distribution in the philippines. economic development and cultural change, 46(2), 329-350. ratha, d. (2012, february). outlook for migration and remittances 2012-14. in tenth coordination meeting on international migration world bank, new york. saboor, a. (2004). agricultural growth, rural poverty and income inequality in pakistan: a time series analysis (doctoral dissertation, university of agriculture, faisalabad, pakistan). sarno, l., taylor, m. p., 1999. hot money, accounting labels and the permanence of capital flows to developing countries: an empirical investigation. journal of development economics, 59, 337-64. http://www3.interscience.wiley.com/journal/118520819/home http://www3.interscience.wiley.com/journal/119172455/issue sana suleman, and ahmed raza cheema 248 solimano, a. (2003). workers remittances to the andean region: mechanisms, costs and development impact. santiago: un-eclac. mimeo, 5-8. stark, o., taylor, j. e., &yitzhaki, s. (1988). migration, remittances and inequality: a sensitivity analysis using the extended gini index. journal of development economics, 28(3), 309-322. suleri, a. q., & savage, k. (2006). remittances in crises: a case study from pakistan. london: sustainable development policy institute. tsakloglou, p. (1988). development and inequality revisited. applied economics, 20(4), 509-531. taylor, j.e., (1992). remittances and inequality reconsidered: direct, indirect, andintertemporal effects. journal of policy model, 14, 187–208. taylor, j.e.,& wyatt, t.j., (1996). the shadow value of migrant remittances, income and inequality in a household farm economy. journal of development studies, 32, 899–912. taylor, j.e., mora, j., adams, r.h., lopez-feldman, a., (2005) remittances, inequality and poverty: evidence from rural mexico. in: dewind, j., holdaway, j. (eds.), migration and development within and across borders: research and policy perspectives on internal and international migration. international organization for migration, geneva, pp. 101–130. taylor, j. e., & lopez-feldman, a. (2010). does migration make rural households more productive? evidence from mexico. the journal of development studies, 46(1), 68-90. todaro, m. p., & smith, s. c. (2002). economic development (8th ed.). reading, ma: addison-wesley. torres-zorrilla, j. (2008). remittances and income distribution in peru. journal of centrum cathedra, 1(1), 64-70. world bank. (2019). migration and development brief 31. world bank. (2019). world development indicators. world banks group, washington. zhu, n., & luo, x., (2008). the impact of remittances on rural poverty and inequality in china.world bank policy research working paper no. 4637. zhu, n.,&xubei, l., (2010). the impact of migration on rural poverty and inequality: a case study in china. agriculture economics. 41 (2), 191–204. appendix figure-1 journal of applied economics and business studies, volume. 4, issue 4 (2020) 237-250 https://doi.org/10.34260/jaebs.4411 249 figure-2 table 2 relationship between income inequality and foreign remittances in pakistan fixed effect results random effect results variables gini ge(0) ge (1) gini ge(0) ge (1) constant 27.546 (61.64)*** 12.8965 (30.69)*** 12.962 (30.72) 27.494 (17.49) 12.84 (9.24) 12.91 (19.19)*** fr -.00056 (-2.94)*** -.0005 (3.03)*** -.0005 (3.19)*** -.0005 (2.70)*** -.0005 (2.78)*** -.0005 (-2.94)*** source: authors’s own calculations sana suleman, and ahmed raza cheema 250 table 3: role of foreign remittances on income inequality (gini coefficient) fixed effect model random effect model 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 fr -.00056 2.94)** -.00052 -2.62)*** -.0005 (2.54)** -.0005 (2.45)** -.0005 (2.40)** -.0005 (2.50)** -.0005 (2.41)** -.00049 (2.36)** -.0005 (2.70)** -.0005 (2.37)** -.0005 (2.31)** -.0004 (2.24)** -.0004 (2.20)** -.00049 (2.31)** -.00048 (-2.26)* -.00047 (-2.21)** income -.0001 (0.70) -.00003 (-0.15) -.00004 (-0.19) -.00003 (-0.17 -.00009 (-0.51) -.00007 (-0.27) -.00003 (-0.17) -.00002 (-0.11) hc .0224 (0.83) .01815 (0.46) .0113 (0.39) .0046 (0.11) pg 9.65 (0.89) 7.992 (0.57) 7.071 (0.62) 6.0318 (0.42 spg 31.71 (1) 28.048 (0.73) 26.178 (0.79) 24.679 (0.62) role of foreign remittances on income inequality (ge (0) of generalized entropy) fr -.0005 (3.03)*** -.0005 (2.76)** -.0005 (2.75)** -.0005 (2.67)** -.0005 (2.63)** -.0005 (2.69)** -.0005 (2.62)** -.0005 (2.58)** -.0005 (2.78)** -.0005 (2.50)** -.0005 (2.50)** -.0005 (2.45)** -.00049 (2.40)** -.0005 (2.48)** -.0005 (2.47)** -.00049 (-2.27)** income -.00007 (0.47) -.00006 (-0.27) -.00005 (-0.25) -.00003 (-0.20) -.00004 (-0.28) -.00009 (-0.39) -.00004 (-0.22) -.00002 (-0.13) hc .0101 (0.39) .0030 (0.08) -.0012 (-0.05) -.0108 (0.28) pg 4.568 (0.45) 2.558 (0.19) 1.901 (0.18) .5280 (0.04) spg 16.404 (0.55) 12.405 (0.34) 10.633 (0.34) 8.889 (0.24) role of foreign remittances on income inequality (ge (1) of generalized entropy) fr -.0005 (3.19)*** -.00053 (2.83)** -.0005 (2.82)** -.0005 (2.75)** -.0005 (2.71)** -.00053 (2.75)** -.00052 (2.69)** -.0005 (2.64)** -.0005 (2.94)** -.00056 (2.97)** -.0005 (2.55)** -.0005 (2.51)** -.0005 (2.48)** -.0004 (2.09)** -.00046 (2.27)** -.000470 (-2.34)** income -.0001 (0.82) -.0001 (-0.47) -.0001 (-0.52) -.00009 (-0.48) .00115 (0.91) .0038 (1.88)* .00289 (1.66)* .0024 (1.55) hc .0175 (0.69) .0050 (0.14) .0055 (0.20) .06283 (1.46) pg 6.876 (0.67 2.659 (0.20) 4.109 (0.48) 21.50 (1.47) spg 22.539 (0.75) 12.72 (0.35) 16.60 (0.53) 55.18 (1.42) source: authors’ own calculations journal of applied economics and business studies, volume. 4, issue 2 (2020) 163-180 https://doi.org/10.34260/jaebs.428 163 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x the performance effect of hpws and relational coordination on branch efficiency: a data envelopment analysis approach muhammad siddique1*, zahoor khan2 & saleem gul3 1assistant professor, institute of management sciences, peshawar. 2assistant professor, institute of management sciences, peshawar 3associate professor, institute of management sciences, peshawar abstract this study is set to analyse the role of high-performance work systems which is mainly aimed to boost knowledge, skills, and abilities of human resources in the form of effective communication and better coordination termed as relational coordination and subsequently improving performance. using the theory of relational coordination that explains the relational aspects of coordination along with effective communication to influence performance outcomes, the study obtained data from 218 branches through survey focusing on employees’ perspectives regarding high-performance work system in anticipating relational coordination among employees in performing focal work process of branch banking functions. the data obtained from officers was analysed through the data envelopment analysis approach to identify efficient bank branches pivotal in formulating optimal policy measures. results indicated that high-performance work systems envisaged relational coordination via the intervention of deposits and profitability. the findings contribute to the literature of high-performance work systems by indicating what interventions in high-performance work systems and the relational coordination may lead to bringing down operations cost or making improvements in the delivery of banking services. the study also implies that compared with managerial perspectives in the field, employees’ perspectives provide significant insights about the relational process in explaining how a well-coordinated approach towards hr systems establish the social context for better hrm practices, enhancing. keywords dea, financial sector, hpws, organizational performance, pakistan, relational coordination. jel classification b26; d2; d21 * email: muhammad.siddique@imsciences.edu.pk muhammad siddique, zahoor khan & saleem gul 164 1. introduction over the last few decades, there has been an increasing interest within management research in recognising the human factor in organizational performance. organizations rely on various forms of resources to accomplish desired objectives, among other resources; human resources can accomplish a significant impact on firm performance. it is based on the premise that enhanced performance is mainly realized through the effective use of human resources in the organization (nien-chi & lin, 2019). the relationship between hrm and performance has been widely reported in the last few decades (e.g. russell, et al., 2018; ogbonnaya & valizade, 2018) as a large number of studies have reported significant associations between hr practices and performance (delery & doty, 1996). recently, the focus is to examine not only the linkage between hr practices but also to explore the process through which human resource management influence performance. researchers have suggested that to understand the relationship between human resource management practices and firm performance, there is a need to distinguish what constitutes as high-performance work systems (hpws), the impact hrm practices as a system make on performance, and most importantly exploring the mechanism and process through which hrm practices make such impact on organizational performance. the hpws is generally understood to mean a set of hrm practices in the hrm literature (methot, rosado, & allen, 2018). extensive research (e.g., kehoe & collins, 2017; zacharatos et al., 2005) has shown that studies have mostly adopted pfeffer (1998) set of practices. these practices have received greater attention in the hpws literature. these include practices that can improve the skills, knowledge, and abilities of employees, enhance their motivation, and can be helpful in the retention of competent employees while advocating poor performers to leave organization. the hr-performance link has been investigated from various perspectives rooted in organizational psychology, industrial relations, sociology and economics. theories such as human capital, rbv of the firm, and the amo framework provided support for understanding the hrm-performance link (boselie, dietz, & boon, 2005). there is a large volume of studies describing the multidimensional nature of organizational performance (boxall, 2012). the seminal work of huselid’ (1995) established empirical hrmperformance link, inspired a series of studies analysing the various form of performance such as economic gains, accounting income, customer satisfaction, and productivity (paauwe, wright, & guest, 2012). across various sectors, there are different measures of organizational performance. the rationale for selecting certain indicators of performance largely depends on the objectives of the specific organization under study (methot, milwani, & rothman, 2017). dyer and reeves (1995) labelled performance outcomes into 4 categories including employee, financial, organizational, and market outcomes. journal of applied economics and business studies, volume. 4, issue 2 (2020) 163-180 https://doi.org/10.34260/jaebs.428 165 despite studies have reported that hpws are positively associated with performance outcomes (e.g., chao, & chih-ting, 2018; liao, toya, lepak, & hong, 2009), researchers have identified that earlier studies have paid little attention to the process linking hr practices to organizational performance (boon, den hartog, & lepak, 2019; malhotra & singh, 2016; han, bartol, & kim, 2015) previous studies have taken up different theories in examining the relationship between hrm practices with performance (boselie et al., 2005). using the concept of social capital, the relationships between employees have been theorised to act an important part in accomplishing high levels of performance (havens, gittell, & vasey, 2018). the theory of relational coordination as one of these relational perspectives describes specific dimensions of relationships that are essential to the coordination of work. relational coordination is considered a developing theory for distinguishing the dynamics of coordination work. the term relational coordination is generally understood to mean coordination of group members in the context of relationships in interdependent work setup (gittell, 2001). according to gittell (2002, p. 301) “relational coordination is a mutually reinforcing process of interaction between communication and relationship carried out for task integration”. relational coordination is an organised process that consists of communication and relationships. in addition to theoretical understanding, scholars have also suggested methodological gaps such as considering both managers’ and employees’ perceptions about hr practices will result in a better understanding of the hr-performance relationship (wright &ulrich, 2017; bowen & ostroff, 2004;). these issues imply that the process of linking highperformance work systems and performance is distant from complete hence, indicating a research gap (fu, bosak, flood, & ma, 2019; gerhart, 2012). therefore, based in the theory of relational coordination, this study focuses on examining the links through which hpws fosters relational coordination among employees and consequently influences organizational performance. the research questions are: 1. to what extent hpws can influence the efficiency of unit-level performance in predicting relational coordination among employees. 2. to what degree the linkages between hpws, relational coordination, and performance is related to the ranking of efficient units. this study used the above arguments to develop a research framework that incorporated the rc theory in hpws-performance nexus, identifying relational coordination might help explain the process linking hpws in positively influencing the efficiency of unit-level performance. this study contributes to the multiple facets of hpws and rc literature (concerning pakistan). firstly, the study used data envelopment analysis (dea) approach based on input muhammad siddique, zahoor khan & saleem gul 166 and output models in which efficiency scores were calculated for three combinations. firstly, to estimate the effects of hpws in predicting relational coordination among employees (hpws as input and relational coordination as output), secondly, to assess the relationship of hpws with branch level performance outcomes (hpws as inputs and performance outcomes as outputs), and thirdly, the effects of relational coordination on performance outcomes (rc as inputs and performance as outputs). thus, inputs for dmu's are divided into two main groups, i.e., hpws & relational coordination. the variables under the umbrella of hpws are js, t&d, employee participation, information sharing, job description, compensation, and appraisal. the variables concerning relational coordination are timely, frequent, accurate, problemsolving, communication, shared knowledge, shared goals, and mutual respect. in terms of output for each decision-making unit (branches as dmu’s) are deposits, advances, and profitability of 120 bank branches (see table 1). the findings of dea suggest that based on the combination of input-output models, certain branches were more efficient on hpws in predicting relational coordination among employees. secondly, some branches were highly efficient on hpws concerning performance outcomes of deposits, advances, and profitability and many branches were efficient in terms of relational coordination and its influence on performance outcomes, which is in line with the literature (gittell, seidner & wimbush, 2010). thirdly, it provides practical implications for policymakers suggesting the importance of employees’ perspectives. fourthly, the empirical evidence supports an important role of key stakeholders in designing hpws frameworks, and increasing the input of relational coordination will further improve performance outcomes in terms of higher levels of deposits, performing loans, and profitability. 2. literature review hrm literature carries a consistent message that hr practices positively influence organizational performance (murphy, torres, ingram, & hutchinson, 2018). in recent years, researchers have aimed at several other issues in hpws. one of the major issues is related to the “black box” in hr literature (sun, xing, yin, & yang, 2018; boselie et al., 2005). to examine the black box, studies have focused on the role of hrm systems in affecting the attitude and behavior of employees (mingqiong, cherrie, dowling, & bartram, 2013; karadas & osman, 2019). for instance, ostroff and bowen (2000) viewed hr practices as critical elements in encouraging employees to behave in manners beneficial to organizational goals. guest (1997) proposed that appropriate hr practices entice the commitment and motivation of employees that have a direct relationship with business results. in doing so, researchers have mainly relied on the resource-based view and amo framework as theoretical foundations of the hpws-performance relationship (paauwe et al., 2012). journal of applied economics and business studies, volume. 4, issue 2 (2020) 163-180 https://doi.org/10.34260/jaebs.428 167 resource-based view: the rbv approach proposes that firms pay greater attention to their resources for competitive advantage. researchers have suggested that the rbv has significantly influenced the field of hrm and considered that a system of hr practices that are valuable in nature, rare, unique, and non-substitutable will lead to competitive advantage (romanow, rai, & keil, 2018). however, a number of studies have found that each hr practice prevailing in the organization cannot be a source of sustained competitive advantage (meuer, 2017; guest, 1997). the amo framework: the amo model (appelbaum, bailey, berg, & kalleberg, 2000) has received more attention from the scholars for several reasons including its main components that enable employees to develop better skills, knowledge, and abilities along with motivation and get better opportunities about involvement in decision making. 2.1 relevance of relational coordination theory several models explain the link between hpws and performance (e.g. guest, 1997; appelbaum et al., 2000; lee, pak, kim, & li, 2019). in view of ostroff and bowen (2000), most models are based on the assumption that superior hr practices are considered to result in more committed, motivated, and better-skilled employees, who, in turn, can be more productive to affect firm performance. so far, the most predominant arguments about causal mechanisms are based on employees’ skills, abilities, motivation, human capital, and commitment. in addition to these arguments, there is a developing view that focuses on relationships between employees and considers it as an essential causal mechanism through which hpws influence performance (collins & smith, 2006). therefore, in this study, a relational view is adopted and a model of hpws has proposed in which each hr practice applies to multiple functions in an interdependent work setting. relational coordination distinguishes the relational dynamics of coordination work, which comprises communication and relationship aspects in the performance of the interdependent task (siddique, procter, & gittell, 2019; gittell, 2002). relational coordination theory assumes that hpws is due to influence the performance outcomes through their effects on the degree of relational coordination. it suggests that because of social relationships and better information processing capacity, relational coordination is expected to result in better communication ties among employees with different expertise across functions. relational coordination brings more consistency in communication that enables employees to reduce the number of errors and time wasted in searching misplaced information, carrying out unnecessary communication, and waiting for a response from other workers (gittell et al., 2010). based on social capital, the relationships between employees have been understood to act an important role in accomplishing performance (faraj & sproull, 2000; adler, kwon, & heckscher, 2008). in general, rc theory is exclusive in determining explicit dimensions of muhammad siddique, zahoor khan & saleem gul 168 relationships that are essential to the coordination of job, precisely going ahead of shared knowledge to encompass mutual respect and shared goals, while concentrating on the improvement of these relationships between roles instead of between specific individuals. relational coordination theory offers a particular approach to conceptualize the relational dynamics of coordination, their anticipated effects, and basic predictors. therefore, this theory is applicable to work processes with a high level of interdependence involving multiple employees to carry out the tasks under uncertain circumstances and time restrictions (gittell et al., 2010). 2.2 conceptual framework in this model, relationships among employees are considered as a link between hpws and performance. this study recommends that hpws is set for improving relational coordination, which is expected to build up branch-level performance. based on the theory of relational coordination, hpws affects relational coordination at the unit level, and relational coordination in turn is linked with the impact of hpws on overall branch performance, indicating a relational process through which hpws work. figure 1: conceptual framework source: the above theoretical framework is adopted form gittell, seidner & wimbush (2010). 3. methods 3.1 research approach the study used a predictive research design. for this purpose, in the first stage, the information related to hpws and rc was obtained from managerial and officer cadre employees working in various functions including operations, cash, and credit departments from 218 branches (babbie, 2004). secondly, data regarding branch-level performance outcomes including deposits advances, and profitability was gathered subsequently after the survey on completion of the financial year. thirdly, based on the inputs and outputs of decision-making units (dmu’s) efficiency and super-efficiency of dmu’s were determined via data envelopment analysis (dea) analysis. these efficiency scores were computed using hpws js t&d employee participation information sharing jd contingent comp appraisal relational coordination communication ties relationship ties unit performance deposits advances profitability journal of applied economics and business studies, volume. 4, issue 2 (2020) 163-180 https://doi.org/10.34260/jaebs.428 169 the variable returns to scale (vrs) approach. the vrs approach allows for the computation eliminating calculation of technical efficiency (te) further, using input-oriented bcc models, super efficiency scores were computed for 120 branches and were ranked according to efficiency scores. 3.2 measurement of variables high-performance work systems and unitary index there has been little agreement as to what constitutes hpws (lepak, liao, chung, & harden, 2006). organizations have specific objectives and connecting them to the design of hpws might present better insights into the effectiveness of hpws (lepak et al., 2006). therefore, for this study, influential studies and its relevance to the study settings were identified which have focused on the effects of high-performance work practices on performance as an inclusion criterion in hpws. these practices include js, t&d, employee participation, jd, information sharing, contingent compensation, and appraisal. these practices were measured with validated scales (delery & doty, 1996; zacharatos et al., 2005; scott & james, 1992) on a 5-point likert scale. the study adopted a widely used additive approach, which is consistent with one of the main principles of strategic human resource management that the effects of practices is better understood by testing the system of high-performance work practices (huselid, 1995; ostroff & bowen, 2000; lepak et al., 2006). following the subscale aggregation additive approach, seven hr practices were aggregated into a unitary index that measures hpws (macky & boxall, 2007). an additive approach is also more appropriate for the present study model that assumes each practice is equally important within the hpws index (gittell et al., 2010). relational coordination: a seven items rc survey, developed by gittell et al. (2010) was adapted for measuring relational coordination. the rc survey has been used to determine cross-functional as well as cross-organizational coordination among employees working in highly interdependent work settings. branch performance measures to measure the branch performance of a bank, paradi et al. (2011) proposed a production, intermediation, and profitability approach. using paradi’s approach, selected branches are analysed about efficiency, liquidity, capital adequacy, and profitability. an identical set of unitlevel performance measures comprising deposits, lending, and profit were gathered across one hundred and twenty bank branches. sampling and data collection procedure this study has been conducted in one of the largest and pioneer banks in pakistan with a nationwide branch network of more than 1300 branches. considering the distinct geographical location of the bank, muhammad siddique, zahoor khan & saleem gul 170 the study considered 755 branches operating and located in the major provinces of punjab, khyber, and federal capital. table 1: sampled bank branches the punjab federal capital kp total branches 360 190 205 755 surveyed 155 78 107 340 hpws and rc 58 68 92 218 matched branches 38 49 33 120 in total, 3400 questionnaires were personally administered in the survey in 218 branches. the survey was sponsored by research funding. participants completed 1830, of which only 1560 were usable. the sample characteristics of the respondents suggested that most of the employees are young professionals in the range of 31 to 40 years and mostly holding master’s in business administration (mba) qualifications with 5 to10 years of experience in the banking sector. instrument reliability and validity this study used validated scales with minor modifications and was evaluated and assessed in the context of reliability and validity. in the first step, an exploratory factor analysis (efa) was conducted for hpws and relational coordination. the factor loading of the items indicated the presence of strong loadings of all variables on only one component. reliability was estimated using a common technique of cronbach alpha coefficient and interrater reliability coefficients of icc1 and icc2 (lebreton & senter, 2008). validity is assessed in terms of interrater agreement (ira) (see table 2). the alpha coefficient values for each scale, icc1 and icc2 were more than the recommended value justifying aggregation of hpws and rc at a branch level. in terms of validity, the average values of interrater agreement rwg(j) of hpws relational coordination for all functions exceed the recommended value of 0.70 (lebreton et al., 2003). table 2: reliability and validity analysis for hpws and rc departments icc1 icc2 intraclass (1) intraclass (2) interrater agreement two-way one-way rwgj hpws manager 0.265 0.904 0.273 0.912 0.93 operations 0.260 0.901 0.281 0.914 0.90 credit 0.238 0.890 0.241 0.902 0.94 cash 0.185 0.855 0.191 0.864 0.91 survey items alpha survey items alpha es 4 0.612 is 3 0.702 training 4 0.739 cc 3 0.602 participation 4 0.813 pa 5 0.853 jd 3 0.835 hpws 26 0.895 journal of applied economics and business studies, volume. 4, issue 2 (2020) 163-180 https://doi.org/10.34260/jaebs.428 171 table 2a: rotated component matrix for hpws components practices items 1 2 3 4 5 6 7 performance appraisal q25 .787 q27 .767 q26 .741 q28 .700 q29 .651 employee participation q10 .813 q9 .738 q11 .722 q12 .651 extensive training q7 .779 q5 .712 q8 .704 q6 .609 role clarity q14 .804 q15 .772 q13 .770 information sharing q19 .730 q17 .708 q20 .624 job security q2 .729 q3 .722 q4 .569 q1 .520 performance based compensation q21 .751 q22 .645 q23 .606 table 2b: component matrix for relational coordination component items factor loading accurately .785 timely .746 problem solving .732 shared goals .726 frequently .713 respect .703 shared knowledge .688 4. results and discussion the study is determined to measure the extent to which hpws can predict the degree of relational coordination; to examine whether there is a positive relationship between the hpws and firm performance in the banking sector of pakistan. data is analyzed using data envelopment analysis (dea) technique to identify the most efficient branches in terms of their muhammad siddique, zahoor khan & saleem gul 172 inputs and output efficiencies for 120 branches. these branches were analyzed through a widely used dea for the financial efficiency of decision-making units (dmu). dea is set up to measure the financial efficiency of units with several inputs and outputs (gutierrez & goitisolo lezama, 2011). to examine the effects of hpws on relational coordination and performance outcomes, dea efficiency scores were calculated through hpws as input for relational coordination and performance outcomes using bcc (banker, charnes, & cooper, 1984) input-based models. these inputs and outputs serve an exclusive combination to report the causes of either efficiency or inefficiency for each branch. tables 3, 4, and 5 provide efficiency scores for 120 branches among the inputs and outputs. based on the model of cinca and molinero (2004), this study calculated dea efficiency scores for three combinations. firstly, to estimate the effects of hpws in predicting relational coordination among employees (hpws as input and relational coordination as output), secondly, to assess the relationship of hpws with branch level performance outcomes (hpws as inputs and performance outcomes as outputs), and thirdly, the effects of relational coordination on performance outcomes (rc as inputs and performance as outputs). finally, for all three combinations, dea super efficiency scores were computed to rank the efficient branches (lovell & rouse, 2003). table 6 presents the ranking of branches in terms of superefficiency scores. the maximum score of an efficient branch is 1suggesting that the branch is fully efficient to convert inputs into outputs. branches that are less than 1 are considered inefficient branches on transforming inputs into outputs (charnes, cooper, & rhodes, 1978). concerning the effects of hpws in predicting relational coordination among employees, shown in table 3, 4 out of 120 branches were 100 percent efficient. these branches include dmu number 58, 109, 102, and 116. similarly, in terms of hpws and its effects on performance, shown in table 4, 6 out of 120 branches were fully efficient. a group of these 100 percent efficient branches comprises dmu 115, 101, 38, 103, and 33 respectively. besides, regarding the effects of relational coordination on performance outcomes, shown in table 5, a total of 6 out of 120 branches were found 100 percent efficient in terms of rc as inputs and performance as outputs. among full efficient branches, include dmu number 115, 103, 101, 38, 33, 45, 3, 109, and 15. results shown in tables suggest that dea efficiency scores of branches vary across a combination of inputs and outputs. several branches were efficient in several combinations. for instance, dmu such as 33, 38, 101, 105, and 109 are efficient branches in terms of using hpws-performance relationships and relational coordination to performance relationships. these findings are consistent with numerous previous studies that examined the linkages and found that hpws is positively related to firm performance (e.g. siddique et al., 2019; liao et al., 2009; gittell et al., 2010). together, these results contribute to the current understating of the linkages between hpws and organizational performance in the service context. this study journal of applied economics and business studies, volume. 4, issue 2 (2020) 163-180 https://doi.org/10.34260/jaebs.428 173 emphasizes the central role of relational coordination among employees as a meaningful element in the relationships between hpws and performance. table 3: super efficiency of branches on bcc input oriented models hpws-rc dmu vrs efficiency score dmu vrs efficiency score dmu vrs efficiency score dmu vrs efficiency score 1 0.56 31 0.32 61 0.46 91 0.67 2 0.41 32 0.49 62 0.53 92 0.56 3 0.45 33 0.58 63 0.62 93 0.73 4 0.43 34 0.54 64 0.72 94 0.68 5 0.71 35 0.55 65 0.58 95 0.62 6 0.58 36 0.63 66 0.61 96 0.51 7 0.42 37 0.42 67 0.56 97 0.84 8 0.67 38 0.50 68 0.34 98 0.55 9 0.48 39 0.43 69 0.39 99 0.72 10 0.45 40 0.46 70 0.55 100 0.37 11 0.57 41 0.36 71 0.52 101 0.39 12 0.47 42 0.32 72 0.52 102 1.00 13 0.47 43 0.42 73 0.57 103 0.83 14 0.57 44 0.49 74 0.45 104 0.56 15 0.63 45 0.52 75 0.70 105 0.49 16 0.42 46 0.35 76 0.73 106 0.63 17 0.50 47 0.52 77 0.57 107 0.63 18 0.48 48 0.54 78 0.52 108 0.70 19 0.44 49 0.63 79 0.62 109 1.00 20 0.61 50 0.52 80 0.44 110 0.60 21 0.47 51 0.60 81 0.44 111 0.74 22 0.52 52 0.69 82 0.57 112 0.53 23 0.57 53 0.56 83 0.42 113 0.56 24 0.67 54 0.44 84 0.44 114 0.51 25 0.58 55 0.62 85 0.49 115 0.71 26 0.50 56 0.66 86 0.52 116 1.00 27 0.40 57 0.50 87 0.56 117 0.43 28 0.65 58 1.00 88 0.58 118 0.58 29 0.60 59 0.44 89 0.54 119 0.83 30 0.44 60 0.67 90 0.58 120 0.96 table 4. super efficiency of branches on bcc input oriented models hpwsperformance dmu efficiency score dmu efficiency score dmu efficiency score dmu efficiency score 1 0.45 31 0.29 61 0.34 91 0.76 2 0.35 32 0.47 62 0.37 92 0.43 3 0.46 33 1.00 63 0.53 93 0.46 4 0.38 34 0.40 64 0.78 94 0.46 5 0.61 35 0.38 65 0.45 95 0.83 6 0.66 36 0.69 66 0.52 96 0.37 7 0.27 37 0.33 67 0.45 97 0.89 8 0.63 38 1.00 68 0.28 98 0.49 9 0.56 39 0.40 69 0.34 99 0.58 muhammad siddique, zahoor khan & saleem gul 174 10 0.34 40 0.42 70 0.41 100 0.29 11 0.68 41 0.27 71 0.40 101 1.00 12 0.33 42 0.32 72 0.51 102 0.53 13 0.41 43 0.36 73 0.38 103 1.00 14 0.44 44 0.51 74 0.38 104 0.43 15 0.69 45 0.96 75 0.40 105 0.45 16 0.40 46 0.30 76 0.55 106 0.48 17 0.46 47 0.43 77 0.56 107 0.43 18 0.34 48 0.34 78 0.37 108 0.59 19 0.38 49 0.39 79 0.43 109 1.00 20 0.47 50 0.32 80 0.31 110 0.57 21 0.33 51 0.39 81 0.31 111 0.59 22 0.38 52 0.46 82 0.39 112 0.34 23 0.37 53 0.43 83 0.42 113 0.43 24 0.52 54 0.31 84 0.38 114 0.42 25 0.41 55 0.70 85 0.46 115 1.00 26 0.38 56 0.37 86 0.50 116 0.56 27 0.32 57 0.34 87 0.32 117 0.35 28 0.54 58 0.46 88 0.27 118 0.48 29 0.75 59 0.29 89 0.27 119 0.58 30 0.35 60 0.44 90 0.36 120 0.86 table 5: super efficiency of branches on bcc input oriented models for rc-performance dmu vrs efficiency score dmu vrs efficiency score dmu vrs efficiency score dmu vrs efficiency score 1 0.68 31 0.55 61 0.28 91 0.53 2 0.51 32 0.54 62 0.25 92 0.40 3 1.00 33 1.00 63 0.43 93 0.24 4 0.36 34 0.30 64 0.63 94 0.26 5 0.72 35 0.27 65 0.40 95 0.55 6 0.44 36 0.45 66 0.38 96 0.28 7 0.25 37 0.33 67 0.43 97 0.40 8 0.36 38 1.00 68 0.41 98 0.36 9 0.45 39 0.51 69 0.44 99 0.44 10 0.31 40 0.80 70 0.37 100 0.39 11 0.59 41 0.40 71 0.39 101 1.00 12 0.28 42 0.90 72 0.55 102 0.16 13 0.54 43 0.52 73 0.27 103 1.00 14 0.57 44 0.59 74 0.50 104 0.37 15 1.00 45 1.00 75 0.21 105 0.60 16 0.77 46 0.47 76 0.28 106 0.33 17 0.62 47 0.29 77 0.66 107 0.29 18 0.28 48 0.26 78 0.34 108 0.35 19 0.52 49 0.27 79 0.27 109 1.00 20 0.38 50 0.22 80 0.32 110 0.39 21 0.27 51 0.24 81 0.29 111 0.38 22 0.32 52 0.22 82 0.23 112 0.36 23 0.28 53 0.30 83 0.94 113 0.33 24 0.34 54 0.29 84 0.63 114 0.41 25 0.26 55 0.46 85 0.68 115 1.00 26 0.30 56 0.22 86 0.43 116 0.22 journal of applied economics and business studies, volume. 4, issue 2 (2020) 163-180 https://doi.org/10.34260/jaebs.428 175 27 0.46 57 0.27 87 0.18 117 0.39 28 0.47 58 0.12 88 0.13 118 0.45 29 0.48 59 0.28 89 0.15 119 0.27 30 0.39 60 0.22 90 0.24 120 0.54 table 6: ranking of efficient branches based on super efficiency hpws-rc hpws-performance rc-performance dmu efficiency score rank dmu efficiency score rank dmu efficiency score rank 58 1.30 1 115 1.52 1 115 1.88 1 109 1.23 2 101 1.47 2 103 1.86 2 102 1.08 3 38 1.43 3 101 1.84 3 116 1.00 4 103 1.40 4 38 1.80 4 33 1.38 5 33 1.71 5 109 1.26 6 45 1.61 6 3 1.51 7 109 1.18 8 15 1.05 9 5. conclusion and implications this study was aimed to assess the effects of hpws in predicting the degree of relational coordination among employees and improving the efficiency of unit-level performance in the banking service sector of pakistan. the data obtained from officer cadre employees of 218 branches about hpws and relational coordination and branch performance outcomes including deposits, advances, and profitability was analysed using dea techniques for efficiency scores of bank branches in terms of the use of inputs in generating efficient output. the superefficiency scores were computed for analysing the effect of hpws and relational coordination in improving the efficiency of branches. using input-oriented bcc data envelopment analysis models, super efficiency scores were computed for 120 branches and the most efficient branches were ranked in terms of three combinations of input-output models including hpwsrc; hpws-performance; and rc-performance. the findings suggested that based on the combination of input-output, certain branches were more efficient on hpws in predicting relational coordination among employees, while some branches were highly efficient on hpws and relational coordination concerning performance outcomes of deposits, advances, and profitability. these results suggest that hpws designed to foster relational coordination among employees will positively influence unit-level performance. the study provides a new understanding of the linkages between hpws and performance. it provides practical implications for policymakers suggesting the importance of employees’ perspectives as the findings present an important indication that there was a relational process, whereby relational coordination was related to an extent of hpws and its impact on unit-level performance outcomes. muhammad siddique, zahoor khan & saleem gul 176 references adler, p., kwon, s. & heckscher, c. (2008). the emergence of collaborative community, organization science, 19(2), 359–376. appelbaum, e., bailey, t., berg, p. & kalleberg, a. l. (2000). manufacturing advantage: why highperformance work systems pay off. washington, d. c.: economic policy institute. babbie, e. (2004). the practice of social research, wadsworth, thomson learning inc. banker, r. d, charnes, a, & cooper, w. (1984). some models for estimating technical and scale inefficiencies in data envelopment analysis. management science, 30(9), 10781092. bing, m., guimei, m., xiaolang l., & lassleben, h, (2020). relationship between a highperformance work system and employee outcomes: a multilevel analysis. social behavior & personality: an international journal, 48(1), 1-9. boon, c., den hartog, d., & lepak, d. (2019). a systematic review of human resource management systems and their measurement. journal of management, 45(6), 2498–37. boselie, p., dietz, g. & boon, c. (2005). commonalities and contradictions in hrm and performance research, human resource management journal, 15(3), 67–94. bowen, d. e. & ostroff, c. (2004). understanding hrm-firm performance linkages: the role of the strength of the hrm system’. academy of management review, 29(2), 203-212. boxall, p. (2012). building highly performing work systems: analysing hr systems and their contribution to performance, in: guest, d., paauwe, j., and wright, p. hrm and performance: achievements and challenges. hoboken: wiley, hoboken. 108-133. chao, w., & chih-ting, s. (2018). customer service-focused hrm systems and firm performance: evidence from the service industry in taiwan, the international journal of human resource management, 29(19), 2804-2826. charnes, a, cooper, w, & rhodes, e. (1978). measuring the efficiency of decision-making units. european journal of operational research, 2(6), 429-444. cinca, c., serrano, & molinero, c. (2004). selecting dea specifications and ranking units via pca. journal of the operational research society, 55(5), 521-528. collins, c. j. & smith, k. g. (2006). knowledge exchange and combinations: the role of hr practices in the performance of high-technology firms’. academy of management journal, 49(3), 544-560. delery, j. & doty, d. (1996). modes of theorizing in strategic human resource management. academy of management journal, 39(4), 802-835. journal of applied economics and business studies, volume. 4, issue 2 (2020) 163-180 https://doi.org/10.34260/jaebs.428 177 dyer, l. & reeves, t. (1995). human resource strategies and firm performance: what do we know and where do we need to go?, international journal of human resource management, 6(3), 656-670. faraj, s. & sproull, l. (2000). coordinating expertise in software development teams, management science, 46(12), 1554–1568. fu, n., bosak, j., flood, p. c., & ma, q. h. (2019). chinese and irish professional service firms compared: linking hpws, organizational coordination, and firm performance. journal of business research, 95, 266–276. gerhart, b. (2012). research on human resources and effectiveness: some methodological challenges, in: guest, d., paauwe, j., and wright, p. hrm and performance: achievements and challenges. hoboken: wiley, hoboken.108-133. gittell, j. (2001). supervisory span, relational coordination and flight departure performance. organization science, 12(4), 468-483. gittell, j. h. (2002). coordinating mechanisms in care provider groups: relational coordination as a mediator and input uncertainty as a moderator of performance effects, management science, 48(11), 1408–1426. gittell, j., seidner, r. & wimbush, j. (2010). a relational model of how high performance work systems work. organization science, 21(2), 490-506. graham, l., zheng, y., gracey, s., & plater, m. (2020). the role of managers’ values in the relationship between high performance work systems and customer satisfaction: a team-level perspective. available at ssrn: https://ssrn.com/abstract=3553721. guest, d. (1997). human resource management and performance: a review and research agenda, international journal of human resource management, 8(3), 263-76. gutierrez, j., & goitisolo lezama, b. (2011). profitability and social performance of microfinance institutions: empirical evidence of relations between different types of variables. revista de economía mundial (27), 189-214. han, j. h., bartol, k. m., & kim, s. (2015). tightening up the performance–pay linkage: roles of contingent reward leadership and profit-sharing in the cross-level influence of individual pay-for-performance. journal of applied psychology, 100(2), 417–430. havens, d. s., gittell, j. h., & vasey, j. (2018). impact of relational coordination on nurse outcomes: achieving the quadruple aim. journal of nursing administration, 48(3), 132-140. https://ssrn.com/abstract=3553721 muhammad siddique, zahoor khan & saleem gul 178 huselid, m. a. (1995). the impact of human resource management practices on turnover, productivity, and corporate financial performance, academy of management journal, 38(3), 635-672. karadas, g., & osman, m. karatepe, (2019). unravelling the black box: the linkage between high-performance work systems and employee outcomes, employee relations, 41(1), 67-83. kehoe, r. r., & collins, c. j. (2017). human resource management and unit performance in knowledge-intensive work. journal of applied psychology, 102, 1222–1236. lebreton, j., burgess, j., kaiser, r., atchley, e., & james, l. (2003). the restriction of variance hypothesis and interrater reliability and agreement. organizational research methods, 6(1), 80-128. lebreton, m. j. & senter, l. j. (2008). answers to 20 questions about interrater reliability and interrater agreement, organizational research methods, 11(4), 815-835. lee, h. w., pak, j., kim, s., & li, l. z. (2019). effects of human resource management systems on employee proactivity and group innovation. journal of management, 45(2), 819–846. lepak, d., liao, h., chung, y., & harden, e. (2006). a conceptual review of human resource management systems in strategic human resource management research. research in personnel and human resource management, 25, 217–271. liao, h., toya, k., lepak, d. p. & hong, y. (2009). do they see eye to eye? management and employee perspectives of high-performance work systems and influence processes on service quality, journal of applied psychology, 94(2), 371-391. lovell, cak, & rouse, apb. (2003). equivalent standard dea models to provide superefficiency scores. journal of the operational research society, 54(1), 101-108. macky, k. & boxall, p. (2007). the relationship between high performance work practices and employee attitudes: an investigation of additive and interaction effects. international journal of human resource management, 18(4), 537-567. malhotra, p., & singh, m. (2016). indirect impact of high performers on the career advancement of their subordinates. human resource management review, 26(3), 209– 226. methot, j. r., melwani, s., & rothman, n. b. (2017). the space between us: a socialfunctional emotions view of ambivalent and indifferent workplace relationships. journal of management, 43(6), 1789–1819. https://www.emeraldinsight.com/author/karadas%2c+georgiana https://www.emeraldinsight.com/author/karatepe%2c+osman+m journal of applied economics and business studies, volume. 4, issue 2 (2020) 163-180 https://doi.org/10.34260/jaebs.428 179 methot, j. r., rosado-solomon, e. h., & allen, d. g. (2018). the network architecture of human capital: a relational identity perspective. academy of management review, 43(4), 723–748. meuer, j. (2017). exploring the complementarities within high-performance work systems: a set theoretic analysis of u.k. firms. human resource management, 56(4), 651–672. mingqiong, z., cherrie, j, dowling, p., & bartram, t. (2013). exploring the effects of highperformance work systems (hpws) on the work-related well-being of chinese hospital employees, the international journal of human resource management, 24(16), 31963212. murphy, k., torres, e., ingram, & hutchinson, j. (2018). a review of high performance work practices (hpwps) literature and recommendations for future research in the hospitality industry. international journal of contemporary hospitality management, 30(1), 365– 388. nien-chi liu & yi-ting lin (2019): high-performance work systems, management team flexibility, employee flexibility and service-oriented organizational citizenship behaviours, the international journal of human resource management, doi: 10.1080/09585192.2019.1651374 ogbonnaya, c., & valizade, d. (2018). high performance work practices, employee outcomes and organizational performance: a 2-1-2 multilevel mediation analysis. the international journal of human resource management, 29(2), 239–259. ostroff, c. & bowen, d. (2000). moving hr to a higher level: hr practices and organizational effectiveness. in klein kj, kozlowski swj (eds). multilevel theory, research, and methods in organizations: foundations, extensions, and new direction. san francisco, ca: jossey-bass; 211–67. paauwe, j., wright, p. & guest, d. (2012). hrm and performance: what do we know and where should we go? in: guest, d., paauwe, j., &wright, p. hrm and performance: achievements & challenges. hoboken: wiley, hoboken. 28-53. paradi, j.c., yang, z. & zhu, h. (2011). assessing bank and bank branch performance modelling considerations and approaches’ in handbook on data envelopment analysis international series in operations research and management science, 164, 315-361. pfeffer, j. (1998). the human equation: building profits by putting people first. boston: harvard business school press. romanow, d. s., rai, a., & keil, m. (2018). cpoe-enabled coordination: appropriation for deep structure use and impacts on patient outcomes. mis quarterly, 42(1), 189-212. http://link.springer.com/book/10.1007/b105307 http://link.springer.com/book/10.1007/b105307 http://link.springer.com/bookseries/6161 muhammad siddique, zahoor khan & saleem gul 180 russell, z. a., steffensen, d. s., ellen, b. p., zhang, l., bishoff, j. d., & ferris, g. r. (2018). high performance work practice implementation and employee impressions of line manager leadership. human resource management review, 28(3), 258–270. scott, a. s. & james, w. d. (1992). integrated manufacturing and human resource management: a human capital perspective’. the academy of management journal, 35(3), 467-504. siddique, m., procter, s. & gittell, j. (2019). the role of relational coordination in the relationship between high-performance work systems (hpws) and organizational performance, journal of organizational effectiveness: people and performance, 6(4), 246-266. sun, j.-m., xing, l., yin, k., & yang, y. (2018). when does high performance work system increase employees' well-being? the role of core self-evaluation and need for achievement. journal of capital university of economics and business, 20(6), 44–53. wright, p., & ulrich, m. (2017). a road well-traveled: the past, present and future journey of strategic human resource management. annual review of organizational psychology and organizational behavior, 4, 45–65. zacharatos, a., barling, j., & iverson, r. (2005). high-performance work systems and occupational safety. journal of applied psychology, 90(1), 77-93. https://www.emerald.com/insight/search?q=muhammad%20siddique https://www.emerald.com/insight/search?q=stephen%20procter https://www.emerald.com/insight/search?q=jody%20hoffer%20gittell https://www.emerald.com/insight/publication/issn/2051-6614 journal of applied economics and business studies, volume. 4, issue 2 (2020) 23-40 https://doi.org/10.34260/jaebs.422 23 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x technological spillovers, manufacturing growth and transboundary pollution in case of pakistan muhammad usman1, gulnaz hameed2*, abdul saboor3 and khuram nawaz sadozai4 1 phd scholar at pmsa-arid agriculture university rawalpindi. 2 assistant professor at pmsa-arid agriculture university rawalpindi 3 professor at pmsa-arid agriculture university rawalpindi 4 assistant professor, department of agricultural & applied economics, the university of agriculture, peshawar. abstract role of technological spillover in manufacturing sector growth and climate change is the running debate in the world to avoid the problem of production inefficiency and environmental damages. environment friendly technological spillover plays pivotal role in manufacturing sector growth which leads to economic growth. in order to investigate the inconclusiveness of the major issues of production inefficiency and climate change in pakistan the current research was aimed at finding the relationship between technological spillovers, manufacturing growth and climate change. to meet the objectives, the study investigated both short run and long run dynamics by employing autoregressive distributive lagged (ardl) model. an annual time series data over the period of 1973 to 2017 was collected for comparative analysis of technological spillover performance in manufacturing sector and environmental condition of the country. the results of cusm test and bound test validated the existence of long run cointegration relationship among estimated models. the results demonstrated that technological spillover performs significantly positive role in manufacturing growth with less absorptive ability. the empirical analysis proved that technological spillover and imported technology have affirmative role in reducing amount of net carbon emission over the long run. it is suggested that the firms should adopt innovative technologies and try to improve absorptive capacity while government must opt country specific policies to control negative externalities. keywords technology spillover, absorptive capacity, manufacturing growth, time series empirics jel classification n7, o31, f18 * corresponding author: gulnaz.hameed@uaar.edu.pk muhammad usman, gulnaz hameed, abdul saboor and khuram nawaz sadozai 24 1. introduction technological spillover has performed straightforward functioning in manufacturing growth, environmental performance in both developed and developing countries (coe, helpman, & hoffmaister, 2009; di maria & smulders, 2017; z. liu et al., 2015). technological spillover execute central role in industrial productivity, new innovation, learning by doing and value addition of manufacturing sector. it is destroying environment severely in developing countries due to large population and lack of precautionary measures adopted for imported technology. however, international trade plays positive role in knowledge spillover, new innovation and introduces efficient product varieties across the region. international trade in technological products increases the market size through innovative commodity varieties. trade openness plays a fundamental role for introducing new varieties as it provides ways to access technical knowledge which subsequently reduces the cost of innovation (coe & helpman, 1995). the innovative technology provides more efficient products with less negative externalities. the firm expenditure on research and development (r&d) spillover and firm’s own innovation perform vital role to reduce the emission of co2 in the environment. the r&d spillover has both direct (improved output performance) and indirect (reduced carbon emission) impacts on industrial performance (jiao, yang, & bai, 2018; lee, 2013). currently, global warming and climate change is attaining more attention to reduce net co2 emission and clean environment to save the ozone layer. so, initiative of iso-9001 certification for quality control accomplish positive role in reduction of negative externalities, innovation and producing market compatible products which attain through domestic and foreign r&d spillovers (alemdar & özyildirim, 2002; bittencourt & giglio, 2013; mayer, 2000). in addition, an open economy extracts larger productivity benefits from foreign r&d expenditure than less open economies through trade liberalization (coe & helpman, 1995; coe et al., 2009). the knowledge based technological spillover performs fundamental role in output performance of host country. however, developing countries can enhance their production through investing in r&d spillover. the technology imports from developed to developing countries not only results in increased productivity but it also improves education level and innovative thinking of host country’s labor force (mingyong, shuijun, & qun, 2006; seck, 2012). furthermore, countries’ disbursements on human capital to learn new knowledge directly improve the education, skill and learning by doing process which leads to raise in output productivity (mayer, 2000). the empirical results of dalgıç & mıhç, 2013; gorkey-aydinoglu suggested that contribution of r&d spillover by the means of foreign technology and domestic innovation has significant impact on economic growth. in addition, imported technology creates externalities which may be journal of applied economics and business studies, volume. 4, issue 2 (2020) 23-40 https://doi.org/10.34260/jaebs.422 25 positive or negative. the technology externalities include the technological transformation and production of co2 emission, manufacturing wastes, and other harmful gases which affect the climate factors. but technology spillover with less negative externalities is more valuable and socially optimal. similarly, youssef, 2009 argued that increase in r&d spillover lead to positive externalities and increased production while reduced the pollution thus raised the society’s welfare. w. liu, xu, yang, zhao, & xing, 2016 highlighted through empirical analysis that technological spillover affected the manufacturing output positively with innovative knowledge creation. in short term it also produces negative externalities (like, co2 emission which damages the environmental very poorly). the pollution spillover has negative impact on total factor productivity (tfp) growth with larger magnitude of country’s own emission control. the rise in pollution is harmful for tfp growth in both developed and developing countries (costantini, mazzanti, & montini, 2013; empora, 2017). new technological innovation and new product varieties lead to unique product to get high market share (w. liu et al., 2016). abatement cost will fall as a result of successful r&d activities and alternative use of knowledge spillover (heal & tarui, 2010). other pioneer studies including that of dalgıç & mıhç, 2013; türker, 2012; vayá, lópez-bazo, moreno, & suriñach, 2004 argued that r&d spillover and innovation capacity of host country is dependent on innovative thinking and absorptive ability. a strategic partnership is required between research institute and domestic firms which increases the system of innovative thinking and technology improvement. the firm gain profit on the basis of unique product in competitive market, this strategy increased customer satisfaction, performance and comparative advantage on other firms (seck, 2012; zarrabi & vahedi, 2012) . the scientific literature shows that in the upcoming decade high temperature and change in climate is burning issue for developing countries, so high temperature and climate change carrying significant increase in amount of hydro events and bring extreme changes in temperature in current century. the average global temperature from last century has accelerated by 0.8 co, whereas changes in precipitations is reducing over the time. currently, global issues are arising such as climate change, deforestation, global warming, depletion of ozone layer, and biodiversity preservation. such issues cause environmental damages and create serious problems for future (fischer, 2008; lei, shangguan, & rui, 2012; z. liu et al., 2015). in case of pakistan both industrial and services sector has more foreign direct investment (fdi) with higher labor productivity share while agriculture sector presenting poor condition of fdi inflow with low labor force productivity. mohammadi, veismoradi, hashemi, akbari, & rostami, 2015; serfraz, 2017 found that agriculture sector requires more technology for increasing muhammad usman, gulnaz hameed, abdul saboor and khuram nawaz sadozai 26 potential level of output to attain the fruit from r&d spillovers. mehmood, chaudhry, tufail, & irfan, 2009 highlighted the probability of new technology adoption in large scale manufacturing in pakistan and suggested that new technology adoption depends on the size and volume of sale, type of ownership, geographical and environmental condition. kuo & young, 2008 concluded that fdi increases the technology spillover by increasing carbon emission in host country. pakistan textile manufacturing industry produces wastewater which is not only demolishing the environment but also reduces the output productivity of textile industry. the treatment of textile wastewater is costly process which generates long lasting effects on climate factors (verma, dash, & bhunia, 2012). technological spillover in manufacturing sector produces harmful, waste water, smoke, and other waste material which produce direct impact on climate change. in pakistan, the research activities and research expenditure were increasing over the last two decades (mehmood et al., 2009). every research activity has positive outcome but unfortunately such research activities are not contributing for economic development of the country. research activities have been promoted but innovative ideas could not further proceed to be implemented for economic development. with the advent of endogenous growth theory, a large segment of economic literature has an eternizing interest in the relationship between r&d spillovers and role of absorptive ability to enhance the long term economic growth (kazmi, 2016; kinda, 2012; mustafa, 2011). like other developing countries, pakistan’s technological knowledge mostly depends on knowledge spillover produced in developed countries (kazmi, 2016; mehmood et al., 2009). so there is a dire need to answer the questions like how the technology spillover contributes in the manufacturing growth? does the manufacturing sector labor force is efficient to absorb the spillover technology? does technological spillover raise environmental pollution in the country? how trade liberalization contributes in the economic growth via r&d spillovers in pakistan? what is the role of technological spillover in manufacturing sector in view of aspects like production efficiency, productivity growth, absorptive capacity of industrial labor, environmental pollution produced by industrial sector, and importance of iso-9001 certification in pollution reduction? to address the above questions the specific objectives of current study are; to examine the role of technological spillover in manufacturing sector growth/output performance through technology diffusion in pakistan, to investigate the absorptive ability of new knowledge spillover in pakistani work force in manufacturing sector, to explore specific foreign r&d channel suitable for manufacturing growth, to estimate the transboundary pollution in pakistan and role of technology spillover in climate change and co2 emission. journal of applied economics and business studies, volume. 4, issue 2 (2020) 23-40 https://doi.org/10.34260/jaebs.422 27 2. research methods 2.1 theoretical background the numerous channels and ways were explored in voluminous stream of research to investigate the national and international technology spillovers impact on country industrial productivity. like globalization, foreign direct investment (fdi), cross country, regional and sectoral technology spillovers effects, knowledge sharing interactions among countries, multinational presence, the coe & helpman, 1995 seminal work explore empirically the contribution of r&d spillovers to country’s tfp level. the imports of intermediate goods (machinery and equipment’s) are the main channel of technology spillovers. “technology consists of a set of processes like how to produce, how to use the tools/equipment’s and experiments that are used to produce the goods and services and how to transfer information from internal and external resources to provide innovative ways for input output process?” (gujarati & dawn, 2010). in development, innovative technology with efficient utilization of the human capital perform key role to absorb and develop new ways of production. human capital comprises of human resources, education, skillful creative and innovative labor, which majorly contribute in development of technology. “in an environment of uncertainty, innovation capacity developed through knowledge transfer, and the macro-level growth models, has emphasized the role of science and technology models” (romer, 1990). the environmental kuznets curve (ekc) shows the postulated connection between environment quality, innovative technology and economic development i.e. high per capita income (hanley, 2001; hanley, shogren, & white, 2016; stern, 2004). the ekc monotonically showed raise in emissions which is evident to decrease the income in long term. according to stern (2004), review of both the theoretical and empirical work on the ekc leads to be skeptical about the existence of a simple and predictable relationship between pollution and per capita income. the higher level of income and sustained environment can be achieved through innovation and adoption of new knowledge spillover. however, the emission reduces the technological innovation which declines the income not only in developing but also in developed countries(stern, 2004; torras & boyce, 1998). in contrast, the innovative technological spillover is helpful in reduction of pollution emission. the important point is why markets and manufacturing sector fails to allocate the resources? the reason behind is negative externalities(hanley et al., 2016). hanley, 2001 argued that analysis of environmental issues in terms of externalities shows fails in the resource allocation. currently, the phenomena of “investment in green technology” is workable which can boost the firm productivity, capacity to get competitive advantages and measures for pollution reduction (galdeano-gómez, cespedes-lorente, & martínez-del-río, 2008). muhammad usman, gulnaz hameed, abdul saboor and khuram nawaz sadozai 28 2.2 data and data sources the time series empirics are based on r&d spillover models concentrate on manufacturing sector’s productivity and environmental factor (coe et al., 2009; s. wang, 2015; x. wang, fang, zhang, & fang, 2018). in order to examine the role of technological spillover through different aspects of manufacturing sector of pakistan, the manufacturing sector productivity and net carbon emission are taken as dependent variables in two different models while net foreign direct investment (fdit), technological imports (techt), trade openness (tot) as main dependent variables. however, human capital (hct), labor force in industrial sector (lft), are taken as control variables in both models, in addition to capture the absorptive ability of foreign knowledge, interactive term of human capital with foreign direct investment (hc*fdit). for empirical outcome, the time series annual data for the period of 1973 to 2017 was gathered from various sources. data relevant to pakistan technological spillover, manufacturing sector performance and climate change is collected from the various issues of pakistan economic survey, federal bureau of pakistan, ministry of finance pakistan, hand book of statistics of pakistan’s economy 2015 published by state bank of pakistan, penn world table 9.1 and world bank, and world development indicators (wdi). 2.3 empirical framework considering the technological augmented model and knowledge spillover as endogenously determined, the level of technology spillover is dependent on both domestic and foreign r&d variables (coe et al., 2009; mohammadi et al., 2015). hence, the equation for technological level can be written as; 𝐴𝑡 = 𝐷𝑅𝐷𝑡 𝛼1 , 𝐹𝑅𝐷𝑡 𝛼2 (1) here the at is total factor productivity (tfp), drdt is domestic r&d while frdt is foreign r&d. taking log and add intercept in equation 1 follows as, 𝑙𝑛𝐴𝑡 = 𝛼0 + 𝛼1𝐷𝑅𝐷𝑡 + 𝛼2𝐹𝑅𝐷𝑡 (2) where, 𝐷𝑅𝐷𝑡 representing the domestic r&d variables and 𝐹𝑅𝐷𝑡 representing foreign r&d variables affects the technological spillover effect on economic growth and absorptive capacity. the study took the intensive form of hicks neutral, (dupuy, 2006) whereas growth per unit of effective labor(𝑦𝑡 = 𝑌𝑡 𝐿𝑡 ⁄ ), technology per unit of effective labor (𝐴𝑡 = 𝐴𝑡 𝐿𝑡 ⁄ ) and capital per unit of effective labor (𝑘𝑡 = 𝐾𝑡 𝐿𝑡 ⁄ ) are given in equations 3. 𝑦𝑡 = 𝐴𝑡 𝑘𝑡 𝛽 (3) journal of applied economics and business studies, volume. 4, issue 2 (2020) 23-40 https://doi.org/10.34260/jaebs.422 29 taking log the equation (3) takes the following form 𝑙𝑛𝑦𝑡 = 𝑙𝑛𝐴𝑡 + 𝛽𝑙𝑛𝑘𝑡 (4) putting value of 𝑙𝑛𝐴𝑡 from equation (2) 𝑙𝑛𝑦𝑡 = 𝛼0 + 𝛼1𝑙𝑛𝐷𝑅𝐷𝑡 + 𝛼2𝑙𝑛𝐹𝑅𝐷𝑡 + 𝛽𝑙𝑛𝑘𝑡 (5) in equation 5, yt is productivity output in manufacturing sector, while domestic 𝑅𝐷𝑡 𝑑 variables and foreign 𝑅𝐷𝑡 𝑓 variables are presented in weighted. the weighted frdt stock is proxy of technological spillover in manufacturing sector (dalgic and michi, 2013) in this regard weighted stock of both domestic and foreign r&d coefficients demonstrate the technological spillover in industrial sector. in the single stage estimation, by including the technological spillover as well as absorptive capacity series of variables, the nonlinear estimation equation become, 𝑙𝑛𝑦𝑡 = 𝛼0 + 𝛼1𝑙𝑛𝑇𝑆𝑡 + 𝛼2(𝑋𝑡 ) + 𝐻𝐶 ∗ 𝑙𝑛𝑇𝑆𝑡 + 𝑡 (6) for empirical analysis manufacturing value addition annual growth and net carbon emission are taken as dependent variables in two different models. the study estimated model 2.6 by adopting the foreign rd spillover indicators. empirical outcomes of ols based autoregressive distributive lag (ardl) model of conditional unrestricted error correction approach is applied for long run cointegration association which was develop by (m hashem pesaran, shin, & smith, 1996). to investigate the long as well as short run appearance of variables under consecration the ardl technique is applied which provides the parametric stability through cumulative sum (cusum), and cumulative sum of square (cusumq) tests. for accomplishment of long run and short run dynamics, the bound test show the existence of long run relationship and error correction model (ecm) coefficient determines the short run equilibrium and time for convergence to stability position. in vector error correction (var) approach the lag structure remains same for concerning variables while ardl technique provide way to select optimal and different lags of each variable. the ardl estimate used the following equation for parametric action: ∆ln(manu)t = β0 + ∑ β1∆ln(manu)t−1 n t=1 + ∑ β2∆ln(fdi)t−1 n t=1 + ∑ β3∆ln(tech)t−1 n t=1 + ∑ β4∆ln(to)t−1 n t=1 + ∑ β 5 ∆ln(lf)t−1 n t=1 ∑ β6∆ln(hc)t−1 n t=1 + ∑ β7∆ln(hc × fdi)t−1 n t=1 + δ1ln(manu)t−1 + δ2ln(fdi)t−1 + δ3ln(tech)t−1 + δ4ln(to)t−1 + δ5ln(hc)t−1 + δ6ln(lf)t−1 + δ7ln(hc ∗ fdi)t−1 + εt (7) the equation given above is representation of short run dynamics of ardl model whereas 𝛽0, 𝛽1, … … … … 𝛽7, are short run parameters, (the variable in change (δ) form are short run parameters) while parameters 𝛿1, 𝛿2, … … … … … 𝛿7 represent long term association and 𝑡 represent the error term in the model. in equation 7, the dependent variable is manufacturing muhammad usman, gulnaz hameed, abdul saboor and khuram nawaz sadozai 30 section value addition (manut) while explanatory variables are net foreign direct investment (fdit), technological imports (techt), trade openness (tot) as main dependent variables. however, human capital (hct), labor force in industrial sector (lft), are taken as control variables in both models, in addition to capture the absorptive ability of foreign knowledge, we used interactive term of human capital with foreign direct investment (hc*fdit) the estimated equation for impact of technology spillover on environmental performance is given below. in equation 8 the independent variables are same as equation 7 while dependent variable is co2 emission produced by manufacturing sector in pakistan. ∆ln(co2)t = β0 + ∑ β1∆ln(manu)t−1 n t=1 + ∑ β 2 ∆ln(fdi)t−1 n t=1 + ∑ β 3 ∆ln(tech)t−1 n t=1 + ∑ β 4 ∆ln(to)t−1 n t=1 + ∑ β 5 ∆ln(lf)t−1 n t=1 ∑ β 6 ∆ln(hc)t−1 n t=1 + ∑ β 7 ∆ln(hc × fdi)t−1 n t=1 + δ1ln(manu)t−1 + δ2ln(fdi)t−1 + δ3ln(tech)t−1 + δ4ln(to)t−1 + δ5ln(hc)t−1 + δ6ln(lf)t−1 + δ7ln(hc ∗ fdi)t−1 + 𝜇𝑡 here 𝜇𝑡 represent the error term in the model. 3. results and discussion before carrying out a formal analysis, to determine the existence of long run relationship among the considered variables, conventional unit root test namely augmented dickey fuller (adf) was applied. 3.1 unit root test results of the unit root test are given in table 1. time series empirics are highly dependent on cointegration relationship but it is mandatory to investigate the order of integration of given variables before final estimation method for co-integration relation to avoid the problem of spurious results. the adf test describes the behavior of given variables and report that given variables can be used in ardl or not. the findings of adf test are given table 3.1 which shows that all the given variables are integrated at level or first difference (i (0) or i (1)), so ardl method of cointegration is applicable for final empirical outcomes. (8) journal of applied economics and business studies, volume. 4, issue 2 (2020) 23-40 https://doi.org/10.34260/jaebs.422 31 table 1: result of unit root test variables level 1stdifference manut -1.179 (0.9015) -6.229 (0.0000)** co2t -1.760 (0.5250) -5.068 (0.0009)** fdit -2.352 ( 0.0197)* -4.306 (0.0001) lft -1.300 (0.8746) -6.937 (0.0000)** techt -1.071 (0.7045) -4.011 (0.0411)** tot -1.013 (0.9316) -6.028 (0.0001)** hct -1.939 (0.6169) -6.465 (0.0000)** *, ** shows the stationary of given variables at level and first difference correspondingly source: authors' own estimation. the optimal lag selection is important for ardl analysis after detecting stationarity of variables. the aic is not as consistent but is usually more efficient, while sbc is usually more consistent but inefficient (brooks, 2008). for final analysis, 4 lags are selected on the bases of the akaike information criterion (aic), and hannanquinn criterion (hqc). the output results are shown in table 2. table 2: var base optimal lag selection criterion *shows the optimal lag length for ardl bound test. source: authors' own estimation. 3.2 technology spillover role in manufacturing growth and absorptive ability this section attempts to accomplish the precise objectives of the study like technology spillover role in manufacturing growth, testing the absorptive ability of manufacturing labor force and highlights the suitable channel for technology spillover in manufacturing sector. the result of adrl bound test is given in table 3, the null hypothesis of bound test is “no long run relationship exist”. on the basis of f-stat calculated value, the null hypothesis is rejected because f-test value fall in rejection rejoin, which is higher than upper bounds critical values. in addition, the result shows the existence of long run cointegration relationship among variables under consideration. lag logl lr fpe aic sc hq 0 -66.94312 na 9.51e-08 3.697156 3.992710 3.804019 1 187.2325 406.6810 3.46e-12 -6.561626 -4.197194* -5.706722 2 239.9014 65.83612 3.63e-12 -6.745070 -2.311762 -5.142126 3 293.4926 48.23204 5.75e-12 -6.974628 -0.472442 -4.623643 4 415.6222 67.17127* 8.51e-13* -10.63111* -2.060045 -7.532082* muhammad usman, gulnaz hameed, abdul saboor and khuram nawaz sadozai 32 table 3: ardl bound test results null hypothesis: no long-run relationships exist test statistic value k f-statistic 8.069472 6 critical value bounds significance i0 bound i1 bound 10% 1.75 2.87 5% 2.04 3.24 2.5% 2.32 3.59 1% 2.66 4.05 source: authors' own estimation. the results of long run relationship among technological spillover and manufacturing growth are given in table 4. in this model, foreign direct investment (fdit), technological imports (techt) and trade openness (tot) are three proxies to capture the technological spillover. the results of foreign direct investment, trade openness, and technology imports are statistically significant with positive sign. so, increase in technological spillover from foreign source leads to raise the manufacturing growth over the long run. the coefficient of foreign direct investment (fdit) has significantly positive impact on manufacturing sector productivity both in short as well as in long run dynamics with long run elasticity value 0.23. in similar line, trade openness (tot) and technological imports (techt) perform significantly positive impact on economic growth in long run with elasticity value 1.46 and 0.15 respectively. the trade openness does not exert any significant impact in short run with contradiction of technological imports. the lag coefficient of manufacturing lector growth has positively significant impact on manufacturing growth which shows technology upgrading is important with previous technology (old technology). the results of the study are consistent with the findings of liu et al. (2016). however, the findings of industrial labor force, and interactive term is significantly negative. the possible justification is that the technology imported from developed countries cannot be absorbed efficiently. the findings of labor productivity in manufacturing sector are consistent with the findings of kuo & young, 2008. labor force working in manufacturing is mostly illiterate, less skilled and inexperienced, so labor force is not capable to absorb the imported technology. so, the interactive term result shows that learning through technological spillover cannot perform significant role to absorption of external knowledge spillover, to enhance internal innovation and innovative thinking the absorptive ability matter (bittencourt & giglio, 2013). research and development spillover perform crucial role to innovate and increase the capability of “learning by doing” and “learning through training” (pavitt, 1984). the coefficient of human capital has positively significant impact on output performance. this shows that investment on education and pays high return on that education. the behavior of estimated coefficients are consistent with the journal of applied economics and business studies, volume. 4, issue 2 (2020) 23-40 https://doi.org/10.34260/jaebs.422 33 findings of kuo & young, 2008; mingyong et al., 2006; seck, 2012 as more open economies extract larger productivity benefits from foreign r&d expenditure than less open economies. in short run, fdit, techt insert significantly positive impact on manufacturing growth while tot, lft, hct, have no significant impact on manufacturing growth. in short run, interactive term (hc*fdi)t has significantly negative impact on manufacturing growth, which shows that technology imported through fdi cannot absorbed effectively. in addition, the lag coefficient of manuf(-1) growth has significantly positive impact on current year performance with 0.35 elastic value. the error correction mechanism ecm (-1) coefficient has significantly negative sign at 1 percent level which shows that variables in given model are cointegrated in long run. the coefficient of ecm is 1.28, which implies that deviation from steady state position in manufacturing sector in estimated model is corrected by 128 percent by the same time period (shittu, 2012). the possible justification of the high value ecm coefficient is dependency of manufacturing sector of pakistan which is based on imported technology, so any shock in manufacturing sector must be corrected 128 percent in same year through imported technology or fdi shocks. the ecm coefficient shows that non-adjustment in manufacturing sector takes place relatively at more quick speed of adjustment. table 4: long run and short run dynamics of ardl model long run coefficients short run coefficients variable coefficient variable coefficient fdit 0.233074 (0.0001)*** d(fdi)t 0.160608 (0.0923)* tot 1.461524 (0.0000)*** d(to)t 0.049690 (0.8134) techt 0.154219 (0.0028)*** d(tech)t 0.032255 (0.0490)** hct 1.239437 (0.0001)*** d(hc)t 0.353613 (0.8457) hc*lfdit -0.583203 (0.0000)*** d(hc*fdi)t -0.290169 (0.0842)* lft -0.075445 (0.0205)** d(lf)t -0.009229 (0.4626) d(manuf(-1))t 0.359062 (0.0830)* ecm(-1) -1.289925 (0.0004)*** the explained variable is lmanu. the variables under consideration are in logarithmic form because data of given variables are not available in same unit. so to enhance the actual outcome log of variables is used. the satiric specifying the significance level “*, **, ***” at 10, 5, and 1 percent respectively. in addition the values with parenthesis are coefficient value and values in parenthesis are probability values of t-statistic. source: authors' own estimation. muhammad usman, gulnaz hameed, abdul saboor and khuram nawaz sadozai 34 the error correction model (ecm(-1)) estimated through mohammad hashem pesaran & pesaran, (1997) suggests that cumulative sum of recursive residuals cusum and cusum square tests to evaluate the parametric consistency. the estimated result of diagnostic tests is given in table 5 which shows that there is no issue of heteroskedasticity and serial correlation in the given data set. furthermore, the graph of cusum test is confirmation of existence of long run relationship among variables and estimated coefficients are structurally stability over the long run (appendix a and b). table 5: diagnostic tests results diagnostic test f-statistic serial correlation 0.229679 (0.9105) heteroskedasticity 0.570524 (0.8746 source: author’s estimation 3.3 technology spillover impact on environmental performance in manufacturing sector, each firm sign an agreement with government for higher r&d spending and degree of net carbon emission production domestically. technology spillover produces negative or positive externalities but social optimum instruments tell the efficient use of r&d technology and suggest the economic impact on the society. however, the co2 emission can be controlled effectively through r&d spillover (technological innovation) tradable permits, and taxes on carbon emission. the internal or external climate agreements relevant to r&d spillover are not synchronized according effectiveness of social cost and social benefits from climate policies. furthermore, the government tax agreement on emission control is highly dependent on technology abetment cost and level of innovation. the results of lag selection criterion are given in table 6 which permit 4 lag to estimate the ardl bound test. table 6: var based selected lag length criterion lag logl lr fpe aic sc hq 0 -66.94312 0 9.51e-08 3.697156 3.992710 3.804019 1 187.2325 406.6810 3.46e-12 -6.561626 -4.197194* -5.706722 2 239.9014 65.83612 3.63e-12 -6.745070 -2.311762 -5.142126 3 293.4926 48.23204 5.75e-12 -6.974628 -0.472442 -4.623643 4 415.6222 67.17127* 8.51e-13* -10.63111* -2.060045 -7.532082* source: authors' own estimations ardl bounds test the results of ardl bound test are given in table 7 which show that estimated value of f-stat is higher than upper bound and in that case null hypothesis of no long run relationship is significantly rejected. in case of technology spillover effect on co2 emission the rejection of null journal of applied economics and business studies, volume. 4, issue 2 (2020) 23-40 https://doi.org/10.34260/jaebs.422 35 hypothesis highlights that there is long run cointegration relationship between technology imports and co2 emission in case of pakistan. the ardl estimates shows technology spillover has long term effect on environmental performance. the empirical results are fairly matching with the findings of (w. liu et al., 2016; z. liu et al., 2015). table 7: estimates of ardl bound test null hypothesis: no long-run relationships exist test statistic value k f-statistic 6.229124 6 critical value bounds significance i0 bound i1 bound 10% 1.75 2.87 5% 2.04 3.24 2.5% 2.32 3.59 1% 2.66 4.05 source: author’s estimation the table 8 shows the estimated results of long run and short run dynamics technology spillovers and its impact on net co2 emission in case of pakistan. the estimated result shows that r&d spillovers have significantly negative impact on net co2 emission. furthermore, the estimates illustrated that imported technology is helpful for net carbon emission reduction in case of pakistan. the results show that improvement in technology and technological innovation has fundamental role in co2 reduction. technological spillovers and technological advancement in pakistan is favourable for co2 emission abetment over the long run. the long run coefficients of technology imports (techt), trade openness (tot) and foreign direct investment inflow (fdit) have significantly negative. the estimates are consistent with the findings of (alemdar & özyildirim 2002; jebli, & youssef 2017; wang, et al., 2018) the technological innovation is helpful in emission reduction and innovative technology is fundamental source to produce less emission pollution. the technology import (techt) has less elasticity as compared to fdit and tot. in short run both foreign direct investment (fdi) trade openness (to) and labor force (lf) have no significant impact on co2 emission while technology imports (tech) has positive significant impact. in addition, the expenditure on education and return on education (hc) also have significantly negative impact on co2 production from manufacturing sector. the result shows that higher the technology imports (r&d spillover) in short run, higher the fraction of co2 emission (transboundary pollution) in case of pakistan. the estimated results are justified as per findings of jebli & youssef, 2017; verma et al., 2012 that technological spillover perform positive role in pollution emission. muhammad usman, gulnaz hameed, abdul saboor and khuram nawaz sadozai 36 table 8: long run and short run dynamic of technology spillover and co2 emission long run coefficients of adrl bound test short run coefficients of ardl bound test variable coefficient variable coefficient lfdi -0.161805 (0.0002)*** d(lfdi) 0.004347 (0.6552) lhc .976285 (0.0011)*** d(lhc) -2.204762 (0.0239)** lto -0.165560 (0.6507)*** d(lto) -0.084566 (0.2773) ltech -0.087630 (0.0452)*** d(ltech) 0.026757 (0.0067)*** lf 0.162251 (0.0118)** d(lf) 0.010012 (0.2208) hc*lfdi -0.340753 (0.0016)*** d(hc*lfdi) 0.002272 (0.8945) d(lco2(-1)) -0.385935 (0.0728) ecm(-1) -0.305342 (0.0101)*** the explained variable is lmanu. all variables under consideration are in logarithmic form because data of given variables are not available in same unit. the satiric indicates the significance level “*, **, ***” at 10, 5, and 1 percent respectively. in addition the values with parenthesis are coefficient value and values in parenthesis are probability values of t-statistic. source: author’s estimation. the results of heteroskedasticity and autocorrelation are given in table 9. the empirics show that there is no problem of heteroskedasticity and serial correlation in given data set. table 9: diagnostic results diagnostic test f-statistic serial correlation 1.880957 (0.1798) heteroskedasticity 0.520304 (0.9281) source: authors' own estimations 4. conclusion and policy recommendation the fundamental role of technology spillover in manufacturing sector is to improve the output performance through efficient and market competitive products with positive r&d spillovers. the crucial impact of technological spillovers is sustainability; minimized environmental damages through maximizing the human well-being. the study concluded that foreign r&d spillovers are journal of applied economics and business studies, volume. 4, issue 2 (2020) 23-40 https://doi.org/10.34260/jaebs.422 37 major channels of transmission through which technology can diffuse across the borders and regions. the empirical results show that the foreign knowledge spillover is essential component for manufacturing growth and it aspires to maximize the technological inflow in pakistan. furthermore, the important contribution of technology spillover is dependent on labor force absorptive ability. results show that labor force has less absorptive ability for efficient use of imported technology while technology spillover has positive impact on manufacturing growth with less absorptive ability. therefore, the firm can attain more and efficient production by improving the innovative research, encouragement to innovative thinking and new ideas, and also by improving the labor force skills. every newly developed technology gets obsolete over the time but innovation makes sense to replace with modern technology. the estimated result of second model shows that the cointegration relationship and technology spillover can decrease net co2 emission in pakistan's manufacturing sector. the fdi inflow, technology imports and technological transformation expand the manufacturing growth and reduction in corbin emission. on the basis of this result it is clinched that the technology spillover has direct as well as indirect consequences on manufacturing sector and co2 emission. directly it increases output, reduces energy consumption and co2 emission. while indirectly technology spillover enhances internal innovation and accelerates technology transformation, absorptive capacity and efficient utilization of new technology. it is recommended for manufacturers to voluntarily increase foreign technology and improve the labor force skills through research activities, research workshops, research seminars, and via encouraging the innovative ideas. in addition, government of pakistan needs to provide incentives to industrialists to upgrade the manufacturing plants, fdi inflow, technology imports and trade openness. government may use such measures to investigate the magnitude of r&d spillover and firm own efforts of innovation. the study suggests that the manufacturing sector should improve the technology spillover from foreign resources for co2 emission reduction. government and firm policy preferences for sustainable output growth and clean environment are highly dependent on green technology spillover. so, government may opt for demoting the carbon taxes to allow clean technology to overhaul the sordid technology. international knowledge spillover creates both negative as well as positive transboundary pollution, but cross country specific policies targeted to attain efficient technology to reduce negative externality increase the long run social welfare. references alemdar, n. m., & özyildirim, s. (2002). knowledge spillovers, transboundary pollution, and growth. oxford economic papers, 54(4), 597-616. bank, w. (2018). world development indicators database. obtenido de gross domestic product: databank. worldbank. org. bittencourt, p. f., & giglio, r. (2013). an empirical analysis of technology absorption capacity of the brazilian industry. cepal review. muhammad usman, gulnaz hameed, abdul saboor and khuram nawaz sadozai 38 coe, d. t., & helpman, e. (1995). international r&d spillovers. european economic review, 39(5), 859887. coe, d. t., helpman, e., & hoffmaister, a. w. (2009). international r&d spillovers and institutions. european economic review, 53(7), 723-741. costantini, v., mazzanti, m., & montini, a. (2013). environmental performance, innovation and spillovers. evidence from a regional namea. ecological economics, 89, 101-114. dalgıç, b., & mıhç, s. (2013). technology spillovers through imports and absorptive capacity in turkey. international journal of humanitie, 3(17), 298-309. di maria, c., & smulders, s. (2017). a paler shade of green: environmental policy under induced technical change. european economic review, 99, 151-169. dupuy, a. (2006). hicks neutral technical change revisited: ces production function and information of general order. the be journal of macroeconomics, 6(2). empora, n. (2017). air pollution spillovers and us state productivity growth. retrieved from fischer, c. (2008). emissions pricing, spillovers, and public investment in environmentally friendly technologies. energy economics, 30(2), 487-502. galdeano-gómez, e., cespedes-lorente, j., & martínez-del-río, j. (2008). environmental performance and spillover effects on productivity: evidence from horticultural firms. journal of environmental management, 88, 1552-1561. doi:10.1016/j.jenvman.2007.07.028 gorkey-aydinoglu, s. international diffusion of technology in the manufacturing industry: emerging countries within the eu and turkey. gujarati, d. n., & dawn, c. (2010). porter.(2009). basic econometrics: mcgraw-hill education. hanley, n. (2001). cost—benefit analysis and environmental policymaking. environment and planning c: government and policy, 19(1), 103-118. hanley, n., shogren, j. f., & white, b. (2016). environmental economics: in theory and practice: macmillan international higher education. heal, g., & tarui, n. (2010). investment and emission control under technology and pollution externalities. resource and energy economics, 32(1), 1-14. jebli, m. b., & youssef, s. b. (2017). the role of renewable energy and agriculture in reducing co2 emissions: evidence for north africa countries. ecological indicators, 74, 295-301. jiao, j., yang, y., & bai, y. (2018). the impact of inter-industry r&d technology spillover on carbon emission in china. natural hazards, 91(3), 913-929. kazmi, s. t. h. (2016). china's journey in renewable energy and its potential spillover effects through the cpec in pakistan. the pakistan development review, 619-631. kinda, t. (2012). foreign ownership, sales to multinationals and firm efficiency: the case of brazil, morocco, pakistan, south africa and vietnam. applied economics letters, 19(6), 551-555. kuo, f. y., & young, m. l. (2008). a study of the intention–action gap in knowledge sharing practices. journal of the american society for information science and technology, 59(8), 1224-1237. lee, j. w. (2013). the contribution of foreign direct investment to clean energy use, carbon emissions and economic growth. energy policy, 55, 483-489. lei, d., shangguan, z.-p., & rui, l. (2012). effects of the grain-for-green program on soil erosion in china. international journal of sediment research, 27(1), 120-127. journal of applied economics and business studies, volume. 4, issue 2 (2020) 23-40 https://doi.org/10.34260/jaebs.422 39 liu, w., xu, x., yang, z., zhao, j., & xing, j. (2016). impacts of fdi renewable energy technology spillover on china’s energy industry performance. sustainability, 8(9), 846. liu, z., guan, d., wei, w., davis, s. j., ciais, p., bai, j., . . . marland, g. (2015). reduced carbon emission estimates from fossil fuel combustion and cement production in china. nature, 524(7565), 335338. mayer, j. (2000). globalization, technology transfer and skill accumulation in low-income countries. globalization, marginalization and development. mehmood, t., chaudhry, m., tufail, m., & irfan, n. (2009). heavy metal pollution from phosphate rock used for the production of fertilizer in pakistan. microchemical journal, 91(1), 94-99. mingyong, l., shuijun, p., & qun, b. (2006). technology spillovers, absorptive capacity and economic growth. china economic review, 17(3), 300-320. mohammadi, a., veismoradi, a., hashemi, h. h., akbari, p., & rostami, r. (2015). the estimation of the share of technological spillover in sectorial-regional growth (case study: 29 production sectors of 30 iran provinces). asian economic and financial review, 5(3), 371. mustafa, z. (2011). climate change and its impact with special focus in pakistan. paper presented at the pakistan engineering congress, symposium. pwt, (2019) pann world table 9.1 https://www.rug.nl/ggdc/productivity/pwt/?lang=en pavitt, k. (1984). sectoral patterns of technical change: towards a taxonomy and a theory. technology, management and systems of innovation, 15-45. pesaran, m. h., & pesaran, b. (1997). working with microfit 4.0: interactive econometric analysis;[windows version]: oxford university press. pesaran, m. h., shin, y., & smith, r. j. (1996). testing for the'existence of a long-run relationship'. retrieved from romer, p. m. (1990). endogenous technological change. journal of political economy, 98(5, part 2), s71-s102. seck, a. (2012). international technology diffusion and economic growth: explaining the spillover benefits to developing countries. structural change and economic dynamics, 23(4), 437-451. serfraz, a. (2017). analyzing short-run and long-run causality between fdi flows, labour productivity and education in pakistan. retrieved from shittu, a. i. (2012). financial intermediation and economic growth in nigeria. british journal of arts and social sciences, 4(2), 164-179. stern, d. i. (2004). the rise and fall of the environmental kuznets curve. world development, 32(8), 1419-1439. torras, m., & boyce, j. k. (1998). income, inequality, and pollution: a reassessment of the environmental kuznets curve. ecological economics, 25(2), 147-160. türker, m. v. (2012). a model proposal oriented to measure technological innovation capabilities of business firms–a research on automotive industry. procedia-social and behavioral sciences, 41, 147-159. vayá, e., lópez-bazo, e., moreno, r., & suriñach, j. (2004). growth and externalities across economies: an empirical analysis using spatial econometrics advances in spatial econometrics (pp. 433-455): springer. verma, a. k., dash, r. r., & bhunia, p. (2012). a review on chemical coagulation/flocculation technologies for removal of colour from textile wastewaters. journal of environmental management, 93(1), 154-168. muhammad usman, gulnaz hameed, abdul saboor and khuram nawaz sadozai 40 wang, s. (2015). research on technology spillover effects on agricultural productivity in china. wang, x., fang, h., zhang, f., & fang, s. (2018). the spatial analysis of regional innovation performance and industry-university-research institution collaborative innovation—an empirical study of chinese provincial data. sustainability, 10(4), 1243. youssef, s. b. (2009). transboundary pollution, r&d spillovers and international trade. the annals of regional science, 43(1), 235-250. zarrabi, f., & vahedi, m. (2012). alignment between technology strategy and leadership. procedia-social and behavioral sciences, 41, 23-28. appendix -.04 -.02 .00 .02 .04 3.8 4.0 4.2 4.4 4.6 1980 1985 1990 1995 2000 2005 2010 2015 residual actual fitted cumulative sum (cusum) -.04 -.03 -.02 -.01 .00 .01 .02 .03 .04 1980 1985 1990 1995 2000 2005 2010 2015 lmanuf residuals cumulative sum of square journal of applied economics and business studies, volume. 5, issue 1 (2021) 47-66 https://doi.org/10.34260/jaebs.513 47 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x a complete picture of spatial disparity in cost of living index: a case study of pakistan’s cities farrukh mahmood1* shumaila hashim2 hina fatima3 1 assistant professor at muhammad ali jinnah university – karachi. 2 scholar, university of karachi, karachi, pakistan 3 assistant porfessor, muhammad ali jinnah university, karachi. abstract this study constructed the cost of living index by using all available data on 488 commodities of the 40 cities of pakistan for the month of may 2019. empirically, results revealed that there is a statistical difference in the cost of living index among cities from the standard of living. based on the national average prices, the islamabad is ranked at first, and mirpurkhas, a city of sindh, is at fortieth. furthermore, province wise highest cost of living is found in nwfp and lowest in sindh. by employing national average prices that have aggregation bias; therefore, it is replaced by province-level prices; the ranking among cities within the province is changed. at province average prices, the highest cost of living index is found in rawalpindi, karachi, abbottabad, and loralai, and the lowest cost of living in gujranwala, mirpurkhas, peshawar and turbat, for the province of punjab, sindh, nwfp, and baluchistan, respectively. this spatial disparity in the cost of living is mainly due to specific factors of production in a specific city as compare to other; quetta is known as “fruit garden in pakistan,” and khuzdar is an agriculture-based city. similarly, karachi and lahore have (i) high per capita income, and (ii) over-population are the factors of the high cost of living. hence, in the light of the present study, it is suggested there is no single rule through which disparity in the cost of living can be overcome. preferably the solution is laying at the micro-level, i.e., the disparity in the cost of living is mainly due to disparity in prices of same goods and services across cities, therefore by controlling prices of goods and services across the cities will suppress this disparity. keywords cost of living, disparity, aggregation bias, microlevel jel classification e21, r31, p22 * farrukh.mahmood@jinnah.edu.pk farrukh mahmood, shumaila hashim and hina fatima 48 1. introduction understating of spatial disparity in the cost of living (hereafter, col) is crucial to recognize inequality in terms of living standards, wage disparity, and amenities, etc. the col is a more appropriate and accurate measure of standard of living [jorgenson and slesnick, 2012; argente and lee, 2015]. this dimension of research, however, has been largely ignored in the literature due to the difficulty of calculating col at disaggregated data. literature has given much attention to the cpi to measure the level of inequality. it is an inappropriate measure of inequality, (i) aggregation bias, (ii) no substitution effect, (iii) easier to understand, but misinterpretation, (iv) market-led substitution, and including many other issues [boskin et al., 1998]. if the col varies across the cities, then the implication of any unique estimate or policy is ineffective, mainly when cpi is used for the construction of policy. several studies are available on this topic, which confirms that the spatial disparity in the col is present (chaudhry and chaudry, 1974; deller et al., 1997; kurre, 2000; feenstra et al., 2017). voicu and lahr (1999) constructed the col index for metropolitan areas through expenditure data. the variation in expenditures can be taken to estimate prices when data on price is not available. the concept behind it is that the variation in expenditure reflects both quantity and prices according to the fundamental law of demand. it is well established now that prices vary across spatial differences (johnston et al., 1996, walden, 1997, deller et al., 1997, handbury, 2013, albouy et al., 2016), and across rural and urban areas as well (mcmahon, 1991, kurre, 2000, timmins, 2006). col index for rural pennsylvania’s cities consistently lower as compared to the urban cities, and this disparity decrease over time [kurre, 2000]. practically, every study reveals that there exist a significant difference between col across cities, which is due to population density (cebula, 1980, langsten et al., 1985, haworth and rasmussen, 1993), climate change (ostrosky, 1983, haworth and rasmussen, 1993), age-wise preferences (cebula, 1998), housing price (hederson, 1999), family and child well-being [chien and mistry, 2013]. per capita income, rightto-work laws, temperature, and negative externalities are the adverse functions of col [cebula and toma, 2008]. the col is lower in more substantial cities due to higher competition among firms [handbury and weinstein 2015, feenstra, et al., 2017]. pasha and pasha (2002) calculated the col index by using 274 commodities of 25 cities in pakistan for different categories.1 they found that the col of karachi is the highest and lowest for jhang. similarly, ahmad and gulzar (2008) constructed the col index by employing 133 commodities of 32 cities for three categories.2 this study concluded that col in turbat is the highest and lowest in karachi. furthermore, province wise col is highest in baluchistan and lowest in sindh. 1 food and beverages, apparel and footwear, fuel and lighting, rent, others. 2 food and beverages, wheat and non-food journal of applied economics and business studies, volume. 5, issue 1 (2021) 47-66 https://doi.org/10.34260/jaebs.513 49 pasha and pasha (2002) and ahmad and gulzar (2008) explored the col for the cities of pakistan. these studies have the following issues; first, these studies used some part of the whole data set on prices for a few commodities and categories. hence, the inference will be biased and misleading, based on the bias data set. furthermore, they used national average prices for the construction of col; these prices have aggregation bias if a good is produced in karachi; like fish at the shore of the sea will cheaper, and will be expensive at the mountain of mingora due to transportation cost and expected margin of profit. the average prices of these two cities will represent neither mingora nor karachi. this aggregation bias also affects the measure of col and hence, the ranking of cities. instead, the more appropriate measure lies at the disaggregated level. lastly, there is no research available on this issue after the study of ahmad and gulzar (2008). in these years, many structural reforms have occurred; therefore, it is time to reestimate the col index. all these issues are analyzed in the study by using all 488 commodities of 40 cities in pakistan. this disaggregated data will present the complete picture of col across cities of pakistan. the study has the following objectives to overcome the above issues; the first objective is to construct and analyze the disparity in the col index across the cities and provinces of pakistan. the second objective is to examine the aggregation bias across national and province-level prices. in the case of pakistan, wage policies are often implemented across provinces on the bases of political interest, which implicitly increases the disparity in the col across different regions. therefore, it is essential to estimate the col because it not only gives information about the disparity in the col across cities but also highlights the ranking across cities. furthermore, this information ultimately is used by the policymaker for the construction of adequate wage policy. in the current situation of pakistan, its importance to construct a col index, where different groups of individuals are demanding higher wages. this study helps the government of pakistan to analyze the disparity in the col index across different cities and also in wage adjustments. the rest of the paper is organized as follows. section 2 describes the data and methodology. section 3 explains the empirical results along with the discussion and comparison with previous studies. the final section concludes the paper. 2. data and methodology 2.1 the data: for the construction of the col index, monthly prices of 40 cities, along with all 488 commodities is used. this data set is collected for the month of may 2019 from the latest available monthly bulletin of statistics, of the pakistan bureau of statistics. out of these 40 cities, nineteen are from punjab; eight are from sindh, six are from nwfp, six are from baluchistan, and islamabad. the data of province-wise share in household consumption by category is used as a proxy of quantities3, which has been taken from 3 for detail see pasha and pasha, 2002. farrukh mahmood, shumaila hashim and hina fatima 50 the latest available household integrated economic survey, of the pakistan bureau of statistics 2015-2016. 2.2 methodology: local consumer prices are required to construct spatial disparity in the col index. it provides information on precise and actual levels of money expenditures and consumption across spatial disparity, respectively. the spatial disparity col index for any region 𝑗 may be written as follows (thomas, 1980): 𝐶𝑜𝐿𝑗 = ∑ ∑ 𝑄𝑗𝑃𝑖𝑗 𝑚 𝑗=1 𝑛 𝑖=1 ∑ ∑ 𝑄𝑗𝑃𝑘 𝑛 𝑘 𝑚 𝑗=1 . 100 … (1) where 𝑄 and 𝑃 are quantities and prices, respectively, for the commodity 𝑖 in region 𝑗, and 𝑘 represents the cross-sectional average price for the 𝑖𝑡ℎ commodity. eq. 1, can be modified in terms of relative prices using expenditure weights instead of quantity weights is given below: 𝐶𝑜𝐿𝑗 = ∑ ∑ ∑ 𝑄𝑗𝑃𝑘 ( 𝑃𝑖𝑗 𝑃𝑘 )𝑛𝑘 𝑚 𝑗=1 𝑛 𝑖=1 ∑ ∑ 𝑄𝑗𝑃𝑘 𝑛 𝑘 𝑚 𝑗=1 . 100 … (2) where, 𝑛 = 488 (commodities) with 𝑗 = 40 cities, therefore eq. 2 can be rewritten as follows: 𝐶𝑜𝐿𝑗 = ∑ ∑ ∑ 𝑄𝑗𝑃𝑘 ( 𝑃𝑖𝑗 𝑃𝑘 )488𝑘 40 𝑗=1 488 𝑖=1 ∑ ∑ 𝑄𝑗𝑃𝑘 488 𝑘 40 𝑗=1 . 100 … (3) eq. 3 represents the col index for the region 𝑗.4 it is used to construct the col index for each category separately based on household integrated income and consumption survey.5 similarly, this procedure is repeated for all other cities. moreover, to analyze the comparison among provinces, this study repeats the procedure by changing the average cross-section price of all cities, with cross-sectional average prices of the province. through this procedure, it can be analyzed that the variation in the col is present among and within provinces. the anova test is applied to analyze statistical variation in col among provinces. furthermore, to examine the variation in col among cities, the one-sample t-test is applied. 3. empirical results and discussion national level average prices: appendix 1 presents the col index for 40 cities with 12 categories at national level average prices. in the case of islamabad, the value of the col index for most of the categories is higher than the 100, which implies that 4 the value of calculated col for the region 𝑗 will be positive. if the calculate value is 100, then it implies the price of region 𝑗 are equal to the national average prices. this point can also be called as standard of living. however, if the value is greater than the 100 then it implies that the prices in region 𝑗 are higher than the national average prices and hence high cost of living, vice versa. 5 there are 12 categories in household integrated economic survey, i.e., food & non alcoholic beverages, alcoholic beverages and tobacco, clothing and foot wear, housing, water, electricity, gas and other fuels, furnishing, household equipment and maintenance, health, transport, communication, recreation & culture, education, restaurants and hotels, and miscellaneous goods and services journal of applied economics and business studies, volume. 5, issue 1 (2021) 47-66 https://doi.org/10.34260/jaebs.513 51 the prices of commodities in islamabad are high as compare to the national level average prices. mainly there are two significant reasons behind; (i) highest per capita income among all other cities (ii) low production units and high demand for luxury living at islamabad put positive pressure on prices. moreover, to satisfy the needs of citizens, goods are imported from other cites, which contain two types of additional prices, i.e., transportation cost and expectation of high profit by the supplier. moreover, in the case of mithi and mirpurkhas, the value of the col is approximately consistently low as compare to all other cities, more specifically in sindh. the col in abbottabad for alcoholic beverages and tobacco is significantly high, as compared to others. khuzdar and turbat are expensive for the communication category. loralai is valuable for restaurants and hotels category. similarly, there is variation in prices within and across provinces for the same category.6 figure 1 presents the col ranking for all 40 cities.7 islamabad is ranked at first, and mripurkhas is ranked, at last, i.e., col is highest in the case of islamabad and lowest for mirpurkhas. as islamabad has high col for most of the categories (for detail, see appendix 1). the col for the category of food, housing, health, education, and recreation & culture is highest as compared to all other cities. which accumulating the over-all col in islamabad and hence have the highest col. it is mainly due to the high per capita income and luxury lifestyle of living in islamabad. and in the case of mirpurkhas, the col for most of the categories is lowest. in mirpurkhas, the average monthly is about 11000 rupees, which significantly less than the minimum wage in pakistan.8 moreover, the overall value of the col is also less than 100, which implies that it is below the standard level of living. in this ranking, abbottabad is ranked second. rawalpindi is at third. lahore is ranked at fourth and karachi is ranked at twenty-one. the over-all picture implies that there is a difference in col among cities. to test this difference statistically among cities, the one-sample t-test is applied. statistically, there is an insignificant difference among all cities (p-value – 0.9000), while it is statistically different from the standard of living (p-value >0.0000).9 it can be visualized as well in figure 1. the anova test is applied to test the statistical difference among provinces. results of the anova test are presented in appendix 2-a. these results imply that the mean of all provinces is approximately the same, i.e., there is no statistical difference among provinces. appendix 3 lists the category-wise over-all ranking of the col index with the province. based on the national level average price islamabad is the most expensive city to live, while sindh is the cheapest to live, reason remains the same behind this ranking. nwfp is ranked second after islamabad. baluchistan is at third, and punjab is at fourth. 6 these results are not discussed in detail because of aggregation bias. 7 this is constructed by taking the average of all the categories for each city. 8 pakistan emergency situational analysis, district mirpurkhas, october 2014. 9 one-sample t-test is applied two times; first, by comparing the sample mean, i.e., 111 with the data series, second, by comparing with the value of 100, which implies the standard of living. farrukh mahmood, shumaila hashim and hina fatima 52 figure 1: over-all cost of the living index with ranking by cities. (national level average prices) islamabad abbotabad rawalpindi lahore mardan peshawar bahawalpur bahawalnagar jhang faisalabad jehlum mianwali sahiwal attock turbat gawadar larkana d.m. jamali khuzdar hyderabad karachi r.y. khan mingora quetta dadu sukkar nawabshah multan d.g. khan sialkot loralai vehari bannu sargodha wazirabad muzafarghar mithi gujranwala d.i. khan mirpurkhas 90 100 110 120 130 140 150 160 journal of applied economics and business studies, volume. 5, issue 1 (2021) 47-66 https://doi.org/10.34260/jaebs.513 53 province level average prices: appendix 4 listed the col index for 40 cities with 12 categories at province level average prices.10 the first category is food and nonalcoholic beverages, in which the col for rawalpindi city is highest, while the lowest for the d.g. khan. notably, there is no significant production unit in rawalpindi; therefore, to fulfill the need for the citizen, goods are imported from other cities. rawalpindi and islamabad are twin cities, and the standard of living in both cities is not only high, but also highly correlated, and have similar consumption behavior patterns. because of this behavior, the col in rawalpindi for most of the categories is high, i.e., health, and recreation and culture. while d.g. khan is the city of punjab where not only is wheat produced, but it is also used as a raw material in many products; bread, rusk, biskets, somasa, nimko chips, among others. this city is also rich in production livestock, dairy products, almonds, raisin, lal beans, onion, tomato, tomato ketchup, carliforwer, bottle ground, pickels, vinegar, garlic, tapal danedar, mosambi, malata, kinu, banana, guava, kharbooza, gram whole, tinda, card, lemon, sugar, gur, honey, milo, and many more. the second category in appendix 4 is listed as alcoholics, beverages, and tobacco. for this category, faislabad has the lowest col, while mainwali has the highest. there are six commodities in this category, where the betel nuts have the highest weight, and it has minimum weight in the case of faislabad and maximum weight for mainwali, according to the demand of these cities, respectively. figure 2 presents the province-wise col for cities. in the case of punjab, the col for rawalpindi is the highest and lowest for gujranwala. the difference between rawalpindi and gujranwala is about 24 percent. the fourth category is housing, water, electricity, gas, and other fuels, the col is highest for lahore and lowest for wazirabad. it is mainly due to the over-population at lahore. thus, an increase in population needs more housing and related material; therefore, it puts upward pressure on col. it is also true for transport and education category in lahore. similarly, for karachi in sindh province. furthermore, the col in sialkot is about 101.55, just close to 100, and it implies that the most favorable city for a living is sialkot regarding this category. so, the col for wazirabad indicates that it is not a desirable city to live in a living standard. in sindh, the col for karachi is highest and lowest for the mirpurkhas, and the difference is about 17 percent. in the case of karachi, the high col is due to the high population, i.e., the population of karachi is about 8 percent of the total population of pakistan, and about 34 percent of sindh.11 this high level of the population puts upward pressure on prices like the housing category among others. 10 in the table islamabad is lost, i.e., now this study considering the average level prices within province, and islamabad is not included in province. or, even we can add it into the table but all values will be equal to 100, which give no information. 11 population census 2017. farrukh mahmood, shumaila hashim and hina fatima 54 in kpk, the highest and lowest col is for abbottabad and d.i. khan, respectively. the difference between abbottabad and peshawar is about 21 percent. abbottabad is the main junction for the tourism places, like; ayubia national park, bara gali, shimla hill, dor river valley (at harnoi). khaira gali, thandiani and nathia gali. abbottabad also links skardu, hunza, pak-china border, etc. tourism generates a high level of income and employment multipliers per unit of visitor spend (slee et al., 1997), and the likelihood of spending is higher due to the luxury effect. contrary to abbottabad, the col of d.i. khan is the lowest in kpk. this is mainly due to the lowest price level for the most categories, like; alcoholic beverages & tobacco, clothing & footwear, recreation & culture, restaurant and miscellaneous goods & services. in baluchistan, the col for turbat is highest and lowest for the loralai, and the difference is about 17 percent. in the case of turbat, the prices are higher for the category of education, recreation & culture, furnishing and miscellaneous goods & services, as compare to all other cities in baluchistan. the col is quite high for most of the categories in abbottabad because it is a junction that connects most of the tourist areas; naran kaghan, skardu, nagan parbat, gilgit, and many others. in other words, it provides a platform for tourists to stay here and then move to other areas for enjoyment. due to this luxury effect, most of the commodities are expensive in this city. in addition to this, most of the population of abbottabad is belong to the transportation and hoteling sector as a profession for their living. similarly, for turbat in baluchistan. the comparison of the transportation category across provinces is on standard, i.e., around the value of 100, for the province of punjab, while there is considerable variation in transportation category can be seen in the case of sindh, nwfp, and baluchistan. it is suggested based on the results that transportation col is on its standard in punjab, other provinces should adopt this model to control transportation prices in their areas. figure 2 presents the col index for each city separately, and cities are ranking within the province, based on the province level average prices. anova test is applied to test the statistical difference among the province. results of the anova test are presented in appendix 2-b. these results imply that the average col in baluchistan is statistically different from all other provinces. in cases of punjab, the col in rawalpindi is highest among other cities, and the reason remains the same, that the behavior of individuals in rawalpindi is approximately the same as in islamabad as defined earlier, hence increase the col. furthermore, most of the well-established housing societies are in rawalpindi, where the luxury effect is dominated, which also increases the col in rawalpindi. the col for sialkot, multan, and vehari is around 100, which implies that these cities are most favorable to live in the province in punjab. wazirabad, sargodha, muzaffargarh, and gujranwala are the cities in which col is below the standard of living. journal of applied economics and business studies, volume. 5, issue 1 (2021) 47-66 https://doi.org/10.34260/jaebs.513 55 figure 2: over-all cost of the living index with ranking by cities. (province level prices) rawalpindi mianwali lahore bahawalpur bahawalnagar jhang jehlum sahiwal attock faisalabad r.y. khan d.g. khan sialkot vehari multan wazirabad sargodha muzafarghar gujranwala karachi larkana dadu hyderabad sukkar nawabshah mithi mirpurkhas abbotabad peshawar mardan mingora bannu di khan turbat quetta d.m. jamali gawadar khuzdar loralai 90 95 100 105 110 115 120 125 farrukh mahmood, shumaila hashim and hina fatima 56 the col for the karachi is highest and lowest for the mipurkhas, for the province of sindh. in the case of karachi, the col is high due to the high per capita income and demand by over-population in this city. while mipurkhas is below the standard of living, it can be easily visualized that the col in mirpurkhas and mithi is significantly different from other cities of sindh, i.e., they are deprived of necessary facilities. in the case of nwfp, abbottabad is ranked at first, and peshawar is last, and in baluchistan, turbat is ranked at first, and loralai is at last. aggregation bias: comparing figures 1 and 2 highlights the issue of aggregation bias. in figure 1, peshawar is ranked after mardan and abbottabad, while in figure 2, abbottabad is at rank, and peshawar is at last for the province of nwfp. similarly, for the province of baluchistan, at national level average prices, quetta is ranked fourth and gawadar at second (figure 1). when national level average prices replaced by provincelevel average prices, gawadar is at fourth, and quetta is at second. 3.1 comparison and discussion: in the literature, it is found that there are different determinants of the col in an economy. this inference is extracted based on the national price level, which has aggregation bias. moreover, before applying any statistical analysis, it is essential to know “what is data, and from where it comes” mahmood (2017). often it is the practice of researchers that they feed data into the system without analyzing data and treat the output of system persuasive; therefore, it could lead to misleading inferences. this study it is analyzed that the col varies across cities with no particular reason. for example, in jhang, the price of the visit to the doctor is high, in abbottabad's cost of transportation and housing is high. in the case of mirpurkhas, the average monthly income is quite low. in karachi, it is over-populated and has high per capita income, and hence high col, similarly in lahore. rawalpindi and islamabad have no significant unit of production; that is why col is high. hence, it can be inferred that there is no single solution through which disparity in col can be removed. pasha and pasha (2002) and ahmad and gulzar (2008) concluded that the population has a positive impact on the col, which is not valid. it is just an anscombe’s effect (for detail see, anscombe, 1973), as shown in figure 3. it can be easily visualized that there is no effect on the population on the col. for example, karachi has the highest population among all other cites but has low col. contrary, islamabad, rawalpindi, and abbottabad have a low population with high col. similarly, lahore has a population about 12 million and has col 121, while peshawar has about 2 million with col 120. the population of peshawar is just 17% of the population of lahore. ahmad and gulzar (2008) concluded that, as the distance from karachi increases, because karachi has a port, the col increases, which is also not true both logically and based on the results of visualization in figure 4. it is also anscombe’s effect and an example of spurious regression. for example, the col in karachi is 111, which has 0journal of applied economics and business studies, volume. 5, issue 1 (2021) 47-66 https://doi.org/10.34260/jaebs.513 57 kilometer distance; contrary, mingora has the most substantial distance from karachi and has less col than karachi, i.e., 108. similarly, mirpukhas has the same col as d.i. khan, but has a distance of 240 and 1067 kilometers from karachi, respectively, fortunately, their col is less than the col in karachi. moreover, gawadar has its working port and has approximately the same col as karachi, but the distance of gawadar from karachi is about 630 kilometers. hence, it can easily be inferred that distance from karachi does not have any impact on col. figure 3: impact of population on the col by cities. figure 4: impact of distance12 from karachi port on the col by cities. 12 the data on distance is collected through google.map, and the unit of the distance is kilometers. faisalabad gujranwala islamabad lahore rawalpindi karachi abbotabad peshawar 93.0 113.0 133.0 153.0 3 6 9 12 15 millions impact of population (city-size) on col islamabad karachi mirpurkhas abbotabad di khan mingora gawadar turbat 85 105 125 145 165 250 500 750 1,000 1,250 1,500 1,750 kilometers impact of distance from karachi port on col farrukh mahmood, shumaila hashim and hina fatima 58 4. conclusion the primary purpose of this study is to construct the col index for the cities of pakistan based on the latest available data set. this research is constructed in the col index for 40 cities with 488 commodities. these 488 commodities are divided into 12 categories. following conclusion can be drawn from the empirical analysis; first, the col for cities is significantly different from the level of standard living. second, the col is different among provinces, and baluchistan is statistically different from others. third, due to aggregation bias, the ranking of cities varies within the province. the difference in col among cities is due to specialization in their given resources like d.g. khan is specialized in food and non-alcoholic beverages, and sialkot is famous for its production in surgical and musical instruments, and sports, leather, and textile goods. while sialkot is the second-largest source of foreign exchange in pakistan. loralai is famous for the production of almond and apple. moreover, the disparity in the col across cites within province exists because the law of one price does not hold. it is also right across provinces. for example, in jhang, the price of a visit to the doctor is high, in abbottabad's cost of transportation and housing is high. in the case of mirpurkhas, the average monthly income is quite low. in karachi, it is over-populated and has a high level of per capita income, and hence high col, similarly in lahore. rawalpindi and islamabad have no significant unit of production that why col is high. hence, it can be inferred that there is no single solution through which disparity in col can be removed. similarly, other cities are specialized in the production of some other goods and services, i.e., few cities are rich in some goods and services, while other cities are rich in others. therefore, it is not possible to suggest a policy at a macro level; to overcome col disparity across cities and provinces; one should have to focus on micro-level issues. preferably the solution is laying at the microlevel, i.e., the disparity in col is mainly due to differences in prices of the same goods and services across cities, therefore, controlling prices of goods and services across the cities will suppress this disparity in col. journal of applied economics and business studies, volume. 5, issue 1 (2021) 47-66 https://doi.org/10.34260/jaebs.513 59 references ahmed, s., & gulzar, a. (2008). inter-city variation in prices. pakistan development review,winter , 877-892. albouy, d., ehrlich, g., & liu, y. (2016). housing demand, cost-of-living inequality, and the affordability crisis (no. w22816). national bureau of economic research. asra, abuzar (1999) urban-rural differences in costs of living and their impact on poverty measures. bulletin of indonesian economic studies 35:3, 51–69. boskin, m. j., dulberger, e. l., gordon, r. j., griliches, z., & jorgenson, d. w. (1998). consumer prices, the consumer price index, and the cost-of-living. journal of economic perspectives, 12(1), 3-26. cebula, r.j., 1980. ‘determinants of geographic cost-of-living differentials in the united states: an empirical note’, land economics, 56 (november), 476-8. chaudhry, m. g., & chaudry, m. a. (1974). cost-of-living indexes for rural labourers in pakistan. the pakistan development review, 13(1), 26-39. chien, n. c., & mistry, r. s. (2013). geographic variations in cost-of-living: associations with family and child well‐being. child development, 84(1), 209225. feenstra, r. c., xu, m., & antoniades, a. (2017). what is the price of tea in china? towards the relative cost of living in chinese and us cities. national bureau of economic research. working paper, no. 23161 handbury, j. (2013). are poor cities cheap for everyone? non-homotheticity and the cost-of-living across us cities. the wharton school research paper, (71), 1-054. handbury, j. and d.e. weinstein (2015). goods prices and availability in cities. the review of economics and statistics, 82(1), 258-296. haworth, c. t. and d. w. rasmussen (1973) determinants of metropolitan cost-ofliving variations. southern economic journal 40:2, 183–192. henderson, j. vernon, zmarak shalizi, and anthony j. vernables (2001) geography and development. journal of economic geography 1:1, 81–105. kurre, j. a. (2003 ). is the cost-of-living less in rural areas? international regional science review,vol 26, issue 1 . langston d., d.w. rasmussen and j.c. simmons, 1985. ‘a note on geographic costof-living differentials’, land economics, vol.61, no.3, 314-318. mahmood, f., (2017). model specification and data problems: a case study of market volatility and retail interest rate pass-through. mpra working paper no. 7934. mcmahon, w. w. (1991). geographical cost-of-living differences: an update. real estate economics, 19(3), 426-450. farrukh mahmood, shumaila hashim and hina fatima 60 ostrosky, a.l., 1983. ‘determinants of geographic cost-of-living differentials in the united states” comment’, land economics, vol.59, no.3, 350-354. pasha, h. a., & aisha, g. p. (2002). cost-of-living index by city of pakistan. social policy and development center, pakistan. research report 43. ravallion, m. and dewalle, d., 1988, ‘urban-rural cost-of-living differentials in a developing economy’, journal of urban economics, 29, 113-127. slee, b., farr, h., & snowdon, p. (1997). the economic impact of alternative types of rural tourism. journal of agricultural economics, 48(1‐3), 179-192. thomas, v., 1980. ‘spatial differences in the cost-of-living’, journal of urban economics, 8, 108-122. timmins, c. (2006). estimating spatial differences in the brazilian cost-of-living with household location choices. journal of development economics, 80(1), 59-83. walden, m. l. (1997). how much income variation" really" exists within a state?. review of regional studies, 27(3), 237-250. appendix 1. col index with categories for each city (national level average prices) cont. city food and non alcoholic beverages alcoholic beverages and tobacco clothing and foot wear housing, water, electricity, gas and other fuels furnishing, household equipment, and maintenance health islamabad 119 108 141 575 117 167 p u n ja b attock 105 88 127 115 110 102 bahawalnagar 102 78 127 102 133 128 bahawalpur 102 90 144 114 122 106 d.g. khan 96 109 82 100 94 103 faisalabad 109 72 107 108 106 133 gujranwala 99 80 98 101 109 96 jehlum 110 81 121 121 106 109 jhang 97 86 151 98 116 135 lahore 109 80 113 173 131 124 mianwali 102 202 95 98 103 115 multan 103 70 108 134 105 99 muzafarghar 101 70 103 123 104 85 rawalpindi 113 101 132 173 114 156 r.y. khan 100 91 112 98 108 115 sahiwal 103 77 116 127 116 109 sargodha 106 84 105 110 103 98 sialkot 106 81 105 108 99 107 vehari 104 83 113 102 111 94 wazirabad 98 82 99 94 104 94 s in d h dadu 101 120 102 103 100 107 hyderabad 105 110 107 93 95 106 karachi 112 89 127 132 98 118 larkana 110 174 92 91 102 104 mirpurkhas 97 119 83 91 90 94 mithi 97 126 98 90 91 97 nawabshah 98 125 111 87 95 117 journal of applied economics and business studies, volume. 5, issue 1 (2021) 47-66 https://doi.org/10.34260/jaebs.513 61 sukkar 103 86 104 96 108 170 n w f p abbotabad 108 328 132 114 120 121 bannu 97 111 111 88 100 92 d.i. khan 105 76 99 92 111 113 mardan 99 243 134 97 107 108 mingora 98 103 135 100 96 116 peshawar 101 194 137 118 109 110 b a lu ch is ta n d.m. jamali 102 222 92 101 98 103 gawadar 106 164 96 115 110 168 khuzdar 110 163 101 95 103 111 loralai 101 103 93 102 83 100 quetta 105 125 98 135 103 112 turbat 106 164 93 106 118 102 farrukh mahmood, shumaila hashim and hina fatima 62 appendix 1. col index with categories for each city (national level average prices) city transpo rt communica tion recreation & culture educati on restaurants and hotels miscellaneous goods and services islamabad 100 96 141 157 126 138 p u n ja b attock 99 94 141 112 127 134 bahawalna gar 99 115 131 109 146 113 bahawalpur 99 162 133 116 100 142 d.g. khan 101 149 97 120 91 93 faisalabad 100 75 126 146 129 158 gujranwala 98 101 83 99 99 101 jehlum 99 103 140 119 136 124 jhang 101 110 119 108 133 118 lahore 102 90 114 160 103 157 mianwali 97 107 112 123 102 107 multan 99 118 118 98 86 99 muzafargha r 99 118 116 87 79 99 rawalpindi 99 94 141 151 127 136 r.y. khan 97 104 128 112 125 135 sahiwal 99 98 115 115 154 131 sargodha 99 84 98 106 108 100 sialkot 99 88 102 113 119 107 vehari 99 134 92 101 96 99 wazirabad 99 134 91 101 96 99 s in d h dadu 168 97 104 103 82 98 hyderabad 103 88 98 217 109 101 karachi 106 99 118 113 109 107 larkana 106 114 99 124 124 103 mirpurkhas 106 76 99 118 79 98 mithi 96 88 102 105 88 100 nawabshah 96 105 114 134 81 108 sukkar 96 95 67 138 119 99 n w f p abbotabad 107 142 104 110 102 101 bannu 107 106 113 114 87 100 d.i. khan 107 89 73 121 70 98 mardan 102 115 110 119 87 121 mingora 102 91 110 113 113 121 peshawar 102 112 123 122 97 112 b a lu ch is ta n d.m. jamali 102 118 98 96 110 101 gawadar 101 88 75 70 156 102 khuzdar 101 166 85 58 148 102 loralai 101 117 89 57 184 98 quetta 102 95 101 106 102 100 turbat 103 135 98 114 106 106 journal of applied economics and business studies, volume. 5, issue 1 (2021) 47-66 https://doi.org/10.34260/jaebs.513 63 appendix 2-a: results of anova test. (national level average prices) multiple comparisons province mean difference std. error sig. 95% confidence interval lower bound upper bound punjab sindh 3.53 3.63 0.77 -6.26 13.31 nwfp -3.47 4.03 0.83 -14.35 7.4 baluchistan -0.31 4.03 1 -11.18 10.57 sindh punjab -3.53 3.63 0.77 -13.31 6.26 nwfp -7 4.65 0.45 -19.54 5.54 baluchistan -3.83 4.65 0.84 -16.38 8.71 nwfp punjab 3.47 4.03 0.83 -7.4 14.35 sindh 7 4.65 0.45 -5.54 19.54 baluchistan 3.17 4.97 0.92 -10.24 16.58 baluchistan punjab 0.31 4.03 1 -10.57 11.18 sindh 3.83 4.65 0.84 -8.71 16.38 nwfp -3.17 4.97 0.92 -16.58 10.24 appendix 2-b: results of anova test. (province level average prices) multiple comparisons province mean difference (i-j) std. error sig. 95% confidence interval lower bound upper bound punjab sindh 1.53 3.31 0.65 -5.20 8.25 nwfp 2.03 3.68 0.58 -5.44 9.51 baluchistan -4.96 3.68 0.19 -12.43 2.52 sindh punjab -1.53 3.31 0.65 -8.25 5.20 nwfp 0.51 4.25 0.91 -8.11 9.13 baluchistan -6.48 4.25 0.14 -15.10 2.14 nwfp punjab -2.03 3.68 0.58 -9.51 5.44 sindh -0.51 4.25 0.91 -9.13 8.11 baluchistan -6.99 4.54 0.13 -16.20 2.23 baluchistan punjab 4.96 3.68 0.19 -2.52 12.43 sindh 6.48 4.25 0.14 -2.14 15.10 nwfp 6.99 4.54 0.13 -2.23 16.20 farrukh mahmood, shumaila hashim and hina fatima 64 appendix 3. the province with category-wise over-all col index, province-wise col, and over-all ranking. category-wise col index over-all col and ranking f o o d a n d n o n a lc o h o li c b e v e ra g e s a lc o h o li c b e v e ra g e s, t o b a c c o c lo th in g a n d f o o t w e a r h o u si n g , w a te r, e le c tr ic it y , g a s a n d o th e r f u e ls f u rn is h in g , h o u se h o ld e q u ip m e n t, a n d m a in te n a n c e h e a lt h t ra n sp o rt c o m m u n ic a ti o n r e c re a ti o n a n d c u lt u re e d u c a ti o n r e st a u ra n ts a n d h o te ls m is c e ll a n e o u s g o o d s a n d s e rv ic e s p ro v in c e -w is e c o l p ro v in c e -w is e r a n k in g is la m a b a d 119.26 107.73 141.27 575.27 117.25 166.94 100.26 95.75 140.53 156.52 125.95 138.49 165.44 1 p u n ja b 103.50 89.64 113.43 115.80 110.09 111.03 99.20 109.26 115.71 115.65 113.36 118.53 109.60 4 s in d h 102.97 118.74 102.86 97.72 97.31 113.98 109.74 95.21 100.05 131.51 99.00 101.61 105.89 5 n w f p 101.39 175.76 124.80 101.54 107.26 109.88 104.62 109.21 105.54 116.47 92.56 108.69 113.14 2 b a lu c h is ta n 105.09 157.01 95.36 109.16 102.61 115.91 101.81 119.58 91.05 83.64 134.39 101.67 109.77 3 journal of applied economics and business studies, volume. 5, issue 1 (2021) 47-66 https://doi.org/10.34260/jaebs.513 65 appendix 4. col index with categories for each city (province level average prices) cont. city food and non alcoholic beverages alcoholic beverages and tobacco clothing and foot wear housing, water, electricity, gas and other fuels furnishing, household equipment, and maintenance health p u n ja b attock 105.33 106.21 119.22 110.10 104.96 98.48 bahawalnag ar 102.38 91.26 121.80 97.14 128.22 129.65 bahawalpur 101.76 112.96 133.93 106.97 116.82 101.97 d.g. khan 96.13 146.31 79.52 94.89 90.87 103.48 faisalabad 107.55 82.72 102.12 101.90 101.29 127.89 gujranwala 97.88 96.13 93.79 96.69 105.03 92.14 jehlum 109.61 96.57 114.75 114.58 101.45 105.27 jhang 96.92 104.84 142.35 94.30 110.89 135.31 lahore 108.90 96.13 106.60 159.94 124.30 116.50 mianwali 101.85 281.50 91.04 94.49 98.98 113.78 multan 101.98 83.06 101.55 126.81 100.10 96.57 muzafargha r 100.52 83.06 97.47 117.10 99.70 83.43 rawalpindi 113.36 126.93 123.74 158.27 109.39 150.54 r.y. khan 99.67 115.72 106.15 94.29 102.80 112.71 sahiwal 102.05 91.59 109.70 119.35 111.15 103.77 sargodha 104.68 100.85 99.86 104.47 98.24 94.75 sialkot 104.90 98.93 99.94 101.90 95.25 104.76 vehari 102.65 100.43 106.63 97.36 106.24 90.90 wazirabad 97.74 100.00 96.02 90.97 100.24 90.90 s in d h dadu 101.87 101.94 103.78 118.26 106.25 104.02 hyderabad 105.13 93.81 108.22 102.38 100.28 100.68 karachi 111.48 78.13 130.91 153.94 104.60 111.65 larkana 108.77 143.56 96.08 99.66 109.98 100.03 mirpurkhas 95.84 100.84 84.04 100.08 94.14 88.24 mithi 96.08 106.95 99.94 98.02 96.18 91.10 nawabshah 98.10 105.51 114.31 94.61 101.09 109.64 sukkar 101.92 77.04 107.08 108.69 116.15 158.96 n w f p abbotabad 109.65 204.76 112.48 121.20 118.18 115.29 bannu 97.85 78.53 96.80 91.43 98.82 88.90 d.i. khan 107.62 61.83 86.77 95.18 110.86 114.11 mardan 99.70 155.66 109.70 100.02 106.04 100.87 mingora 98.17 72.89 112.62 102.17 93.86 105.34 peshawar 101.68 125.93 112.99 127.37 107.46 103.02 b a lu c h is ta n d.m. jamali 101.89 134.40 107.23 102.29 102.75 97.64 gawadar 104.48 99.55 105.24 115.28 112.92 168.42 khuzdar 109.02 99.98 112.45 96.69 104.75 105.52 loralai 100.00 70.75 108.27 102.54 86.39 96.13 quetta 106.23 79.59 108.15 132.15 109.75 108.32 turbat 104.40 99.55 102.83 105.82 129.95 103.65 farrukh mahmood, shumaila hashim and hina fatima 66 appendix 4. col index with categories for each city (province level average prices) city transport communication recrea-tion and culture educati on restaurants and hotels miscellaneous goods and services p u n ja b attock 101.06 90.11 131.71 103.24 121.49 116.13 bahawalna gar 100.83 110.86 122.64 100.12 142.56 105.75 bahawalpu r 100.69 156.38 125.72 109.20 95.37 119.89 d.g. khan 102.50 162.37 92.49 116.26 88.21 90.31 faisalabad 101.63 74.07 120.95 135.05 122.44 129.72 gujranwala 100.21 101.25 79.28 89.01 94.61 98.25 jehlum 101.06 100.54 131.08 108.49 127.96 108.71 jhang 102.54 108.40 112.31 99.26 127.38 105.84 lahore 103.73 88.45 106.54 147.74 98.56 129.28 mianwali 98.79 105.20 105.31 112.94 99.50 100.65 multan 100.46 114.21 111.27 90.33 83.11 95.87 muzafargh ar 100.46 112.14 109.98 79.37 75.33 95.84 rawalpindi 101.09 91.08 132.36 141.22 120.81 118.64 r.y. khan 99.11 102.32 120.35 102.60 118.60 117.79 sahiwal 100.38 97.14 110.83 106.50 145.32 115.12 sargodha 100.86 81.05 91.07 97.18 103.78 96.91 sialkot 100.89 84.91 94.75 104.59 113.92 102.82 vehari 100.85 129.88 85.98 95.81 92.99 96.76 wazirabad 100.85 129.88 85.30 95.81 92.99 96.76 s in d h dadu 160.80 105.92 106.14 121.56 87.81 98.34 hyderabad 98.81 96.08 99.53 169.64 `112.81 101.34 karachi 102.28 108.81 120.37 118.02 115.12 107.85 larkana 101.77 125.86 99.78 111.78 127.95 103.45 mirpurkhas 101.78 82.73 100.29 104.03 82.71 98.58 mithi 93.18 96.10 102.15 97.46 92.27 100.55 nawabsha h 93.10 114.71 113.99 118.65 86.15 109.02 sukkar 93.16 104.94 69.19 118.04 124.86 98.98 n w f p abbotabad 102.25 126.81 106.42 102.93 109.95 100.11 bannu 102.24 98.01 110.31 116.74 94.44 97.20 d.i. khan 102.21 81.96 74.12 123.99 75.83 97.00 mardan 97.91 106.58 109.87 109.84 93.83 115.51 mingora 97.89 84.10 102.09 104.58 123.73 114.68 peshawar 97.93 103.86 117.10 107.71 104.62 107.13 b a lu c h is ta n d.m. jamali 100.05 116.27 110.51 151.49 93.77 103.34 gawadar 99.99 79.79 85.74 109.40 129.32 105.96 khuzdar 100.02 137.67 97.35 100.31 125.17 104.96 loralai 100.00 104.61 102.11 101.07 160.23 99.40 quetta 100.03 94.45 119.15 276.78 88.07 102.78 turbat 102.15 126.58 118.92 284.26 93.20 109.87 journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.521 1 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x tobacco farming and illness induced poverty in pakistan muhammad shahzad1, anwar shah2, frank joseph chaloupka3 1 ph.d. scholar, school of economics, quaid i azam university, islamabad 2 associate professor of economics, school of economics, quaid i azam university islamabad 3 director health policy research center, university of illinois at chicago, usa abstract tobacco has been projected for its economic prosperity in pakistan and many other developing countries while the opportunity cost of domestic labor, health issues and associated health cost related with tobacco farming are often overlooked. various health conditions associated with tobacco farming result in catastrophic health expenditures which not only increase the chances of poverty head counts but also deepens it further. taking into account the opportunity cost of domestic labor and health cost associated with tobacco induced illnesses obscure the tobacco prosperity rhetoric. this study examined effect of incremental health cost associated with tobacco farming on poverty head counts and severity. using survey data from the tobacco producing districts in khyber pakhtunkhwa, pakistan, this study found that tobacco farmers experienced higher proportion of sever health hazards and illnesses like cvd and respiratory issues. the severe nature of ailments caused them higher share of out of pocket expenditures as compared to non-tobacco farmers and general population. increase in health expenditures not only increased their poverty head counts by four and half percent but also severity of poverty worsened further by 8 percentage points using the cost of basic needs (cbn) approach for measuring poverty. this study concludes that it is not only tobacco consumption associated with various health conditions but also tobacco farming. it is recommended that health cost associated with tobacco farming be considered both in setting up of minimum indicative prices for tobacco and consideration of health cost can be used as a tool against prosperity rhetoric which is used to block tobacco control policy. keywords tobacco, health cost, illness, poverty, policy jel classification h51, i30 muhammad shahzad, anwar shah, frank joseph chaloupka 2 1. introduction tobacco industry has been promoting tobacco as a panacea for the economic distress of developing countries (campaign for tobacco free kids, 2001). country after country in developing world has been succumbing to this projection and viewed tobacco as panacea for alleviating poverty, unemployment and balance of payment problem (tomson, et. al., 2009; & barraclough & morrow, 2010). this tendency has led to cultivation of tobacco on 10.5 million acre of land in almost 124 countries globally, mostly in the lower middle income countries (lmics) like china, the leading producer followed by brazil and india (faostat, 2013). tobacco employs thousands of farmers and agricultural workers on country basis while total employment reaches to over 33 million labor in growing and initial processing of tobacco while including other tobacco related activities the employment magnitude rises to over 100 million. (ilo, 2014). many of asian countries are hard target of the tobacco industry to increase their profits due to favorable climatic conditions and cheap labor (van minh, et. al., 2009). tobacco occupies an important position in terms of profitability, revenue and employment in pakistan also. tobacco was introduced in the indian sub-continent by the portuguese in the 16th century but its regular cultivation did not begin until the first decades of the 17th century (bhatti, 1992). production of tobacco in pakistan was first tried in sindh, then in north-west plains of punjab and finally in parts of the then north west frontier province (nwfp) now called khyber pakhtunkhwa. tobacco history in khyber pakhtunkhwa can be traced back to 1912 when tobacco was cultivated for experimental purposes at the agricultural research station tarnab in peshawar (muhammad, 1975). due to its economic significance, pakistan tobacco board (ptb) was established in 1968 with the purpose of promoting cultivation, manufacturing of cigarettes, exports of tobacco and tobacco made products, marketing of tobacco products and fixation of tobacco prices along with other related information. to facilitate the growers of tobacco ptb has been carrying out field experiments at their research stations to impart the latest knowledge regarding tobacco technologies for better yields and income. before 1968 tobacco produced in the country had poor quality and quality tobacco was mostly imported. establishment of ptb enabled the country to improve tobacco quality and become self-sufficient in tobacco production. the efforts of ptb have also enabled the country to achieve tobacco yield at par with the developed countries (ali, et al., 2015). the economic significance of tobacco crop in pakistan economy can be inferred from the workforce of 350,000, it employs, which generates 300 billion rupees of revenue and livelihoods for 1.2 million people annually. there are almost 75,000 tobacco growers producing about 80 to 85 million kilograms of fluecured virginia (fcv) tobacco each year. the production of fcv is mostly concentrated in the province of khyber pakhtunkhwa province in pakistan. the province has a population of 45000 tobacco farmers cultivating 25,500 hectares of land and producing 75 million kg of tobacco mostly used in cigarettes. tobacco contributed journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.521 3 nearly 89 billion rupees to national treasury in the form of various taxes in 2013-14 (ptb, 2018). governments of tobacco producing countries see tobacco as source of tax revenues, employment creation and earning income for deprived segment of their population while overlooking health of labor and productivity loss due to premature mortality (who, 2004; & alderete, et. al., 2020). though, tobacco related revenue and employment holds ground however, its consumption has established relation with poor health conditions and is considered as a leading risk factor of six of the eight preventable causes of morbidity and pre-mature mortality (barraclough & morrow, 2010). moreover, perils of tobacco are not limited to the smokers only, but creates economic distress for the 8 million tobacco associated death victims’ families and 1436 billion usd cost to the global economy in form of health seeking cost and loss in productivity (anh, et al., 2016 & goodchild, et al., 2018). along with consumption related health catastrophe, tobacco cultivation also exposes farmers and workers to poor working conditions including exposure to chemicals, dermal absorption of nicotine, high temperature in tobacco barns, and tobacco dust during the process of tobacco curing (ballard et. al., 1995 &arcury et. al., 2003). moreover, tobacco workers are prone to dermal and respiratory absorption of poisonous substances causing health damages like poisoning, skin & eye irritations, respiratory, kidney problems, and neuropsychiatric issues due to pesticide application for plant protection purposes (cox, 1992; cox, 1995, lonsway et. al., 1997; salvi et. al., 2003 & ngajilo, et. al., 2018). these health issues not only raise their upfront treatment cost, but also deteriorate long term health outcomes of farming communities. in addition, seeking treatments for various health conditions causes a drain on the financial resources of those who could hardly manage above the poverty line. in addition to human and economic costs, tobacco diverts useful resources from food and other essential needs along with inverse relation between income and tobacco consumption, which exacerbates poverty situation in poor countries (who, 2011; parera et al., 2017). many studies have maintained that tobacco is associated with vicious circle of poverty through different channels like poor health, loss in productivity, environmental hazards, diversion of resources from essentials of food security and child labor (who, 2004; kaying, et. al., 2005, adeioetomo, et. al., 2005, & who, 2008). the current study is aimed to quantify difference in health cost of tobacco and nontobacco farmers and associated impact on deepening of poverty in both groups. this study hypothesized that tobacco farmers are at higher risk of developing poor health conditions and incur higher expenditures on seeking treatments. using survey data from major tobacco producing districts of khyber pakhtunkhwa, pakistan results of this study indicate that health costs produce more adverse effects on tobacco farmers compared to non-tobacco farmers. tobacco farming poverty head counts and poverty depth registered higher magnitudes due to host of health issues. rest of the work follows theoretical framework in section 2, data and methodology in section 3 while results and muhammad shahzad, anwar shah, frank joseph chaloupka 4 discussion are presented in section 4 and 5 respectively. section 6 concludes the paper with conclusion and recommendations. 2. theoretical framework catastrophic illness though affects relatively small proportion of population, yet it accounts for a substantial share of out of pocket expenditures of poor households. expenditure for medical care becomes catastrophic when it endangers the standard of living of a family as family is a basic spending unit and consumption of each member is interdependent. health expenditures divert resources from other basic chores of life necessary for a better standard of living. working conditions have strong implication for health and the effects of poor working conditions on health are accumulative. poor health and poverty have a two-way relationship. poverty increases the risk of illness and illness in turn increases the chances to fall below the poverty line by forgoing the earnings and out of pocket expenditures (majra & gur, 2009).these health expenditures cause drain on farmers’ resources and decline their ability of acquiring other goods and services of urgent needs. 3. data and methodology this study is based on cross sectional data collected from three major tobacco producing districts of swabi, mardan and charsadda in khyber pakhtunkhwa, province of pakistan. taking into account the financial and time constraints, we selected a sample of 330 farming households. we used 318 farmers’ data for analysis as 12 farmers reported incomplete data. among 318 farmers 201 were tobacco growers whereas, rest of 117 did not grow tobacco crop and termed as non-tobacco group. we used stratified multistage sampling technique for the selection of sample from the respective districts. in the first stage we selected khyber pakhtunkhwa province on the basis of higher share of tobacco production. in the subsequent stage we selected three districts including swabi, mardan, and charsadda purposively on the basis of their share in total production. in third stage we selected one tehsil each in the charsadda and mardani.e tangi and takhtbhai respectively, while two tehsils chota lahore and razzar in the swabi district on the basis of their share in total production. among the districts sample was distributed proportionately on the basis of tobacco share in total production in each district. the districts of charsadda and mardan share 15 and 25 percent in total tobacco production (nasrullah, et. al., 2019). the sub samples collected from the swabi, mardan and charsadda districts had 190, 90, and 50 farmers, respectively. data were solicited by using a well-structured and pre-tested questionnaire and conducting face to face personal interview with the farmers at their farms/ residences. after collection of data from field we transferred the data into excel sheet and used stata 12 for analysis of the data. we carried out descriptive as well as econometric analysis to test our hypothesis. journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.521 5 3.1occupational hazards we termed occupational hazard as condition experienced by farmers while performing farming chores and required medical intervention to get relief. we classified the hazards into four groups as sun stroke, chemical exposure, cut/injury and allergy/skin rashes. we added the number of times farmers experienced these conditions and calculated the average score for all these conditions in both groups of farmers using t-test analysis. ℎ𝑎𝑧𝑖𝑗 = ∑𝑓ℎ𝑎𝑧𝑖𝑗 ∑ℎℎ𝑗 …………. (1) where ℎ𝑎𝑧𝑖𝑗 𝑠𝑡𝑎𝑛𝑑𝑠 𝑓𝑜𝑟 𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑠𝑐𝑜𝑟𝑒 𝑜𝑓 ℎ𝑎𝑧𝑎𝑟𝑑𝑖 𝑖𝑛 𝑔𝑟𝑜𝑢𝑝𝑗 𝑓𝑎𝑟𝑚𝑒𝑟𝑠 ∑𝑓ℎ𝑎𝑧𝑖𝑗 𝑆𝑡𝑎𝑛𝑑𝑠 𝑓𝑜𝑟 𝑠𝑢𝑚𝑚𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 ℎ𝑎𝑧𝑖 𝑖𝑛 𝑔𝑟𝑜𝑢𝑝𝑗 𝑓𝑎𝑟𝑚𝑒𝑟𝑠 ∑ℎℎ𝑗 𝑠𝑡𝑎𝑛𝑑𝑠 𝑓𝑜𝑟 𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 ℎ𝑜𝑢𝑠𝑒ℎ𝑜𝑙𝑑𝑠 𝑖𝑛 𝑔𝑟𝑜𝑢𝑝𝑗 3.2 health conditions of households we measured the health condition of households by the inverse of disease prevalence. we examined the households for diseases like high blood pressure, diabetes, heart diseases, chest & respiratory issues and cancer. we aggregated the number of cases for these diseases and compared the mean scores of all individual cases in both group of farmers. 𝐷𝑖𝑠𝑒𝑎𝑠𝑒𝑖𝑗 = ∑𝑓𝐷𝑖𝑠𝑒𝑎𝑠𝑒𝑖𝑗 ∑ℎℎ𝑗 ……… (2) where 𝐷𝑖𝑠𝑒𝑎𝑠𝑒𝑖𝑗 𝑠𝑡𝑎𝑛𝑑𝑠 𝑓𝑜𝑟 𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑠𝑐𝑜𝑟𝑒 𝑜𝑓 𝐷𝑖𝑠𝑒𝑎𝑠𝑒𝑖 𝑖𝑛 𝑔𝑟𝑜𝑢𝑝𝑗 , 𝑖 𝑓𝑜𝑟 ℎ𝑖𝑔ℎ 𝑏𝑙𝑜𝑜𝑑 𝑝𝑟𝑒𝑠𝑠𝑢𝑟𝑒, 𝑑𝑖𝑎𝑏𝑒𝑡𝑠, 𝐻𝑒𝑎𝑟𝑡 𝑖𝑠𝑠𝑢𝑒𝑠, 𝐶ℎ𝑒𝑠𝑡 𝑎𝑛𝑑 𝑟𝑒𝑠𝑝𝑖𝑟𝑎𝑡𝑜𝑟𝑦 𝑖𝑠𝑠𝑢𝑒𝑠 𝑎𝑛𝑑 𝑐𝑎𝑛𝑐𝑒𝑟 𝑗 𝑠𝑡𝑎𝑛𝑑𝑠 𝑓𝑜𝑟 𝑓𝑎𝑟𝑚𝑒𝑟𝑠 𝑔𝑟𝑜𝑢𝑝 𝑖. 𝑒 𝑡𝑜𝑏𝑎𝑐𝑐𝑜 𝑓𝑎𝑟𝑚𝑒𝑟𝑠 𝑎𝑛𝑑 𝑛𝑜𝑛 − 𝑡𝑜𝑏𝑎𝑐𝑐𝑜 𝑓𝑎𝑟𝑚𝑒𝑟𝑠 ∑𝑓𝐷𝑖𝑠𝑒𝑎𝑠𝑒𝑖𝑗 𝑆𝑡𝑎𝑛𝑑𝑠 𝑓𝑜𝑟 𝑠𝑢𝑚𝑚𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 𝑐𝑎𝑠𝑒𝑠 𝑜𝑓 𝐷𝑖𝑠𝑒𝑎𝑠𝑒𝑠𝑖 𝑖𝑛 𝑔𝑟𝑜𝑢𝑝𝑗 ∑ℎℎ𝑗 𝑆𝑡𝑎𝑛𝑑𝑠 𝑓𝑜𝑟 𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 ℎ𝑜𝑢𝑠𝑒ℎ𝑜𝑙𝑑𝑠 𝑖𝑛 𝑔𝑟𝑜𝑢𝑝𝑗 3.3 health cost we estimated health cost of farmers by aggregating their expenditure on seeking medical interventions. we aggregated the expenditure incurred on transportation used to medical facility, doctor fee and expenditure on medicine and other medical equipment’s and procedures on the annual basis. we classified health costs into three categories i.e. expenditure on treatment of occupational hazards experienced during farming chores, minor and frequently occurring issues like flu, fever, cuts, allergy and major long term health issues including blood pressure, diabetes, cardiovascular diseases, chest & respiratory issues and cancers of various types. estimation of health cost is given below muhammad shahzad, anwar shah, frank joseph chaloupka 6 ℎ𝑎𝑧𝐶𝑗 = ∑𝐶ℎ𝑎𝑧𝑖𝑗……….. (4) 𝑆𝑅𝐷𝑖𝑠𝑒𝑎𝑠𝑒𝑠𝐶𝑗 = ∑𝐶𝑆𝑅𝐷𝑖𝑠𝑒𝑎𝑠𝑒𝑖𝑗 …….(5) 𝐿𝑅𝐷𝑖𝑠𝑒𝑎𝑠𝑒𝐶𝑗 = ∑𝐶𝐿𝑅𝐷𝑖𝑠𝑒𝑎𝑠𝑒𝑖𝑗……. (6) 𝑇𝐻𝐶𝑗 = ℎ𝑎𝑧𝐶𝑗 + 𝑆𝑅𝐷𝑖𝑠𝑒𝑎𝑠𝑒𝑠𝐶𝑗 + 𝐿𝑅𝐷𝑖𝑠𝑒𝑎𝑠𝑒𝐶𝑗 ………. (7) where ℎ𝑎𝑧𝐶𝑗 𝑠𝑡𝑎𝑛𝑑𝑠 𝑓𝑜𝑟 𝑡𝑜𝑡𝑎𝑙 ℎ𝑎𝑧𝑎𝑟𝑑𝑠 𝑐𝑜𝑠𝑡 𝑖𝑛𝑐𝑢𝑟𝑟𝑒𝑑 𝑜𝑛 𝑡𝑟𝑒𝑎𝑡𝑚𝑒𝑛𝑡 𝑜𝑓 𝑎𝑙𝑙 ℎ𝑎𝑧𝑎𝑟𝑑𝑠 𝑜𝑓 ℎ𝑜𝑢𝑠𝑒ℎ𝑜𝑙𝑑𝑠 𝑆𝑅𝐷𝑖𝑠𝑒𝑎𝑠𝑒𝑠𝐶𝑗 𝑆𝑡𝑎𝑛𝑑𝑠 𝑓𝑜𝑟 𝑡𝑜𝑡𝑎𝑙 𝑐𝑜𝑠𝑡 𝑖𝑛𝑐𝑢𝑟𝑟𝑒𝑑 𝑜𝑛 𝑡𝑟𝑒𝑎𝑡𝑚𝑒𝑛𝑡 𝑜𝑓 𝑚𝑖𝑛𝑜𝑟 ℎ𝑒𝑎𝑙𝑡ℎ 𝑖𝑠𝑠𝑢𝑒𝑠 𝑜𝑓 ℎ𝑜𝑢𝑠𝑒ℎ𝑜𝑙𝑑 𝐿𝑅𝐷𝑖𝑠𝑒𝑎𝑠𝑒𝐶𝑗 𝑠𝑡𝑎𝑛𝑑𝑠 𝑓𝑜𝑟 𝑒𝑥𝑝𝑒𝑛𝑑𝑖𝑡𝑢𝑟𝑒𝑠 𝑖𝑛𝑐𝑢𝑟𝑟𝑒𝑑 𝑜𝑛 𝑡𝑟𝑒𝑎𝑡𝑚𝑒𝑛𝑡 𝑜𝑓 𝑚𝑎𝑗𝑜𝑟 ℎ𝑒𝑎𝑙𝑡ℎ 𝑖𝑠𝑠𝑢𝑒𝑠 𝑜𝑓 ℎ𝑜𝑢𝑠𝑒ℎ𝑜𝑙𝑑 𝑇𝐻𝐶𝑗 𝑠𝑡𝑎𝑛𝑑𝑠 𝑓𝑜𝑟 𝑡𝑜𝑡𝑎𝑙 ℎ𝑒𝑎𝑙𝑡ℎ 𝑐𝑜𝑠𝑡 𝑜𝑓 ℎ𝑜𝑢𝑠𝑒ℎ𝑜𝑙𝑑 we also compared the means of household’s health costs incurred on the abovementioned conditions of both the farming groups using t-test analysis. 3.4 estimation of poverty to estimate the poverty different measures have been in vogue like arbitrary benchmark, food energy intake (fei) and cost of basic needs (cbn) in different time periods. some pioneering works on poverty used arbitrary benchmark approach using a fixed income or expenditure required for household’s basic needs fulfillment. however, by mid 1970s the focus of researchers shifted to food energy intake to calculate food calorie intake as measure of poverty. with the passage of time embedding non-food essential needs into caloric intake provided a more comprehensive measure of poverty used by different researchers like malik (1988) jaffri and khattak (1995) quraishi and arif (2001) and bashir & idrees (2018) in the context of pakistan. we took cost of basic needs approach and determined the threshold level of income 43,822 rupees per capita, which can cover the cost of basic food and non-food items necessary. we found the head count poverty in pre health expenditure and post health expenditure for both groups of farmers to see the effect of health cost on poverty ratio. we also estimated deviation from poverty line in pre and post health expenditure scenarios for both groups to show the severity condition in both groups. we compared the deviations in both groups using t-test analysis. 4. results 4.1 descriptive analysis socio-economic descriptive statistics show that households’ heads fall in the same age category irrespective of farmer type. however, poor households’ heads were older. regarding family size this study data show that tobacco farmers have larger families in general and in poor farming group in particular. likewise, households with less numbers journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.521 7 of earning members are more likely to be poor whereas, difference in number of earners for overall and poor farmers is 0.31 and 0.60, respectively. similarly, tobacco farmers have large farm sizes in both categories, however poor farmers have small farms as compared to overall category. likewise, poor households are living in congested houses where family size to room ratio is higher reflecting the typical developing countries characteristics (todaro & smith, 2014). moreover, farmers have non-significant difference in per capita income. contrary to the difference in per capita income, tobacco farmers incur more health expenditures and spend higher share of their income on health issues than non-tobacco farmers. mean values and their differences for various socioeconomic characteristics are given in table 1. table 1 socio-economic characteristics of farmers variable farming groups over all poor farmers tobacco nontobacco diff sig tob nontob diff sig age (years) 42.31 42.73 -0.48 0.29<0.70 46.12 47.94 1.82* .64 >.52 education (years) 8.56 7.50 1.06* 2.72>0.007 7.88 8.07 0.19 .36<.71 household size (no.) 13.54 12.40 1.14* 1.19>0.24 14.04 13.60 0.44 .28<.78 earners (no.) 3.19 2.88 0.31* 1.59>0.11 3.20 2.6 0.60* 1.65>.10 farm size (jerib) 13.75 10.02 3.73* 3.40>0.001 10.95 8.01 2.94* 2.26>.03 no of rooms 4.50 4.32 1.18* 0.58>0.56 3.83 3.48 0.35 .90> .37 per capita income (rs.) 104747 100323 4421 .31<.75 33978 27041 6937* 2.54>.01 per capita health expenditure (rs.) 6383 4296 2087* 1.80>.07 6985 3109 3876* 1.18>.24 source: field survey, 2018-19 4.2 hazards exposure farming is a hazardous occupation on overall basis, but some crops pose more exposure to various types of hazards. this study’ survey data show that tobacco farmers experienced significantly higher exposure to sunstroke as compared to non-tobacco farmers. similarly, pesticides and cut/injury adverse effects and exposure are more than twice as those of non-tobacco farmers, while allergy difference was found nonsignificant. findings of this study illustrate that tobacco farmers are more likely to develop poor health outcomes. hazards experience and differences are presented in table 2. muhammad shahzad, anwar shah, frank joseph chaloupka 8 table 2 hazards scores of tobacco and non-tobacco farmers hazard type average score by farmer type difference significance tobacco non-tobacco sun stroke 1.29 (0.16) 0.08 (0.03) 1.21* (0.21) 5.84 >0.00 pesticides affect 0.25 (0.07) 0.10 (0.04) 0.15* (0.09) 1.63 >0.04 cuts/injury 0.07 (0.03) 0.03 (0.02) 0.04* (0.04) 1.75 >0.45 allergy 0.37 (0.16) 0.36 (0.15) 0.01 (0.24) 0.05 < 0.96 source: field survey, 2018-19 4.3 hazards’ intensity perception farmers reported different poor health conditions and ranked their experience. average rank scores for the mentioned conditions were significantly high in tobacco farmers. tobacco farmers reported almost 50 percent more fatigue, heart burn (salivation) and perspiration as compared to non-tobacco farmers, whereas average scores for vomiting and poor appetite was more than twice. however, experiencing of skin rashes was ranked same in both groups of farmers. average rank scores and their differences are presented in table 3. table 3hazards ranking scores hazard type farmer type difference significance tobacco non-tobacco fatigue 3.58 (0.07) 2.33 (0.09) 1.25* (0.11) 10.91>0.00 vomiting 2.78 (0.08) 1.34 (0.11) 1.44* (0.14) 10.55>0.00 perspiration 3.73 (0.07) 2.34 (0.08) 1.39* (0.11) 12.33>0.00 heart burn (salivation) 2.75 (0.08) 1.49 (0.11) 1.26* (0.13) 9.54>0.00 chill 2.17 (0.08) 1.63 (0.14) 0.54* (0.14) 3.89>0.00 poor appetite 2.37 (0.07) 0.91 (0.09) 1.46* (0.12) 12.15 > 0.00 skin rashes 2.37 (0.08) 2.34 (0.10) 0.03 (0.13) 0.22<0.81 source: field survey, 2018-19 4.4 health condition of households tobacco farmers are more likely to develop sever health conditions in comparison with their non-tobacco counterparts. these conditions include high blood pressure, diabetes, cardio-vascular diseases (cvd), illness of chest and respiratory system, and cancers of various types. this study found that high blood pressure has almost same prevalence irrespective of farming group. however, diabetes, cvd, and respiratory journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.521 9 issues have higher prevalence in tobacco farmers. findings from this study’s survey show that almost every third of tobacco farming household has a diabetes and cvd patient, while these conditions prevail in every fifth of nontobacco farmers. similarly, respiratory issues also, have higher prevalence in the tobacco producing districts as every second of tobacco and every third of non-tobacco farming households reported the problem. average cases of these mentioned conditions and differences in means of the groups are presented in table 4. table 4 prevalence of various health conditions across the farming groups illness type farmer type difference significance tobacco non-tobacco high blood pressure 0.94 (0.07) 0.90 (0.10) 0.04 (0.12) 0.36<0.76 diabetes 0.33 (0.04) 0.21 (0.04) 0.12 * (0.06) 1.97>0.05 cvd 0.28 (0.04) 0.20 (0.05) 0.08* (0.06) 1.31>0.19 chest & respiratory 0.45 (0.06) 0.35 (0.07) 0.10* (0.09) 1.05>0.29 cancer 0. 03 (0.01) 0.02 (0.01) 0.01* (0.02) 0.66>0.50 source: field survey, 2018-19 4.5 health expenditure health expenditures depend on frequency and nature of ailments. tobacco farmers face more working hazards and incur more expenditure on treatment in general, but poor tobacco farming households are more prone to hazardous environment and incur more expenditure in particular. similarly, for all other categories of health care expenditure tobacco farmers spend significantly more money on seeking health. on overall basis, tobacco households spent 30,724 rupees more than non-tobacco farmers. however, the difference between the two groups but poor farmers was higher by 50 percent. table 5 presents the average values and differences of means for tobacco and non-tobacco overall and poor households. table 5 health expenditures of various types of health condition across the farming groups treatment nature group status farmer type difference significance tobacco non-tobacco hazards (rs.) overall 1,915 (367) 256 (85) 1,659* (490) 3.38>0.00 poor 2,518 (1,082) 248 (105) 2,270* (1,280) 1.77>0.07 minor diseases (rs.) overall 32,746 (3,132) 23,889 (2,257) 8,857* (4,484) 1.98>0.05 poor 29,880 (3,102) 23,083 (4,177) 6,797* (5,091) 1.33 >0.18 sever diseases (rs.) overall 36,173 (8,079) 15,966 (3,866) 20,207* (11,091) 1.82 > 0.07 poor 48,520 (26,126) 12,083 (3,673) 36,437* (30,497) 1.18 > 0.24 muhammad shahzad, anwar shah, frank joseph chaloupka 10 overall diseases(rs.) overall 70,834 (9,258) 40,110 (5,039) 30,724* (12,839) 2.39 > 0.02 poor 80,918 (26,900) 35,414 (5,267) 45,503* (32,067) 1.42 >0.16 source: field survey, 2018-19 4.6 poor health and poverty head counts the table below shows the effect of health severity on poverty head counts. though tobacco farmers have low head counts than non-tobacco farmers, but poor health conditions affect them adversely as compared to non-tobacco farmers. before incurring health expenditures,19.90 percent of the tobacco farmers were unable to have income sufficient to meet their basic needs whereas, the proportion for non-tobacco group was 29.06 percent. however, incurring health expenditures, the proportion of tobacco poor rose by 4.48 percent while that of non-tobacco farmers rose by 0.85 percent. without incurring health expenditures 23.89 percent of farmers fall below poverty line, while health cost increases incidence of falling below poverty line to 26.42 percent on overall basis. proportion of headcounts and the impact of health cost on poverty head-counts are presented in table 6. table 6 poverty head counts with and without health expenditures group poverty head counts pre health expenditure (%) post health expenditure (%) change (%) tobacco farmers 40 (19.90) 49 (24.38) 4.48 non-tobacco farmers 34 (29.06) 35 (29.91) 0.85 total 74 (23.89) 84 (26.42) 2.53 source: field survey, 2018-19 4.7 health expenditure share farmers face different health issues of short term and long-term nature. they use both public health facilities and private health care providers’ services. tobacco farmers needed more health care services due to associated conditions. they spent on average 8.56 percent of their total income on acquiring health care services, while non-tobacco farmers spent around 7 percent. both groups have significant difference of 1.51 in health spending which reflect the severity of health outcomes of tobacco farmer. however, health care expenditures affect tobacco and non-tobacco farmers differently and causing higher proportion of tobacco farmers to fall below the poverty line in terms of fulfilling their basic needs. among the poor farming groups tobacco farmers spent almost 3.90 percent higher on health-related conditions compared to non-tobacco farmers. shares and their differences for tobacco and non-tobacco farmers are presented in table 7. journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.521 11 table 7 share of health expenditures in farming groups scenario share of health expenditure tobacco farmer (se) non-tobacco (se) difference (se) significance overall farming groups 8.56 (1.03) 7.05 (0.08) 1.51* (1.48) 1.02>0.31 poor farmers groups 17.52 (4.74) 13.62 (2.14) 3.90* (4.67) 0.83>0.40 source: field survey, 2018-19 4.8 illness effect on shrinking of income of poor farmers as shown above tobacco farmers face more health adversities than non-tobacco farmers. they incurred more expenditures, which left them with less income to meet other basic needs. illness caused their income to further fall by over 20.56 and 11.50 percent, for tobacco and non-tobacco farmers, respectively. illness caused a shrink in income by 6,986 rupees on average basis for tobacco farmers while non-tobacco farmers face shrinkage of 3,109 rupee on per capita basis. overall, farmers experienced shrinkage of 5,371 rupees per capita due to poor health conditions and incurring health expenditure. average values of pre and post health expenditures and their differences are presented in table 8. table 8income shrinkage of tobacco and non-tobacco farmers farming group per capita income pre expenditure (rs.) post expenditure (rs.) shrinkage (rs.) (%) significance tobacco 33,978 (1,867) 26,992 (2,316) 6,986* (20.56) 1.85>0.69 non-tobacco 27,041 (1,888) 23,931 (1,889) 3,109* (11.50) 1.16>0.25 overall 31,088 (1,388) 25,717 (1,564) 5,371* (17.28) 2.57>0.01 source: field survey, 2018-19 4.9 poor health and severity of poverty poor health not only affects earning capability but also diverts resources from acquiring goods and services necessary for sustenance of life. tobacco farming is associated with frequent exposure to severe health conditions which causes rise in health expenditure. the incurring of health expenditures intensifies the severity of poverty with higher proportion in tobacco farmers as compared to non-tobacco farmers. in the pre health expenditure scenario tobacco farmers were short of poverty line by 31 percent,while non-tobacco farmers had nearly 40 percent deviation. however, poor health of tobacco farmers caused a further fall by 8.61 percentage points while nontobacco farmers experienced further deviation of over 5.68 percentage points. it indicates that tobacco farmers are hit adversely by poor health outcomes as compared to non-tobacco farmers. overall poor health conditions worsen the situation of poverty for all farmers but tobacco farmers face burden with higher proportions. details of muhammad shahzad, anwar shah, frank joseph chaloupka 12 deviations from poverty lines in pre and post health expenditure scenario are presented in table 9. table 9 health cost and deviations from poverty line group deviation from poverty line pre health expenditure (%) post health expenditure (%) illness effect (%) significance tobacco 31.30 (3.56) 39.91 (3.89) 8.61 (5.28) 1.62>0.10 non-tobacco 39.70(4.19) 45.40(4.32) 5.68 (6.01) 0.94>0.34 total 35.16(2.75) 42.87 (2.90) 7.71 (3.99) 1.93 >0.06 source: field survey, 2018-19 5. discussion historically farming has been an important source of livelihood and employment in developing countries (rao, et al., 2005). at the same time, it also exposes farmers and workers to risks including extreme weather, hazardous chemicals, allergies, cuts and injuries due to agricultural tools and machineries. however, some crops pose higher degree of exposure than others. the difference arises due to variation in use of chemicals, machinery operations, exposure to weather conditions, and physical contact with crop (cordes & foster, 1988 & kidane, et al., 2013). tobacco has been reported to pose higher risks to farmers’ and workers’ health due to host of reasons. tobacco crop involves more plant protection and processing, and exposure to hazards increases with each additional process. more chemicals in the form of pesticides/insecticides, suckericides, and other growth inhibitors are used for plant protection and achieving higher productivity (damalas, et al., 2006). however, the use of pesticides associated with poor handling techniques is associated with adverse health outcomes for farmers and workers especially in developing countries (peres et. al., 2006 & khan et. al.,2010). similarly, the frequency of physical contact with tobacco crop is high and workers are exposed to extreme environment. furthermore, the presence of nicotine enhances the chances of green tobacco sickness (gts) while other crops involve rare physical contact, minor exposure to extreme environmental factors, less use of pesticide and absence of nicotine (gosh et. al., 1986 & mcbride, 1998). difference in exposure and severity of health conditions causes difference in health expenditures. tobacco farmers face hazards of severe nature and had to seek proper medication in their farm vicinity whereas, the more severe cases had admission in hospital. they had to bear the cost of various diagnostic tests and medicine while non-tobacco crops’ hazards were mostly handled through self-medication and traditional home remedies. farmers view tobacco hazards more deadly than non-tobacco hazards as per their own and neighborhood past experience and therefore responded differently. difference in response and expenditures can find support in literature from indian gujrat and veitnam (parikh et. al., 2005 & von minh, et. al., 2009). tobacco farmers had higher proportions for heart problem, chest & respiratory issues and cancer. though these diseases have been reported to have high association journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.521 13 with tobacco consumption, but we witnessed no significant difference in tobacco consumption of these farming groups. however, tobacco growers and workers absorb nicotine from green tobacco leaves through their bodies’ pores from green tobacco while they face tobacco dust in curing and further processing stages. moreover, exposure to smoke of tobacco burning stalks used as fuel for energy increases the likelihood of heart and respiratory problems (mackay & eriksen, 2005;schenker et. al., 2005; & ngajilio, et. al., 2018). dusty environment in tobacco barns and non-use of protective masks increases risk of chest and respiratory problems (caze, et al., 2019).the dust particles are accumulated during curing process in the tiny sponge tissues in respiratory system especially when farmers avoid use of protective masks and duration of exposure is high (osim et. al., 1998; &arcury&quandt, 2006). higher prevalence of chest and respiratory problems confirms the hypothesis that the minute particles are absorbed and causing emphysema even in non-smoker tobacco workers (ghosh et. al., 1980). findings of our study reflect on the link of poor health and poverty. though tobacco has been projected as panacea for eradication of poverty in developing countries, but the illness associated with tobacco crops has been rarely considered. this study findings show that tobacco farmers are at higher risk of various hazardous situation and various health conditions. they spend a significantly higher share of their income on seeking health. the health cost of tobacco farmers increases likelihood of their net income after paying for health to fall short of the poverty line on one hand and severing the situation of already poor on other. out of pocket health expenditures increase poverty head count of tobacco farmers by 4.48 percent, while non-tobacco headcount experienced a rise of 0.85 percent. out of pocket health expenditures have been reported with increased head counts in many countries like india, china, and bangladesh (chowdhury, 2015; pryer, et. al., 2005 &sun, et. al., 2010). though large-scale tobacco farmers are able to earn livelihood above poverty level but small and poor farmers not only earn low income from tobacco but are more prone to health issues and spent higher share of their income on health (who, 2004). poor farmers in tobacco group spent almost 18 percent of their income on health-related issues, while on overall tobacco farmers spent 8 percent as compared to 7 percent of non-tobacco farmers. higher share of health expenditures support claim of the who and are in contrast with the no link of tobacco farming and poverty (pain, et. al., 2012). out of pocket health expenditure share for tobacco farmers is higher than the national average of 6.51 percent and higher share of out of pocket health expenditures of tobacco farmers as compared to non-tobacco farmers are in conformity with literature (kaying et al., 2005 &adioetomo, et. al., 2005). 6. conclusion and recommendations tobacco consumption has established relation with detrimental health conditions and illness induced poverty, but even then, has been projected as panacea for poverty alleviation of farming community. however, we found that tobacco farmers are at higher risk of developing poor health conditions as well as incur high proportion of out muhammad shahzad, anwar shah, frank joseph chaloupka 14 of pocket health expenditure. tobacco farmers have significantly higher proportion of cases for diabetes, cvd and respiratory issues and spent 80,918 rupees per household on seeking treatment for various hazardous situations and diseases of minor and sever nature. on the other hand non-tobacco farmers spent only 35,414 rupees per household on average. health expenditures increased the head count poverty by 4.48 and 0.85 percent, for tobacco and non-tobacco farming groups, respectively. furthermore, the increase in severity of poverty was also high in tobacco group. the study recommends provision of personal protective dressings for tobacco farmers and workers to reduce physical contact with hazardous chemicals and environment. ensuring effective plant protection chemicals to reduce the frequency of chemical application will not only decrease physical contact with tobacco plant but also adverse pesticide effects. we also recommend that health cost of tobacco farming must be considered, while determining minimum indicative price (mip) for tobacco leaf, so that small farmers must be compensated for bearing the adverse health issues associated with tobacco farming. furthermore, alternative livelihood sources should be developed for tobacco farmers to ensure switching from tobacco to non-tobacco sources. solving marketing issues associated with horticultural crops will encourage speedy switching from health hazardous tobacco farming. references adioetomo, s. m., djutaharta, t., &hendratno. (2005). cigarette consumption, taxation, and household income: indonesia case study. hnp discussion paper, economics of tobacco control paper no.26. alderete, e., livaudais-toman, j., kaplan, c., gregorich, s. e., mejía, r., & pérez-stable, e. j. (2020). youth working in tobacco farming: effects on smoking behavior and association with health status. bmc public health, 20(1), 84. anh, p. t. h., ross, h., anh, n. q., linh, b. n., & minh, n. t. (2016). direct and indirect costs of smoking in vietnam. tobacco control, 25(1), 96-100. arcury, t. a., and quandt, s. a. (2006). health and social impacts of tobacco production. journal of agromedicine, 11(3-4), 71-81. arcury, t. a., quandt, s. a., preisser, j. s., bernert, j. t., norton, d., & wang, j. (2003). high levels of transdermal nicotine exposure produce green tobacco sickness in latino farmworkers. nicotine & tobacco research, 5(3), 315-321. available at https://www.ncbi.nlm.nih.gov/pubmed/19274899 barraclough, s., & morrow, m. (2010). the political economy of tobacco and poverty alleviation in southeast asia: contradictions in the role of the state. global health promotion, 17(1_suppl), 4050. bashir, f., & idrees, m. (2018). addressing the pitfalls of single poverty line in the estimation of poverty in different regions of pakistan. forman journal of economic studies, 14. cazé, a. m. d. b., lacerda, a. b. m. d., lüders, d., conto, j. d., marques, j., & leroux, t. (2019). perception of the quality of life of tobacco growers exposed to pesticides: emphasis on health, hearing, and working conditions. international archives of otorhinolaryngology, 23(1), 50-59. chowdhury, s. (2015). public retreat, private expenses, and penury: a study of illness induced impoverishment in urban india. journal of developing societies, 31(2), 153-183. cordes, d. h., & foster, d. (1988). health hazards of farming. american family physician, 38(4), 233244. https://www.ncbi.nlm.nih.gov/pubmed/19274899 journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.521 15 cox, c. (1992).1,3—dichloropropene. journal of pesticide reform.https://d3n8a8pro7vhmx.cloudfront.net/ncap/pages/26/attachments/original/14284233 63/dichloropropene.pdf?1428423363 cox, c. (1995). chlorpyrifos factsheet, part 2. journal of pesticide reform.http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.200.4271&rep=rep1&type= pdf damalas, c. a., georgiou, e. b., &theodorou, m. g. (2006). pesticide use and safety practices among greek tobacco farmers: a survey. international journal of environmental health research, 16(5), 339-348. fao, x. (2003). projections of tobacco production, consumption and trade to the year 2010.http://www.fao.org/3/a-y4956e.pdf faostat.(2013). world unmanufactured tobacco production, world annual unmanufactured tobacco producer price per tonne in u.s. dollars; 1970-2013. food and agriculture organization of the united nations, statistics division. available from: http://www.faostat3.fao.org. fassa, a. g., faria, n. m., meucci, r. d., fiori, n. s., miranda, v. i., &facchini, l. a. (2014). green tobacco sickness among tobacco farmers in southern brazil. american journal of industrial medicine, 57(6), 726-735. fotedar, s., &fotedar, v. (2017). green tobacco sickness: a brief review. indian journal of occupational and environmental medicine, 21(3), 101. ghosh, s. k., parikh, j. r., gokani, v. n., rao, m. n., kashyap, s. k., & chatterjee, s. k. (1980). studies on occupational health problems in agricultural tobacco workers. occupational medicine, 30(3), 113-117. ghosh, s. k., saiyed, h. n., gokani, v. n., &thakker, m. u. (1986). occupational health problems among workers handling virginia tobacco. international archives of occupational and environmental health, 58(1), 47-52. goodchild, m., nargis, n., &d'espaignet, e. t. (2018). global economic cost of smoking-attributable diseases. tobacco control, 27(1), 58-64. goodchild, m., nargis, n., &d'espaignet, e. t. 2018. global economic cost of smoking-attributable diseases. tobacco control, 27(1), 58-64. ilo. (2014). tobacco sector employment statistical updates. international labour organization, sectoral policies department. international labour office ch-2211, geneva, 22, switzerland. jafri, s. y., &khattak, a. (1995). income inequality and poverty in pakistan. pakistan economic and social review, 33(1), 37-58 khan, d. a., shabbir, s., maiid, m., ahad, k., nagvi, t. a & khan, f. a. (2010). risk assessment of pesticide exposure on health of pakistani tobacco farmers. j expo sci environ epidemiol.20(2),196-204. kidane, a., hepelwa, a., tingum, e., & hu, t. w. (2013). agricultural inputs and efficiency in tanzania small scale agriculture: a comparative analysis of tobacco and selected food crops. tanzanian economic review, 3(1-2), 1. kids, t. f. (2001). golden leaf, barren harvest: the costs of tobacco farming. campaign for tobaccofree kids. kyaing n.n., perucic, a.m., & rahman, k. (2005). study on poverty alleviation and tobacco control in myanmar, hnp discussion paper no. 31, washington dc: world bank and who, 2005 lonsway, j. a., byers, m. e., dowla, h. a., panemangalore, m., &antonious, g. f. (1997). dermal and respiratory exposure of mixers/sprayers to acephate, methamidophos, and endosulfan during tobacco production. bulletin of environmental contamination and toxicology, 59(2), 179-186 mackay j, eriksen m. (2005). the tobacco atlas. geneva: world health organization. majra, j. p., & gur, a. (2009). poverty, tobacco, and health: an indian scenario. journal of health, population, and nutrition, 27(3), 316. https://d3n8a8pro7vhmx.cloudfront.net/ncap/pages/26/attachments/original/1428423363/dichloropropene.pdf?1428423363 https://d3n8a8pro7vhmx.cloudfront.net/ncap/pages/26/attachments/original/1428423363/dichloropropene.pdf?1428423363 http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.200.4271&rep=rep1&type=pdf http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.200.4271&rep=rep1&type=pdf http://www.fao.org/3/a-y4956e.pdf muhammad shahzad, anwar shah, frank joseph chaloupka 16 malik, m. h. (1988). some new evidence on the incidence of poverty in pakistan. the pakistan development review, 27(4), 509-515. mcbride, j. s., altman, d. g., klein, m., & white, w. (1998). green tobacco sickness. tobacco control, 7(3), 294-298. nasrullah, m., chang, l., saddozai, k.n., khalid, a.o., bayisenge, r. & hameed, g.(2019). cost and net return of tobacco growers:a case study of district mardan (kppakistan). sarhad journal of agriculture, 35(2): 565-571. ngajilio, d., adams, s., &jeebhay, m. f. (2018). occupational allergy and asthama in tobacco farmers: a review of literature. current allergy & clinical immunology, 31(2), 88-95. ngajilo, d., adams, s., &jeebhay, m. f. (2018). occupational allergy and asthma in tobacco farmers: a review of literature. current allergy & clinical immunology, 31(2), 88-95. osim, e. e., musabayane, c. t., &mufunda, j. (1998). lung function of zimbabwean farm workers exposed to flue curing and stacking of tobacco leaves. south african medical journal, 88(9), 1127-1131 pain, a., hancock, i., eden-green, s., & everett, b. (2012). research and evidence collection on issues related to articles 17 and 18 of the framework convention on tobacco control. report published by dd international for british american tobacco. online: http://ddinternational. org. uk/viewproject. php. perera, k. m. n., guruge, g. d., &jayawardana, p. l. (2017). household expenditure on tobacco consumption in a poverty-stricken rural district in sri lanka. asia pacific journal of public health, 29(2), 140-148. peres, f., moreira, j. c., rodrigues, k. m., & claudio, l. (2006). risk perception and communication regarding pesticide use in rural work: a case study in rio de janeiro state, brazil. international journal of occupational and environmental health, 12(4), 400-407. pryer, j., rogers, s., & rahman, a. (2005). work-disabling illness as a shock for livelihoods and poverty in dhaka slums, bangladesh. international planning studies, 10(1), 69-80. qureshi, s. k., &arif, g. m. (2001). profile of poverty in pakistan, 1998-99. pakistan institute of development economics (pide). rao, d. p., coelli, t. j., &alauddin, m. (2005). agricultural productivity growth, employment and poverty in developing countries, 1970-2000. geneva: international labour office. salvi, r. m., lara, d. r., ghisolfi, e. s., portela, l. v., dias, r. d., & souza, d. o. (2003). neuropsychiatric evaluation in subjects chronically exposed to organophosphate pesticides. toxicological sciences, 72(2), 267-271. schenker, m. b., farrar, j. a., mitchell, d. c., green, r. s., samuels, s. j., lawson, r. j., & mccurdy, s. a. (2005). agricultural dust exposure and respiratory symptoms among california farm operators. journal of occupational and environmental medicine, 47(11), 1157-1166. sun, x., sleigh, a. c., carmichael, g. a., & jackson, s. (2010). health payment-induced poverty under china’s new cooperative medical scheme in rural shandong. health policy and planning, 25(5), 419-426. van minh, h., giang, k. b., bich, n. n., & huong, n. t. (2009). tobacco farming in rural vietnam: questionable economic gain but evident health risks. bmc public health, 9(1), 24. world health organization. (2008). tobacco and poverty in the philippines. national tobacco control team, department of health, manila college of public health, university of the philippines, manila philippine college of medical researchers foundation, inc., and the tobacco free initiative of the world health organization geneva. world health organization.(2004). tobacco and poverty: a vicious cycle. https://apps.who.int/iris/handle/10665/68704 https://apps.who.int/iris/handle/10665/68704 journal of applied economics and business studies, volume. 4, issue 3 (2020) 35-54 https://doi.org/10.34260/jaebs.432 35 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x competitive pressure and firm-level innovation: the perspective of a developing economy muhammad zeeshan younas*1, faiz-ur-rehman2 1research scholar at school of economics, quaid-i-azam university islamabad, pakistan 2 assistant professor at school of economics quaid-i-azam university islamabad, pakistan abstract the connection between firm-level innovation and competition has received scholarly attention for a long time before now. this paper attempts to shed light on this complex relationship from a novel perspective where a detailed firm-level dataset of private manufacturing pakistani firms spanning from 2002 to 2015 is used. we test whether the non-linearity estimate of aghion et al. (2005) is sustained by our firm-level data. a multivariate probit estimation technique indicates that higher competition leads to a more probability of innovation, but at a declining rate as competitor numbers increase. moderate confirmation of an inverted-u relationship between competition and innovation is found, especially in process and organizational innovation. the findings have practical implications for policymakers in the area of market structure and firm-level innovation. keywords innovation; competition; firm performance; inverted-u shaped; developing economy jel classification l1; l10; o31; o32 1 introduction innovation and rivalry are the key driving forces of productivity growth both at the micro and macro level (beneito et al., 2017). the connection between innovation and competition traces back to schumpeter research (1934, 1942), and is a substantial theoretical and conceptual area that has drawn the attention of the researcher for decades. owing to its significance for the policymaking process, researchers worked extensively on this area but a closer look at the literature reveals that the relationship is still not always crystal clear (moen et al., 2018; bessonova and gonchar, 2019). the tie muhammad zeeshan younas & faiz-ur-rehman 36 between innovation and competition has a complex and strong idiosyncratic character that is hard to generalize across circumstances and firms (mulkay, 2019; tammi, 2020). with this in mind, the current study re-examines this relationship using a panel dataset of a developing economy and finds clear nonlinearities in the form of an inverted-u shape. there exists a considerable body of literature on the relationship between innovation and competition, seminal contributions are made by schumpeter (1942), arrow (1962) and aghion et al. (2005). according to schumpeter (1942), the lack of competition has a beneficial effect on the firm's engagement in r&d activities. due to the availability of more stable funds and facing less market uncertainty, monopolistic firms are more willing to perform innovative practices. arrow (1962) provides a counter-argument and concludes that the competitive environment spurs r&d. he argues that the monopolist firm does not need to innovate due to the possibility to slack. this is also documented as the arrow-schumpeter debate in the literature. besides, aghion et al. (2005) maintain that the association between innovation and competition is neither a schumpeterian or arrow but an inverted-u shaped relationship. two different research directions can be identified from these discrepancies in literature; the first is opposing views of the arrow-schumpeter debate and the second is the integration of both to examine the dual impacts of rivalry (negassi et al., 2019; bonfatti and pisano, 2020). a strand of literature (tang, 2006; vives, 2008; darai et al., 2010; castellacci, 2011) provides empirical evidence on the relevance of the arrowschumpeter debate. following (scherer, 1967a), research in the second direction includes plenty of studies demonstrate that the dual impact of competition on innovation is characterized by non-linearity (kamien and schwartz, 1976; scott, 1984; boone, 2001; aghion et al., 2005). these empirical validations, especially aghion et al. (2005, 2018) studies, revitalized the attention of researchers towards innovation and competition interaction. the central goal here is to re-examine how persistent is the association between firm-level innovation and competition. all previous studies on the interplay between innovation and competition are conducted with the perspective of western and other advanced asian economies but there is no serious attempt has yet been made that provides a firm-level analysis of developing economies like pakistan. study in hand fill this research gap and contribute to the current literature in different ways. first of all, unlike the majority of the previous studies, it is based on the domestically focused firms which is an under-researched category. secondly, more firm-level based research is journal of applied economics and business studies, volume. 4, issue 3 (2020) 35-54 https://doi.org/10.34260/jaebs.432 37 required for developing economies if we acknowledge that “one size fits all” industrial policy is not valid in diversified global markets. thirdly, along with examining the linearity of the relationship this paper also considers the spatial aspects. the study revolves around a key research question; how business-level innovation of domestically focused firms of a developing economy affected by competition? this question is empirically examined with the multivariate probit model which is the best choice when we have more than one dependent variable and these variables are correlated with each other. 1.1 objective of the study the association between innovation and competition has remained a mystery in industrial economics. thanks to the better data availability and motivation from novel theoretical models, the research agenda has added thrust in the last couple of years. the fundamental objective of this study is to find the answer to a research question: “do the linkages between competition and innovation vary across industries in pakistan?” in other words, as suggested by aghion et a. (2005), is the relationship between innovation and market competition in pakistan inverted-u shaped or not? the remaining manuscript is organized as follows. section 2 deals with the review of previous works. section 3 presents the data and variable construction. section 4 belongs to the econometrical method. section 5 provides empirical results. section 6 concludes the paper with suitable policy recommendations. 2 related literature as is clear from the works of karl marx and adam smith, and is nowadays part of the universal consensus among economists, that innovation plays a key role in enlightening the dynamic properties of the economic system, industries, and firms (mulkay, 2019; tammi, 2020). joseph schumpeter (1934), an economist from the austrian school, further elaborated this mechanism who provides innovation a special room in his renowned monograph the theory of economic development. he is recognized as the most radical economist of the 20th century who emphasized entrepreneurship, science, and technology in explaining the economic growth heterogeneity among economies both at the macro and micro levels. according to him, innovation can be divided into five different categories: new forms of organization, new markets, new production methods, new products, and new sources of supply. he endogenized the innovation role in his work and postulate that the dynamic efficiency of industries and firms is based on the concept of “creative destruction” which means something new kills an old thing. muhammad zeeshan younas & faiz-ur-rehman 38 the empirical literature on innovation and competition appears to point in several directions and has caused debates among scholars. although the earlier research of schumpeter (1934) proposes that new and small firms are key sources of innovation but his later work capitalism, socialism, and democracy (1942) transformed the attention and emphasis on the competitive advantage of large-sized firms over smaller ones. he quotes that the larger sized firms have economies of scale in r&d, risk spread, access to finance, and management capabilities which provide a comparative lead to enhance product development and exploit new technologies. there is a bulk of studies available claiming that advancement and change are necessary for firm growth in modern competitive markets. to understand the theoretical foundation underpinning research and development, it is critical to grasp innovation in the context of market competition. arrow and schumpeter have been depicted as rivals in the literature and the complementarities between both have largely been ignored by the researchers till the 2000s. some prominent studies in favor of schumpeter's negative relation between innovation and competition are (horowitz and clemens, 1962; dixit and stiglitz, 1977; nelson and winter, 1982; romer, 1990; hashmi and van biesebroeck, 2010). in contrast, arrow's contribution of a positive association between competition and innovation is empirically supported by (gilbert and newbery, 1982; reinganum, 1983; geroski, 1995; nickell, 1996; blundell et al., 1999). but with the publication of aghion et al. (2005) the empirical literature on the interplay between innovation and competition gets a new direction. they developed an insightful model of competition that describes that there is a u-shaped association exists between both variables and claims that the relationship is a mixture of arrow and schumpeter rather than onedirectional. in other words, innovation takes place when the firm faces more competition, or in a market that has too little or do not have too much competition. the non-linear association between innovation and competition was first spotted by scherer (1967) and empirically tested by kamien and schwartz (1976). aghion et al. (2005) re-estimate the shape of this relationship and find that the competition effect dominates at the low levels of competition leads to a positive association, whereas the schumpeterian effect dominates at a high level of competition leads to a negative effect. generally, this provides us an inverted u-shaped relationship. for simplicity, they divided industries into two groups; the first is neck-and-neck where firms have the same technology and the second is leader-follower where the technological gap among firms is high. in order to improve their odds of getting ahead of their competitor firms invest in r&d practices. the firms need the profit to participate in innovation but also need competition to get motivation for investing (mulkay, 2019). journal of applied economics and business studies, volume. 4, issue 3 (2020) 35-54 https://doi.org/10.34260/jaebs.432 39 according to aghion et al. (2005), two opposite impacts on the firm-level innovation take place when the competition increases in the industry; 1) there is less incentive to innovate in a competitive environment as more competition leads to less profit, known as the schumpeterian effect. 2) the second effect is recognized as the escape competition effect in which firms engage in r&d activities to escape competition from competitors. so we can say that competition and innovation are negatively associated due to the schumpeterian effect and positively connected due to the escape competition effect. they conclude that the escape competition effect dominates the schumpeterian effect in neck-and-neck industries while the schumpeterian effect takes over the escape competition effect in leader-follower industries. when rivalry is too low or too high, the innovation level is low. when rivalry falls in between low and high, the innovation level is high (tammi, 2020). if we combine these propositions we get an inverted-u relationship between innovation and competition. the steepness of the inverted-u relationship is affected by the degree to which industries are technologically homogenous. the literature provides empirical support to the inverted u-shaped relationship between innovation and competition via using different proxies for both variables (sacco and schmutzler, 2011; felisberto, 2012; hashmi, 2013; bento, 2014). however, the findings of these studies are the same. in this study, we re-examine this relationship from the perspective of a developing country pakistan. to sum up the discussion, this paper is an attempt to verify that the non-linear relationship between innovation and competition exists in manufacturing firms of a developing country like pakistan. a distinctive feature of this study is that it emphasizes an under-researched category the domestically focused firms. there is a plethora of literature available which postulating that the exporting firms are extra prospective to innovate but there is no serious attempt has yet been made on the domestically focused firms of a developing economy like pakistan. to the best of our knowledge, this is the first-ever study of its nature which investigating these non-linearities and markets structure impacts on firm-level innovation in the case of pakistan or any other south asian country. muhammad zeeshan younas & faiz-ur-rehman 40 3 data and variable construction the dataset of this research is constructed by merging three recent waves of firmlevel data (investment climate survey 2002-04, enterprise survey 2007-09, enterprise survey 2013-15) obtained from world bank microdata library. following the commonly applied benchmarks of size, geographical location and sector, these surveys include a broad range of subjects associated with the business environment, for instance, performance measures, innovation and technology, competition, crime, infrastructure, corruption and access to finance. due to the stratified random sampling technique, each firm has the same odds of being selected in these surveys. all population units were clustered in different homogenous groups and then firms are selected from each group by using a simple random sampling technique. identical questionnaires were used in all three waves and we merged these datasets after excluding firms with extreme observations or having incomplete information which left us the sample of 718 manufacturing firms. a detailed description of the variables selected for this study is presented in the appendix section. table 1: summary statistics n percent mean std.dev product new to firm 717 27 0.27 0.44 process innovation 717 20 0.20 0.40 organization innovation 717 13 0.13 0.33 product new to market 717 15 0.15 0.36 number of competitors (log) 717 --1.19 0.90 competitors squared (log) 717 --2.22 1.90 r&d active 717 21 0.21 0.35 total employment (log) 717 --1.99 0.57 investment 717 22 0.22 0.41 firm age 717 --21.27 14.10 firm age (log) 717 --1.24 0.29 domestic 717 95 0.95 0.22 multi plant 717 10 0.10 0.30 low-tech 717 65 0.65 0.48 medium tech 717 20 0.20 0.40 high tech 717 15 0.15 0.35 market locality 717 59 0.59 0.49 city over 1m pop 717 60 0.60 0.49 city with 250k to 1m 717 22 0.22 0.42 city with 50k to 250k 717 14 0.14 0.35 city with less than 50k 717 4 0.03 0.17 low competition 717 35 0.35 0.48 medium competition 717 53 0.53 0.50 high competition 717 13 0.12 0.32 table source: authors own calculations based on es survey journal of applied economics and business studies, volume. 4, issue 3 (2020) 35-54 https://doi.org/10.34260/jaebs.432 41 table 1 provides descriptive statistics of our sample. the highest percentage of occurrence among all types of innovations is 27%, belong to the products that are new to the firm. 15% of manufacturing firms allied to the high technology sector and 85% are concentrated in low and medium-level technology businesses. similarly, 95% of sample firms are domestically owned and 90% are operating with single plant capacity. as far as spatial aspects are concerned, almost 60% of firms are situated in the capital or city with over one million population, while 18% of firms are located in smaller cities suggesting that the majority of the domestically focused firms have a tendency towards capital or big cities like islamabad, karachi, lahore, sialkot, and faisalabad. lastly but most importantly, 53% of domestically focused businesses fall in the medium level competition category and 35% belong to the low-level competition markets. table 2: correlation coefficients among innovation activities variables (1) (2) (3) (4) (5) (1) r&d active 1.000 (2) new to firm product 0.233 1.000 (3) process innovation 0.377 0.598 1.000 (4) new to market product 0.292 0.705 0.652 1.000 (5) organizational innovation 0.464 0.292 0.451 0.350 1.000 table source: authors own calculations based on es survey to certify the integrity of our results, table 2 highlights the correlation coefficients among five different types of innovation practices. research and development (r&d) have the highest link with organizational innovation, signifying that the leading purpose of r&d is to promote organizational innovations. new to the firm product is significantly associated with products new to the markets, suggesting that the businesses introducing new products to the firms are also tend to engage in activities leading new products to the markets as well. furthermore, the process innovation is highly correlated with product innovation new to the market. table 3: different levels of competition organizational product (ntf) product (ntm) process low level competition 39 43 44 48 medium level competition 50 44 50 47 high level competition 11 13 6 note: ntf stands for new to firm, and ntm stands for new to market. table source: authors own calculations based on es survey muhammad zeeshan younas & faiz-ur-rehman 42 as mentioned earlier, we divide competition level into three different categories where firms with 0 to 10 competitors fall in low-level competition, 11 to 99 belong to medium level competition and above 100 are considered as highly competitive firms. table 3 indicates a summary of the percentage of firms introducing innovation in terms of different competition levels. in all types of innovations, the highest percentage of firms have their place in the medium level competition category. the table also provides compelling evidence in support of an inverted-u shaped relationship between competition and innovation because the high-level competition is associated with decreasing innovation levels. 4 model and econometrical specification 4.1 inverted u-shaped relationship the association between firm-level innovation and competition has always remained a puzzle for researchers. but the investigation agenda has gained momentum during the last decade, thanks to the stimulation from new theoretical models and the availability of improved datasets. no general harmony has developed despite the seminal contributions by renowned scholars (schumpeter, 1942; arrow, 1962; aghion et al., 2005; crowley and jordan, 2016; chernyshev, 2016) and the prime question is still stand: is augmented competition level obstructive or conducive to the firm-level innovation? in order to find the empirical answer to this puzzle, a study in hand employs an innovation production function that helps to evaluate the impacts of competition level, company-specific factors and various innovation inputs on firm-level innovation performance. the production function stated below defines the linkages between several key explanatory variables and the probability of a firm to participate in innovation actions (following mansury and love, 2008; doran et al., 2012; crowley and jordan, 2016). as far as econometrical methodology is concerned, a multivariate probit model is used for the estimation of innovation production function which is the best choice when we have more than one dependent variables and these variables are more likely to correlate (galia and legros, 2004; gordon and mccann, 2005). innov𝑖𝑗 = 𝛼𝑜 + 𝛼1𝐶𝑖𝑗 + 𝛼2𝑅𝑖𝑗 + 𝛼3 𝑂𝑖𝑗 + 𝜀𝑖 (1) innov𝑖𝑗 is a dummy variable that represents the firm i engagement in four types of innovation and j denotes the innovation type. 𝐶𝑖 indicates the competition level reported by firm i in terms of a log of total competitors of the main product. in order to check the linearity of the relationship between competition and innovation, we use a log of the number of competitors along with the squared of this term. furthermore, to examine the journal of applied economics and business studies, volume. 4, issue 3 (2020) 35-54 https://doi.org/10.34260/jaebs.432 43 aghion et al. (2005) proposed inverted u-shaped relationship, a categorical variable is generated which defines the competition intensity in terms of different levels i.e low level (0-10), medium level (11-99), and high level (above 100) competitors. the relationship between innovation and engagement in r&d events is assumed to be positive and documented as a stylized fact in the literature. 𝑅𝑖 specifies the predicted value of r&d of firm i for innovation type j which we obtained after estimating a probit model defined as the r&d active binary variable as a function of different factors like firm size, product diversification, education level of the employees, and foreign competition. while 𝑂𝑖 indicates a vector of remaining variables including technological segmentation of the firm1, market locality, firm size, firm age, plant capacity, spatial aspect in terms of city location and investment propensities towards types of machinery and equipment. in addition, we re-defined our basic model and further estimate two equations to empirically evaluate the inverted u-shaped connection. in the first reformation, we replace the number of competitors with a categorical variable which we defined in terms of low, medium, and high competition levels, and re-estimated the basic model. in the second reformation, we introduced an interactive term via multiplying competition level with the market locality to examine the impacts of competition level by primary market location, the technology sector, and urban scale. these analyses are based on the last three waves of rich panel datasets of pakistani manufacturing firms (ics 2002-04, es 2007-09, es 2013-15). 5 empirical findings 5.1 inverted u-shaped relationship using a rich panel dataset, we estimate a multivariate probit model for pakistani manufacturing firms, and the results for each type of innovation outcome are presented in table 4. all types of innovations are positively influenced by competition level except organizational innovation. but squared competition level term has significantly negative sign advocates that the market competition has decreasing returns. these outcomes are consistent with a strand of the empirical literature (tingvall and poldahl, 2006; peneder, 2012; bos et al., 2013; correa and ornaghi, 2014; halpern and muraközy, 2015; negassi et al., 2019) where competition level has a positive impact on innovation, however, with additional market contestants this association is non-linear in nature and the probability of introducing new product and/or process grows at a diminishing rate. 1for more details: https://ec.europa.eu/eurostat muhammad zeeshan younas & faiz-ur-rehman 44 understanding this relationship can better equip the policymakers to device fruitful interventions, for instance, it suggests that the propensity to invest in innovation practices by additional entrants decreases with market growth. table 4: determinants of firm-level innovation (multivariate probit model estimation) innovation type product product process organization (new to firm) (new to market) competition level 1.540*** 0.595** 1.427*** 0.354 (0.318) (0.323) (0.327) (0.362) squared competition level -0.701*** -0.320** -0.699*** -0.242 (0.146) (0.150) (0.152) (0.169) domestic 0.129 0.033 -0.046 -0.436 (0.277) (0.279) (0.284) (0.298) firm age 0.759*** 0.247 0.753*** -0.002 (0.200) (0.214) (0.211) (0.244) investment -0.262* -0.155 -0.093 0.282** (0.143) (0.146) (0.147) (0.149) firm size 0.395*** 0.475*** 0.396*** 0.226* (0.107) (0.116) (0.115) (0.125) r&d active 0.472*** 0.707*** 0.986*** 1.331*** (0.157) (0.153) (0.161) (0.162) multi-plant -0.275 -0.180 -0.501** -0.545** (0.189) (0.205) (0.226) (0.288) low tech˟ -0.369** -0.253 -0.314* 0.042 (0.165) (0.163) (0.172) (0.193) low to medium tech˟ 0.024 0.238 0.260 -0.293 (0.202) (0.214) (0.208) (0.260) market locality 0.222** 0.373** 0.51*** 0.005 (0.115) (0.125) (0.124) (0.139) city with pop over 1 millionʺ -0.427 0.212 -0.016 -0.121 (0.291) (0.318) (0.300) (0.345) city over 250 to 1 millionʺ -1.478*** -0.631** -0.670** -0.434 (0.327) (0.355) (0.326) (0.383) city with 50,000 to 2,50000ʺ -1.103*** -0.418 -1.069** -0.018 (0.327) (0.362) (0.366) (0.375) constant -2.03* -2.59* -2.79* -1.30** (0.549) (0.564) (0.578) (0.610) wald chi-square (p-value) 387.9 (0.00) log likelihood -1010.99 no of observations 717 note: *** significant at 1%, ** at 5%, * at 10% level. the parentheses contain standard errors. ˟ high tech is the reference category. ʺ city with a population of less than 50000 is the reference category. table source: authors own calculations based on es survey journal of applied economics and business studies, volume. 4, issue 3 (2020) 35-54 https://doi.org/10.34260/jaebs.432 45 table 4 further test the baseline hypothesis and shows that the firms active in terms of research and development activities are more likely to be innovative (crepon et al., 1998; revilla and fernandez 2012; doloreux et al., 2015; younas and rehman, 2020). schumpeter (1942) claims that the size of the firm plays a key role in firm-level innovation is also confirmed by this study. domestically owned businesses do not influence any type of innovation whereas market locality exerts positive impacts on products and process innovation. in addition, we find no evidence of multi-plant firms' connection with the possibility of product innovation, although the negative impact on the process and organizational innovation signifying that these types of firms are facing some coordination conflict challenge. literature suggests that the agglomeration effects of large cities, localization, urbanization externalities and other spatial aspects of clustering have a strong association with firm-level innovation. the study in hand finds that the firm located in cities with a population of less than 1 million negatively affect the product and process innovation outcomes. as far as technological segmentation of the firms is concerned, low tech firms lower down the introduction of new products to the firm and process innovation. table 5: firm-level innovation at different competition levels (multivariate probit model) innovation type product product process organization (new to firm) (new to market) low level competition® 0.1903*** 0.237*** 0.244*** 0.264*** (0.123) (0.133) (0.128) (0.147) high level competition® 0.147 0.161 -0.610** -0.656** (0.203) (0.235) (0.259) (0.310) constant -2.13* -2.90* -3.03* -1.52** (0.545) (0.573) (0.587) (0.629) wald chi-square (p-value) 296.9 (0.00) log likelihood -824.45 no of observations 717 note: we report only the coefficients of the replaced variable here as remaining all variable coefficients and their significance is the same as table 4. *** significant at 1%, ** at 5%, * at 10% level. ® medium level competition is the reference category. the parentheses contain standard errors. table source: authors own calculations based on es survey to ensure the reliability of our results about the u-shape relationship, a new categorical variable is created which divides the level of competition into three different categories including low level (0-10 competitors), medium level (11-99 competitors) muhammad zeeshan younas & faiz-ur-rehman 46 and high-level competition (above 100 competitors). we replaced the number of competitor variables with a series of competition dummies and re-estimate our base equation after controlling all other variables. the basic idea behind this variable replacement is to test whether the relationship between innovation and competition is inverted u-shaped or not. the simple guideline here is that relative to the reference category of medium level competition the coefficients of low and high-level competition should be significant. table 5 provides the multivariate probit estimation results of this model. manufacturing firms are expected to introduce new products, processes, or organizational innovation at a low level of competition. however, high-level competition is connected with a lower likelihood of process and organizational innovation. so we can conclude that there is some indication of an inverted-u relationship between innovation and competition is found especially in process innovation and organizational innovation. after combining equation 1 and equation 2 results we can say that the firms are more prospective to innovate at a lower level but this continues until the market grows up to a “tipping point as mentioned by aghion et al., (2005)” and after that, the probability to invest in r&d by the new entrants grows at a diminishing rate. additionally, we re-estimate our main model to examine the impacts of competition level by the technology sector and find results differ by the neckand-neck phenomenon proposed by aghion et al. (2005). one justification for these diverging results is the nature of our dataset as almost every pakistani firm belongs to either the low tech or the medium-tech sector. table 6: multivariate probit model estimation for local market competition type of innovation product product process organization (new to firm) (new to market) competition in local markets 0.084 0.055 0.158** -0.079 (0.062) (0.068) (0.065) (0.081) constant -1.871* -2.386* -2.426* -1.346** (0.527) (0.548) (0.545) (0.590) wald chi-square (p-value) 274.5 (0.00) log likelihood -842.83 no of observations 718 note: we report only the coefficients of the replaced variable here as remaining all variable coefficients and their significance is the same as table 4. *** significant at 1%, ** at 5%, * at 10% level. the parentheses contain standard errors. table source: authors own calculations based on es survey journal of applied economics and business studies, volume. 4, issue 3 (2020) 35-54 https://doi.org/10.34260/jaebs.432 47 we further extend the analysis to check how competition in local markets affects different types of innovation outcomes. for this purpose, we introduce an interactive term by combining the market locality variable with several competitors and re-estimate equation 1. multivariate probit model estimation results for the local market competition is presented in table 6 highlights that the competition in the local market has a positive influence on process innovation. 6 conclusions and policy implication there is no universal consensus on how firm-level innovation gets affected by market competition as different papers come to diverging outcomes. this study seeks to contribute to the existing literature on firm-level innovation by re-examining the relationship between competition and innovation in the pakistani manufacturing firms, from a novel perspective of domestically focused businesses in a developing country. it is an attempt to answering the research question; how business-level innovation of domestically focused firms of a developing economy affected by competition? our results, moderately consistent with aghion et al. (2005) and others, specify an invertedu shaped relation between innovation and competition in the pakistani manufacturing firms. to ensure the reliability of our results about the u-shape relationship, a new categorical variable is created which divides the level of competition into three different categories including low, medium and high-level competition. the multivariate probit estimation results of this model suggest that the firms are more likely to introduce new products, processes, or organizational innovation at a low level of competition. however, high-level competition is associated with a lower probability of process and organizational innovation. so we can conclude that there is some clear evidence of an inverted u-shaped relationship between competition and innovation is found especially in process innovation and organizational innovation. we further extend our analysis to check how competition in local pakistani markets affects different types of innovation outcomes and concludes that it has a positive influence on process innovation only. these findings are robust to different model specifications using different sample periods, instruments and inclusion/exclusion of control variables. the overall empirical analysis points to a need to reconsider the regulatory changes by the government as the competition encouraging strategies in pakistan would tend to inverse decay in firm-level innovations. understanding this relationship can better equip the policymakers to device fruitful interventions like it suggests that the propensity to invest in innovation practices by additional entrants decreases with market growth. due to the inverted-u shaped muhammad zeeshan younas & faiz-ur-rehman 48 relation, maximizing competition policy may lead to a reduction in firm-level innovation. 6.1 generalization of findings to assess the generalization of our findings, it would be exciting to see if these patterns are applicable to other developing economies. the empirical results found in this study may not certainly hold for all developing economies because they display dissimilar innovation and industrial settings. moreover, there is a geographical restriction as well as a restriction of organizational size, infrastructure and institutional environment. however, if the local business environment of any developing country is the same as pakistan for example, bangladesh, nepal, sri lanka, etc. then we can generalize these findings to that particular economy. also, the generalization can further be confirmed by using world bank enterprise surveys for any specific country because the world bank is using the same questionnaire and definition of the term “manufacturing” for all developing economies. 6.2 study limitations and future recommendations with regard to empirical analysis, the major caveat to our findings is the limited availability of data. the findings of this study must be used with some caution as the analysis is limited to a binary nature of innovation variables. the binary structure of the dependent variables places some limitations on our understanding of the scope of the innovation. we do not know how many new products were introduced by each firm and have no information on the quality or complexity of these products. therefore, complementarities in these dimensions cannot be ruled out. another sample related important limitation is the lack of information to identify domestic and multinational subsidiaries to compare more appropriately the innovation performance of foreign subsidiaries. our results point to several further themes which also need to be studied in future research. most importantly, it would be critical to identify what magnitude the research and development investments of micro organizations turn out to be profitable in terms of higher net revenues in the short, medium, and long-run in comparison to those micro organizations not making such investments. upcoming research should also examine the question of whether those companies who are not entering the path of innovation are determining against it because they intentionally objective to evade the risks associated with this choice or because they face liquidity limitations. likewise, we expect that the implications of our study are also held in environments outside of pakistan. in order to assess the generalization of our findings, it would be exciting to journal of applied economics and business studies, volume. 4, issue 3 (2020) 35-54 https://doi.org/10.34260/jaebs.432 49 see if these patterns are applicable to other developing economies. this can be done as a replication of this study or other empirical literature on the same area with new datasets. this would not only help in the robustness checking of our findings but also assist to examine the amount of heterogeneity across developing economies with the perspective of firm-level innovation and competition interplay. we, therefore, look forward to upcoming research on this area. references adamou, a., & sasidharan, s. (2007). the impact of r&d and fdi on firm growth in emerging-developing countries: evidence from indian manufacturing industries. available at ssrn: https://ssrn.com/abstract=987024 aghion, p., bergeaud, a., lequien, m., & melitz, m. j. (2018). the impact of exports on innovation: theory and evidence (no. w24600). national bureau of economic research. aghion, p., bloom, n., blundell, r., griffith, r., & howitt, p. (2005). competition and innovation: an inverted-u relationship. the quarterly journal of economics, 120(2), 701-728. amara, n., d'este, p., landry, r., & doloreux, d. (2016). impacts of obstacles on innovation patterns in kibs firms. journal of business research, 69(10), 40654073. arrow, k. (1962). economic welfare and the allocation of resources for invention. in c. k. rowley (ed.), readings in industrial economics(pp. 219-236). london: palgrave. audretsch, d. b. (1995). innovation, growth and survival. international journal of industrial organization, 13(4), 441-457. beneito, p., coscollá‐girona, p., rochina‐barrachina, m. e., & sanchis, a. (2015). competitive pressure and innovation at the firm level. the journal of industrial economics, 63(3), 422-457. beneito, p., rochina-barrachina, m. e., & sanchis, a. (2017). competition and innovation with selective exit: an inverted-u shape relationship?. oxford economic papers, 69(4), 1032-1053. bento, p. (2014). competition as a discovery procedure: schumpeter meets hayek in a model of innovation. american economic journal: macroeconomics, 6(3), 124-52. https://ssrn.com/abstract=987024 muhammad zeeshan younas & faiz-ur-rehman 50 bessonova, e., & gonchar, k. (2019). how the innovation-competition link is shaped by technology distance in a high-barrier catch-up economy. technovation, 86, 15-32. blundell, r., griffith, r., & van reenen, j. (1999). market share, market value and innovation in a panel of british manufacturing firms. the review of economic studies, 66(3), 529-554. bonfatti, r., & pisano, l. (2020). credit constraints and the inverted‐u relationship between competition and innovation. economica, 87(346), 442-469. boone, j. (2001). intensity of competition and the incentive to innovate. international journal of industrial organization, 19(5), 705-726. bos, j. w., kolari, j. w., & van lamoen, r. c. (2013). competition and innovation: evidence from financial services. journal of banking & finance, 37(5), 15901601. bruland, k., & mowery, d. c. (2004). innovation through time. georgia institute of technology castellacci, f. (2011). how does competition affect the relationship between innovation and productivity? estimation of a cdm model for norway. economics of innovation and new technology, 20(7), 637-658. chernyshev, n. (2016). inverted-u relationship between r&d and competition: reconciling theory and evidence. mimeo, university of st. andrews. coad, a., & rao, r. (2008). innovation and firm growth in high-tech sectors: a quantile regression approach. research policy, 37(4), 633-648. cohen, w. m. (2010). fifty years of empirical studies of innovative activity and performance. in handbook of the economics of innovation (vol. 1, pp. 129213). north-holland. correa, j. a., & ornaghi, c. (2014). competition & innovation: evidence from us patent and productivity data. the journal of industrial economics, 62(2), 258285. criscuolo, c. (2009). innovation and productivity: estimating the core model across 18 countries. innovation in firms: a microeconomic perspective, edited by oecd. oecd publishing. crowley, f., & jordan, d. (2017). does more competition increase business-level innovation? evidence from domestically focused firms in emerging economies. economics of innovation and new technology, 26(5), 477-488. journal of applied economics and business studies, volume. 4, issue 3 (2020) 35-54 https://doi.org/10.34260/jaebs.432 51 darai, d., sacco, d., & schmutzler, a. (2010). competition and innovation: an experimental investigation. experimental economics, 13(4), 439-460. dixit, a. k., & stiglitz, j. e. (1977). monopolistic competition and optimum product diversity. the american economic review, 67(3), 297-308. doran, j., jordan, d., & o'leary, e. (2012). the effects of national and international interaction on innovation: evidence from the irish cis: 2004–06. industry and innovation, 19(5), 371-390. felisberto, c. (2012). the relationship between competition and the incumbent’s innovation. journal of industry, competition, and trade, 12(1), 21-46. galia, f., & legros, d. (2004). complementarities between obstacles to innovation: evidence from france. research policy, 33(8), 1185-1199. geroski, p. a. (1990). innovation, technological opportunity, and market structure. oxford economic papers, 42(3), 586-602. gilbert, r. j., & newbery, d. m. (1982). preemptive patenting and the persistence of monopoly. the american economic review, 514-526. gordon, i. r., & mccann, p. (2005). innovation, agglomeration, and regional development. journal of economic geography, 5(5), 523-543. griffith, r., huergo, e., mairesse, j., & peters, b. (2006). innovation and productivity across four european countries. oxford review of economic policy, 22(4), 483498. hall, b. h., lotti, f., & mairesse, j. (2009). innovation and productivity in smes: empirical evidence for italy. small business economics, 33(1), 13-33. halpern, l., & muraközy, b. (2015). the relationship between competition and r&d: theoretical approaches and quantitative results, 113-137. hashi, i., & stojčić, n. (2013). the impact of innovation activities on firm performance using a multi-stage model: evidence from the community innovation survey 4. research policy, 42(2), 353-366. hashmi, a. r. (2013). competition and innovation: the inverted-u relationship revisited. review of economics and statistics, 95(5), 1653-1668. hashmi, a. r., & van biesebroeck, j. (2010). market structure and innovation: a dynamic analysis of the global automobile industry (no. w15959). national bureau of economic research. horowitz, i., & clemens, e. w. (1962). firm size and research activity. southern economic journal (pre-1986), 28(3), 298. muhammad zeeshan younas & faiz-ur-rehman 52 kamien, m. i., & schwartz, n. l. (1976). on the degree of rivalry for maximum innovative activity. the quarterly journal of economics, 90(2), 245-260. klette, t. j., & kortum, s. (2004). innovating firms and aggregate innovation. journal of political economy, 112(5), 986-1018. lachenmaier, s., & wößmann, l. (2006). does innovation cause exports? evidence from exogenous innovation impulses and obstacles using german microdata. oxford economic papers, 58(2), 317-350. lööf, h., & heshmati, a. (2006). on the relationship between innovation and performance: a sensitivity analysis. economics of innovation and new technology, 15(4-5), 317-344. mairesse, j., mohnen, p., zhao, y., & zhen, f. (2012). globalization, innovation, and productivity in manufacturing firms: a study of four sectors of china. eria discussion paper, 10. mansury, m. a., & love, j. h. (2008). innovation, productivity and growth in us business services: a firm-level analysis. technovation, 28(1-2), 52-62. moen, ø., tvedten, t., & wold, a. (2018). exploring the relationship between competition and innovation in norwegian smes. cogent business & management, 5(1), 1564167. mulkay, b. (2019). how does competition affect innovation behaviour in french firms?. structural change and economic dynamics, 51, 237-251. negassi, s., lhuillery, s., sattin, j. f., hung, t. y., & pratlong, f. (2019). does the relationship between innovation and competition vary across industries? comparison of public and private research enterprises. economics of innovation and new technology, 28(5), 465-482. nelson, r. r., & winter, s. g. (1982). the schumpeterian tradeoff revisited. the american economic review, 72(1), 114-132. nickell, s. j. (1996). competition and corporate performance. journal of political economy, 104(4), 724-746. peneder, m. (2012). competition and innovation: revisiting the inverted-u relationship. journal of industry, competition, and trade, 12(1), 1-5. reinganum, j. f. (1983). uncertain innovation and the persistence of monopoly. the american economic review, 73(4), 741-748. revilla, a. j., & fernández, z. (2012). the relation between firm size and r&d productivity in different technological regimes. technovation, 32(11), 609-623. journal of applied economics and business studies, volume. 4, issue 3 (2020) 35-54 https://doi.org/10.34260/jaebs.432 53 romer, p. m. (1990). endogenous technological change. journal of political economy, 98(5-2), 71-102. sacco, d., & schmutzler, a. (2011). is there a u-shaped relation between competition and investment?. international journal of industrial organization, 29(1), 65-73. scherer, f. m. (1967). market structure and the employment of scientists and engineers. the american economic review, 57(3), 524-531. scherer, f. m. (1967). research and development resource allocation under rivalry. the quarterly journal of economics, 81(3), 359-394. schumpeter, j. a. (1934). the theory of economic development: an inquiry into profits, capital, credit, interest, and the business cycle, volume 55. schumpeter, j. a. (1942). capitalism, socialism, and democracy. hoboken: new jersey; taylor and francis. scott, j. (1984). firm versus industry variability in r&d intensity. in r&d, patents, and productivity (pp. 233-248). university of chicago press. shearmur, r., doloreux, d., & laperrière, a. (2015). is the degree of internationalization associated with the use of knowledge-intensive services or with innovation?. international business review, 24(3), 457-465. tammi, t., saastamoinen, j., & reijonen, h. (2020). public procurement as a vehicle of innovation–what does the inverted-u relationship between competition and innovativeness tell us?. technological forecasting and social change, 153, 119922. tang, j. (2006). competition and innovation behavior. research policy, 35(1), 68-82. tingvall, p. g., & poldahl, a. (2006). is there really an inverted u-shaped relation between competition and r&d?. economics of innovation and new technology, 15(2), 101-118. transaction publishers vives, x. (2008). innovation and competitive pressure. the journal of industrial economics, 56(3), 419-469. younas, m. z., & rehman, f. u. (2020). exploring the nexus between innovation and firm performance: new evidences from manufacturing innovation survey of pakistan. asian journal of technology innovation, 1-36. muhammad zeeshan younas & faiz-ur-rehman 54 appendix a table 7: variables description name definition competition level number of competitors in the main market in which this establishment sold its main product (log) firm age number of years since the beginning of the operation (log) firm size total employment in the firm (log) domestic =1 if the organization is owned by a majority of domestic people ntf innovation =1 if the firm introduce new to the firm product ntm innovation =1 if the firm introduce new to the market product process innovation =1 if the firm introduce any innovative methods of manufacturing products or offering services organization innovation =1 if the firm introduce any new or improved organizational activity active in r&d =1 if the firm invested during the last fiscal year in the acquisition of; (a) external knowledge, (b) training, (c) machinery (d) external and internal r&d city over 1m population =1 if situated in capital or a city with over 1 mil. population a city with 250k to 1m =1 if situated in a city with over 250000 but less than 1 million population a city with 50k to 250k =1 if situated in a city with over 50000 but less than 250000 population a city with less than 50k =1 if situated in a city with less than 50000 population investment =1 if made any investment during the last fiscal year multi-plant =1 if part of a larger establishment market locality =1 if the main market of the firm is local low tech =1 if belongs to the low technology industry medium tech =1 if belongs to the medium technology industry high tech =1 if belongs to the high technology industry product diversification =1 if the firm has only one product i.e. if firm sales share is 100% for one product and zero otherwise us and europe =1 if export to the us and europe south asia =1 if export to south asian only journal of applied economics and business studies, volume. 4, issue 4 (2020) 117-134 https://doi.org/10.34260/jaebs.446 117 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x consumers’ willingness to pay for fresh-water grown leafy vegetables salah udin buzdar1, syed muhammad khair2 and syed munawar shah3* 1government postgraduate college saryab road, quetta, pakistan 2department of economics, buitems university, city campus jinnah town quetta 3department of economics, buitems university, city campus jinnah town quetta abstract wastewater-grown vegetables are responsible for various diseases whereas freshwater-grown vegetables are comparatively expensive. this paper examines consumers’ willingness to pay for freshwater grown leafy vegetables, such as, cauliflower, spinach and salad. for this purpose, a survey was conducted in five selected areas of quetta city using contingent valuation method and 255 vegetable-consumers were questioned directly with the help of a semi-structured questionnaire. the results indicate that most of the vegetable consumers (87-88%) are willing to pay extra amounts for fwgv. on average, a vegetable consumer is willing to pay 27.4%, 43% and 52% more than the original price for the three leafy vegetables (cauliflowers, salads and spinaches, respectively). the factors such as household income, level of education, awareness, use of freshwater-grown vegetables and employment were identified as determinants of consumers’ willingness to pay for fwgv. the study results reveal a high willingness to pay for fwgv by the vegetable consumers and suggests a comprehensive strategy on the part of the government to overcome the problem of wastewater-grown vegetables (wwgv). keywords wastewatergrown vegetables, freshwatergrown vegetables, contingent valuation method, quetta, pesticides jel classification n1,n50, q00, q2 1. introduction water is a basic human need. it is used in different ways, e.g. drinking, domestic and irrigation purposes. the water has been decreasing in quantity and deteriorating in quality, continuously over the past decades. even though water resources are important for the working of an economy, these are continuously being depleted and damaged at * syed.munawar@buitms.edu.pk salah udin buzdar, syed muhammad khair and syed munawar shah 118 an unsustainable rate (birol et al., 2006). in this situation, when good-quality water is not available, marginal-quality water _sewage water, industrial wastewater, or any other water that is not fit for drinking_ is used for irrigation. wastewater is used for irrigation in many areas of the world where freshwater is scarce. it (use of wastewater in irrigation) is a common reality in three fourth of asia, africa, and latin america (gupta et al., 2008). the use of wastewater in agriculture is rising due to water shortage, rise in population, and urbanization which lead to the production of more wastewater in urban areas (carr et al., 2004). in north and south africa, me, south america, southwest america, southern europe, mexico, and a big part of central and east asia, at least 20m hectares of land is irrigated with wastewater, mostly for vegetables’ cultivation (bigdeli and seilsepour, 2008). pakistan is also no exception. ensink et al. (2004) estimated that 26% of the total production of vegetables in pakistan is irrigated with wastewater. wastewater irrigation is also a common practice in peri-urban and urban areas of quetta city. a survey result in 2002 shows that 84% of about 100 farmers farming in and around the city and almost all farmers in the city used wastewater for irrigation, at least in the dry season (khalil, 2011). on the other hand, the use of wastewater for irrigation for longer periods may lead to the concentration of heavy metals in soil and plants (singh et al., 2010). according to (khan et al. 2008), food crops grown on such soils uptake these heavy metals, which may create a direct threat to human health (bigdeli and seilsepour, 2008; khan et al., 2008; muchewati et al., 2006; and singh et al., 2010), when they use these food crops. this contamination of the food chain is one of the key reasons for the toxic pollutants to enter the human body (khan et al., 2008). on analyzing the quality of municipalwastewater used for canola plants in quetta city, kakar et al. (2010) found sulfate, chloride, fluoride, ammonia, cadmium, chromium, lead, mercury, nickel, zinc, barium, iron, manganese, chlorine, calcium, sodium and copper exceeded the limit given by national environmental quality standards (neqs) for municipal wastewaters of pakistan. the health impacts of these metals and toxic pollutants can be chronic and severe. according to singh et al. (2010), the intake of toxic and contaminated metals through the diet may lead to very serious diseases. toxicity may cause pulmonary and renal effects in the body (singh et al., 2010). exposure to lead (pb) reduces hemoglobin composition, creates a disturbance in the working of kidneys, joints, reproductive and cardiovascular systems and severely damages the central and peripheral nervous system (ogweugbu and muhanga, 2005) and high accumulation of zink results in weakening of growth and reproduction (noolan, 1983). jarup (2003) explained the severe impacts of the constant consumption of contaminated foodstuffs on human organs. when anything has severe health impacts, it is intuitive for the people to desire and demand its improvement. this motivates researchers to explore and determine the demand for improvements. many researchers have tried to find out such demands journal of applied economics and business studies, volume. 4, issue 4 (2020) 117-134 https://doi.org/10.34260/jaebs.446 119 through the willingness to pay (wtp) of the consumers of different goods (boccaletti and nardella, 2000; coulibaly et al. 2011; asim and lohano, 2015; bogale and urgessa, 2012; lagerkvist et al. 2014). in the same way, when most of the vegetables in quetta city are irrigated with wastewater, the demand for freshwater-grown vegetables (fwgv) becomes inevitable on the part of the consumers. therefore, this paper tries to empirically explore the consumers’ willingness to pay for freshwater-grown vegetables. for this purpose, a survey based on the contingent valuation method (cvm) and the multiple linear regression (mlr) model is conducted in quetta at different locations. the researchers’ interest in quetta is mainly for two reasons, first, an ample of research addresses the issue of willingness to pay of the consumers for water-quality improvements (butt and khair, 2015) and improved solid-waste management (khan et al., 2018). however, no work is undertaken yet to estimate the consumers’ perception and their demand for fresh and risk-free vegetables in quetta city. second, quetta city is facing a serious water shortage and is declared in the drought-affected areas. 2. materials and methods the willingness to pay measurements are based on utility theory and an individual is willing to accept a change if the utility is positive (coulibaly et al. 2011). in microeconomics theory, following the textbooks’ examples (nicholson, & snyder, 2011), an economic model is developed by assuming the utility function of a consumer as: 𝑈 = 𝑈 (𝑄, 𝑍) (1) where q is the quality of vegetables and z is the set of quantities of other commodities consumed by the consumer. practically, in applied microeconomics theory, an indirect approach is proposed to be adopted. so, the corresponding indirect utility function is: 𝑉 = 𝑉 (𝑄, 𝑍, 𝐼) (2) where “i” represents the disposable income of the household. so, the indirect utility function depends on the quality of vegetables (q), household disposable income (i), and optimal quantities of other commodities (z). both (direct and indirect) methods give the same result, but the indirect approach is much richer in terms of economics it holds (nicholson and snyder, 2011). moreover, indirect utility approaches have been also used and proposed by chatterjee et al (2017), mustafa et al. (2014), asim and lohano (2015), khan (2014), and xie and zhao (2017). that is why the study uses the indirect utility function. now suppose, consumers are told that the quality of vegetables will increase from q0 to q1 with the presumption that the cost of the fwgv will be higher than the vegetables currently available in the market. then the consumers must choose whether they are willing to pay more for the fwgv or not. as stated above, the consumers pay only if the utility is positive from the inclusion of q1 to q0. in other words, the willingness to pay would be greater than zero when the payment does not decrease the household’s overall economic utility. objectively, it can be said that wtp > 0 if and salah udin buzdar, syed muhammad khair and syed munawar shah 120 only if wtp ≤ q1 – q0. in indirect utility functions, the income of a household is assumed to be constant. thus, the willingness to pay can be expressed as: 𝑉 (𝑄0, 𝐼, 𝑍) = 𝑉 (𝑄1, 𝐼 − 𝑊𝑇𝑃, 𝑍) (3) where q0 represents the quality of vegetables produced by using wastewater and q1 is the quality of vegetables produced by using fresh water and wtp is the maximum amount of money the consumers would like to pay for fwgv. however, the improved quality of vegetables is represented by a hypothetical scenario in the questionnaire of this paper. the scenario says that water sources used in growing the vegetables are within acceptable levels from the perspective of human health. moreover, the vegetables are washed properly in the market and are hygienically stored and presented. 2.1 estimation of wtp for the estimation of willingness to pay function, the specification of an econometric model is necessary. let’s consider that all the individuals face the same improved quality of vegetables (q1). then, wtp variations across respondents depend on the household income (i), the existing quality of the vegetables (q0), and other relevant socio-economic characteristics of the respondents. for this purpose, the econometric model used is the multiple linear regression (mlr) model. 𝑊𝑇𝑃𝑖 = 𝛼 + 𝛽𝑖 𝑋𝑖 + 𝜀𝑖 (4) where xi is the vector of independent variables and βi is the vector of unknown parameters and this equation (4) is extended in equation (5). the paper employed a multiple linear regression model to examine whether consumers have positive attitudes towards (are willing to pay for) fwgv and evaluate the factors which affect their willingness to pay for that. when consumers have made their decisions of whether they are willing to pay or not for the fwgv, they were asked: how much you are willing to pay in excess. the answer to this question came in a specified amount which is considered the maximum amount, the consumers are willing to pay. to determine the relationship of these maximum amounts with different explanatory variables, a multiple regression model is used. wtp is the dependent variable in this study. the proposed multiple linear regression (mlr) model takes the form: 𝑊𝑇𝑃𝑖 = 𝛼 + 𝛽1𝐸𝐷𝑈14−16𝑌𝑟𝑠 + 𝛽2𝐸𝐷𝑈16+𝑌𝑟𝑠 + 𝛽3𝐴𝐺𝐸 + 𝛽4𝐻𝐻𝑆 + 𝛽5𝐼𝑁𝐶𝑂𝑀𝐸𝐿𝑀 + 𝛽6𝐼𝑁𝐶𝑂𝑀𝐸𝑀 + 𝛽7𝐼𝑁𝐶𝑂𝑀𝐸𝑈𝑀 + 𝛽8𝐼𝑁𝐶𝑂𝑀𝐸𝐻 + 𝛽9 𝐹𝑊𝐺𝑉_𝑈𝑆𝐸𝑂𝐶𝐶𝐴𝑆𝐼𝑂𝑁𝐴𝐿 + 𝛽10𝐹𝑊𝐺𝑉_𝑈𝑆𝐸𝑂𝐹𝑇𝐸𝑁 + 𝛽11𝐴𝑊𝐴𝑅𝐸𝑁𝐸𝑆𝑆𝑊𝑊𝐼 + 𝛽12𝐸𝑀𝑃𝐻𝐻𝐻 + 𝛽13𝑃𝑃𝑉𝐸𝐺 + 𝛽14𝐴𝑅𝐸𝐴 + 𝜀𝑖 (5) where α is the intercept and ε is the error term which is normally distributed. the other variables are defined in the following section. journal of applied economics and business studies, volume. 4, issue 4 (2020) 117-134 https://doi.org/10.34260/jaebs.446 121 2.3 data collection for the above econometric model, the dependent variable is a willingness to pay for fwgv. the independent variables included in this paper are education, income, age, household size, respondents’ information and perception for wastewater grown (existing quality of) vegetables, their perception of the use of fwgv, employment type of the household head, purchase-place of vegetables and area (town/road). all the variables are explained, one by one by using their short forms. the explanation gives details about how each variable is measured and included in the model of the paper. willingness to pay (wtp) refers to the highest amount in pakistani rupees (pkr), a consumer is willing to pay to get fwgv. the rupees in addition to a basic price are taken into consideration in this paper. willingness to pay was measured separately for three types of vegetables, namely: cauliflowers/cabbages, spinaches, and salads. education refers to the educational levels of the respondents. for the analysis, this variable was divided into three categories with intermediate level (edulow), 14-16 years (edu14-16yrs), and higher than 16 years (edu16+yrs) and set edulow as the reference category. the age of the respondents is measured in years and men above 18 years of age were considered. household size (hhs) represents a number of family members living in one house and sharing the same kitchen. income refers to a total household monthly income of the respondent in pkr and it is taken as a categorical variable and recorded in five groups. the first group consists of the respondents (incomel) with incomes of less than pkr 20,000, the second group (incomelm) of pkr 20,000-40,000, the third group (incomem) of pkr 40,000-80,000, the fourth group (incomeum) of pkr 80,000-150,000 and the fifth income group (incomeh) of pkr 150,000 and above. the first group, incomel, is kept as the reference category. freshwater-grown vegetables’ use (fwgv_use) shows the respondents’ answers about how much they have been using fwgv before this survey. three dummy variables have been created. the first category (fwgv_userare) is not included in the model and is kept as the reference category. the other two categories are fwgv_useoccasional and fwgv_useoften. awareness about wastewater-grown vegetables’ irrigation (awarenesswwi) reveals how much the respondents are aware of the wastewater irrigation of vegetables in quetta city. it is also included as a categorical variable. two options were included in the questionnaire to know the awareness of the respondents about wastewater-grown vegetables (wwgv) in quetta city and for it the dummy variable, ‘0’ shows that the respondent knows nothing or a little about the wastewater irrigation of vegetables and ‘1’ denotes that he has a fair amount of knowledge or a great deal of knowledge and awareness about the wastewater irrigation of vegetables. employment or occupation of the household head (emphhh) is measured as a dummy variable where ‘1’ = emphhh is a government employee or has his own business and ‘0’ = emphhh has any other type of employment. purchase-place of the vegetables (ppveg) refers to the place where vegetables are mostly bought by the respondent and his family. it is also measured as a dummy variable where ‘1’ shows that salah udin buzdar, syed muhammad khair and syed munawar shah 122 the respondent buys vegetables from convenience shops and ‘0’ denotes that he makes purchases from any other place. area refers to the area of the respondent. it is included as a dummy variable. it is ‘1’ if the respondent belongs to the site where vegetables are mostly grown by using wastewater (brewery road in this case) and ‘0’ otherwise. 2.4 description of study area quetta is the largest city and the capital of balochistan and covers an area of 2653km2 with a total population of 2,275,699 people and a total number of households are 276,711 of which 148,093 are residing in rural areas and 128,618 are residing in urban areas (population census, 2017). according to the development statistics of balochistan 2015-16, different kinds of vegetables are produced in quetta. these are ladyfinger, tinda, radish, spinach, turnip, broad bean, cabbage, carrot, gourds, pumpkin, cauliflower, beetroot, brinjal, lufa, cucumber, etc. in these vegetables, cauliflower, cabbage, spinach, and leafy salads are produced on a bigger stage by using wastewater in the city. the area selected for this paper is quetta city and the data is collected from 5 different sites. data was collected through convenience sampling techniques with the following two stages. at the first stage, five sub-areas in the city were chosen. then, different sites were selected in each of these sub-areas. in the end, data were collected from the consumers who buy vegetables for their household kitchen. the respondents were approached personally, and their responses were recorded face-to-face via questionnaires. only people above 18 years of age were considered on the assumption that children and under-age would not be able to understand the aims of the research and the questions included in the questionnaire. in a total of 260 individuals were talked to for their opinions. the answers of the 255 individuals were recorded and analyzed in the paper. the 5 of the remaining questionnaires were not included in the sample because of having responses considered as incomplete or outliers. the area-wise distribution of the sample is as follows: saryab road 53; brewery road 51; samungli road 50; kasi road 51; and alamdar road 50. the number of respondents in each area is listed (table 1). table 1: number of respondents interviewed area wise name of area saryab road brewery road samungli road kasi road mari abad total no of respondents 53 51 50 51 50 255 source: authors’ survey, 2018 2.5 questionnaire the data were collected through a semi-structured questionnaire. the questionnaire was made gone through the stages of both the pre-testing and pilot survey. a hypothetical bias which sometimes happens in some contingent valuation method (cvm) studies, is not likely to happen here as most of the respondents are familiar with journal of applied economics and business studies, volume. 4, issue 4 (2020) 117-134 https://doi.org/10.34260/jaebs.446 123 the service in question. on the other hand, the format in which the wtp question is stated is the main factor that affects the outcome of cvm surveys (oerlemans et al., 2016). to avoid these all, an introductory paragraph was added in the beginning so that the respondents could understand the goals of the study and feel comfortable to answer the questions. moreover, each question was also translated into urdu to make it easy for the respondents to understand the questions. the literature in determining consumers’ willingness to pay for various service improvements as chatterjee et al., (2017); xie and zhao, (2017); carlsson and johansson, (2000); mustafa et al. (2014); asim and lohano (2015); khan (2014); coulibaly et al., (2011); boccaletti and nardella, (2000); and yahaya, (2008) are consulted in designing the questionnaire for this study. the popular methods for recording the answers of the respondents are open-ended questions, bidding games, dichotomous choice, payment cards, and others (frew et al., 2003). however, in this paper, both price bids and open-ended questions were used. open-ended techniques give a point estimate of the interviewee’s maximum wtp (frew et al., 2003; chatterjee et al., 2017). 3. results and discussions the results include estimations about the socioeconomic characteristics, perception, and awareness, and consumers’ willingness to pay for fwgv in quetta. the results are based on the data of 255 respondents through the questionnaires collected in the field survey in may-june 2018. the results of socioeconomic characteristics of education, income levels, and occupations of the household heads are given in table 2. in education, most of the respondents (51%) are found to have between 14 and 16 years of education. among the others, 8% of the respondents had no/primary education, 8% had an education level of 18 years or above, and 33% of the respondents had education levels of matric/intermediate. the results show that most of the respondents (60%) have incomes between rs. 20,000 and rs. 80,000. while 14% of the respondents have an income of less than rs. 20,000/and 10% have above than rs. 150,000/-, 27% of the respondents have incomes between rs. 20,000 to 40,000/-, 33% have income between rs. 40,000 to 80,000/and 16% between rs. 80,000 to 150,000. employment categories data show, 64% of the household heads are government employees or having their businesses and 36% had other types of employment. the results of the survey regarding education, income, and employment are given in table-2. table 2: socioeconomic characteristics of the sampled respondents variables options percentages frequency educational levels 1. no education/primary 8% 20 2. matric/intermediate 33% 83 3. 14 yrs. to 16 yrs. 51% 131 4. 18 yrs. or above 8% 21 total 100% 255 monthly hh income 1. less than rs. 20,000/14% 37 salah udin buzdar, syed muhammad khair and syed munawar shah 124 2. rs. 20,000-rs. 40,000 27% 68 3. rs. 40,000-rs. 80,000 33% 83 4. rs. 80,000-rs. 150,000 16% 41 5. rs. 150,000 and above 10% 26 total 100% 255 occupation of hh head 1. salaried worker/own business 64% 164 2. other types of employment (unemployed, street vendor/ small informal business, private employee, and others) 36% 91 total 100% 255 source: authors’ survey, 2018 the mean age of the respondents was found to be 30.8 years and the standard deviation is 10.3 years. the minimum age was 18 years and the maximum age was 65 years. the higher mean age is because children were not included in the survey. the mean household size was almost 10 persons and the standard deviation was 4.27. the minimum household size was 2 persons and the maximum number of persons living in a household was found 24 in the survey. further explanation of age and household size is given in table 3. table 3: sample averages of age and household size source: authors’ survey, 2018 3.1 consumer’s ranking of the food items based on usage figure-1 shows the results of the rankings of the main food items from the respondents. figure 1: consumers’ ranking of the food items according to their usage source: authors’ survey, 2018 2 6 .7 0 % 5 6 .1 0 % 1 4 .5 0 % 2 .7 0 % 3 6 .0 0 % 3 0 .2 0 % 2 9 .0 0 % 4 .8 0 % 3 1 .4 0 % 1 0 .2 0 % 3 5 .0 0 % 2 3 .4 0 % 5 .9 0 % 3 .5 0 % 2 1 .5 0 % 6 9 .1 0 % m e at v e g e ta b l e s p u l s e s ot h e r s 1st 2nd 3rd 4th options age (yrs.) hh size mean 30.76 10 s.d. 10.2759 4.27 min value 18 2 max value 65 24 journal of applied economics and business studies, volume. 4, issue 4 (2020) 117-134 https://doi.org/10.34260/jaebs.446 125 in the survey, some 56% of the respondents ranked vegetables as their main food item, 30% ranked vegetables at 2nd, 10% ranked vegetables at 3rd and 3.5% ranked vegetables at 4th in their preferences in food items consumptions. the meat usage preferences are shown by 26.7% of the households, pulses preferences are shown by 14.5% of the respondents in the survey (figure 1). and others 2.7%. 3.2 purchase-place of vegetables the respondents were asked in the survey about their usual and regular buying-place of vegetables. some 52% of the respondents buy vegetables from convenience shops, followed by 24% in vegetable markets, 23% in particular-shops, and only 1% from farm places. the high percentage of respondents buying vegetables from convenience shops might be because there is no well-established vegetable market in quetta. moreover, the prices of vegetables are generally high in established shops as compared to roadside sellers. the results are shown in figure-2. figure 2: purchase-place of vegetables source: authors’ survey, 2018 3.3 diseases recorded in last one year in the families of the respondents as it is clear from singh et al., (2010), ogweugbu and muhanga, (2005), noolan, (1983) bluementhal and peasy (2002) that wwgv cause diarrhea, kidney, and lungs related diseases, cancer, weakening of growth in children and deficiency of blood. the survey results show diarrhea as the most frequent disease faced by the families of the respondents and 50% of the respondents had cases of this disease in their families followed by 35% of the respondents who reported kidneys-related diseases and 32% of the respondents have a report of weak growth in children. among others, 25% of the respondents stated lung-related diseases and 24% deficiency of blood in their family members in the last year (figure 3). cancer was found in only 7% of the households in the survey. however, there are certainly many other factors that cause these diseases. 0% 10% 20% 30% 40% 50% 60% vegetable market (sabzi-mandi) convenience shop a particular shop farm-place 24% 52% 23% 1% salah udin buzdar, syed muhammad khair and syed munawar shah 126 that’s why it is difficult to separate the role which wwgv plays in causing these diseases. figure 3: diseases recorded in last one year in the families of the respondents source: authors’ survey, 2018 3.4 perception and awareness awareness about wastewater-grown vegetables (wwgv) the survey results revealed that most of the respondents were very much aware of the health hazards and negative impacts of the wwgv in quetta city. when they were asked as to how much they saw, heard, or read about wwgv in the city, the data show that 21% of the respondents have a great deal of knowledge about wwgv while 56% have a fair amount of information, 20% has a little information and only 3% of the respondents know nothing about the wastewater irrigation in vegetables’ production (table 4). table 4: awareness about wastewater-grown vegetables in quetta options percentages frequencies nothing 3% 7 a little 20% 52 a fair amount 56% 142 a great deal 21% 54 total 100% 255 source: authors’ survey, 2018 freshwater-grown vegetables (fwgv)’s use of the respondents the respondents were asked about how often they have been using freshwater-grown cauliflowers, spinaches, or salads before this survey to empirically know the perception of the consumers regarding the current state of vegetables. the data show that only 4% 35% 25% 50% 7% 32% 24% k i d n e y r e l a t e d l u n g s r e l a t e d d i a r r h e a c a n c e r w e a k g r o w t h i n c h i l d r e n d e f i c i e n c y o f b l o o d journal of applied economics and business studies, volume. 4, issue 4 (2020) 117-134 https://doi.org/10.34260/jaebs.446 127 of the respondents stated that they always use the three types of fwgv, 36% of the respondents often use three types of fwgv of such vegetables, 50% of the consumers revealed that they have been occasionally using freshwater-grown cauliflowers, spinaches or salads, and 10% of the respondents have disclosed that they have been rarely using these three types of fwgv (table 5). the results show that most of the consumers know or at least think that the vegetables they are using are not grown in freshwater. but they still use such vegetables because they are left with no option other than this. table 5: freshwater-grown cauliflowers, spinaches and salads use of the respondents options percentages frequencies rarely 10% 26 occasionally 50% 128 often 36% 92 always 4% 9 total 100% 255 source: authors’ survey, 2018 3.5 consumers’ wtp the results reveal that consumers are willing to pay for freshwater-grown and healthy vegetables. out of a total of 255 respondents, 222 respondents (87%) are found willing to pay extra for cauliflowers and spinaches each, and 224 representing (88%) of the respondents for salads only. they were ready to pay extra amounts to enjoy fwgv. on the other hand, only 13% and 12% of the respondents responded that they are not ready to pay any extra amount for these vegetables, respectively. the reasons for not paying an extra amount for fwgv are shown in the table 6, 15% of the respondents said that they are satisfied with whatever existing state of vegetables, and 6% of the respondents said that they did not use fwgv vegetables. on the other hand, 58% of the respondents expressed that they could not afford to pay any extra for fwgv and 21% of them said they just do not want to pay extra for fwgv (table 6). table 6: reasons for non-willingness to pay options frequency percentages satisfied with the existing state of vegetables 5 15% cannot afford 19 58% do not want to pay 7 21% do not use these vegetables 2 6% source: authors’ survey, 2018 in the table-7, the responses of those who were willing to pay extra for fwgv are described. the respondents were willing to pay an extra amount of pkr. 13.71 on the average for cauliflowers/cabbages over the basic price. in this way, the mean wtp for spinaches, and salads were shown about pkr. 13 and pkr. 17, respectively. the standard deviation was found to be 8.48, 7.91, and 10.2 for cauliflowers, spinaches, and salads, respectively. the maximum value for wtp was found to be pkr. 35 for salah udin buzdar, syed muhammad khair and syed munawar shah 128 cauliflower and spinaches each, and pkr. 45 for salads. skewness was found to be ≤10% (less than 10%) in wtps of all the vegetables. on analysis, the results in table-7 show that salads are the most used form of vegetable. fwgv of salad got the highest average extra bid price (i.e. pkr 17 extra on its unit price of pkr 40) and the least number of non-responses and the highest individual payment amount. in other words, the consumers are willing to pay 43% more of the basic price if fwg salad is made available to them. similarly, the average extra amount of pkr. 13.71 and pkr. 13 on each kilogram and bundle for cauliflowers and spinaches, respectively, are shown in data and the results prove that consumers are ready to purchase fwgv if prices are higher than the non-fwgv for consumption (table 7). table 7: descriptive statistics of the wtp for different vegetables in pkrs options wtp for cauliflower/cabbage wtp for spinach wtp for salad average 13.71 12.94 16.84 s.d. 8.48 7.91 10.20 max value 35 35 45 min value 0 0 0 skewness(a) 0.075 0.104 0.049 (a): the value of skewness lies between 0 and 1. source: authors’ survey, 2018 3.6 factors affecting consumers’ wtp for freshwater-grown vegetables the estimates of the parameters for the wtp of the consumers on different socioeconomic and other factors were acquired by estimating the multiple linear regression (mlr) model specified in equation 5. the results were obtained to know the impact of different factors (independent variables) on the consumers’ willingness to pay for fwgv (dependent variable). the results were obtained separately for each of the three vegetable types and are given in table-8. before going for the analysis, some diagnostic tests were performed and most of the tests’ statistics (jarque-bera test, shapiro-wilk w test, ramsey reset test, breusch-pagan/cook-weisberg test, and vif) confirm the usage of mlr for estimating the parameters of the interest of this study. the results of the paper show that higher levels of education, all categories of income, fwgv’s use, awareness about wwgv’s irrigation, and the employment of householdhead have significant effects on the consumers’ wtp for fwgv, and education levels of less than 16 years and area were found insignificant for all the three types of fwgv vegetables. respondents’ age, household size, and purchase-place of the vegetables were found significant in the case of salads and insignificant in the case of cauliflowers and spinaches. in the first model given in table-8 where the dependent variable is wtp for cauliflowers, the r2 is 67.70% and the adjusted r2 is 65.82%. in the second model where, the dependent variable is wtp for spinach, the r2 and the adjusted r2 of the model are 67.90% and 66.03%, respectively. in the third and final model, the dependent variable is wtp for salad. the r2 is 74.75% and the adjusted r2 is 73.28%. these values for r2 and adjusted r2 show that all the independent variables are combinedly journal of applied economics and business studies, volume. 4, issue 4 (2020) 117-134 https://doi.org/10.34260/jaebs.446 129 able to explain significant portions of the variations in the dependent variables (wtp for cauliflower, spinach and salad). the f-statistics of 35.94, 36.27 and 50.75 for the three models are also satisfactory. moreover, all the f-statistics are significant at 1% which assures the overall significance of the models by rejecting the hypothesis that the coefficients of all the independent variables are zero (table 8). 3.7 discussion on the estimated parameters of mlr there is a positive relationship found between wtp and education levels. wtp increases as the level of education increases. on average, the consumers with the edu1416yrs and above are willing to pay extra/more amounts for fwgv as compared to the edulow for all the three types of fwg vegetables, with the exception that consumers with edu14-16yrs are insignificant in all the three fwg vegetables models. this positive relationship of education with the willingness to pay is in line with other studies. as, mustafa et al (2014) mention that people having some level of education are willing to pay about 25% more than those who have no or less education. similarly, respondents with high levels of education have a higher wtp (chatterjee et al., 2017; and coulibaly et al. 2011). age was found to be insignificant in the case of cauliflowers and spinaches while the significant in case of salads. its coefficients range from -0.02 to -0.06. it is also affecting the willingness to pay inversely against the expected positive relationship. the negative coefficients show that younger people are more cautious about their health than older people. but, it is still in accordance with some studies. the relationship of age with wtp was reported differently in the literature. age was found significant and negative by xie and zhao (2017) and chatterjee et al. (2017) but insignificant and positive by boccaletti and nardella, (2000). household size has a negative coefficient, as expected before the survey, for all the vegetables and is insignificant in cauliflowers and spinaches but significant in salads. its coefficients go from -0.1 to -0.14. the negative sign shows that large households are willing to pay less amount than small households. this can be attributed to the fact that larger households are likely to have low per head incomes and, also, their consuming quantities are higher and, hence, have to pay bigger portions of their incomes. the results are in line with the studies of xie and zhao (2017) and coulibaly et al. (2011). income was found one of the most effective explanatory variables in this study. respondents with the level of monthly household incomes between pkrs 20,000 and 40, 000 (incomelm) have, on average, a higher mean of wtp of 4.20, 3.86, and 5.63 for cauliflowers, spinaches, and salads, respectively, as compared to those who have monthly incomes of less than pkr 20,000 (lowest income group). the difference grows larger with an increase in income groups. the respondents in the category of the incomeh are, on average, shown willing to pay pkrs 18.55, 17.20, and 24.35 more than the reference category (i.e. lowest income group) for cauliflowers, spinaches, and salah udin buzdar, syed muhammad khair and syed munawar shah 130 salads, respectively. the positive coefficients of income indicate that consumers with higher incomes are willing to pay more amount than those who have lower incomes. the reason for higher wtp from high-income groups confirms the affordability hypothesis that they can afford to pay more. the results are also in accordance with the hypothesized relationship between income and willingness to pay. the results confirm the results of the studies of boccaletti and nardella (2000), mustafa et al. (2014), xie and zhao (2017), carson and mitchell (1993), and asim and lohano (2015) for showing a positive relationship between income and wtp. the categories of freshwater-grown vegetables (fwgv)’s use has negative coefficients which are significant at 1% for the wtp of all the three vegetables. fwgv_userare was kept as the reference category. the negative coefficients of the other two categories (fwgv_useoccasional and fwgv_useoften) show that if a respondent is already using (or he thinks that he is using) fwg vegetables, then he pays the normal price and may not be liable to pay an extra amount. the more he is already using fwgv, the less he is likely to pay an extra amount for fwgv. if he rarely uses such fwg vegetables, he would be ready to pay more to get these vegetables and if he is always using such vegetables, he would be willing to pay no or less amount for such fwg vegetables. in this case (say), he is already safe from the negative impacts of wastewater-grown vegetables, and hence, no need to pay more. no use of this variable was found in the literature. it was included in the model after testing it in the pilot study, and the results are in accordance with the variables’ hypothesized inverse relationship with willingness to pay. the awareness about wastewater-grown vegetables (wwgv) is the other most important factor after income affecting the willingness to pay of the consumers. the awareness and knowledge about wwgv were hypothesized to be positively affecting the consumers’ willingness to pay. it is highly significant and has positive coefficients of 4.52 in cauliflower, 4.17 in spinaches, and 6.03 in salads. the consumers who had a great deal of knowledge about wwgv were willing to pay more while those who know nothing or have little knowledge were willing to pay less. the consumers with greater knowledge and awareness are aware of the negative impacts of wwgv. they are likely to pay more because they think that the opportunity cost of negative impacts is higher than the amount they pay for freshwater-grown vegetables. the results of the paper are in line with coulibaly et al. (2011) and boccaletti and nardella (2000). the employment categories of household heads showed a positive coefficient of 2.20 in cauliflower, 1.62 in spinach, and 2.18 in salads and are significant at 1% in cauliflower and salads and 5% in spinach. it implies that government employees or business owner is willing to pay more than those who have any other type of employment. job security attributes of government employee and higher-income earning expectations of a business owner may prompt consumers to pay extra for fwg vegetables in quetta city. on the contrary to that, household heads who are unemployed or having an insecure job journal of applied economics and business studies, volume. 4, issue 4 (2020) 117-134 https://doi.org/10.34260/jaebs.446 131 as a private employee, and unskilled labor on daily wages tend to pay less for fwg vegetables as their income echelons could not afford extra payment for fwgv in quetta city. the results are in line with the results of yahaya (2008). the other two variables included in the model as explanatory variables are the purchase place of the vegetables and the residence of the respondents. purchase-place of vegetables is only significant in salads, at 10%, and insignificant in the other two vegetables. the results of residence of the respondents are found insignificant in all the three-vegetable-based mlr models (table 8). table 8: results of the parameter estimates for the wtp of freshwater-grown vegetables variable names cauliflower spinach salad coefficient t (p>|t|) coefficient t (p>|t|) coefficient t (p>|t|) edu14-16yrs 1.12 1.57 (.117) 0.94 1.41 (.160) 1.08 1.42 (.157) edu16+yrs 3.70*** 2.83 (.005) 4.37*** 3.61 (.000) 5.64*** 4.07 (.000) age -0.04 -1.06 (.291) -0.02 -0.61(.542) -0.06* -1.76 (.08) hhs -0.10 -1.17 (.243) -0.10 -1.35(.179) -0.14* -1.71 (.088) incomelm 4.20*** 3.98 (.000) 3.86*** 3.93 (.000) 5.63*** 5.01 (.000) incomem 9.43*** 8.57(.000) 8.96*** 8.77 (.000) 12.55*** 10.73 (.000) incomeum 13.50*** 10.31 (.000) 12.83*** 10.55 (.000) 17.25*** 12.4 (.000) incomeh 18.55*** 12.57 (.000) 17.20*** 12.52 (.000) 24.35*** 15.51 (.000) fwgv_useoccasional -3.65*** -3.32 (.001) -2.92*** -2.86 (.005) -3.60*** -3.08 (.002) fwgv_useoften -4.76*** -3.86 (.000) -3.90*** 3.40 (.001) -4.68*** -3.58 (.000) awarenesswwi 4.52*** 4.82 (.000) 4.17 4.78 (.000) 6.03*** 6.04 (.000) emphhh 2.20*** 2.97 (.003) 1.62** 2.36 (.019) 2.18*** 2.78 (.006) ppveg 0.60 0.9 (.371) 0.93 1.51 (.132) 1.46** 2.08 (.039) area -0.46 -0.58 (.565) -0.95 -1.29 (.199) -1.25 -1.48 (.14) constant 5.20*** 2.67 (.008) 4.37** 2.42 (.016) 5.44*** 2.63 (.009) r2 67.70% r2 67.90% r2 74.75% adjusted r2 65.82% adjusted r2 66.03% adjusted r2 73.28% f (14,240) 35.94 f (14,240) 36.27 f (14,240) 50.75 prob. > f 0.000*** prob. > f 0.000*** prob. > f 0.000*** no. of observations 255 * significant at 10%, ** significant at 5%, and *** significant at 1% and () shows p-values source: authors’ survey, 2018 4. conclusions to get safe food is getting wide importance in the lives of consumers and many producers have tried to consider the supply of such vegetables in the market. the paper used a survey-based contingent valuation method (cvm) to record the amounts respondents were willing to pay in addition to the original price of the vegetables available in the market to-date. mlr model is used to estimate the factors which affect the wtp for fwgv in quetta city. vegetables, being the most used food item among the consumers in quetta city, are associated with many factors of health, income, employment, freshness, grown with freshwater or wastewater, and another factor to give perceptions and preferences for its wtp. it is shown that consumers are aware of both vegetables grown by fresh and wastewater in quetta city. consumers with higher levels of information about the hazards of wwgv, more awareness, higher education levels, salah udin buzdar, syed muhammad khair and syed munawar shah 132 more earnings and income levels, and younger are more likely to pay an extra amount than the market price for the fwgv in quetta city. while, consumers with lower levels of education, knowing less about the probable health damages of wwgv, less awareness, lower levels of income, and having unskilled jobs or non-governmental jobs on contracts are less likely to pay extra amount for fwgv in quetta city. among the vegetable types of this study, salad has been given more considerations by the explanatory factors of education, age, income, employment categories of government employment, and own businesses as compared to the other two types of vegetables in quetta city. moreover, the results show that consumers have a high demand for fwgv as depicted by their wtp responses. these high demands have great importance for the producers and policymakers. considering this high demand, the city and provincial governments should encourage the production of such vegetables through public awareness campaigns, media, and moral suasion. the government should also look at the quality of the wastewater being used in irrigation under the supervision of biological, chemical, and medical experts and economists. only, in this way, she would be able to know the health hazards related to wastewater-grown vegetables. farmers and vegetable growers should use different treatment options for the wastewater before finally using it for irrigation. the high demand is also a sign of investment opportunities in the production of such vegetables. private or public-sector organizations should start this as a test in the beginning and go forward with a plan. for such policies to be successful, different price options should be tested as the prices might be higher than the current market prices. certificates should be given to both the vegetable producers and sellers who produce and sell freshwater-grown vegetables by the city government so that the consumers could differentiate between wastewater-grown and freshwater-grown vegetables. references asim, s., lohano, h. d., & ahmad, i. (2015). households' willingness to pay for improved tap water services in karachi, pakistan/comments. pakistan development review, 54(4), 507. bigdeli, m., & seilsepour, m. (2008). investigation of metals accumulation in some vegetables irrigated with waste water in shahre rey-iran and toxicological implications. in american eurasian j. agric. environ. sci. birol, e., karousakis, k., & koundouri, p. (2006). using economic valuation techniques to inform water resources management: a survey and critical appraisal of available techniques and an application. science of the total environment, 365(1), 105-122. blumenthal, u. j., & peasey, a. (2002). critical review of epidemiological evidence of the health effects of wastewater and excreta use in agriculture. unpublished journal of applied economics and business studies, volume. 4, issue 4 (2020) 117-134 https://doi.org/10.34260/jaebs.446 133 document prepared for world health organization, geneva, www. who. int/water_sanitation_ health/wastewater/whocriticalrev. pdf. boccaletti, s., & nardella, m. (2000). consumer willingness to pay for pesticide-free fresh fruit and vegetables in italy. the international food and agribusiness management review, 3(3), 297-310. bogale, a., & urgessa, b. (2012). households' willingness to pay for improved rural water service provision: application of contingent valuation method in eastern ethiopia. journal of human ecology, 38(2), 145-154. butt, m., & khair, s. m. (2016). cost of illness of water-borne diseases and willingness to pay for quality water: a case study of quetta. buitems, unpublished thesis carlsson, f., & johansson-stenman, o. (2000). willingness to pay for improved air quality in sweden. applied economics, 32(6), 661-669. carr, r. m., blumenthal, u. j., & mara, d. d. (2004). guidelines for the safe use of wastewater in agriculture: revisiting who guidelines. water science and technology, 50(2), 31. chatterjee, c., triplett, r., johnson, c. k., & ahmed, p. (2017). willingness to pay for safe drinking water: a contingent valuation study in jacksonville, fl. journal of environmental management, 203, 413-421. coulibaly, o., nouhoheflin, t., aitchedji, c. c., cherry, a. j., & adegbola, p. (2011). consumers' perceptions and willingness to pay for organically grown vegetables. international journal of vegetable science, 17(4), 349-362. ensink, j. h., mahmood, t., van der hoek, w., raschid-sally, l., & amerasinghe, f. p. (2004). a nationwide assessment of wastewater uses in pakistan: an obscure activity or a vitally important one? water policy, 6(3), 197-206. frew, e. j., whynes, d. k., & wolstenholme, j. l. (2003). eliciting willingness to pay: comparing closed-ended with open-ended and payment scale formats. medical decision making, 23(2), 150-159. government of balochistan. development statistics of balochistan 2015-16. http://balochistan.gov.pk/index.php?option=com_docman&task=cat_view&gid =1686&itemid=677 gupta, n., khan, d. k., & santra, s. c. (2008). an assessment of heavy metal contamination in vegetables grown in wastewater-irrigated areas of titagarh, west bengal, india. bulletin of environmental contamination and toxicology, 80(2), 115-118. järup, l. (2003). hazards of heavy metal contamination. british medical bulletin, 68(1), 167-182. khan, g., khan, m. jawad, & khan, m. ayaz. (2018) willingness to pay for improved solid waste management in quetta. buitems, unpublished thesis salah udin buzdar, syed muhammad khair and syed munawar shah 134 kakar, s. r., wahid, a., tareen, r. b., kakar, s. a., tariq, m., & kayani, s. a. (2010). impact of municipal wastewater of quetta city on biomass, physiology and yield of canola (brassica napus l.). pak. j. bot, 42(1), 317-328. khalil, s. (2011). agricultural use of untreated urban wastewater in pakistan. asian journal of agriculture and rural development, 1(1), 21. khan, h. (2014). estimating elasticities of demand and willingness to pay for clean drinking water: empirical evidence from a household level survey in northern pakistan. water and environment journal, 28(1), 145-152. khan, s., cao, q., zheng, y. m., huang, y. z., & zhu, y. g. (2008). health risks of heavy metals in contaminated soils and food crops irrigated with wastewater in beijing, china. environmental pollution, 152(3), 686-692. lagerkvist, c. j., hess, s., okello, j., & karanja, n. (2013). consumer willingness to pay for safer vegetables in urban markets of a developing country: the case of kale in nairobi, kenya. the journal of development studies, 49(3), 365-382. muchuweti, m., birkett, j. w., chinyanga, e., zvauya, r., scrimshaw, m. d., & lester, j. n. (2006). heavy metal content of vegetables irrigated with mixtures of wastewater and sewage sludge in zimbabwe: implications for human health. agriculture, ecosystems & environment, 112(1), 41-48. mustafa, u., ahmad, i., & ul haq, m. (2014). capturing willingness to pay and its determinants for improved solid waste management (no. 2014: 110). pakistan institute of development economics. nicholson, w., & snyder, c. (2011). microeconomic theory: basic principles and extensions. nelson education. nolan, k. r. (1983). copper toxicity syndrome. journal of orthomolecular psychiatry, 12(4), 270-282. ogwuegbu, m. o. c., & muhanga, w. (2005). investigation of lead concentration in the blood of people in the copperbelt province of zambia. j environ, 1, 66-75. pakistan bureau of statistics. population census, 2017. retrieved on 17th september 2018. http://www.pbscensus.gov.pk/ singh, a., sharma, r. k., agrawal, m., & marshall, f. m. (2010). risk assessment of heavy metal toxicity through contaminated vegetables from wastewater irrigated area of varanasi, india. tropical ecology, 51(2), 375-387. xie, b. c., zhao, w., & liu, l. q. (2017, june). willingness to pay for green electricity based on contingent valuation method: case study of tianjin, china. in meeting the energy demands of emerging economies, 40th iaee international conference, june 18-21, 2017. international association for energy economics. yahaya, i. (2008). measuring consumers’ willingness to pay for “safer” vegetables in urban and peri-urban ghana (doctoral dissertation). journal of applied economics and business studies, volume. 2, issue 1 (2018) 1-18 https://doi.org/10.34260/jaebs.211 1 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x household food demand in pakistan: a provincial analysis sajjad1*, zahoor ul haq2 & javed iqbal31 1 assistant professor of economics at higher education department, khyber pakhtunkhwa. 2 professor, department of economics, pakhtunkhwa economic policy research institute, abdul wali khan university mardan, kp, pakistan 3 assistant professor, department of economics, abdul wali khan university mardan abstract in developing countries of the world, the pattern of food consumption is usually used as a benchmark for measuring the living standard of people. this paper examines and compares the food expenditure pattern of households across provinces in pakistan using 2011-12 household integrated economic survey (hies) data. the study used the linear approximate almost ideal demand system (la-aids) to estimate food demand elasticities. the patterns of food consumption and expenditure for eight food groups: wheat flour, rice, dairy, pulses, meats, fruits and vegetables, cooking oil and other foods are examined. the study observed differences in household consumption patterns across provinces. the estimated results show that all food groups have negative own price elasticities and are consistent with economic theory. all the expenditure elasticities are positive and significant indicating that all food groups are normal. the expenditure elasticities estimated show that dairy and meats are luxury foods in all provinces, while wheat flour, pulses, cooking oil, and other foods are necessities in the diet of the pakistani households. in all provinces, a household spends most of its food expenditure on dairy products, wheat flour, cooking oil, fruit and vegetables. the estimates suggest that policymakers in pakistan should ensure and monitor the availability of these essential food items in order to reduce undernourishment and food insecurity in the country. keywords food demand; elasticities; laaids; pakistan jel classification d11; d12; q10; r22 copyright © 2018 saebr all rights reserved 1. introduction food is a basic need of life and essential for the existence of every human being. a balance diet helps in maintaining good health. meeting basic needs such as food remain a major priority of the poor population and they spend a larger portion of their money on it (begum et al., 2010). majority of the undernourished and poor people live in the developing world. in a recent estimate, 821 million people worldwide live in chronic hunger, in which 1*corresponding author: sajjadeconomist2013@gmail.com open access sajjad, zahoor ul haq & javed iqbal 2 515.1 million reside in asian countries (fao, 2018). people in pakistan also suffers from severe hunger. pakistan ranks 106th out of 119 qualifying countries, lagging behind bangladesh and neighbouring india (ghi, 2018). in pakistan, about 60 percent of the population is food insecure, with highly prevalent undernourishment. in addition, 44% of children under the age of 5 are stunted, chronically malnourished, and 15% are acutely malnourished (usaid, 2018). pakistan has made considerable development in getting higher per capita availability of all important food commodities such as meat, cereals, eggs, milk and sugar. the calorie intake of an average pakistani has increased from 2078 calories in 1949-50 to 2450 calories in 2012-13. however, the problem is that half the country's population is still unable to meet its required caloric intake (malik et al., 2014; 2015). in pakistan, many households consume insufficient quantities of calories, protein and other nutrients. rather than a balanced diet, households consume more high-energy food products. furthermore, households have limited dietary diversity, as only dairy products and wheat account for 40% of food expenditure on average. the lowest income quintile households spend most of their food expenditure on wheat, while top income quintile households spend more of their income on meat and dairy products (karim et al., 2018). in developed countries, people spend more of their money on non-food items and services, whereas, in developing countries, households spend a major portion of their income on food items. people of developed countries such as canada, holland and usa spent 13.7, 14.4 and 10.4 percent of their income on food items, respectively; however, in developing countries like india, philippines, sudan, more than 50 percent of household money is spent on food items (begum et al., 2010). the situation in pakistan is not different, and the households spend 37.05 percent of its income on food (gop, 2016). in the last few decades, the analysis of consumer demand has attracted attention both in pakistan and internationally. in pakistan, unprecedented poverty, hunger and undernourishment have revived interest in estimating food demand. although, household consumption patterns in pakistan have been investigated in a number of studies (burney and khan, 1991; aziz et al., 2011; begum et al., 2012; haider and zaidi, 2017; irfanullah et al., 2018), still very little is known about the province wise comparison of complete food demand in pakistan. quantifying household responses to income and price changes requires careful econometric analysis of the patterns of household consumption. most of the previous studies estimated food demand for a single province or food demand for the country's rich and poor households. this study compares the consumption patterns of households by estimating the marshallian, hicksian, and expenditure elasticities in four pakistani provinces. the study uses linear approximate almost ideal demand system (la-aids) for the estimation of elasticities. the analysis includes eight food commodity groups: (i) wheat flour; (ii) all kinds of rice; (iii) dairy (iv) pulses (v) meats (beef, mutton, fish and chicken); (vi) fruits and vegetables; (vii) ghee and oil; (viii) other foods. journal of applied economics and business studies, volume. 2, issue 1 (2018) 1-18 https://doi.org/10.34260/jaebs.211 3 2. literature review an important element of any type of research work is the literature review. it guides researchers in finding a gap in the relevant area to be investigated. this section of the paper examines some important studies already conducted on food consumption and household expenditure pattern. dynamic almost ideal demand system (aids) was used by fan et al. (1994) to investigate patterns of food demand for china’s rural households using household survey data. they found that the estimated expenditure elasticities of all food groups have positive signs indicating that the food items used by household are normal. furthermore, the estimated expenditure elasticity showed that wheat, grain and rice were necessary food items in household consumption pattern, while meat, tobacco, vegetables and alcohol were luxury items. the result also indicated that the demand for these food items is also increasing with the increase in population and per capita income of the household. in order to avoid food shortages, the study proposed that china should increase its food production. jiang and davis (2007) used the la-aids model with a three-stage budgeting procedure to study food consumption patterns of rural household in jillin province of northern china, using household survey data. in particular, the study was conducted to investigate the demand for livestock products. the data were divided into four groups of food commodities, namely grain, animal products, vegetables and other food products, for estimation purposes. they estimated the expenditure elasticities of the food groups and mainly focused on animal products. the result showed that grain expenditure elasticity was 0.64, indicating that household demand for grain will continue to grow because of the increase in per capita income in that region. furthermore, the results indicated that the conditional and total expenditure elasticity values of animal products were 1.22 and 0.76 respectively. haq et al. (2011) analyzed household food demand pattern in pakistan's punjab province using data from the 2004-2005 household income and expenditure survey. food products were categorized into eight groups. they found that households with farming profession consumed more wheat and less of all other products. furthermore, households with an educated head consumed additional food in rural and urban areas of punjab, except wheat and vegetables. the estimated expenditure elasticities had a positive sign, indicating that all food groups are normal. the study concluded that demand was inelastic for all food groups, but found that demand for wheat was more inelastic in both regions. alexandria, pauna & luca (2015) used household budget survey data for estimating food demand in romania using the almost ideal demand system (aids). they estimated food demand elasticities for urban and rural areas of romania. they found that all the estimated own-price elasticities had expected negative signs. the study concluded that the sajjad, zahoor ul haq & javed iqbal 4 estimated expenditure elasticities in rural areas were higher than in urban areas, mainly because of the low cash income of rural people. chen et al. (2016) carried out a meta-analysis for the estimation of food and agricultural elasticities in china. the study found that, as per capita increases, the expenditure elasticities of many food products decreases. one of the possible reasons for this may be economic development, which has increased food supply chains, giving people more choices regarding food products that lead to more substitution. the estimated results suggested that china's demands for dairy and meat and, in turn, for animal feed will continue to grow strongly. they also suggested that the demand for these food products be monitored to ensure food security, especially due to the growing population and tight domestic food supplies in china. aziz et al. (2017) studied own price and income elasticities of various consumption quintiles for urban and rural households in pakistan using household survey data. they estimated the parameters of aggregate food commodity groups using the linear approximate almost ideal demand system (la-aids). the study found that households increased vegetable, fruit, milk, and meat consumption with higher incomes. the result also showed that the elasticity of expenditure in urban areas is smaller than in rural areas and that expenditure on most food groups increases at a decreasing rate as household income increases. the estimated expenditure elasticities for all food groups were less than one, with the exception of meat, fruits and milk with elasticities greater than one. the marshallian own price estimated elasticities for all food groups were found negative and less than one with the exception of meat greater than one indicating elastic demand for meat. the study examined the values of cross price elasticities and found substitution relationships for all the selected food groups. haider and zaidi (2017) studied changes in consumption patterns of household in pakistan by dividing food expenditures into eleven composite groups. the study used household income expenditure survey (hies) from 2000-01 to 2013-14 and used quadratic almost ideal demand system (quaids) for empirical estimation. in addition to differences in calories and consumption bundles, the study also estimated household's response to changes in income and prices. empirical results indicated that the patterns of food consumption vary across regions as well as between provinces. in spite of the increase in food supply availability and per capita income increase, per adult equivalent average calorie intake in the country remains below 2350 kcal. furthermore, 30% of children under the age of 5 were found underweight and 45% of children in pakistan were stunted. they concluded that if the present scenario persists, it can increase the vulnerability of poverty, the burden of diseases throughout the country and low levels of productivity. irfanullah et al. (2018) studied household food consumption decision making in pakistan using the 2011-12 household integrated economic survey and the la-aids model for journal of applied economics and business studies, volume. 2, issue 1 (2018) 1-18 https://doi.org/10.34260/jaebs.211 5 estimation. they investigated the effects of price and income changes on food demand in pakistan. the study found that the marshallian own price elasticities of milk, fruits, meat and rice are more elastic as compared to other groups. the study concluded that the imposition of taxes on personal income of the household could reduce the consumption of food. the study suggested that prices of essential food groups be maintained at reasonable in order to protect low and middle income groups from food insecurity. 3. research methods the almost ideal demand system (aids) model introduced by deaton and muellbauer (1980) is used for this study. the aids model has several advantages over the translog and the rotterdam models such as it satisfies exactly the axioms of choice, simple to estimate, and testing the empirical validity of the restrictions of symmetry and homogeneity is also easy. its functional form is consistent with known household budget data. although, both the translog and rotterdam models have some of these properties too but neither of them possesses all the theoretical properties (deaton & muellbauer, 1980a). deaton and muellbaur (1980a,b) specified the following demand equation in budget share form for la-aids. 𝜔𝑖=𝛼𝑖+ ∑𝑗 𝑦𝑖𝑗 ln 𝑝𝑗 + 𝛽𝑖ln ( 𝑥 𝑝 ) (1) where 𝜔𝑖 is the budget share of good 𝑖, 𝑝𝑗 is the price of good 𝑗, 𝑥 is the expenditure and 𝑝 is price index approximated by the stone’s price index. the following theoretical properties of demand are imposed on equation (1). adding up ∑𝑖 𝛼𝑖 = 1 ∑𝑖 𝛽𝑖 = 0 ∑𝑖 𝑦𝑖 = 0, homogeneity: ∑𝑗 𝑦𝑖𝑗 = 0, and symmetry: 𝑦𝑖𝑗 = 𝑦𝑗𝑖 the parameters of the estimated model are used to derive the elasticities using the following relationships. marshallian: ℰ𝑖𝑗 = 𝑦𝑖𝑗−𝛽𝑖𝜔𝑖 𝜔𝑖 − 𝛿𝑖𝑗 (2) hicksian elasticity: 𝑒𝑖𝑗 = 𝑦𝑖𝑗 𝜔𝑖 + 𝜔𝑗 − 𝛿𝑖𝑗 (3) expenditure elasticity: 𝜂𝑖 = 𝛽𝑖 𝜔𝑖 + 1 (4) where 𝛿𝑖𝑗 is kronecker delta which is one for own price and 0 for cross prices. in order to implement this approach, the study estimates the own price, cross-price and expenditure elasticities of separate food commodity on the basis of household consumption / expenditure data. sajjad, zahoor ul haq & javed iqbal 6 the data used in this paper are derived from the hies 2011/2012 and were used for household food expenditure patterns and elasticities estimation. the hies 2011-12 covers 15807 households that were selected from the urban and rural areas of all provinces of pakistan. the survey adopted a two-stage stratified random sample design for the selection of a household. in the main stage of the survey, 1158 primary sampling units in rural and urban areas of all pakistani provinces were selected. sample 15807 was randomly selected from these primary sampling units in the second stage of the survey. a random systematic sampling system with a random start was selected from each primary sampling unit for 12 or 16 households. the hies survey gathers detailed information about the value and quantity of a variety of food items consumed. the hies collects data on patterns of consumption, income of the household by source, characteristics of household and social indicators. this detailed information enables us to study the budget share of different food items to estimate the la-aids system. 4. results and discussion 4.1 household monthly food expenditures and food budget shares the patterns of food consumption in pakistan have changed over the last two decades. the share of expenditure measures the proportion of income in a specific food group in relation to the total expenditure on food. descriptive statistics in table 1 explains province wise monthly food budget shares of various food items in pakistan. it is observed from estimation that dairy products and wheat flour are the major food consumption groups having average budget share more than 40 percent of household expenditure whereas fruits and vegetables, other food, meats, cooking oil and rice having average budget shares of 13.0, 15.8, 11.8, 11.4, and 4.6 respectively. the high expenditures share of wheat flour and dairy products reflects a high consumption level of these food commodities. table 1 also explains that how consumption and expenditure pattern of the household differ across provinces. per household expenditures on the selected food groups are divided into eight groups. the average expenditures per household has been calculated for each food group as well as for each province in the country. the average food expenditures on aggregate food groups is rs. 9755/month per household in the pakistan. baluchistan province has the highest expenditures on food which is rs.11892 per household per month. the household monthly average expenditure on wheat flour are much higher for the baluchistan province amounting to rs. 2415 reflecting a higher expenditure as compare to other provinces of the country. in the sindh province the expenditure on rice commodity group is much higher which is rs. 681/ month per household. the expenditure on dairy products is higher in punjab province as compare to other provinces, which is rs. 2744. table 1: household monthly average food expenditures and food budget shares in pakistan journal of applied economics and business studies, volume. 2, issue 1 (2018) 1-18 https://doi.org/10.34260/jaebs.211 7 source: author’s calculation using 2011-12 hies data. the figures in parentheses are budget share of food item. * show estimates are statistically significant at 1%. 4.2 monthly average food expenditure by urban and rural household in pakistan descriptive statistics in table 2 explain rural and urban household differences in monthly food expenditure for various food items in pakistan. it is observed that rural household expenditure on wheat flour is higher than urban households in all the provinces. the expenditure on other food groups in urban areas is higher in all provinces except khyber pakhtunkhwa. the expenditure on dairy products in the urban regions of punjab, sindh and baluchistan is higher compared to the rural regions in those provinces. the reason for low expenditure on dairy products in rural areas is the easy availability of dairy products and the price differences between urban and rural areas. households in urban areas consume more meat than rural regions in all provinces. households in urban punjab and baluchistan spend more on fruit and vegetables as compared to other provinces in pakistan. table 2: monthly average food expenditure for urban and rural household in pakistan food group punjab sindh khyber pakhtunkhwa baluchistan pakistan urban wheat flour 1214 1278 1660 2172 1405 rice 371 642 288 412 436 dairy 2896 2620 2210 1688 2576 pulses 231 259 281 375 262 meats 1293 1589 1299 1961 1440 fruits/vegetables 1398 1247 1458 1591 1385 cooking oil 1069 1136 1152 1567 1150 other food 1433 2039 1694 1940 1700 rural food group punjab sindh khyber pakhtunkhwa baluchistan pakistan f wheat flour 1298(14.2 ) 1356(13.8 ) 1804(17.9) 2415(20.3) 1519(15.6 ) 838.50* rice 336(3.7) 681(6.9) 385(3.8) 413(3.5) 445(4.6) 382.61* dairy 2744(30.1 ) 2431(24.8 ) 2277(22.6) 1648(13.8) 2466(25.3 ) 136.73* pulses 222(2.4) 229(2.3) 284(2.8) 360(3.0) 249(2.5) 307.68* meats 1007(11.0 ) 1181(12.0 ) 1107(11.0) 1872(15.7) 1152(11.8 ) 148.52* fruits/vegetables 1237(13.6 ) 1116(11.4 ) 1381(13.7) 1662(13.9) 1273(13.0 ) 216.69* cooking oil 1016(11.1 ) 1077(11.0 ) 1125(11.1) 1627(13.7) 1110(11.4 ) 488.16* other food 1258(13.8 ) 1745(17.8 ) 1723(17.1) 1895 (15.9) 1541(15.8 ) 339.73* total 9118 9816 10086 11892 9755 122.96* no of households 6903 4195 3276 1424 15798 sajjad, zahoor ul haq & javed iqbal 8 wheat flour 1360 1420 1895 2607 1604 rice 309 713 446 413 451 dairy 2632 2275 2320 1616 2383 pulses 215 204 286 348 240 meats 794 844 985 1801 938 fruits/vegetable 1117 1008 1331 1718 1190 cooking oil 977 1029 1108 1674 1081 other food 1128 1502 1742 1860 1423 source: author’s calculation using 2011-12 hies data. 4.3 parameters estimates of the la-aids model the study used the seemingly unrelated regression (sur) of zellner (1963) for the estimation of la-aids equations system. the delta method (stata, 2005) were used for deriving the statistical significance of the estimated elasticities. by imposing the restriction of additivity on the expenditure function makes the problem that the variance and covariance matrix become singular due to which one of the equations of the demand system needs to be dropped. in this study the other food equation was selected for omission. the coefficients that are estimated are given in table 3, 4,5,6,7. the estimated coefficient included in the system are mostly significant at the 95 and 99 percent level of significance. the parameters of the dropped equation (other food) are derived using the condition of adding up. the estimated low values of 𝑅2 are not uncommon when using cross-sectional data. table 3: parameters estimates of the la-aids model for punjab province explanatory variable wheat flour rice dairy pulses meats fruits/ve g oils other log of price of wheat flour 0.087 (0.002)* 0.004 (0.001)* -0.044 (0.005)* 0.004 (0.001)* -0.007 (0.004)*** -0.009 (0.002)* 0.015 (0.002)* -0.051 (0.002)* log of price of rice -0.009 0.011 -0.006 0.000 -0.000 0.004 -0.000 0.001 (0.001)* (0.000)* (0.001)* (0.000) (0.001) (0.000)* (0.000) (0.00)** * log of price of dairy -0.015 0.001 0.038 -0.000 -0.003 -0.007 -0.004 -0.007 (0.001)* (0.001) (0.003)* (.000) (0.002)*** (0.001)* (0.001)* (0.001)* log of price of pulses 0.005 -0.002 -0.010 0.007 -0.006 0.003 0.007 -0.003 (0.001)* (0.000)* (0.002)* (0.000)* (0.002)* (0.001)* (0.001)* (.001)* log of price of meats -0.006 (0.000)* -0.002 (0.000)* -0.006 (0.001)* -0.001 (0.000)* 0.019 (0.000)* -0.001 (0.000)* -0.003 (0.000)* 0.001 (0.000)* log of price of fruits/vegetables -0.010 (0.001)* 0.003 (0.001)* -0.026 (0.003)* 0.004 (0.000)* 0.014 (0.002)* 0.018 (0.001)* -0.001 (0.001) -0.001 (0.001) log of price of oils 0.006 0.000 -0.020 -0.002 -0.007 -0.000 0.026 -0.003 (0.001)* (0.001) (0.003)* (0.000)* (0.002)* (0.001) (0.001)* (0.001)* log of price of other food 0.000 (0.003) 0.001 (0.001) -0.061 (0.005)* 0.005 (0.001)* 0.016 (.004)* -0.002 (0.002) 0.011 (0.002)* 0.029 (0.002)* household size 0.012 0.001 -0.014 0.001 -0.006 -0.001 0.005 0.001 (0.000)* (0.000)* (0.000)* (0.000)* (.000)* (0.000)* (0.000)* (0.000)* household education -0.000 0.000 -0.001 0.000 0.000 -0.000 0.000 0.000 (0.000) (0.000)* (0.000)** (0.000)** (0.000)** (0.000) (0.000)* * (0.000) household age 0.000 0.000 0-.000 0.000 0.000 -0.000 0.000 0.000 (0.000) (0.000) (0.000)* (0.000)** (0.000) (0.000)** (0.000) (0.000) constant 0.453 0.005 -0.061 0.019 -0.317 0.228 0.240 0.431 0.015 (0.007) (0.026)** (0.004)* (0.019)* (0.012)* (0.009)* (0 .012)* number of observations 6903 6903 6903 6903 6903 6903 6903 6903 r-squared 0.431 0.170 0.257 0.188 0.333 0.067 0.402 0.155 journal of applied economics and business studies, volume. 2, issue 1 (2018) 1-18 https://doi.org/10.34260/jaebs.211 9 chi 5239.5* 1415.5* 2390.7* 1599.5* 3449.9* 500.2* 4645.9* 1267.3* source: author’s estimation using 2011-12 hies data. values in parentheses are standard errors. * significant at 1%, ** significant at 5% and *** significant at 10%. sajjad, zahoor ul haq & javed iqbal 10 table 4: estimated parameters of la-aids model for sindh province explanatory variable wheat flour rice dairy pulses meats fruits/veg oils other log of price of wheat flour 0.054 (0.002)* 0.021 (0.003)* -0.014 (0.004)* 0.000 (0.000) -0.008 (0.003)* -0.002 (0 .001) 0.010 (0.001)* -0.017 (0.002)* log of price of rice -0.011 0.019 -0.005 -0.000 -0.004 0.000 0.001 -0.008 (0.001)* (0.001)* (0.002)* (0.000)** (0.001)* (0.001) (0.001)* * (0.001)* log of price of dairy -0.024 0.011 0.026 0.002 0.024 0.000 -0.000 -0.018 (0.002)* (0.003)* (0.004)* (0.000)* (0.003)* (0.000) (0.001) (0.003)* log of price of pulses 0.003 0.003 -0.019 0.009 -0.002 0.003 0.003 -0.010 (0.002)* (0.002) (0.003)* (0.000)* (0.002) (0.001)* (0.001)* (0.002)* log of price of meats -0.008 -0.002 -0.003 -0.000 0.019 -0.003 -0.001 0.001 (0.000)* (0.001)* (0.001)* (0.000)* (0.000)* (0.000)* (0.000)* (0.001) log of price of fruits/vegetables -0.009 (0.002)* -0.012 (0.002)* -0.015 (0.003)* 0.001 (0.000)*** 0.005 (0.002)* * 0.020 (0.001)* -0.004 (0.001)* 0.024 (0.002)* log of price of oils 0.002 0.007 -0.038 -0.001 -0.007 0.004 0.029 -0.007 (0.002)* (0.003)* (0.004)* (0.000)** (0.003)* (0.001)* (0.001)* (0.002)** log of price of other food 0.003 0.004 0.039 (0.005)* -0.028 (0.006)* 0.010 (0.001)* 0.026 (0.004)* 0.005 (0.002)** 0.012 (0.002) 0.020 (0.005)* household size 0.006 0.004 -0.007 0.000 -0.004 0.000 0.002 -0.002 (0.000)* (0.000)* (0.000)* (0.000)* (0.000)* (0.000)*** (0.000)* (0.000)* household education -0.000 0.000 0.000 0.000 0.000 -0.000 0.000 -0.000 (0.000)*** (0.000) (0.000)* * (0.000) (0.000)* ** (0.000) (0.000) (0.000)* household age -0.000 -0.000 -0.000 0.000 0.000 -0.000 -0.000 0.000 (0.000) (0.000)* (0.000)* (0.000)* (0.000)* (0.000)** (0.000)* * (0.000)* constant 0.320 0.221 0.079 0.001 -0.313 0.100 0.140 0.449 (0.017)* (0.021)* (0.027)* (0.003) (0.019)* (0.011)* (0.010)* (0 .021)* number of observations 4195 4195 4195 4195 4195 4195 4195 4195 r-squared 0.363 0.152 0.174 0.184 0.3783 0.119 0.227 0.143 chi 2395.2* 754.3* 882.8* 947.9* 2552.9* 568.8* 1236.4* 698.9* source: author’s estimation using 2011-12 hies data. values in parentheses are standard errors. * significant at 1%, ** significant at 5% and *** significant at 10%. table 5: parameters estimates of the la-aids model for khyber pakhtunkhwa explanatory variable wheat flour rice dairy pulses meats fruits/ve g oils other food log of price of wheat flour 0.064 (0.002)* 0.004 (0.002)** -0.010 (0.003)* 0.001 (0.001) -0.017 (0.003)* -0.000 (0 .002) 0.007 (0.001)* -0.049 (0.002)* log of price of rice -0.008 0.009 -0.003 0.001 0.002 0.002 -0.002 -0.001 (0.001)* (0.001)* (0.001)** (0.000)* (0.001) (0.001)* (0.000)* (0.001)*** log of price of dairy -0.005 0.004 0.021 0.003 -0.000 -0.009 0.001 -0.015 (0.002)** (0.001)* (0.003)* (0.000)* (0.002)* (0.002)* (0.001) (0.002)* log of price of pulses 0.004 (0.001)* -0.003 (0.001)* -0.003 (0.002)** * 0.008 (0.000)* -0.000 (0.001) 0.002 (0.001)** 0.001 (0.001) -0.009 (0.001)* log of price of meats -0.006 -0.002 -0.005 -0.001 0.020 -0.001 -0.000 -0.003 (0.000)* (0.000)* (0.001)* (0.000)* (0.001)* (0.000)** (0.000) (0.000)* log of price of fruits/vegetables -0.003 (0.002) -0.003 (0.001)** 0.002 (0.003) -0.002 (0.001)* 0.018 (0.003)* 0.014 (0.002)* -0.004 (0.001)* -0.021 (0.002)* log of price of oils 0.003 0.000 -0.010 -0.001 -0.006 -0.001 0.028 -0.013 (0.002)*** (0.001) (0.002)* (0.000)* * (0.002)* (0.001) (0.001)* (0.002)* journal of applied economics and business studies, volume. 2, issue 1 (2018) 1-18 https://doi.org/10.34260/jaebs.211 11 log of price of other food 0.032 (0.004)* -0.009 (0.003)* -0.029 (0.005)* 0.002 (0.001)* * 0.009 (.005)** * 0.008 (0.003)** 0.006 (0.002)* * -0.018 (0.004)* household size 0.009 -0.000 -0.006 0.000 -0.004 -0.002 0.003 0.001 (0.000)* (0.000) (0.000)* (0.000)* (0.000)* (0.000)* (0.000)* (0.000)* household education 0.000 -0.000 -0.000 0.000 -0.000 0.000 0.000 -0.000 (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) household age 0.000 -0.000 0.000 0.000 0.000 -0.000 0.000 0.000 (0.000) (0.000)* (0.000) (0.000) (0.000) (0.000)** * (0.000)* * (0.000) constant 0.232 0.009 -0.068 -0.004 -0.161 0.090 0.141 0.760 (0.019)* (0.012) (0.025)* (0.005) (0.024)* (0.015)* (0.010)* (0 .018)* number of observations 3276 3276 3276 3276 3276 3276 3276 3276 r-squared 0.334 0.086 0.125 0.218 0.289 0.056 0.279 0.224 chi 1644.4* 310.2* 467.4* 911.8* 1331.1* 192.9* 1268.1* 946.8* source: author’s estimation using 2011-12 hies data. values in parentheses are standard errors. * significant at 1%, ** significant at 5% and *** significant at 10%. table 6: parameters estimates of the la-aids model for baluchistan province explanatory variable wheat flour rice dairy pulses meats fruits/v eg oils other food log of price of wheat flour 0.050 (0.005)* -0.017 (0.003)* 0.002 (0.006) 0.000 (0.001) -0.004 (0.005) -0.005 (0 .003) 0.008 (0.003)** -0.035 (0.004)* log of price of rice -0.005 0.007 -0.001 0.000 0.003 0.002 0.004 -0.010 (0.001)* (0.001)* (0.001) (0.000) (0.001)** (0.001)* (0.001)* (0.001)* log of price of dairy -0.017 0.003 0.034 0.001 0.002 -0.010 -0.009 -0.004 (0.001)* (0.000)* (0.001)* (0.000) * (0.001)** * (0.001)* (0.001)* (0.001)* log of price of pulses 0.006 (0.006) 0.000 (0.004) -0.041 (0.007)* 0.012 (0.001) * 0.012 (0.007)** * -0.001 (0.004) 0.003 (0.004) 0.008 (0.004)*** log of price of meats -0.005 (0.002)* -0.007 (0.001)* -0.021 (0.002)* 0.001 (0.000) * 0.035 (0.002)* -0.009 (0.001)* 0.002 (0.001) 0.005 (0.001)* log of price of fruits/vegetables 0.028 (0.004)* -0.015 (0.002)* -0.028 (0.005)* 0.003 (0.001) * 0.007 (0.005) 0.012 (0.003)* 0.008 (0.003)* -0.014 (0.003)* log of price of oils -0.006 (0.004) -0.003 (0.002) -0.005 (0.005) -0.000 (0.000) -0.005 (0.004) -0.003 (0.003) 0.025 (0.003)* -0.003 (0.003) log of price of other food -0.016 (0.007)** 0.010 (0.004)* * 0.015 (0.009) -0.006 (0.001) * -0.041 (.008)* 0.033 (0.005)* -0.003 (0.005) 0.008 (0.005) household size 0.007 0.001 -0.006 0.001 -0.003 -0.001 0.001 -0.001 (0.000)* (0.000)* (0.001)* (0.000) * (0.001)* (0.000)* (0.000)* (0.000)** household education 0.000 -0.001 -0.001 0.000 0.001 0.001 0.000 -0.000 (0.000)*** (0.000)* (0.000)* * (0.000) (0.000)** (0.000)* (0.000) (0.000) household age 0.000 -0.000 -0.000 0.000 0.000 -0.000 0.000 0.000 (0.000)* (0.000)* (0.000)* (0.000) * (0.000)** (0.000) (0.000)* (0.000) constant 0.176 0.226 0.0 0.131 -0.078 0.144 0.112 0.250 (0.059)* (0.037)* (0.027)* (0.012) * (0.069) (0.042)* (0.044)** (0 .046)* number of observations 1424 1424 1424 1424 1424 1424 1424 1424 r-squared 0.409 0.213 0.427 0.297 0.207 0.235 0.204 0.188 chi 986.8* 385.7* 1063.0* 601.9* 372.9* 439.5* 364.9* 330.4* sajjad, zahoor ul haq & javed iqbal 12 source: author’s estimation using 2011-12 hies data. values in parentheses are standard errors. * significant at 1%, ** significant at 5% and *** significant at 10%. journal of applied economics and business studies, volume. 2, issue 1 (2018) 1-18 https://doi.org/10.34260/jaebs.211 13 table 7: parameters estimates of the la-aids model for pakistan explanatory variable wheat flour rice dairy pulses meats fruits/ veg oils other log of price of wheat flour 0.070 (0.001)* -0.001 (0.001) -0.041 (0.002)* 0.002 (0.000) * -0.007 (0.002) * -0.004 (0 .001)* 0.010 (0.001) * -0.029 (0.001)* log of price of rice -0.008 0.010 -0.002 0.000 0.000 0.003 0.000 -0.004 (0.000)* (0.000)* (0.001)* (0.000) (0.000) (0.000)* (0.000) (0.000)* log of price of dairy -0.014 0.003 0.034 0.001 -0.001 -0.009 -0.008 -0.004 (0.001)* (0.000)* (0.001)* (0.000) * (0.001) (0.000)* (0.000) * (0.001)* log of price of pulses 0.005 -0.001 -0.009 0.008 -0.002 0.002 0.004 -0.007 (0.001)* (0.001) (0.001)* (0.000) * (0.001) ** (0.001)* (0.000) * (0.001)* log of price of meats -0.007 -0.001 -0.008 -0.001 0.021 -0.002 -0.002 0.001 (0.000)* (0.000)* (0.000)* (0.000) * (0.000) * (0.000)* (0.000) * (0.000)* log of price of fruits/vegetables -0.004 (0.001)* -0.005 (0.001)* -0.019 (0.002)* 0.002 (0.000) * 0.013 (0.001) * 0.019 (0.001)* -0.004 (0.001) * -0.001 (0.001) log of price of oils 0.006 0.001 -0.013 -0.001 -0.008 -0.000 0.027 -0.011 (0.001)* (0.001)** * (0.002)* (0.000) * (0.001) * (0.001) (0.000) * (0.001)* log of price of other food -0.002 (0.002) 0.002 (0.001) -0.031 (0.003)* 0.003 (0.001) * 0.011 (0.002) * -0.002 (0.001) 0.007 (0.001) * 0.010 (0.002)* household size 0.010 0.002 -0.012 0.000 -0.004 -0.001 0.003 0.001 (0.000)* (0.000)* (0.000)* (0.000) * (0.000) * (0.000)* (0.000) * (0.000)* household education -0.000 0.000 0.000 -0.000 0.000 -0.000 -0.000 -0.000 (0.000)* (0.000)** (0.000)* (0.000) (0.000) (0.000)* (0.000) (0.000) household age 0.000 -0.000 -0.000 0.000 0.000 -0.000 0.000 0.000 (0.000)* (0.000)* (0.000)* (0.000) * (0.000) * (0.000)** (0.000) (0.000) constant 0.385 0.032 0.068 0.019 -0.255 0.179 0.199 0.507 (0.009)* (0.007)* (0.017)* (0.002) * (0.012) * (0.007)* (0.005) * (0.009)* number of observations 15806 15806 15806 15806 15806 15806 15806 15806 r-squared 0.368 0.072 0.208 0.152 0.325 0.075 0.296 0.089 chi 9200.6* 1232.1* 4141.5* 2827.9 * 7619.9 * 1296.8* 6654.7* 1558.3* source: author’s estimation using 2011-12 hies data. values in parentheses are standard errors. * significant at 1%, ** significant at 5% and *** significant at 10%. 4.4 estimated expenditure elasticities for various food groups in pakistan consumers generally adjust their consumption in response to price and income changes (mudassar, aziz, & anwar, 2012). the expenditure elasticities for eight food commodity groups consumed by households across provinces are reported in table 8. the expenditure elasticities are positive and significant indicating that all food groups are normal. the estimated expenditure elasticities for all the provinces range from 0.526 for wheat flour to 1.501 for meats. the expenditure elasticities estimated shown that dairy and meats are sajjad, zahoor ul haq & javed iqbal 14 luxury food items in all provinces, while wheat flour, pulses, cooking oil, and other foods are necessities in the diet of pakistani household. our results are similar to what haq et al. (2008) found. the rice is necessity food item in punjab and baluchistan while luxury in sindh and khyber pakhtunkhwa. table 8: expenditure elasticities for various food groups in pakistan source: author’s estimation using 2011-12 hies data. values in parentheses are standard errors. * significant at 1%, ** significant at 5% and *** significant at 10%. 4.5 estimated elasticities marshllian and hicksian's own price and cross price elasticities show the response of consumer to changes in prices. keeping the level of utility constant, the estimated marshallian elasticities of demand describes changes in the quantity of demand as a result of price changes capturing both substitution and income effects, whereas, the estimated hicksian elasticity represents only the substitution effect as a result of changes in prices. all estimated own price elasticities of the marshallian and hicksian demand have the expected negative sign and are statistically significant, indicating that the price of a commodity itself have negative effect on its quantity demanded. the own price elasticity of dairy in all provinces is more elastic as compare to other food groups, showing high responsiveness to changes in price. the study of cross price elasticity is helpful in defining the nature of the food commodity for their substitutability and complementarity. the exact nature of the cross price elasticity for eight food groups for all the provinces of pakistan has also been estimated in this study. if the estimated cross price elasticity for any two food groups are positive, the two commodities are said to be substitute, while both may be called complementary if cross price elasticity have a negative sign. in punjab province, out of the 56 marshallian cross price elasticities, 24 elasticities are negative indicating that these goods are complementary, and the remaining 32 elasticities are positive indicating gross substitutes. on the other hand, out of 56 hicksian cross price elasticities, 6 are negative which are indicating gross complements and the remaining fifty elasticities have positive sign indicating gross substitutes. in sindh province total estimated marshallian cross price elasticities are 56 in which 30 elasticities are negative indicating that these goods are complementary consumer goods, and the remaining 26 elasticities have positive sign indicating gross substitute. on the other hand, out of 56 hicksian cross price elasticities, 7 food group punjab sindh khyber pakhtunkhwa baluchistan pakistan wheat flour 0.526* 0.717* 0.704* 0.903* 0.652* rice 0.884* 1.121* 1.185* 0.672* 0.933* dairy 1.478* 1.365* 1.302* 1.406* 1.411* pulses 0.628* 0.601* 0.945* 0.356* 0.721* meats 1.501* 1.196* 1.231* 1.144* 1.341* fruits/vegetables 0.908* 0.888* 1.012* 0.939* 0.953* cooking oil 0.584* 0.777* 0.742* 0.830* 0.688* other food 0.813* 0.854* 0.933* 1.043* 0.852* journal of applied economics and business studies, volume. 2, issue 1 (2018) 1-18 https://doi.org/10.34260/jaebs.211 15 are negative which are indicating gross complements and the remaining forty nine elasticities have positive sign indicating gross substitutes. in khyber pakhtunkhwa province total estimated marshallian cross price elasticities are 56 in which, 31 elasticities have negative sign indicating that these goods are complementary consumer goods, and the remaining 25 elasticities are positive indicating that these are gross substitute. further, there are total 56 hicksian cross price elasticities, four are negative which are indicating gross complements and the remaining fifty two elasticities are positive indicating that these commodities are gross substitutes. in baluchistan province the total number of estimated marshallian cross price elasticities are 56 in which, 30 elasticities have negative sign indicating that these goods are complementary consumer goods, and the other 26 elasticities are positive indicating that these are gross substitute. finally, there are total 56 hicksian cross price elasticities, nine are negative which are indicating gross complements and the remaining forty seven elasticities are positive indicating that these commodities are gross substitutes. table 9: estimated marshallian and hicksian own-price and cross-price elasticities of demand for punjab food group wheat flour rice dairy pulses meats fruits /vegetable oils others uncompensated wheat flour -0.376 0.018 -0.022 0.112 0.035 0.013 0.116 0.079 rice 0.124 -0.697 0.031 -0.066 -0.057 0.086 0.017 0.021 dairy -0.293 -0.156 -0.999 -0.172 -0.158 -0.230 -0.206 -0.351 pulses 0.177 0.012 -0.011 -0.686 -0.048 0.154 -0.055 0.206 meats -0.122 -0.049 -0.087 -0.087 -0.824 0.114 -0.125 0.143 fruits/vegetables -0.050 0.042 -0.038 0.032 0.005 -0.854 0.009 -0.004 cooking oil 0.176 0.047 0 .014 0.106 0.027 0.040 -0.734 0.142 other food -0.339 0.033 -0.028 -0.000 0.035 0.017 0.000 -0.764 compensated wheat flour -0.291 -0.021 0.185 0.062 0.048 0.077 0.162 0.144 rice 0.281 -0.665 0.310 -0.044 0.027 0.222 0.135 0.158 dairy 0.002 0.015 -0.580 -0.010 0.065 0.045 0.051 -0.075 pulses 0.328 0.038 0.262 -0.670 0.030 0.284 0.057 0.337 meats 0.082 0.030 0.240 -0.045 -0.691 0.298 0.042 0.329 fruits/vegetables 0.097 0.065 0.232 0.045 0.081 -0.727 0.119 0.124 cooking oil 0.286 0.033 0.246 0.081 0.065 0.129 -0.663 0.232 other food -0.224 0.148 0.085 0.114 0.149 0.132 0.114 -0.649 source: author’s estimation using 2011-12 hies data. values in parentheses are standard errors. * significant at 1%, ** significant at 5% and *** significant at 10%. table 10: estimated compensated and uncompensated own-price and cross-price elasticities of demand for sindh food group wheat flour rice dairy pulses meats fruits /vegetable oils other s uncompensated wheat flour -0.590 0.034 0.122 0.134 -0.012 -0.021 0.124 0.060 rice -0.329 0.723 0.176 0.032 -0.044 -0.192 0.102 -0.585 sajjad, zahoor ul haq & javed iqbal 16 dairy -0.148 0.109 0.976 -0.168 -0.099 -0.148 0.247 -0.205 pulses 0.010 0.011 0.129 -0.622 -0.002 0.039 0.037 0.418 meats -0.102 0.016 0.203 -0.037 -0.844 0.026 0.088 0.221 fruits/vegetables -0.009 0.016 0.017 0.043 -0.014 -0.808 0.054 0.062 cooking oil 0.117 0.036 0.024 0.052 0.015 -0.099 0.718 0.043 other food -0.068 0.016 0.075 -0.034 0.021 0.160 0.014 -0.859 compensated wheat flour -0.484 0.008 0.074 0.116 0.053 0.053 0.197 0.198 rice -0.173 0.646 0.070 0.065 0.071 -0.066 0.225 -0.397 dairy 0.086 0.045 0.651 -0.056 0.095 0.055 0.045 0.061 pulses 0.149 0.046 0.358 -0.607 0.095 0.146 0.067 0.588 meats 0.066 0.106 0 .462 0.008 -0.714 0.165 0.046 0.423 fruits/vegetables 0.126 0.071 0.243 0.054 0.080 -0.704 0.156 0.229 cooking oil 0.240 0.078 0 .237 0.051 0.097 -0.007 0.628 0.198 other food 0.085 0.136 0 .078 0 .119 0 .175 0.313 0.138 -0.705 source: author’s estimation using 2011-12 hies data. values in parentheses are standard errors. * significant at 1%, ** significant at 5% and *** significant at 10%. table 11: marshallian and hicksian cross and own-price elasticities of demand for khyber pakhtunkhwa food group wheat flour rice dairy pulses meats fruits /vegetable oils others uncompensated wheat flour -0.599 0.013 0.029 0.074 0.023 0.039 0.071 0.231 rice 0.091 -0.774 0.102 -0.084 -0.069 -0.096 -0.003 -0.250 dairy -0.113 -0.079 -0.966 -0.081 -0.090 -0.058 -0.114 -0.203 pulses 0.040 0.050 0.108 -0.717 -0.065 -0.068 -0.037 0.080 meats -0.193 -0.005 -0.028 -0.024 -0.814 0.158 -0.084 0.066 fruits/vegetables -0.005 0.016 -0.066 0.016 -0.009 -0.898 -0.006 0.053 cooking oil 0.094 0.011 0.035 0.037 0.025 -0.007 -0.723 0.081 other food -0.272 0.001 -0.077 -0.038 -0.009 -0.113 -0.064 -1.095 compensated wheat flour -0.468 -0.003 0.193 0.048 0.067 0.124 0.132 0.348 rice 0.284 -0.730 0.329 -0.049 0.036 0.049 0.119 -0.071 dairy 0.139 0.024 -0.681 0.013 0.074 .147 0.067 0.035 pulses 0.225 0.086 0.325 -0.690 0.032 0.069 0.076 0.251 meats 0.016 0.054 0.213 0.026 -0.692 0.320 0.053 0.262 fruits/vegetables 0.182 0.055 0.154 0.047 0.090 -0.757 0.110 0.227 cooking oil 0.251 0.019 0.224 0.035 0.094 0.102 -0.757 0.223 other food -0.111 0.163 0.084 0.122 0.151 0.047 0.096 -0.934 source: author’s estimation using 2011-12 hies data. values in parentheses are standard errors. * significant at 1%, ** significant at 5% and *** significant at 10%. table 12: estimated cross-price and own-price marshallian and hicksian elasticities of demand for baluchistan food group wheat flour rice dairy pulses meats fruits /vegetable oils others uncompensated wheat flour -0.727 -0.006 -0.064 0.049 -0.007 0.160 -0.009 -0.062 journal of applied economics and business studies, volume. 2, issue 1 (2018) 1-18 https://doi.org/10.34260/jaebs.211 17 rice -0.471 -0.784 0.089 0.019 -0.196 -0.403 -0.079 0.292 dairy -0.039 -0.065 -0.799 -0.361 -0.215 -0.265 -0.096 0.056 pulses 0.027 0.023 0.045 -0.580 0.067 0.105 0.019 -0.182 meats -0.047 -0.003 0.008 0.055 -0.790 0.022 -0.054 -0.294 fruits/vegetables -0.025 0.024 -0.061 0.004 -0.053 -0.910 -0.011 0.238 cooking oil 0.086 0.052 -0.043 0.048 0.037 0.087 -0.791 0.001 other food -0.228 -0.070 -0.034 0.041 0.026 -0.097 -0.027 -0.952 compensated wheat flour -0.547 0.010 0.052 0.063 0.124 0.285 0.111 0.078 rice -0.283 -0.759 0.213 0.039 -0.057 -0.270 0.048 0.440 dairy 0.215 0.026 -0.607 -0.273 -0.009 -0.065 0.098 0.272 pulses 0.206 0.039 0.161 -0.568 0.198 0.229 0.138 -0.043 meats 0.174 0.054 0.149 0 .110 -0.617 0.188 0.107 -0.112 fruits/vegetables 0.166 0.051 0.066 0 .028 0.089 -0.774 0.119 0.389 cooking oil 0.262 0.065 0.069 0.056 0.165 0.208 -0.675 0.137 other food -0.061 0.097 0.132 0.209 0.193 0.069 0.139 -0.785 source: author’s estimation using 2011-12 hies data. values in parentheses are standard errors. * significant at 1%, ** significant at 5% and *** significant at 10%. table 13: compensated and uncompensated cross-price and own-price elasticities of demand for pakistan food group wheat flour rice dairy pulses meats fruits /vegetable oils others uncompensated wheat flour -0.517 0.009 -0.030 0.090 0.014 0.029 0.095 0.045 rice -0.024 -0.769 0.064 -0.018 -0.031 -0.105 0.030 0.047 dairy -0.271 -0.109 -0.958 -0.139 -0.135 -0.179 -0.156 -0.227 pulses 0.100 0.013 0.032 -0.680 -0.041 0.089 -0.040 0.128 meats -0.106 -0.031 -0.043 -0.059 -0.827 0.096 -0.120 0.083 fruits/vegetables -0.024 0.031 -0.067 0.022 -0.008 -0.854 0.005 -0.011 cooking oil 0.128 0.039 -0.029 0.075 0.020 0.005 -0.731 0.101 other food -0.161 -0.001 -0.006 -0.023 0.032 0.013 -0.049 -0.910 compensated wheat flour -0.408 -0.003 0.156 0.059 0.058 0.106 0.158 0.147 rice 0.138 -0.727 0.306 0.005 0.066 0.025 0.148 0.204 dairy -0.004 0.035 -0.612 -0.012 0.067 0.055 0.065 0.032 pulses 0.258 0.051 0.269 -0.661 0.052 0.216 0.072 0.281 meats 0.094 0.048 0.236 0.001 -0.690 0.265 0.034 0.277 fruits/vegetables 0.135 0.070 0.170 0.042 0.086 -0.726 0.119 0.142 cooking oil 0.256 0.046 0.178 0.065 0.084 0.101 -0.648 0.223 other food -0.025 0.134 0.129 0.113 0.168 0.150 0.086 -0.774 source: author’s estimation using 2011-12 hies data. values in parentheses are standard errors. * significant at 1%, ** significant at 5% and *** significant at 10%. 5. conclusion the paper examined complete food demand for households in pakistan using la-aids model. pakistan is a developing country, and, like other developing countries, the people of sajjad, zahoor ul haq & javed iqbal 18 the country consumes a major portion of their income on food. the la-aids model was estimated for eight food groups. the food items were categorized into eight food groups, including wheat flour, rice, dairy, pulses, meats, fruits and vegetables, cooking oil, and other food. results suggested that baluchistan, which is the poorest province, has the highest monthly food expenditure per household. the own price elasticities of all food groups are found to be negative and consistent with economic theory. all the expenditure elasticities are positive and significant, indicating that all food groups are normal. in addition, the expenditure elasticities estimated show that dairy and meats are luxury food items in all provinces, while wheat flour, pulses, cooking oil, and other foods are necessities in the diet of pakistani household. in all provinces, a household spends most of its food expenditure on dairy products, wheat flour, cooking oil, fruit and vegetables. the estimated food demand in this paper can be very useful in the formulation of the country’s food and agricultural policies. in pakistan, households spend a major proportion of their food expenditures on dairy products, wheat flour, cooking oil and vegetables, so the government is suggested to ensure and monitor the availability of these essential food items in order to reduce undernourishment and food insecurity in the country. furthermore, baluchistan, the poorest province, has the highest monthly household food expenditure, so the government is suggested to provide targeted food subsidies and initiate various projects that would increase people's purchasing power to protect them against further poverty. references ahmad, n., sheikh, m. r., & saeed, k. (2015). rural urban food consumption analysis in pakistan: expenditure elasticities approach. pakistan journal of commerce and social sciences, 9(1), 159-170. alexandri, c., păuna, b., & luca, l. (2015). an estimation of food demand system in romania: implications for population's food security. procedia economics and finance, 22, 577-586. aziz, b., mudassar, k., iqbal, z., & hussain, i. (2011). estimating food demand elasticities in pakistan: an application of almost ideal demand system. forman journal of economic studies, 7, 1-24. begum, s., khan, m., farooq, m., begum, n., & shah, i. u. (2010). socio economic factors affecting food consumption pattern in rural area of district nowshera, pakistan. sarhad journal of agriculture, 26(4), 649-653. burney, n., & khan, a. h. (1991). household consumption patterns in pakistan: an urbanrural comparison using micro data. the pakistan development review, 30(2), 145171. journal of applied economics and business studies, volume. 2, issue 1 (2018) 1-18 https://doi.org/10.34260/jaebs.211 19 chen, d., abler, d., zhou, d., & yu, x. (2016). a meta-analysis of food demand elasticities for china. applied economic perspectives and policy, 38(1), 50-72. deaton, a., & muellbauer, j. (1980a). an almost ideal demand system. the american economic review, 70(3), 312-326. deaton, a., & muellbauer, j. (1980b). economics and consumer behavior. university press, cambridge, uk. fan, s., cramer, g., & wailes, e. (1994). food demand in rural china: evidence from rural household survey. agricultural economics, 11, 61-69. fao, i., w. ifad, & unicef (2018). the state of food security and nutrition in the world 2018. building resilience for peace and food security. rome: fao. ghi. 2018. global hunger index results. global, regional, and national trends. https://www.globalhungerindex.org/results/ government of pakistan (gop). 2011-12. household integrated economic survey. federal bureau of statistics, islamabad. government of pakistan (gop). 2015-16. household integrated economic survey. federal bureau of statistics, islamabad. haider, a., & zaidi, m. (2017). food consumption patterns and nutrition disparity in pakistan. mpra paper 83522, university library of munich, germany. haq, z., nazli, h., & meilke, k. (2008). implications of high food prices for poverty in pakistan. agricultural economics, 39, 477–484. haq, z., nazli, h., meilke, k., ishaq, m., khattak, a., hashmi, a. h., & rehman, f. (2011). food demand patterns in pakistani punjab. . sarhad j. agric, 27(2), 305-311. irfanullah., jan, a. u., shah, s. a., ishaq, m., & ali, g. (2018). food consumption in pakistan: application of linear approximate almost ideal demand system. biomedical journal of scientific & technical research, 5(5), 1-8. jiang, b., & davis , j. (2007). household food demand in rural china. applied economics, 39(3), 373-380. karim, r., khan, s., ibrahim., m., & nasrullah. (2018). estimation of per capita food consumption patterns and related poverty in kabal. business and economics journal, 9(2), 1-6. sajjad, zahoor ul haq & javed iqbal 20 malik, s. j., nazli, h., & whitney, e. (2014). food consumption patterns and implications for poverty reduction in pakistan. conference of pakistan society of development economists, islamabad, pakistan, (pp. 1-25). malik, s. j., nazli, h., & whitney, e. (2015). food consumption patterns and implications for poverty reduction in pakistan. the pakistan development review, 54(4), 651-670. mudassar, k., aziz, b., & anwar, a. (2012). estimating consumer demand of major food items in pakistan: a micro data analysis. pakistan journal of life and social sciences, 10(1), 53-58. usaid. 2018. food assistance fact sheet pakistan. https://www.usaid.gov/sites/ default/files/ documents/1866/ ffp_fact_sheet_pakistan.pdf zellner, a. (1963). estimators for seemingly unrelated regression equations: some exact finite sample results. journal of the american statistical association, 58(304), 58, 977-992. journal of applied economics and business studies, volume. 4, issue 4 (2020) 97-116 https://doi.org/10.34260/jaebs.445 97 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x how election process is rent-seeking in pakistan? a case study of local bodies election in district pakpattan anwar shah1, nasrullah habib2 and muhammad zeeshan younas3* 1 associate professor, school of economics, quaid-e-azam university islamabad. 2 m.phil. graduate, school of economics, quaid-i-azam university islamabad. 3 ph.d. scholar, school of economics, quaid-i-azam university islamabad. abstract this study examines the difference between the total spending on the election campaign in 54 union councils during local bodies’ election and the expected developmental fund for the same area. the study is based on primary data collected from 324 contesting candidates for the chairmanship/vice-chairmanship of 54 union councils (ucs) in district pakpattan during the local bodies’ election of 2015 in punjab, pakistan. per local bodies council rules, a winning chairman of uc has to allocate developmental funds in the area. so, the whole process of local bodies’ election is for the finalization of a person among the people, who could be given the mandate to use the developmental funds. in this regard, we assume that the prevailing process of election is rent-seeking for society, in the sense that the total cost of the election process for society is greater than the developmental funds for which the same process is conducted. candidates consume more than the allowed limit set by the election commission of pakistan (ecp). hence, society bears the monetary cost of spending more than the set limit and non-monetary costs like rivalry, foes, conflicts, etc. the analysis of data shows that the total amount of campaign spending is 473.27 million rupees which are significantly more than the allowed limit, set by ecp as well as the total allocated developmental funds (128 million) of 54 union councils. the study shows that the whole process of local bodies’ election is putting rent of 345 million on society. apart from monetary cost, this process brings those people to the power corridors who themselves break ecp rules. the study suggests comprehensive reforms in the process of the election campaign for saving society from its associated cost. keywords election campaign, political process, local bodies’ election, rent-seeking jel classification c24, i21, i24 * mzeeshan@eco.qau.edu.pk mailto:mzeeshan@eco.qau.edu.pk anwar shah, nasrullah habib and muhammad zeeshan younas 98 1. introduction we know that market failure rationalizes the intervention and role of a government in the allocation of resources in an economy (buchanan, 1968; menon & yao, 2017; sun et al., 2020). in this regard, the debate regarding the process of formation of the government has converged on democracy (downs, 1995; sagar, 2019). in a democratic society, every adult, on average, is entitled to contest the election and become part of the government. likewise, every adult has the opportunity to elect people, whom they wish and consider eligible for running the affairs of government. to let people, whom they should choose for running the affairs of government, the candidates are provided an opportunity to inform the general people, before the election, about the program they wish to implement after having the reign of government. such dissemination of information is done through the election campaign. hence, the process of the election campaign is an integral part of the formation of government in a democracy at all three tiers of government, that is, federal, provincial, and local (saward, 2008). the efficiency of the process of election campaign requires that the associated cost must not exceed the return that the ensuing government is supposed to provide to society. in other words, the net returns of the process of election must not be negative. moreover, the private cost of each contesting candidate on an election campaign must not exceed the allowed limit for the same purpose set by constitutional bodies such as the election commission of pakistan (ecp). in case this is not followed by the letter and spirit, then one can claim that the process of election campaign involves rent-seeking.1 in this paper, we are interested to examine whether the process of the election campaign is rent-seeking. we take local bodies' election as a testbed for examining the answer to our question. we examine the amount of rent conferred on the society due to the process of the election campaign in local bodies election of 2015 held in district pakkpattn, punjab pakistan. what is rent-seeking? according to lambsdorff (2002), rent-seeking involves transfer costs. that is the rent-seeking activity is for transferring the fixed rent and such activities do not increase the size of the prize. hence, in an effort to get the prize, if the cost of effort for transferring the prize gets bigger and larger, then, society falls to a disadvantaged position. lambsdorff (2002) is of the view that this transfer cost of resources is rent dissipating. tullock (1967) has explained such rent-seeking in terms of the administrative cost of transferring tax and tariffs revenue to state account from people account. such administrative cost includes salaries of tax inspectors and customs officials as well as public costs for containing smuggling and tax evasion. such costs lower welfare and is a sort of deadweight loss and waste. they are not aimed at increasing production or lowering costs to maximize profit. tullock (1975) considers the effort for obtaining subsidies a pure waste that neither raises production nor bears 1the literature covers such discussion in the context of rent seeking. see, for example, tullock (1967), krueger (1974), posner (1975) and dougan, (1991). journal of applied economics and business studies, volume. 4, issue 4 (2020) 97-116 https://doi.org/10.34260/jaebs.445 99 any other fruit for third parties. tullock (1975) is of the view that: “the transfer itself may be cost less, but the prospect of the transfer leads individuals and groups to invest resources in either attempting to obtain a transfer or to resist a transfer away from themselves. p-109” based on the above discussion we consider the money spent on the election campaign a transfer cost for having some fixed developmental funds. to our knowledge, the literature on election campaigns mostly discusses the relationship between campaigning cost and votes gained (see, for example, ansolabehere & snyder, 1996; goldstein & freedman, 2000). in addition, we find literature related to political contests for rents and transfer (hillman & riley, 1989; fu & wu, 2019). however, there is a lack of empirical analysis related to how political contest through the process of election campaigning becomes a transfer cost for the allocation of developmental funds. this paper aims to fill this gap by examining the transfer cost of developmental funds from the government exchequer to the doorsteps of common people in the 54 union council of district pakpattan. we collect primary data from 324 candidates, who contested election for the seat of chairmanship/vice chairmanship of union council in district pakpattan in the local bodies’ election of 2015 in the province of punjab, pakistan. the data is still valid as till today no new local election has been held. it is pertinent to mention that there are 54 union councils in district pakpattan. per rules of the election commission of pakistan (f.4 (10)/2015-lge (p)) a candidate for chairmanship/vice-chairman can spend up to pkr 100,000 (one hundred thousand) in the election campaign.2 per rationality, the transfer cost of the developmental fund will be low if candidates avoid deviation in spending and follow the rules of ecp. however, spending on election campaigns is like n-players prisoner dilemma game. each candidate in the election is like a player in this game. in case, each player sticks to the limit set by ecp, then transfer cost will be low and we will have a socially optimal solution. however, each player has an incentive for deviation from the rules of ecp if they think that doing so can attract more votes and increase their probability of a win. if every player plays the dominant strategy, nash equilibrium will be sub-optimal, thus putting a huge cost on society. based on this theory we expect a significant amount of rent-seeking activity, thus enhancing transfer cost, in the process of election campaigning during local bodies’ election of 2015 in pakkpattn, held in punjab, pakistan. the analysis of the collected data shows that on average every candidate for chairmanship/vice chairmanship of union council consume1.56 million pakistani rupees 2the relevant clauses is 17(d) per code of conduct of ecp for contesting local bodies election (f.4 (10)/2015-lge (p) is: the contesting candidates of chairman and vice chairman of union council can spend 100000 rupees for election campaign. anwar shah, nasrullah habib and muhammad zeeshan younas 100 (pkr) on the election campaign. some candidates spend more than pkr 5.2 million. furthermore, the cost of each subhead is multiple times the limit set by ecp. the society as a whole bears the rent of pkr 345 million for the expected developmental fund of pkr 32 million per year (pkr 208 million for 4 years). it is worth mentioning that we do not take into account the campaigning cost other than chairmanship/vicechairmanship in a union council. in addition, we do not take into account the time and other non-material costs, when calculating this amount of rent. the rest of the paper is organized as follows. section 2 provides a brief review of the literature. section 3 explains the structure of local government, election procedure and code of conduct. section 4 describes the theoretical framework and explains the hypotheses of the study. section 5 provides details about data collection. the details of the findings are provided in section 6, while section 7 concludes. 2. literature review in the literature, we find studies examining the relationship between a candidate’s campaign expenditures and his or her vote share in u.s. politics (e.g., ansolabehere & snyder, 1997; erickson & palfrey, 1996; green & krasno, 1988; jacobson, 1980, 1985, 1990). likewise, we find literature on the same phenomenon for britain (pattie et al., 1995). the intuition behind these studies is that more spending ought to benefit the spender—that spending and vote share ought to be correlated; other things being equal (see also, cox & thies, 2000). the empirical findings on the effects of election campaigning at the local level support the view that campaign spending by parties or candidates put an impact on votes gained (green & krasno, 1988; johnston & pattie, 1997; pattie et al, 1995). evidence for this comes from analysis both of single elections and of election series in canada (palda, 1973), the united states (grier, 1989; jacobson, 1980; 1985; welch & hibbing, 1984), and britain (johnston et al, 1988; pattie et al, 1995). in addition, we find studies that examine how rent-seeking behavior affect the political dynamics in african countries (mbaku, 1994; mbaku & kimenyi, 2015). a series of papers have explained various aspects of rent and rent-seeking after tullock (1967). for example, nitzan (1994) provides a review of models explaining the behavior of rent-seeking agents. the paper provides how various work such as baik & shogren (1992) model the rent-seeking process for risk-neutral agents having an equal probability of winning the rent. other papers discuss how rent is modeled for the riskaverse agents (hillman & katz, 1984) and the asymmetric valuation of the prize (baye et al., 1992). likewise, nitzan (1991) shows that when the competition for winning a prize is between two groups instead of two individual rent-seekers, the amount of rent dissipation drops. this is due to a decline in the intensity of the level of competition and free-riding by some contenders in the group. hillamn & riley (1989) develop a general model of politically contestable rents and transfers which includes asymmetric valuation for the prizes. in such cases, the outlays made in contesting a prize are substantially journal of applied economics and business studies, volume. 4, issue 4 (2020) 97-116 https://doi.org/10.34260/jaebs.445 101 below the value of the prize. the reason is that asymmetric valuation acts as a barrier to entry and the number of agents actively contesting a politically allocated rent or transfer gets small. zupan (2015) discusses the supply side of the rent and examines how rulers, political leaders, and public employees can take advantage of their positions. the insiders in the government have the motive, political power, and opportunity of rent-seeking and hence are more likely to grab rent from people. for example, in china, 203 of the 1,271 richest people in the country (about one in seven) are delegates to the nation’s parliament or its advisory body. the delegates’ combined net worth is nearly usd 500 billion. 18 of them have assets exceeding the combined wealth of all 535 members of the us congress, ex-president obama’s cabinet, and all nine members of the supreme court. officials in the government have imperfect accountability and no competition on the supply side, hence, they earn more rent. patnaik (2015) shows that firms and industries across the world regularly engage with political and government players to obtain economic rents through a variety of strategic tools. rovinaruand (2014) show that, for years, the us and several eu countries have provided government support that was found to be against wto rules to airplane manufacturers boeing and airbus, respectively. even in modern democratic countries, the costs imposed by rent-seeking – both indirectly through welfare losses and directly through the costs rent-seekers have to expand to secure their rents (luchsinger & mueller, 2003). more related work with our papers is of laband & sophocleus (1992). they examine the expenditures on non-exchange, non-charity transfer activity in the united states in 1985. by non-exchange transfer, the study implies that the gain in wealth by the recipient(s) comes at the expense of one or more individuals from whom the transfer is originated. that is, the transfer is involuntary from the perspective of one or more individuals whose wealth is targeted. such non-exchange and non-charity transfer activity include, but is not limited to, theft and theft prevention, fraud, and governmentsponsored transfer programs. the study reports expenditure that is designed to facilitate and inhibit non-exchange transfers, executed privately or through the state. the numbers indicate that individuals plausibly invested nearly a trillion dollars in transfer activity that year. the nominal gnp of the usa in 1985 was just over $4 trillion, which includes numerous transfer-related resource investments that arguably should be subtracted out. transfer activity thus apparently constitutes a much larger fraction of all economic activity conducted in the united states than previously recognized. the study terms such expenditure as rent-seeking as here a person is investing in protecting the gains which they have already achieved through some sources. we take help from the literature of rent-seeking and want to show empirically; that to what extent expenditure on election campaign in a local body election is a transfer cost of developmental funds from the government to people. based on the findings, we anwar shah, nasrullah habib and muhammad zeeshan younas 102 aim to show that each of the candidates is after some rents associated with a win (frank & cook, 2013). 3. the structure of local government, election procedure and code of conduct according to punjab local government act 2013, the local government has three tiers and consists of the district, tehsil, and union councils.3 the government at the level of union council has 13 members including the chairman and vice chairman as shown in table 1. chairman and vice chairman contest election on one symbol. each union council is further divided into six wards. each ward contains 2000 to 3000 registered votes. there are 6 general counselors in each union council. these 6 general counselors are elected from 6 wards of the union council. there are 2 lady councilors, 1 minority councilor, 1 labor councilor and 1 youth member in each union council. one of the lady councilors is elected from the first three wards, while the second lady councilor is elected from the remaining three wards. on the other hand, the constituency of the labor councilor, minority councilor and youth member consists of all 6 wards of the union council. table 1: composition of the candidate in a union council chairman vice chairman general councilor lady councilor youth councilor minority councilor labor councilor total 1 1 6 2 1 1 1 13 the election commission of pakistan (ecp) appoints “district returning officer” for every district and “returning officer (ro)” for every constituency. per rules, ecp announces the schedule of an election. after the announcement people submit nomination papers for contesting an election. the nomination papers are checked by ro and announce decisions about its acceptance or rejection. an appeal against the acceptance or rejection of the nomination paper can be filed with the appellate tribunal, which makes the final decision. next, the final list of the contesting candidates is prepared by ro and each is allocated election symbols according to party affiliation or per independent capacity of the candidate. after allotting of election symbol candidates start a formal campaign, which lasts till 48 hours before the day of the election. ecp provides copies of the voters’ lists to ro, who divide these copies among presiding officers per polling station. a voter can cast vote after confirming his name in the voter list, while polling is continuing from 8.00 am to 5.00 pm, without any break. finally, the presiding officer counts the casted votes in the presence of the candidate or the agents of the candidates and makes a statement in which the details of all the votes of candidates are given. this statement along with unused material is sent to ro. in addition, the presiding officers announce the results of the polling stations and give its copy to relevant contesting candidates. after receiving results from all polling stations, 3see for detail of the act http://punjablaws.gov.pk/laws/2542.html journal of applied economics and business studies, volume. 4, issue 4 (2020) 97-116 https://doi.org/10.34260/jaebs.445 103 ro compiles the unofficial results and sends them to ecp on the prescribed form. the official results are announced by ecp. 4. theoretical framework and hypotheses we assume that a contestant candidate for chairmanship/vice-chairmanship is a rational economic agent (sugden, 1991; vriend, 1996). the objective of such an agent is to maximize private gains, which is winning the election. however, winning the election is not an end; rather it is a means to an end, which is to decide on the allocation of funds that are set aside in the annual budget of the punjab government for each union council in a district. the dilemma is that making a decision about the allocation of funds for the developmental purpose of a union council is a social objective and the candidates are less likely to receive private rewards from this process. if other members of society receive benefits, he will also receive the same in the capacity of a member of the same society. per economic theory a selfish person thinks about the private rewards, hence the maximization of the social objective cannot be the primary goal of a selfish person. in other words, every candidate must consider earning private rewards in the guise of promoting social objectives. thus, each candidate expects to receive back some private returns from spending. all private returns are in the form of rents associated with a win (frank & cook, 2013). such rents included but not limited to, granting approval to benefit his/her near and kins, receiving a kickback from contractors, becoming influential in giving/selling jobs etc. (krauger, 1974; hillman & riley 1989) according to zuban (2015), government insiders capture rents as they face imperfect accountability. financial rents can be secured through the diversion of public funds or from payments received from other sources (legal as well as illegal). the various potential channels for material remuneration include campaign contributions, honoraria, in-kind gifts, revolving-door positions in the industry, and payments to family members and family businesses. the rents may be derived from relevant interest groups in return for promoting particular policies. they also may accrue on account of politicians’ threatening to undertake activities that will make an interest group worse off. in addition, zuban (2015) shows that people in government have power, opportunity, and motive for involving in rent-seeking. the “motive” is symbolized by the level of election spending. we do not capture the total value of rent in the current study. however, we use the amount of spending over and above the allowed limit as a proxy for rent. we assume that a person is more rent seeker if the amount he spends during an election campaign is higher than the allowed limit. he must be expecting to get a return for this money. in case, he is not entitled to the salary against chairman/vice-chairman, this means that the marginal productivity of his job for the society is worth zero. in case, he is entitled to a salary, then the marginal productivity is worth the amount of salary. hence, we can say that the consumption beyond such amount of salary is rent on the society, which the society will bear but is less likely to receive any reward for the same. anwar shah, nasrullah habib and muhammad zeeshan younas 104 based on the above discussion, we assume that the incentives associated with a win are likely to drag candidates to spend more on election campaigns than the set amount of ecp. the reason is that the election campaign process is like the n-players prisoner dilemma game. every candidate should stick to the limit per set rules of ecp, however, every candidate has an incentive for deviation to consume more to attract more voters. hence, hypothesis 1 of the study is: hypothesis 1: every candidate is likely to deviate from the required standard of the allowed limit of the election campaign in order to earn the rent associated with a win. in addition to the above, a candidate who belongs to the ruling party is more likely to consume more. the reason is that such candidates have the support of the ruling party. hence, they know that after a win they can ask for extra funds and thus can reap more rents from kickbacks. also, they can earn rent through various tactics such as giving jobs, releasing people from police custody, lodging and dislodging fir with police, etc. such candidates can easily realize people that he is indispensable for resolving their issues at the local and provincial level. moreover, such candidates can go for returns against the law as they are less likely to face actions for such kind of actions from their opponents as the machinery of government is on average not cooperative with their opponents. the second hypothesis of the study, thus, is: hypothesis 2: a candidate affiliated with the incumbent ruling party in the province is likely to consume more as compared to other candidates 5. methodology and data the secondary data on spending related to the election campaign by each candidate is not available. hence, we designed a detailed questionnaire (see appendix a) and conduct a survey in “district pakpattan”, which is located about 210 kilometers away from lahore, the capital city of punjab. according to the 2017 census of pakistan, the district pakpattan had a population of 1.82 million of which about 16 percent lives in urban areas. most of the people of pakpattan belong to 200 different small and large clans. however, the dominant clans are wattoo, doger, gujjar, rajpoot, jutt, khagga, bodla, aryin, sukhera, hottiyana and rath. most of these clans have contested candidates at all tiers of an election. the main language in pakpattan is punjabi, which is spoken by 96 percent population. agriculture is the main source of income for most of the population of pakpattan. the unit of analysis per questionnaire is a candidate who contested election for the post of chairmanship or vice-chairmanship in any of 54 union councils of district pakpattan during local bodies’ election held in 2015.4 per ecp data 340 candidates contested election in 54 union councils of district pakpattan. out of these 340 candidates, we conducted interviews with 324 candidates. out of 324 candidates, only 4 it is to be noted that, we would have included fresh data, but new local election has not been held after 2015, till today journal of applied economics and business studies, volume. 4, issue 4 (2020) 97-116 https://doi.org/10.34260/jaebs.445 105 4 candidates were interviewed through their kin, the rest of all candidates were interviewed in person directly. our respondents include candidates from all mainstream political parties like pakistan muslim league nawaz (pml-n), pakistan tehreek insaff (pti), pakistan people party (ppp), jamat islami (ji), pakistan awami tahreek (pat) and independent. 6. results we present results in various sub-section. in this regard, section 6.1 provides an overview of total campaigning cost. on the other hand, sections 6.2 and 6.3 compare campaigning costs across all union councils and position holders, respectively. the analysis of campaigning cost across political parties is given in section 6.4. the other related details are given in section 6.5 6.1 an overview of findings table 6.1 summarizes the total cost of the election campaign in the local bodies election of 2015 in district pakpattan. table 6.1: an overview of total spending on election campaign nature of constituency number of seats nature of candidature number of candidates respondent candidates total campaigning cost in millions average campaigning cost in millions union council 54 chairman and vice chairman 340 324 473.27 1.46 table 6.1 clearly shows that the contested candidates for chairmanship and vicechairmanship in local bodies’ election deviated from the prescribed rules of ecp regarding election spending. this deviation from the prescribed rules of ecp shows support in favor of our first hypothesis. everyone in the election is better off if every candidate stays within the limit of ecp rules. however, the objective of winning an election incentivizes deviation from the socially optimal outcome. it is pertinent to mention that the estimated budget for 54 union councils which contain 84 urban wards is 32 million rupees per our discussions with the relevant authorities in the local bodies office in district pakpattan. likewise, these union councils and municipal wards generate some revenues. this generated revenue is also used for developmental purposes. nevertheless, the contested candidates of local bodies’ election spent 473.27 million rupees on the election campaign. this does not include the cost of the election campaign by candidates for councilors at wards levels. in other words, we have calculated the cost of a campaign for one seat of the union council, while there is a candidate who contests election for 12 other seats in a union council. anwar shah, nasrullah habib and muhammad zeeshan younas 106 this shows that to obtain authority for spending developmental funds of about128 million rupees, candidates for one seat only spend 473.27 million rupees. hence, the whole process of election campaign add in transfer cost and show that candidates consume more so that to earn rents associated with a win. we also examine whether election campaign spending varies across the union councils or not. figure 6.1 depicts the average campaigning cost of candidates across union councils. it is evident from the figure, that rarely the average campaigning cost per candidate in a union council is less than a million rupees. the lowest and highest average campaigning cost ranges between 0.8 and 5.2 million rupees, respectively. the figure also shows that candidates of union council 11, 36 and 54 spent on average more amount than other ucs on the election campaign. per our analysis and interviews, the main reason for this high expenditure is that candidates of these union councils are interested in contesting elections at a further higher level such as district government. based on the above discussion, we can say that all candidates are players in this rentseeking process. they don’t follow the defined rules regarding the election campaign spending set by ecp, thus hurting society negatively. figure 6.1: average campaigning cost across union councils 6.2 campaigning cost and winning position across all ucs in order to examine whether the cost of campaigning plays any role in determining the position of a candidate in the list of winners, we present figure 6.2. the figure compares the expenditure on campaigning across the first three positions holders in a union council.5 it is evident from the figure that the winner on average consumes more than the second and third position holders. the average expenditure of first position 5it is to be noted that in some union councils the context was held between two candidates. hence, third bar does not exist for such union councils in the graph. 0 1 2 3 4 5 6 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 c o st i n m il li o n union councils journal of applied economics and business studies, volume. 4, issue 4 (2020) 97-116 https://doi.org/10.34260/jaebs.445 107 holder is pkr 4.63 million while the average expenditure of second and third position holders is pkr 4.00 million and 1.48 million, respectively. the analysis shows that expenditure on campaigning does play a role in determining the position. this finding is in line with the early literature. we can say that the incentive for winning attracts candidates for deviation from socially optimal equilibrium, which is sticking to the rules of ecp. however, the question is that what drags a candidate for consuming such a high level of expenditure for a seat of chairman/vice-chairman which does not carry any monetary incentive? there is no answer to this question other than that each candidate is after the rents associated with winning. the society bears a huge transfer cost for assigning the mandate of using the developmental fund. figure 6.2: campaigning cost across position holders of all union councils 6.3 campaigning cost across candidates of various political parties table 6.2 provides information about campaigning costs by candidates affiliated with various mainstream political parties. 0 5 10 15 20 25 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 c o st i n m il li o n union councils 1st position 2nd position 3rd position anwar shah, nasrullah habib and muhammad zeeshan younas 108 table 6.2: campaigning cost across candidate affiliated with various political parties party name number of union council total number of candidates total expenditure (in million) average expenditure (in million) pml(n) 48 96 167.06 1.74 independent 41 142 161.95 1.14 pti 38 76 135.64 1.78 ppp 03 06 8.47 1.41 pat 01 02 0.017 0.008 ji 01 02 0.13 0.66 total 324 473.27 1.46 the table shows that the average cost of the candidates of pti is more than the average cost of the rest of the political parties which is not in line with hypothesis 2. however, there is no significant difference across the mean expenditure of candidates from pml (n) and pti (p=0.41). hence, we can claim that data provide some support in favor of hypothesis 2. one can ask why candidates of pti consume more than the candidates of pml (n). a possible answer is that the expenditure from the candidates of pti is high as they have to compete with candidates from the ruling party of pml (n). peer pressure from competing candidates could also force them for spending high on the election campaign. the closer the competition, the closer is likely to be the expenditure. this conjecture is also supported by figure 6.3 which depicts the expenditure of the first three position holders across pml (n), pti and independent. the figure indicates that on average the expenditure on campaigning by position holder is higher and close to the corresponding candidates on the same position. in the diagram, ppp shows a single bar as there were no candidates in the rank of second and third position holder from this party. similar is the case of ji and pat, whose candidates were not on the list of the first three positions. figure 6.3: average campaigning cost of candidates with respect to position across candidates affiliated with various political parties 0 2 4 6 8 10 pml(n) pti indepen ppp pat ji c o st i n m il li o n 1st position avg cost 2nd position avg cost 3rd position avg cost journal of applied economics and business studies, volume. 4, issue 4 (2020) 97-116 https://doi.org/10.34260/jaebs.445 109 table 6.3 provides details of the expenditure on the main three heads of an election campaign. the table helps us to know which category is costlier for the candidates. the table shows that candidates spent the highest amount on the provision of gifts and donations to voters. the second major head of election spending is an advertisement, while spending on the public meeting is the third major head for candidates. in nutshell, we can say that candidates on average consume more than the allowed limit on all main heads of the election campaign. table 6.3: campaign cost across various heads category total cost (millions) gifts and donations 171.40 advertisement 121.78 transportation cost 99.20 7. conclusion and policy implication: the main objective of an election process in general and at the local level, in particular, is to assign power and funds related to the collective objectives of a society. in this regard, all individuals are the same; hence all can claim the right to political leadership. in order to decide who among the people be assigned the collective responsibility, almost all civilized societies of the world have adopted the process of one vote-one person. the societies that are following this process are called democratic societies. in a democratic society, every adult, on average, is entitled to contest an election and become part of the government. likewise, every adult has the opportunity to elect people, whom they wish and consider eligible for running the affairs of government. to let people, whom they should choose for running the affairs of government, the candidates are provided an opportunity to inform the general people, before the election, about their program. such dissemination of information is done through the election campaign. hence, an election campaign is an integral part of the formation of government at all three tiers of government, that is, federal, provincial and local. the efficiency of the process of election campaign requires that the associated cost must not exceed the return that the ensuing government is supposed to provide to society. if this is not the case, then the process of election campaign reflects a huge transfer cost and shows a rent-seeking activity on the part of candidates. this paper examines the transfer cost and rent-seeking aspect of the process of the election campaign in a local bodies’ election at district pakpattan, in the province of punjab, pakistan. rent-seeking is like the transfer costs of some funds or commodity. we find in the literature that transfer cost decrease welfare of the society as mentioned by tullock (1967) that taxation and tariffs shift part of consumers’ rent to the state but require administrative costs in the form of salaries for tax inspectors or customs officials. such anwar shah, nasrullah habib and muhammad zeeshan younas 110 costs lower welfare and must be added to the harberger triangle that is deadweight loss and is a form of waste. based on the same example, an election campaign is like a transfer cost of some developmental funds from the government to the public. this is contrary to the transfer cost of taxation where funds come from the public to the government. to our knowledge, the currently available literature on election campaign examines the relationship between campaigning cost and votes gained (see for example, ansolabehere & snyder, 1996; erickson & palfrey, 1996) as well as election campaign and rent-seeking. this paper takes help from the available literature and empirically estimates the transfer cost and rent-seeking based on data of campaigning cost in 54 union councils of district pakpattan. we collect data from 324 candidates, who contested election for the seat of chairmanship or vice-chairmanship in the 54 ucs of district pakpattan. based on the referred literature, we assume that the campaigning cost for the transfer of developmental funds will be low if candidates avoid deviation in spending and follow the rules of ecp. however, the process of an election campaign is like an n-players prisoner dilemma game for the candidates. each candidate is a player in the game. in case, each player sticks to the limit set by ecp, then campaigning cost will low, giving us a socially optimal solution. however, each player has an incentive for deviation as doing so can attract more votes. hence, each player plays the dominant strategy, putting a huge cost on society. based on this theory we expect a significant amount of rentseeking activity in the election campaigning during local bodies’ election of 2015 in pakpattan, held in punjab, pakistan. our data shows that on average a candidate for chairman/vice chairman in a union council consumes 1.56 million rupees on the election campaign. we also find that some candidate spends more than 5.2 million rupees. furthermore, the cost of each subhead is multiple times the limit given by ecp. the society as a whole bears the transfer cost of pkr345 million for the expected developmental fund of pkr 32 million per year (128 million for 4 years). it is worth mentioning that we do not take into account the campaigning cost of the other 12 members of a union council as well as the time and other non-material costs while calculating this amount of transfer cost. in addition, the paper lack to quantify the rents, which each candidate is after. we do show that transfer cost is equal to election campaigning cost, which is like a rent-seeking cost, as it does not increase profit for the society. however, we could not show the exact amount of such prize or rent. our work offers grounds for reforms in the election process. first of all, the institution of ecp must show its presence with full strength, so that they can implement and impose their power related to the election campaign. also, there should be support from all other institutions such as police, local administration with ecp in implementing their rules. moreover, there is a need for serious discussion related to the cost associated with an election campaign at policy-making forums such as provincial and federal journal of applied economics and business studies, volume. 4, issue 4 (2020) 97-116 https://doi.org/10.34260/jaebs.445 111 parliament. the current process of election and the election campaign is not only inefficient but put a huge monetary and non-monetary transfer cost on the society. hence, the members of the parliament of pakistan should seriously consider the transfer cost and renting aspect of an election campaign and come up with a substitute that is democratic as well as cost-efficient. references ansolabehere, s., & snyder, j. m. (1997). the inter election dynamics of campaign finance: us house elections, 1980-94. in legislative studies quarterly (vol. 22, no. 1, pp. 124-124). university of iowa, w307 seashore hall, iowa city, ia 52242-1409: comparative legislative research center. baik, k. h., &shogren, j. f. (1992). strategic behavior in contests: comment. the american economic review, 82(1), 359-362. baye, m. r., kovenock, d., & de vries, c. g. (1992). it takes two to tango: equilibria in a model of sales. games and economic behavior, 4(4), 493-510. blankart, c. b., & ehmke, d. c. (2015).overcoming the rent-seeking defect in regional policy: time to re-think the institutional design. dice report, 13(3), 44. buchanan, j. m. (1968). the demand and supply of public goods (vol. 5). chicago: rand mcnally. cox, g. w., &thies, m. f. (2000). how much does money matter?“buying” votes in japan, 1967-1990. comparative political studies, 33(1), 37-57. dougan, w. r. (1991). the cost of rent seeking: is gnp negative?. journal of political economy, 99(3), 660-664. downs, a. (1995). the origins of an economic theory of democracy anthony downs. information, participation, and choice: an economic theory of democracy in perspective, 197-240. eagles, m. (1993). money and votes in canada: campaign spending and parliamentary election outcomes, 1984 and 1988. canadian public policy/analyse de politiques, 432-449. forrest, j., johnston, r. j., & pattie, c. j. (1999). the effectiveness of constituency campaign spending in australian state elections during times of electoral volatility: the new south frank, r. h., & cook, p. j. (2013). winner-take-all markets. studies in microeconomics, 1(2), 131-154. fu, q., & wu, z. (2019). contests: theory and topics. in oxford research encyclopedia of economics and finance. anwar shah, nasrullah habib and muhammad zeeshan younas 112 goldstein, k., & freedman, p. (2000). new evidence for new arguments: money and advertising in the 1996 senate elections. the journal of politics, 62(4), 10871108. green, d. p., &krasno, j. s. (1988). salvation for the spendthrift incumbent: reestimating the effects of campaign spending in house elections. american journal of political science, 884-907. grier, k. b. (1989). on the existence of a political monetary cycle. american journal of political science, 376-389. hillman, a. l., & katz, e. (1984). risk-averse rent seekers and the social cost of monopoly power. the economic journal, 94(373), 104-110. hillman, a. l., & riley, j. g. (1989). politically contestable rents and transfers. economics & politics, 1(1), 17-39. jacobson, g. c. (1980). money in congressional elections.yale university press. jacobson, g. c. (1985). money and votes reconsidered: congressional elections, 1972– 1982. public choice, 47(1), 7-62. jacobson, g. c. (1990). the effects of campaign spending in house elections: new evidence for old arguments. american journal of political science, 334-362. johnston, r. j., & pattie, c. j. (1995). the impact of spending on party constituency campaigns at recent british general elections. party politics, 1(2), 261-273. johnston, r. j., & pattie, c. j. (1997). where's the difference? decomposing the impact of local election campaigns in great britain. electoral studies, 16(2), 165-174. johnston, r. j., pattie, c. j., & allsopp, j. g. (1988). a nation dividing?: the electoral map of great britain, 1979-1987. longman; new york: wiley. krueger, a. o. (1974). the political economy of the rent-seeking society. the american economic review, 64(3), 291-303. laband, d. n., & sophocleus, j. p. (1988). the social cost of rent-seeking: first estimates. public choice, 58(3), 269-275. laband, d. n., & sophocleus, j. p. (1992). an estimate of resource expenditures on transfer activity in the united states. the quarterly journal of economics, 107(3), 959-983. lambsdorff, j. g. (2002). corruption and rent-seeking. public choice, 113(1-2), 97-125. luchsinger, c., & mueller, a. (2003). incentive compatible extraction of natural resource rent. in cepe working paper no. 21, centre for energy policy and economics (cepe. menon, a. r., & yao, d. a. (2017). rationalizing outcomes: mental-model-guided learning in competitive markets. harvard business school. millner, e. l., & pratt, m. d. (1989). an experimental investigation of efficient rentseeking. public choice, 62(2), 139-151. journal of applied economics and business studies, volume. 4, issue 4 (2020) 97-116 https://doi.org/10.34260/jaebs.445 113 mbaku, j. m., & kimenyi, m. s. (2015). rents and development failure in africa. in companion to the political economy of rent seeking. edward elgar publishing. mbaku, j. m. (1994). military coups as rent-seeking behavior. jpms: journal of political and military sociology, 22(2), 241. nitzan, s. (1991). collective rent dissipation. the economic journal, 101(409), 15221534. nitzan, s. (1994). modelling rent-seeking contests. european journal of political economy, 10(1), 41-60. palda, k. s. (1973). does advertising influence votes? an analysis of the 1966 and 1970 quebec elections. canadian journal of political science/revue canadienne de science politique, 6(4), 638-655. patnaik, s. (2015). rent-seeking and public policy. cesifo dice report, 13(3), 3-5. pattie, c. j., johnston, r. j., & fieldhouse, e. a. (1995).winning the local vote: the effectiveness of constituency campaign spending in great britain, 1983– 1992. american political science review, 89(4), 969-983. randall, a. (1983). the problem of market failure. natural resources journal, 23(1), 131-148. rovinaru, m., & pop, l. (2014). the subsidies-an instrument of trade policy. case study: the boeing-airbus" war". studies in business & economics, 9(3). sagar, p. (2019). the opinion of mankind: sociability and the theory of the state from hobbes to smith. princeton university press. saward, m. (2008). representation and democracy: revisions and possibilities. sociology compass, 2(3), 1000-1013. sommerville, j. p. (1992). thomas hobbes: political ideas in historical context. macmillan international higher education. sugden, r. (1991). rational choice: a survey of contributions from economics and philosophy. the economic journal, 101(407), 751-785. sun, j., schram, a., & sloof, r. (2020). elections under biased candidate endorsements—an experimental study. games and economic behavior, 125, 141-158. tullock, g. (1967). the welfare costs of tariffs, monopolies, and theft. economic inquiry, 5(3), 224-232. tullock, g. (1971). subsidized housing in a competitive market: comment. american economic review, 61(1), 218-219. tullock, g. (1975). competing monies. journal of money, credit and banking, 7(4), 491-497. anwar shah, nasrullah habib and muhammad zeeshan younas 114 vriend, n. j. (1996). rational behavior and economic theory. journal of economic behavior & organization, 29(2), 263-285. wales case, 1988–95. environment and planning a, 31(6), 1119-1128. welch, s., & hibbing, j. r. (1984). hispanic representation in the us congress. social science quarterly, 65(2), 328. zupan, m. a. (2015). rent-seeking on the supply side of politics. cesifo dice report, 13(3), 6-14. appendix questionnaire dear respondent, we are from the school of economics quaid-i-azam university, islamabad and working on a research project related to the election campaign. in this regard, we need some information. we shall be grateful if you kindly provide me the following information to the best of your knowledge. we assure you that all information will remain private and used for the purpose of this research. thanks in advance for your cooperation and valuable time. part a: general profile of the respondent name of the respondent (optional) please tick the box next to the answer of your choice. 1). the name of your district is 1) pakpattan 2) okara 2). the nature of your union council is 1) urban 2) rural 3). which seat did you contest for? 1) chairman 2) vice-chairman 4). which party ticket did you hold? 1) pti 2) pmln3) ppp 4) independent 5) other (please specify) 5). which brothery do you belong to? 6). what is your age? 7). since how long are you active politically? 1) since forefathers 2) since----------year 8). did you contest the election before the 2015 local bodies poll? 1) yes 2) no 9). what is your education level? 1) school 2) college 3) university 10). do you have a spiritual relationship with any person/durbar? 1) yes 2) no if yes, then kindly mention the name 11). did you get any support during the election campaign from your peer? 1) yes 2) no part b: election turn over 12). total registered voters as per ecp in the constituency journal of applied economics and business studies, volume. 4, issue 4 (2020) 97-116 https://doi.org/10.34260/jaebs.445 115 13). total casted votes in your constituency 14). how many votes did you receive? 15). total number of contested candidates 16). your position in the final outcome of election 17). how many votes did you receive from brothery? part c: details about advertisement cost during the election campaign 18). did you give any advertisement on t.v channels during an election campaign? 1) yes2) no if yes, then how much is the approximate total cost of such an advertisement? 19). did you give any advertisement in the newspaper? 1) yes 2) no if yes, then how much is the approximate total cost? 20). did you give any advertisement on cable channels? 1) yes 2) no if yes, then how much is the approximate total cost? 21). can you please tell us that the approximate total cost of pena flexes (billboards)? 22). can you please tell us the approximate total cost of all types of banners? 23). can you please tell us the approximate total cost of all types of posters? 24). can you please tell us the approximate total cost of all types of badges? 25). can you please tell us the approximate total cost of any kind of flushing types of bill boards? 26). can you please tell us about the approximate total cost of all types of advertisement including the one we have asked above as well as the one we have not asked for? part d: details about transportation cost during the election campaign 27). what is the approximate total cost of transportation during the election campaign before the day of the election? 28). what is the approximate total transportation cost on the day of the election? part e: public meetings during the election campaign 29). how many corner meetings did you conduct during an election campaign? 30). what was the approximately average cost of each corner meeting? 31). how many corner meetings were sponsored by other people? 32). how many large public gatherings did you conduct during an election campaign? 33). what was the average cost of each large public gathering? part f: cost on agents and other staff during the election campaign 34). did you hire any staff or agent for the purpose of an election campaign?1) yes2) no if yes, how many agents/staff? 35). what is the approximate total cost of an all agent/staff? part g: unsolicited materials cost anwar shah, nasrullah habib and muhammad zeeshan younas 116 36). can you kindly tell us about the total cost of items other than what we have asked for above? 37). did anyone ask (demand) for financial help before the election? 1) yes 2)no if yes, then how much help did you make? part h: cost on the day of election at each polling station 38). how many polling stations are in your union council? 39). how much is the average cost of each polling station on the day of the election? part i: personal expenses during the election campaign 40). did you buy or hire any special vehicle during an election campaign? 1) yes 2)no if yes, then how much is the total cost? 41). did you hire a special gunman for security purposes during an election campaign? 1) yes 2) no if yes, then how much is the total cost? 42). did you buy any weapon for security purposes during an election campaign? 1)yes 2)no if yes, then how much is the total cost? 43). did you do any personal shopping during an election campaign? 1) yes 2) no if yes, then how much is the total cost? 44). any other expense which we did not ask in the above question? 45). did you become the chief guest in any type of sports tournament during the election campaign? 1) yes 2) no if yes then, how many times? 46). did you announce any prize or sponsorship in the capacity of chief guest? 1) yes 2) no if yes then, how much? part j: donations 47). did you receive any donations? 1) yes 2) no if yes, then how much? 48). did you reject any offer of donation? 1) yes 2) no if yes, then how much? part k: source of livelihood 49). what is the nature of your job 1) lawyer 2) businessman 3) zamindari 4) other 50). did you inherit any assets from your forefathers? 1) yes 2) no if yes then, how much? journal of applied economics and business studies, volume. 5, issue 1 (2021) 83-108 https://doi.org/10.34260/jaebs.515 83 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x an empirical analysis of altruistic consumer behaviour and economic welfare of households in pakistan saira saeed1*, abdul saboor 2, & gulnaz hameed3 1 phd scholar economics, uaar 2 professor, dean faculty of social sciences, uaar 3 assistant professor, faculty of social sciences, uaar abstract the recent literature supports the fact that choices of individual are not always rational. they are bounded rational not just because of lack of complete information but due to involvement of individual personal attributes in decision making. this study is aimed at evaluating the effect of altruistic behaviour on the welfare of some selected households. cluster-based sampling has been used with a sample size of 1000 household covering 36 major districts of punjab. the analysis revealed that the traditional utility maximization model positively impacted charitable donations on household welfare. this study found a high tendency of altruistic consumer behaviour among households. altruism on the model of an individual to individual connection can improve community welfare in the long run. in this regard, the benefits of giving charity should be widely demonstrated through a formal government system, which will help improve the social welfare of society. by publicizing the services of charity for donor, the contribution will increase across all income groups. the government needs to develop specific policies that can use charity for community development. by mobilizing individuals' donations, the burden on the part of the government can be shared in terms of projects based on the social safety net. keywords altruistic behaviour, economic welfare, household, pakistan jel classification e70, i31, h31 1. introduction richard thaler nobel laureate 2016 played important role in development of behavioral economics. his first contribution was the validation of deviations from rational behavior in economic decision making and giving theory of mental accounting * drabdul.saboor@uaar.edu.pk saira saeed, abdul saboor & gulnaz hameed 84 to explain the cognitive operations used by economic individual to evaluate his/her economic activities. his second contribution is related to self-control problems which prevent an individual to take optimal decisions. thaler explained the saving behavior of individual in context of planner-doer model. the work of thaler’s on limited cognition and self-control has major policy implications. the third contribution was explaining the importance of social preferences in making economic decisions. the social preferences were defined in term of dictator game. his work explained that people are more concerned about fairness both in consumer and labor market. finally, his work on providing evidence of psychological aspects in economic decisions makes him eligible for nobel prize (committee, 2017). he is considered as pioneer of behavioral finance, particularly explaining investor psychology while making investment decisions. this journey motivated to determine the social preferences of individuals and how it’s going to create economic welfare for individual which was previously defined purely in context of rational decision making. recent empirical and theoretical literature has provided substantial evidence suggesting that fairness motives affect the behavior of many people. most theoretical papers describe reciprocal altruism and equilibrium behavior by considering psychological game theory which usually gives different predictions with respect to the standard notions of equilibrium in games (marco, 2010). choices studied in behavioral game theory are not always rational and do not always represent the utility maximizing choice. rushton, (1981) developed a scale to report self-reported altruism. this scale is used to analyze the altruistic behavior of an individual. it is not just limited to giving charity to others but it also include giving directions to some unknown individual, help in carrying belonging of some unknown person, holding door for someone, allowing someone to go in front of you in line, giving neighbor some valuable item of yours, helping a handicap to cross the road and giving seat to someone in bus who is standing. high scores clearly exhibit that individual ha high altruism. the theoretical analysis revealed that individual gifts to charity are interdependent. the individual contribution to charitable organizations is highly influenced by contributions from other individuals belonging to same group (andreoni, 1998). the evidence is neither significant nor large. this means that standard models which ignore such kind of interdependence of preferences may not be misleading. in case of pakistan, the reports reveal that in pakistan the total charity collection is equal to one percent of total gdp. and that is what reported by the people. lot of charity goes to the people which is not properly reported or donor want to remain anonymous. it is observed that 98 percent of total pakistani population takes part in charity abased activities as reported by pakistan center for philanthropy (pcp). in a study conducted by pcp it was revealed that total $2 billion which are equal to rs. 240 billion are yearly donated by people of pakistan. pakistan is ranked amongst uk and canada where total donations is 1.3 percent and 1.2 percent of total gdp respectively. it is believed that the main reason of such donations is based on religion. according to basic teaching of islam journal of applied economics and business studies, volume. 5, issue 1 (2021) 83-108 https://doi.org/10.34260/jaebs.515 85 every well-established person is bound to give part of their wealth to poor in terms of zakaat. the survey reported that pakistanis prefer to give donations on their own because this gives them more satisfaction. it was suggested that, to have productive use of resources, the donations should be collected in a structured way. out of sample selected two-third of individuals reported that they give donations to the needy by their own hand. while one-third prefer to give charities to the poor. the reported reason of not giving charity donations to charitable organization is lack of trust on charity based organizations. the other factors identified for donations in pakistan are religion and feeling of sympathy for others. the lower middle class and middle class of pakistan is directly linked to the needy so ofhen prefer to give donations by their own hands. while the rich class is not directly linked to the poor’s, so they prefer to give to the organizations. the rationale of the study is to analyze how altruistic activities can bring economic wellbeing for an individual. by promoting such activities at national level, government would be able to solve the problems of local community at large. the social safety net provided by the government has few limitation which can be shared by the people who are willing to take responsibility of others. the current analysis will help to promote altruistic activities and develop litertsure in this respect which create space for theorys of irrationality in economics. 2. literature review pigou define economic welfare as “that part of social welfare which can be brought, directly or indirectly, into relation with the measuring-rod of money” and the phenomena latterly defined as “the economic welfare of a community consists in the balance of satisfactions from the use of the national dividend over the dissatisfactions involved in the making of it” (hicks, 1975). the economic wellbeing is defined by different indicators. the major indicators of economic and social wellbeing are, measure of economic welfare (mew), genuine progress indicator (gpi), index of economic well-being (iewb), index of social health (ish), fraser institute index of living standards (ils), human development index (hdi), quality of life index (qol) and index of social progress (isp) (sharpe, 1999). the economic wellbeing will be used to see whether altruism has any impact on it. there are few barriers in measuring economic welfare. context factors which is embedded in policy factor while indicator factors which mainly depend on level of experience and expertise of the users of indicators and the institutional culture in which they operate (bleys, 2015). the value of different measures of economic welfare can be increased by updating methodological framework, extending macroeconomic models to incorporate more welfare related items, improved communication and researcher skills of entrepreneurship. the measure of economic welfare should include health, social and environmental statistics to measure overall wellbeing of an individual. saira saeed, abdul saboor & gulnaz hameed 86 the research reveals that subjective wellbeing is positively and significantly correlated with social activities. the inverse causality prevails which means that satisfaction in life induces more social behavior in an individual. the phenomena can be estimated through standard ols and ordered probit estimates (becchetti, 2013). the basic model still remains same but the preference of individual should be reported more realistically. the only significant difference prevails is the difference in type of motivation which vary from individual to individual. the phenomena also vary across different cultures and age group because of different set of norms. the paradox of happiness reveals that the quality of genuine relational life that is more important as compare to income while estimating subjective wellbeing. in economics, the relational goods have positive impact on life satisfaction. the relational ties, voluntary activity and number of people whom respondent can rely are positively and significantly related to life satisfaction (becchetti, 2015). the direct causality hypothesis if supported in different cultures, support that high quality of relationship should bring higher subjective wellbeing. in short, individual donate time and money in order to have strong relationships and without having strong relations, economic stability is not possible because economic vitality depend on quality of interpersonal relationships. there are different approaches through which charitable giving can be studied. the first approach is giving as individual economic decision. in this approach the supply of donations is determined in traditional utility maximization framework subject to budget constraint. the second approach is giving as strategic interaction. in this approach the role of donors, fund raising organizations, government and charitable foundations is significant. the third approach is giving as social exchange. in this scenario presence of overt request is necessary which may come from a friend, a colleague, and a door to door solicitor, a call from fund raising organization, a media campaign, and any natural disaster. the very recent approach is giving as response to empathic, moral, or cultural urges (andreoni, 2013). the analysis of all approaches will help to identify the social and economic wellbeing of a particular household. the economic gain is primary motive as explained by basic economic theories. an empirically grounded theory can be used to explain other motives like altruism and organizational identifications (simon, 2018). according to that theory one can assume that human motives change over time. the main reason is that individuals don’t form their preferences in isolation but they are response of public events happening around them. the basic economic theory need to be altered to explain the motives other than economic welfare which determine the choice set of an individual. at organizational level, altruism induces loyalty which brings positive impact on performance of organization. not only in organizational setup but in all kind of networks, altruism play significant role in developing social capital. in a fixed network, individual also care about wellbeing of their neighbors. people prefer to help their poor friends if they find them in any kind of financial crisis. in this context, nash equilibrium can be explained by a concave function. bourles, (2017) established existence of unique equilibrium journal of applied economics and business studies, volume. 5, issue 1 (2021) 83-108 https://doi.org/10.34260/jaebs.515 87 consumption and equilibrium transfers. the whole analysis helped to state that a positive shock for an individual is going to benefits everyone around him. any kind of small change in income depends on individual current financial status. surprisingly the analysis also revealed that expansion of the altruism network may increase consumption inequality. in japan, a study was conducted to analyze the relationship of altruism with consumption of staple food like rice. the analysis found the altruism and other attitudes have a significant influence on food consumption (ujiie, 2011). these findings suggest that a pipeline to provide public goods via daily food consumption might exist. altruism promotes the efficient working of markets and allows them to function regardless of imperfect information. this revealed that altruism is real and help to promote social integration. altruism increases as much as the social complexities of society increase. for the betterment of society, the world demand social man over economic man (kennett, 1980). it’s not just the society which gets benefited by altruism but there are other benefits as well like mental, psychological, emotional, and even physical. analyzing in respect, positive impact of altruism is consistent across each segment. particularly taking about improved health, altruism improve mental health for both male and female but addition to that, female also experience improved physical health as a result of altruistic activities (fechter, 2016). community connection increases if societies promote altruism which play significant role in community development. there is a direct and positive relationship between the amount of time an individual spends volunteering and the resulting increase in well-being. volunteering increase positive effects on wellbeing of an individual and decrease negative effects like depression and stress. the volunteer behavior is more dominant in low income group. while explaining behavior of individual, certain imperfections may come in that way (povey, 2014). there are lots of reasons for these imperfections. the main imperfection is punishment system which weakens the ability of altruistic preferences. the purchase of product in international market has major impact on welfare of country from which those products belong. the employment opportunity increase for the product whose purchase is high. the altruism can explain such behavior in term of consumer ethnocentrism, cognitive moral development and pro-social behavior. the consumer ethnocentrism varies based on region of origin and pro-social behavior varies based on methods of acquisition and region of origin (powers, 2006). social capital is the set of relationship an individual have, including his/her friends, family, neighbors and co-workers. altruism and social capital may not exhibit strong association but surprisingly revealed that altruism cause an increase in social capital (theurer, 2010). in multivariate analysis altruistic behavior has statistically significant association with happiness and life satisfaction. the hedonic and eudemonic benefits of altruistic behavior appear to be shaped by the more general dimensions of social capital. when an individual prefer a over b the phenomena is explained by hedonic principle in economics. but current stream of research focus demographic, social and psychological aspects of a and b before giving explanation of behavior of a (unger, 1981). this saira saeed, abdul saboor & gulnaz hameed 88 direct to a new fertile area of research in behavioral economics in which connection between a certain type of behavior and wellbeing is explored. like why poor and rich behave differently can have reason other than what explained by commodity specific and income specific statistical contrivances. altruism is declared to be rational, if it’s captured by quasi-concave utility functions for individuals (andreoni & miller, 2002). all individuals are heterogeneous. the percentage of fairness adopted by individual is not fixed for each individual. for that reason, fairness should be analyzed at individual level. because of individual heterogeneity, a single model would fail to predict aggregate behavior of individuals. in order to capture all choices of individuals and aggregating these choices will help to understand altruism for both individuals and markets. it is claimed that non-economic behavior is beyond economic analysis. as economist it is believed that individual behavior is consistent with self-interest. in this self-interest is defined as choices that are complete, reflexive and transitive hence utility function can be used to explain behavior. what variables are included in that self-interest are explained by people itself, by their actions and choices (andreoni & miller, 1998). this means that unselfish behavior can be captured by model of self-interested agents but self-interested agents are not always money maximizing. those unselfish acts can be described and predicted with the standard neoclassical model of choice. the basic economics model assumption is that people are self-interested. this means that some choices of individual following underlying preference ordering like complete, reflexive and transitive. but this failed to explain what economic variable operate in whole framework. people from their actions; define those variables which formulate their utility function. this means that unselfish behavior can be explained by self-interested framework but that self-interested framework is not always money maximizing. the unselfish actions of an individual can be explained by neo-classical model of choice (andreoni, 2008). during last part of 20th century, there came a fundamental paradigm shift in social sciences. in the previous literature, the altruism and true generosity was not intellectually acceptable. this kind of behavior can be explained with comparison to selfish consumer in economics. the recent development in literature creates lot of room for existence of true generosity and explained that it is an important characteristic of human being. such kind of behavior is fundamentally opposite to the basic ideas of utility theory which take selfishness as central motivation for consumer behavior. the point is that generosity conflict with basic economic theory and it is different at each level due to number of factors. the present utility maximization models fail to provide explanation of human behavior at these different levels. secondly, the preferences are endogenous which are highly responsive to social and economic circumstances around them. this means that the final consumption output cannot incorporate the true picture of human welfare. thirdly, social norms are the major determinants of human behavior, which means that human behavior has certain kind of flexibility which is not incorporated in basic economic models (karacuka, 2012). thus neoclassical model of journal of applied economics and business studies, volume. 5, issue 1 (2021) 83-108 https://doi.org/10.34260/jaebs.515 89 choices is recommended explaining the concept of altruism and true generosity. the individual who have higher education level usually provide unpaid assistance and emotional support. however the financial support is not associated with higher education level. the people belonging to older age group are less likely to give emotional support and unpaid assistance but no significant association was found with financial support (fujiwara, 2009). people who have some association with broken families are more likely to provide financial assistance. the gender and employment level don’t have any significant association with having empathy for others. the dictator game has been analyzed for altruistic behavior using large random sample size in case of netherlands. in the whole survey six percent of respondent has donated money. the analysis revealed that generosity increase with age, education, income, trust and pro-social value orientation (rene, 2007). the pro-social behavior has been analyzed to understand behavior of students in dictator and ultimatum games. one student sample has been used which has reciprocity and lower level of trust. prisoner’s dilemma games can be analyzed in the current context. four different sections have been used for analysis. among them, one group exhibit higher pro-social behavior and was an outlier. women showed more cooperation in prisoner’s dilemma but lower trust in trust game. men and women both behaved same in dictator and ultimatum games (chaudhry, 2011). but phenomena were observed in experimental environment that need to be analyzed in realistic environment. people gain utility from the act of giving, they also gain more utility by increasing its supply. the impure altruism model leads to predictions that are intuitive (andreoni, 1990). the redistribution to more altruistic people from less altruistic people will increase total provision, which may lead to desired effect of giving subsidies. the altruism decline with income and when altruism is considered, the effect of policy sometime got reversed. the conventional view of charitable giving may be inaccurate but the importance of developing interdependence of preferences and empirical models that incorporate impure altruism is still there. charitable giving is perennial topic in current economic research. it is very challenging task to determine the motives of giving, the operations of charitable institutions for policy analyst to measure benefits of charity and for experimenters to explore market of charitable giving. over the period of time the government become more reliant to private institutions for providing public goods and donations to charity has become increasingly sophisticated at raising money (andreoni, 2007). the demanders of charity will become essential for calculating the social costs and benefits of charitable institutions. it is revealed that altruistic preferences come from culture (andreoni, 2007). they are seems to be acquired through psychological development and socialization using children as mode of transmission. the third argument is that people are innately wired to care. the mri images studied by many scientists revealed that human brain has isolated centers involved in altruism. the phenomena are not just limited to experimental boundaries but it’s valid outside experimental boundaries in real term. saira saeed, abdul saboor & gulnaz hameed 90 the analysis of determinants of charitable giving in a study based on netherlands revealed that richer, religious and female respondents are more likely to give. no significant impact of perceptual and attitudinal factors was found. the donations in term of money, in-kind and time are not significantly influenced by altruism, warm glow, prestige and reciprocity, and government policies. the only major motivator found was family which implies that if current generation has habit of giving, that would result in giving habit of future generation as well (nu, 2015). some people believe that either people give more because it is cheap to give or individual has large income. but charity is sensitive to temporary or permanent changes in variables. researchers used classical model of charitable giving to determine benefits of it based on public and private characteristics. the empirical analysis revealed that private benefits are the primary motive for giving. the classical model mislead when data is interpreted because empirical analysis give limited evidence to support the common belief that donors give because they care about the nonprofit’s output (vesterlund, 2006). 3. methodology public economics has remained active to develop a comprehensive literature on justifying charitable giving’s in the framework of economic theory. in 1970s, the first policy study was published on given topic by martin feldstein. currently charitable giving’s has been discussed under different areas of economics including, economic theory, game theory, applied econometrics, experimental and behavioral economics. the economic theory is trying to identify how charitable giving is influencing of associated with economic variables. the form of preferences in case of charitable giving is critical. it is assumed that charity has same model as public goods but it is observed that despite income level, people prefer to give charity which means that motives other than charitable output are involved in decision regarding giving charity. a comprehensive research on theoretical, experimental and behavioral aspect of giving is being done to identify the motives of charitable giving’s. to explain individual preferences for charitable giving’s current study considers an individual 𝑖 having an income 𝑚𝑖 who have consumption of private good 𝑥𝑖 and donate 𝑔𝑖 to charity. for 𝑛 individual, let’s assume 𝑮 as the total contribution to charity. in case of pure altruism (andreoni, 1989) in which an individual gain utility form final output of charity then 𝑮 is a public good. so the first model which defined utility as function of altruism assumed preferences as 𝑈𝑖 = 𝑢𝑖 (𝑥𝑖 , 𝑮) but such a model may lead to meaningless conclusions like may be only a part of total population contributes to charity as discussed by varian (1986) and andreoni (1988). so any model which is based on empirical findings, need to experience higher utility from their charities as compare to others that is 𝑈𝑖 = 𝑢𝑖 (𝑥𝑖 , 𝑮, 𝑔𝑖 ) journal of applied economics and business studies, volume. 5, issue 1 (2021) 83-108 https://doi.org/10.34260/jaebs.515 91 where utility is increasing in third argument in which it is assumed that only warm glow component of utility is going to have significant impact on an individual giving decision. this impure altruism was given by andreoni in 1990 in his theory of warm glow giving. may studies in the past, has giving strong evidence regarding existence of warm-glow. different laboratory experiments have shown response to manipulations giving support for existence of basic concept of “joy of giving”. a study in this respect has shown that 57 percent of subject has given share of their income to charity. although the contribution was not that much high neither it increased the overall contribution by any significant amount still the study has given strong evidence in support of warm glow (crumpler & grossman, 2008). the main analysis of the study is based on analyzing the possible connection between altruism and economic status of household. the simple regression model is used to predict the relationship between two variables. to make the analysis more specific different income proxies has been used to predict the accurate relationship between the two variables. 3.1 relation between income of household and altruism the first regression is based on predicting the relationship between economic status of household and altruism by taking income of household as proxy of economic status. the regression would take functional form of log-linear in which log of yearly income (2017-18) would be taken as dependent variable while the behavior of given donation as proxy of altruism is taken as independent variable along with a set of controlled variables. the model estimated to predict the relation between altruism and net income is as follows. 𝐿𝑜𝑔(𝐼𝑛𝑐) = 𝛼(𝐷𝑜𝑛𝑜𝑟) + 𝛽1(𝐻𝐻𝑠𝑖𝑧𝑒) + 𝛽2(𝐸𝑚𝑝𝑙𝑜𝑦𝑒𝑒𝑑 𝑖𝑛 𝐻𝐻) + 𝛽3(𝐿𝑖𝑡𝑒𝑟𝑎𝑡𝑒 𝑖𝑛 𝐻𝐻) + 𝜀 where, log(inc) (yearly income 2017-18) is taken as dependent variable. the independent variables include possibility of being donor which take one if the person is charity donor and zero otherwise. the set of controlled variables include hhsize (household size), employed in hh (total number of employed individual in household) and literate in hh (total number of literate individual in household). the proposed hypothesis claims that there is positive relationship between income and altruism. for set of controlled variables, it is assumed that higher household size would lead to negative pressure on economic welfare of household, while more number of employed and literate members in household will lead to positive impact on welfare of household. 3.2 relation between expenditure of household and altruism the second regression is based on predicting the relationship between economic status of household and altruistic behavior of an individual by taking expenditure of household as proxy of economic status. the regression would take functional form of log-linear in which log of yearly expenditure (2017-18) would be taken as dependent saira saeed, abdul saboor & gulnaz hameed 92 variable while the behavior of given donation as proxy of altruistic behavior is taken as independent variable. the model estimated to predict the relation between altruism and net income is as follows. 𝐿𝑜𝑔(𝐸𝑥𝑝) = 𝛼(𝐷𝑜𝑛𝑜𝑟) + 𝛽1(𝐻𝐻𝑠𝑖𝑧𝑒) + 𝛽2(𝐸𝑚𝑝𝑙𝑜𝑦𝑒𝑒𝑑 𝑖𝑛 𝐻𝐻) + 𝛽3(𝐿𝑖𝑡𝑒𝑟𝑎𝑡𝑒 𝑖𝑛 𝐻𝐻) + 𝜀 where, log(exp) (yearly expenditure 2017-18) is taken as dependent variable. the independent variables include possibility of being donor which take one if the person is charity donor and zero otherwise. the set of controlled variables include hhsize (household size), employed in hh (total number of employed individual in household) and literate in hh (total number of literate individual in household). the proposed hypothesis claims that there is positive relationship between expenditure and altruism. for set of controlled variables, it is assumed that higher household size would lead to negative pressure on economic welfare of household, while more number of employed and literate members in household will lead to positive impact on welfare of household. 3.3 instrumentation of data collection and sampling method the questionnaire has been developed to test the desired relationship between the variables. the questionnaire was designed to incorporate all the necessary information required for analysis. the first part of questionnaire is based on income profile of household in which income for period 2007-08 and 2017-18 has been asked. same for the expenditure profile of household. the households overall expenditure of health was incorporated in separate section. the next section was developed to incorporate altruistic behavior of respondent. the question regarding altruism, general perception of respondents regarding generosity and identification of passive donor has been incorporated. the probability sampling was used to cover the province of punjab which comprised of 36 major districts. based on previous literature and statistical analysis a sample of 1000 respondents was finalized based on cluster base sampling. the population proportion of 36 major districts of punjab was used to identify sample size for each district. (sample detail attached at appendix-ii) the multiple indicator cluster survey 2014 published by bureau of statistics punjab, (2014) has been used to identify population proportion in each district. (questionnaire attached at appendix –i) 4. results the analysis of the study is based on analyzing the possible connection between altruism and economic status of household. the simple regression model is used to predict the relationship between two variables. to make the analysis more specific different income proxies like household income and household expenditure has been used to predict the accurate relationship between the two variables. journal of applied economics and business studies, volume. 5, issue 1 (2021) 83-108 https://doi.org/10.34260/jaebs.515 93 4.1 descriptive statistics income profile the mean monthly salary of households for the period 2007-08 is 24965. the deviation is huge like the minimum monthly salary is 2000 while maximum is 400000. the rent form property is 21414 on average at monthly basis. however earning from saving account, remittance from abroad and financial assistance from government is not major source of income. at maximum 15 respondents reported that they receive remittance form abroad. the mean monthly salary reported for the period of 2017-18 is 50865. the deviation is again highest. the minimum amount of salary per month is 5000 while maximum is 1500000. the rent from property on monthly basis is reported 36297 on average. the income from other sources like earning from saving account, remittance from abroad and financial assistance from government is on maximum availed by 41 respondents. the mean yearly income of household reported for the period 2017-18 is 672979. table 1: descriptive statistics income profile descriptive statistics (2007-08) (2017-18) income profile mean std. dev. mean std. dev. salary/main business (monthly income) 24965 32129 50865 81768 rent from any property (monthly income) 21414 27336 36297 50857 earning from saving account (yearly) 260000 473127 34259 42647 remittance from abroad (yearly) 2553333 242343 339365 440795 financial assistance from government (yearly) 13600 13575 18936 17153 others 56000 97020 149228 236927 total income of household in base year 320018 403250 672979 1125589 4.2 descriptive statistics expenditure profile the mean monthly expenditure for food consumption reported by household in 2007-08 is 10576. the items included under food are spices, sugar, grains, cereals, oil, fats, tea, milk, baked and fried products. the mean yearly expenditure for non-food items like travelling, clothing, health, housing, education, taxes, furniture, utility bills and loan installments is 72235. the selected durable items owned by household worth 16169 on average. the durable items include refrigerator, freezer, ac, air cooler, geyser, washing machine, car, motor cycle, laptop and mobile phone. the total yearly expenditure of household for the period 2007-08 is 213485 on average. table 2: descriptive statistics expenditure profile descriptive statistics (2007-08) (2017-18) expenditure profile mean std. dev. mean std. dev. food consumed by household (monthly) 10576 13402 20015 19823 non-food items consumed by household (yearly) 72235 107016 147267 177914 selected durable items hold by household 16169 30840 60489 78416 total expenditure of household 213485 232803 458568 618069 saira saeed, abdul saboor & gulnaz hameed 94 the mean monthly expenditure for food consumption reported by household in 2017-18 is 20015. the items included under food are spices, sugar, grains, cereals, oil, fats, tea, milk, baked and fried products. the mean yearly expenditure for non-food items like travelling, clothing, health, housing, education, taxes, furniture, utility bills and loan installments is 147267. the increase is mainly due to high prices of electricity and utility bills. the cost of schooling also increases during last decade. the selected durable items owned by household worth 60489 on average. the highest increase is mainly because of the fact that pakistan has experienced highest rate of inflation during last decade resulting in significant increase prices of electronic items. the durable items include refrigerator, freezer, ac, air cooler, geyser, washing machine, car, motor cycle, laptop and mobile phone. the total yearly expenditure of household for the period 201718 is 458568 on average. 4.3 relation between income of household and altruism table 3: log-linear regression results (income) number of observations 1000 f (4,995) 54.12 prob > f 0.0000 r-squared 0.1787 adj. r-squared 0.1754 root mse 0.72588 log income 2017t coefficient tvalue donor charity 0.1973308*** 3.14 household size -0.0964194*** -6.36 no. of employed in hh 0.1259648*** 5.64 no. of literate in hh 0.1655475*** 11.52 constant 12.48966*** 144.81 notes: (***) significant at 1% level, (**) 5% level, and (*) 10% level. the log-linear regression having log of yearly income (2017-18) as the dependent variable while the behavior of given donation as proxy of altruism is taken as an independent variable along with a set of controlled variables revealed that the coefficient of donor charity is 0.19 which is positive and highly significant considering t-stats and p-value. the coefficient of household size is negative and highly significant which means that higher household size lead to negative impact on income of household. the coefficient of number of employed in a household and number of literate in a household are positive and significant. on basis of this it can be stated that more employed and educated individuals will lead to positive impact of economic welfare of household. journal of applied economics and business studies, volume. 5, issue 1 (2021) 83-108 https://doi.org/10.34260/jaebs.515 95 figure 1: dot plot income of household the dot plot of income of household is given in graph below revealed that the dependent variable is normally distributed. this is in line with the assumption of basic linear regression model which require that the variable should be distributed normally. 4.4 relation between expenditure of household and altruism the regression is based on predicting the relationship between economic status of household and altruistic behavior of an individual by taking expenditure of household as proxy of economic status. the log-linear regression is used in which log of yearly expenditure (2017-18) has been taken as dependent variable while the behavior of given donation as proxy of altruistic behavior is taken as independent variable. the set of controlled variables include household size (household size), employed in household (total number of employed individual in household) and literate in household (total number of literate individual in household). the coefficient of donor charity is 0.139 which is positive and highly significant considering t-stats and p-value.the coefficient of household size is negative and highly significant which means that higher household size lead to negative impact on income of household. the coefficient of number of employed in a household and number of literate in a household are positive and significant. on basis of the results it can be stated that altruism has positive impact on expenditure (taken as proxy of economic welfare) of household. table 4: log-linear regression results (expenditure) number of observations 1000 f (4,995) 51.66 prob > f 0.0000 r-squared 0.1720 adj. r-squared 0.1686 root mse 0.63816 log expense 2017t coefficient tvalue 11 .5 12 12 .5 13 13 .5 14 fi tte d va lu es 0 50 100 150 frequency saira saeed, abdul saboor & gulnaz hameed 96 donor charity 0.1395278** 2.53 household size -0.0504473*** -3.78 no. of employed in hh 0.08163*** 4.16 no. of literate in hh 0.1421246*** 11.25 constant 12.19507*** 160.83 notes: (***) significant at 1% level, (**) 5% level, and (*) 10% level. the dot plot of income of household is given in graph below clearly revealed that the dependent variable is normally distributed. this is in line with the assumption of basic linear regression model which require that the variable should be distributed normally. figure 2: dot plot expenditure of household 5. conclusion the final regression based on log linear regression tried to explore the relationship between altruism and economic welfare by taking income and expenditure as proxies of economic status of an individual. the analysis revealed that welfare level of a household like income and expenditure have positive relation with altruism. so it can be claimed that altruistic behavior induce economic welfare. the other controlled variables like household size has negative impact on economic welfare of household but more number of educated and employed individuals increase economic status of household. on the basis of this it can be suggested that the benefits of altruism should be communicated at national level so that the behavior can be adopted at national level in order to share the burden of the government. the analysis gives major policy implication for developing countries especially countries like pakistan who face major debt crisis. altruism on model of individual to 11 .5 12 12 .5 13 13 .5 fi tte d va lu es 0 20 40 60 80 100 frequency journal of applied economics and business studies, volume. 5, issue 1 (2021) 83-108 https://doi.org/10.34260/jaebs.515 97 individual connection can improve community welfare in long run. in this regards, the benefits of giving charity should be widely demonstrated through formal government system which in return will help to improve social welfare of society. by publicizing, the benefits of charity for donor, contribution will increase across all income groups. government need to develop certain policies which can use charity for community development. by mobilizing the donations of individuals, the burden on part of government can be shared in terms of projects based on welfare of general public. saira saeed, abdul saboor & gulnaz hameed 98 references andreoni, j. (1998). toward a theory of charitable fund-raising. journal of political economy, 28. andreoni, j. (1989). giving with impure alturism: application to charity and ricardian equivalence. the journal of political economy, 1447-1458. andreoni, j. (1990). impure altruism and donations to public goods: a theory of warm-glow giving. the economic journal, 464-477. andreoni, j. (2007). giving gifts to groups: how altruism depends on the number of recipients. journal of public economics, 1731–1749. brent bleys, a. (2015). barriers and opportunities for alternative measures of economic welfare. ecological economics, 162–172. crumpler, h., & j.grossman, p. (2008). an experimental test of warm glow giving. journal of public economics. fechter, l. (2016). altruism and well-being. western oregon university honors senior theses/projects, 62. fujiwara, t. (2009). is altruistic behavior associated with major depression onset? plos one. james andreoni, w. t. (2007). altruism in experiments. new palgrave dictionary of economics,. james andreoni, j. m. (2002). giving according to garp: an experimental test of the consistency of preferences foraltruism. econometrica, 737-753. james andreoni, j. m. (1998). an econometric analysis of charitable giving with independent preferences. economic inquiry, 410-428. kennett, d. a. (1980). altruism and economic behavior: ii: private charity and public policy. the american journal of economics and sociology, 337-353. kristine theurer, a. w. (2010). altruistic behaviour and social capital as predictors of well-being among older canadians. ageing and society, 157 181. leonardo becchetti, l. b. (2015). human values, civil economy and subjective well being . world happiness report. leonardo becchetti, l. c. (2013). sociability, altruism and subjective well-being. center for economic and international studies, 43. mehmet karacuka, a. z. (2012). the empirical evidence against neoclassical utility theory: a review of the literature. international journal pluralism and economics education, 366–414. nu, n. n. (2015). economics of charitable giving: understanding the motivation of donation behavior. institute of social studies the hague the netherlands. povey, r. (2014). the limits to altruism a survey. hertford college, oxford university. renaud bourles, y. b.-r. (2017). alturism in networks. econometrica, 675–689. rene, b. (2007). measuring altruistic behavior in surveys: the all-or-nothing dictator game. survey research methods, 139-144. sharpe, a. (1999). a survey of indicators of economic and social well-being. centre for the study of living standards, 72. simon, h. a. (2018). altruism and economics. american economic association, 156-161. thomas l. powers, r. a. (2006). alturism and consumer purchase behaviour. journal of international consumer marketing, 13. theodore suranyi-unger, j. (1981). consumer behavior and consumer well-being: an economist's digest. journal of consumer research, 132-143. journal of applied economics and business studies, volume. 5, issue 1 (2021) 83-108 https://doi.org/10.34260/jaebs.515 99 theresa thompson chaudhry, m. s. (2011). norms of cooperation, trust, altruism, and fairness: evidence from lab experiments on pakistani students. the lahore journal of economics, 347375. ujiie, k. (2011). the effect of altruism on consumer behavior in japan: an analysis on rice consumption using scanner data. eaae 2011 congress change and uncertainty: challenges for agriculture, food and natural resources, 10. vesterlund, l. (2006). why do people give? the nonprofit sector a research handbook . saira saeed, abdul saboor & gulnaz hameed 100 appendix –i dear participant, my name is saira saeed. i am a student of phd economics (hec indigenous scholar) at university of arid agriculture rawalpindi. here is my final thesis to finish the degree; my area of research is regarding behavioral economics. the purpose of the research is to figure out the empirical connections between altruism and economic welfare of individual donor. therefore, i would like to invite you to participate in this research study by fulfilling the survey. all of your responses are only used for research purpose. moreover, you do not need to provide your name for anonymous and confidential information. please complete the questionnaires honestly and as best as you can. it will take maximum 30-40 minutes. thank you so much for providing useful information and, even, supporting me to finish an important part of this project. if you have any questions or feedbacks about this project, please ask me via the following contract information. sincerely yours, saira saeed journal of applied economics and business studies, volume. 5, issue 1 (2021) 83-108 https://doi.org/10.34260/jaebs.515 101 the empirical connections of altruistic consumer behavior to economic welfare: empirical evidence from punjab, pakistan personal information regarding household id code relation with head age marital status gender (male=1) (else=0) employment education (highest) 01 02 03 04 05 06 07 08 09 10 11 12 relation with head employment marital status head =1 working – paid employed = 1 married =1 spouse = 2 working – self employed = 2 widowed =2 son/daughter = 3 not working – temporary leave = 3 divorced = 3 grand son =4 not working – looking for work = 4 separated = 4 father/mother = 5 not working – retired = 5 never married = 5 brother/sister = 6 not working – disabled = 6 nikkah (no rukhsati) = 6 other = 7 not working – other = 7 other = 7 income profile of household source of income (collective*) base year 200708 current year 2017-18 salary/main business (monthly income) rent from any property (monthly income) earning from saving account (yearly) remittance from abroad (yearly) financial assistance from government (yearly) other (like receivables from others) *if more than one person is involved in economic activity then enter aggregate income of all members of household expenditure profile of household main heads of expenditure base year 2007-08 current year 2017-18 saira saeed, abdul saboor & gulnaz hameed 102 food (monthly) spices, sugar and juices, pulses grains and cereals, edible oil and fats, tea and coffee, baked and fried products nonfood (yearly) • travelling • clothing&footwear • personal effect and health • housing • education • utility bill • maintenance charges • loan installments selected durable consumption items (mark) • refrigerator • freezer • air cooler • air conditioner • geyser • washing machine • car • motor cycle • laptop • others according to your expenditure pattern, which head take larger portion of your income? 1) food 2) luxury activities (like going for dinner, shopping and movies) 3) luxury items (like purchasing car, laptop and mobiles) 4) utility bills and house maintenance 5) children education journal of applied economics and business studies, volume. 5, issue 1 (2021) 83-108 https://doi.org/10.34260/jaebs.515 103 health profile of household household member name health issue faced during last 6 months duration till full recovery cost incurred economic welfare of respondent monetary status questions yes no do you or your spouse have any saving or current account? did you applied for any credit during last year? bank/ lender approved your application easily? do you experience any financial hardship such as a job loss, drop in income, health emergency, divorce, or loss of your home? did you sustain that hard time? over the past 12 months have you or your household received any financial assistance from your family or a friend to cover expenses after a financial hardship? over the past 12 months have you or your household provided any financial assistance to a friend or family member to cover expenses after a financial hardship? do you have more than one source of income? perception for future financial status 5 4 3 2 1 do you prefer doing government job? do you prefer doing business rather than job? did you opt for any business?(if you are a government employee) did you faced any constraint doing business?(related to previous question) did that constraint altered your motivation level? (connected to previous question) do you have a behaviour of taking new initiatives in life?(related to income generating activities) do you have defined plan to expand your income in future? do you think the decision making should be decentralized in a household? do you have decentralized decision making in your house? do you think you are managing your financial requirements efficiently? do you prefer to have your own house if you can afford? saira saeed, abdul saboor & gulnaz hameed 104 do you have retirement plan? do you think you will have sufficient amount at your retirement? do you think you have enough saving? if you were to lose your main source of income (e.g. job, government benefits), could you cover your expenses for 3 months? (5 = yes, 4 = may be yes, 3 = don’t know, 2 = may be no, 1 = no) perception about relative financial status better off same worse off compared to 12 months ago, would you say that you (and your family living with you) are better off, the same, or worse off financially? think of your friends and social circle around you. would you say you (and your family living with you) are better, the same, or worse off financially than they are? think of rest of world. would you say you (and your family living with you) are better, the same, or worse off financially than they are? think of your parents when they were your age. would you say you (and your family living with you) are better, the same, or worse off financially than they were? think about the next generation of your family (e.g. your children, nieces, nephews, etc.). when they are your age, do you think that they will be better off, the same, or worse off financially than you are today? think of strong financial status. would you say you (and your family living with you) are better, the same, or worse off if you have strong financial status which may lack peace of mind? think of living welfare oriented life. would you say you (and your family living with you) are better, the same, or worse off if you have peaceful life rather than higher financial status? altruistic behavior of respondent please answer the following question about your charity/ altruistic activities for the last 12 months. a profile of reflection on altruism 5 4 3 2 1 1 it is always peaceful to help others 2 in kind help is better than in cash 3 helping others may shake their ego 4 donation play key role in welfare of society 5 practicing charity enhance your welfare level 6 we can practice charity by offering gifts 7 i consider it important to help the poor and the needy 8 i often feel concern for people who are less fortunate materially than me 9 i am often touched by what other people go through 10 the world/ society/country demands responsible citizens (5 = strongly agree, 4 = agree, 3 = neutral, 2 = disagree, 1 = strongly disagree) journal of applied economics and business studies, volume. 5, issue 1 (2021) 83-108 https://doi.org/10.34260/jaebs.515 105 time use profile base year 2007-08 current year 2017-18 sleeping hours (daily) work time (daily) leisure and pleasure (daily) transport (daily) time with parents (weekly) visiting friends (weekly) visit to hospital (weekly) helping others (weekly) time consumed in non-economic activities like cleaning (weekly) analysis of general perception 5 4 3 2 1 giving is better than taking most of people prefer to get their own benefit time is an important source of wealth and happiness the size of our basket of good is going to be increased. it has become difficult to meet our needs in present economic situation, helping others has become difficult life has become a luxury for us our forefathers were happier than us. our behavior is related to our expenditure those who help other always lead a happy life some people are habitual of weeping their economic distress. a desire to help others is a blessing facilitating others in any respect gives satisfaction. people need to care about wellbeing of other as they care about themselves. tragedies motive us to be kind to others. being optimistic is a family trait. respect can be earned by helping others tax benefits help to promote charity people earn local prestige by giving more donations facilitating others bring peace of mind charity can improve community development charity based development projects are good for economy giving to charity organizations is not secure (5 = strongly agree, 4 = agree, 3 = neutral, 2 = disagree, 1 = strongly disagree) saira saeed, abdul saboor & gulnaz hameed 106 charity taking household profile 5 4 3 2 1 1 i prefer to work for my own welfare rather than that of others. 2 social safety net helps to improve condition of poor. 3 the initial help in any form help individual to stand on his/her own feet. 4 prolonged dependence on charity is bad. 5 to help others it’s important that you (helper) don’t need it first. 6 only those who have financial stability in their lives, help others. 7 to avoid misuse of charity, monitoring is necessary. 8 if someone helps you, you need to help someone else so that chain of help continue. 9 desire to help others is dominating in those who once receive it. 10 societies grow if people help each other without expecting reward. (5 = strongly agree, 4 = agree, 3 = neutral, 2 = disagree, 1 = strongly disagree) inter-generational welfare profile relation education employment tangible property (worth today) bank account (yes / no) land holding agri.land house, building, plaza, shops parents father mother grand parents (from father side) grand father grand mother 3.1do you prefer donation anonymous? 1. yes 2. no 3.2the reason why you choose for your donation to appear anonymous? 1. i don’t want to be recognized 2. i don’t want any others to know i gave less 3. i don’t want any others to know i gave more 4. i don’t want any others to know my amount donation 5. other reasons: … 3.3please specify your charity performance in the last 12 months, in case of donation performance. in which form you usually donate: 1. direct cash 2. bank wire transfer 3. text mobile phone message 4. property or goods 3.4how often you have given money to a charity? 1. once in the past year 2. once a month 3. every 2 – 3 months 4. once or twice a week journal of applied economics and business studies, volume. 5, issue 1 (2021) 83-108 https://doi.org/10.34260/jaebs.515 107 5. every 2 – 3 weeks 3.5 explain which form of in kind donation you opt for in last 12 months? 1. household goods 2. used clothing 3. food 4. used furniture 5. medicines 6. book, journal 7. other 3.6which charity sectors did you choose to support? multiple responses 1. medical 2. children and teenagers 3. hospital 4. religious 5. disables 6. environment protection 7. education and scholarship 8. disaster 9. others 3.7did you choose a particular charity organization to donate or volunteer? 1. yes (please give name:…………………………..) 2. no 3.8 would you take require any financial assistance from government? 1. yes (suggest amount)………………..(monthly) 2. no 3.9 do you think benazir income support program is giving sufficient among? 1. yes 2. no (suggest amount)……………….. (monthly) 3.10 did you faced any severe health issue in last 12 months 1. yes (if yes, kindly give detail:………………………………….) 2. no thank you for your time. enumerator detail name:…………………………………………………………………………….…… district: ………………………………………………………………………………. district code:…………………………………………………………………………. tehsil:………………………………………………………………………………… tehsil code: ……………………………………………………………………………. household code:……………………………………………………………………… rural/ urban: ……………………………………….……………………………… date:……………………………………………………….…………………………. starting time: ……………………… ending time:………………….………… saira saeed, abdul saboor & gulnaz hameed 108 appendix -ii district tehsil sample size bahawalpur bahawalpur 34 107 bahawalnagar 28 rahim yar khan 45 dg khan dg khan 24 90 layyah 16 muzaffargarh 34 rajanpur 16 multan multan 48 121 khanewal 29 lodhran 17 vehari 27 faisalabad faisalabad 71 127 chiniot 13 jhang 23 tt singh 20 gujrawala gujrawala 41 145 gujrat 27 hafizabad 11 mandi bahauddin 15 narowal 17 sialkot 34 lahore lahore 94 172 kasur 30 nankana sahib 15 sheikhupura 33 sahiwal sahiwal 22 69 pakpattan 19 okara 28 rawalpindi rawalpindi 50 95 attock 18 chakwal 15 jhelum 12 sargodha sargodha 34 74 bhakkar 14 khushab 12 mianwali 14 total 1000 1000 journal of applied economics and business studies, volume. 5, issue 3 (2021) 19-42 https://doi.org/10.34260/jaebs.532 19 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x livelihood assets and outcomes of rural farm households in central khyber pakhtunkhwa of pakistan abdul hassan*1, shahnaz akhtar2 and muhammad ishaq3 1 phd scholar, the university of agriculture, peshawar, pakistan. 2 professor, the university of agriculture, peshawar, pakistan. 3 director, ssd, pakistan agricultural research council, islamabad. abstract this study explores the factors of livelihood assets possessed by small farm households in central khyber pakhtunkhwa of pakistan that determines the livelihood outcomes using sustainable livelihood framework. primary data were collected from 349 small farm households using well-structured pre-tested questionnaire having both closed and open-ended questions. the study first measured the livelihood assets worth through composite indices followed by the factors that influencing the livelihood outcomes using multiple regression model. the overall value of livelihood assets of small farm households in the study area was 0.297. the area small farmers were lacked in livelihood assets along with low level of living standards as well as economic development in the area. the empirical findings of regression model revealed that all the five capitals of livelihood asset had significant positive effect on livelihood outcomes. additionally, household active labour and education of labor earners of human capital, family land of natural capital; livestock and access to formal financial credit of financial capital, distance to public services of physical capital and membership in mfscs, access to service providers of social capital had significant positive effect on the livelihood outcomes. the study suggests that the livelihood asset should be upgraded in all capitals followed by changing the approach of agriculture departments and other allied stakeholders for developing agriculture sector and rural economy. keywords livelihood, assets, outcomes, rural, farm, pakistan jel classification n5, q0, q13 * ahassan7796@gmail.com mailto:ahassan7796@gmail.com abdul hassan, shahnaz akhtar and muhammad ishaq 20 1. introduction asset are the households’ endowment of resources that makes living. its acquirement or creation needs time and money investments (galab et al., 2006). five capitals of asset (human, natural, financial, physical and social) are identified by the department for international development (dfid) that represents the building blocks of livelihood. these capitals could be partially substituted for each other. being people first and foremost, the livelihood approach is deemed for necessitating on choices of capitals for positive livelihood outcomes. a single category of capital is not enough for yielding the entire and different livelihood outcomes that people seek (dfid, 1999 a, b, c). human capital means knowledge, skills, labor ability and good health that depend on quantity and quality of available labor (dfid, 1999c). access to and investment in education and health sectors is important in stimulating the agricultural and nonagricultural activities in rural areas that influence livelihood opportunities of household and returns on other capitals in rural areas (siegel, 2005). natural capital is the resource flows and services originated from natural resource stocks. it is vital for those where the livelihood depends mainly or partly on natural resource-based activities. infrastructure and producer goods are the physical capital necessary for supporting livelihoods (dfid, 1999c) and is considered as a critical asset by influencing the availability and accessibility of goods and services (dominique van de walle, 2000a; 2000b; siegel, 2009). financial capitals are the available stocks like savings and financial resources and regular inflows of money. it is multipurpose of the five capitals (dfid, 1999c). access to financial capital empowers people to generate stable and productive lives (meinzen-dick et al., 2011). social capital impacted the other capitals of asset. it is the networks, formalized groups membership and trust relationships, reciprocity and exchanges. this capital facilitated the co-operation and reduction of transaction costs that might be the provision as a base for informal safety nets among poor. livelihood outcomes are the output of livelihood strategies. it could be categorized as improved food security, higher income, reduction in vulnerability, increased well-being and usage of natural resources in more sustainable way subject to circumstances (dfid, 1999c). in pakistan, as well as in khyber pakhtunkhwa province, the rural economy has been highly dependent on agriculture as perceived by many researchers and policymakers. recently in the literature this view was revised and an emerging nonfarm segment has been reported with a vibrant role at an increasing trend (adams, 1993; urrehman et al., 2008; ali et al., 2014; farooq, 2014 and kanwal et al., 2016). the rural farm households were resorting to farm and non-farm sources for their livelihood journal of applied economics and business studies, volume. 5, issue 3 (2021) 19-42 https://doi.org/10.34260/jaebs.532 21 (israr and khan, 2010). the available asset with farmers has a significant base for their survival to attain decent livelihood. in khyber pakhtunkhwa, majority (81 percent excluding newly merged districts) of the population is living in rural areas (gop, 2017). more than half (57.8 percent) of the rural population are identified as multidimensionally poor (mpi). of the rural population 42.8 percent is engaged in agriculture activities to earn their livelihood mostly at subsistence level (gokp, 2010). agriculture sector on one side has not performed in line with its potential (gokp, 2015) while on the other side it is most susceptible to the negative climate change impacts (gokp, 2016). studies from khyber pakhtunkhwa showed that the rural inhabitants were not solely relying on farming but supplemented their farm income with off-farm income to secure their livelihood (israr and khan, 2010). therefore, the strategies to cope with the available asset of rural dwellers for securing their livelihood indicated scope for further research. the available evidence on the importance of asset and understanding the strategies of rural livelihood are scarce in pakistan in general and khyber pakhtunkhwa in particular. mostly available studies were concerned to livelihood sources (israr and khan, 2010), livelihood strategies (urrehman et al., 2008), livelihood diversification (israr et al., 2014). few studies to the theme have focused on income diversification (ikram, 2016), non-farm income diversification (kanwal et al., 2016; ping et al., 2016) and also crop diversification (shahbaz et al., 2017). the issues discussed in those studies were crucial; however, the evidence of the household asset that have effect on total household income of farm households was not documented. this study tried to assess the existing livelihood asset, livelihood outcome and their relationship in the framework of sustainable livelihood to help policy makers in initiating future rural development programs. the study identified main factors of farm household capitals that determines the current household incomes generated by smallholder farmers by utilizing their capitals/asset. this provides study rationale where the outcome would help in making recommendations to policy makers in terms of household capitals/asset for future designs. 2. materials and methods 2.1. description of the study area this study has been conducted in central valley plains of khyber pakhtunkhwa province of pakistan. this zone comprises 42 percent of the total household of khyber pakhtunkhwa wherein 71 percent of the households are rural dwellers while the rest 29 percent are urban households. according to latest available agriculture census (2010), abdul hassan, shahnaz akhtar and muhammad ishaq 22 the valley has 22 percent and 25 percent of the farm and cultivated area respectively (gop, 2012). in the climate regime, the valley will receive less precipitation than before that will probably become water stressed region in the coming years and contribute to droughts. the vulnerable sectors in the central valley plains are the agriculture, water and biodiversity (gokp, 2016). 2.2. sampling frame multistage sampling techniques was applied during this study in order to cover the full spectrum of small farm households and to meet the study objectives. in stage i, two districts, peshawar and nowshera were selected from the central valley plains followed by selection of one tehsil from each district in stage ii, two union councils from each tehsil were selected in stage iii and in stage iv two villages were randomly selected from each union council of the selected tehsils (table 1). a list of small farm households of each village was prepared in the study area. the total numbers of farm households in the selected eight villages was selected by utilizing the solwin sampling procedure used and referred by khan (2014) following mwakaje (2013): nr = n 1+ne2 …………………………..…………. (1) where, nr is the representative sample size, n is the population of total farm households in the study area and e is the desired margin of error. applying solwin sampling procedure with n population of 627 and desired margin error e of 5 percent, the representative sample size nr obtained was 244 households (table 1). after selecting the representative sample size for this study, there is a need to have a high response rate from the respondents during the survey. however, there may be non-responses. these non-responses may be due to (1) respondent refusal, (2) respondent ineligibility, (3) lack of ability for locating respondent and (4) respondent positioned but not able to contact. as a result, data collected will not represent the total population and will be biased. therefore, it is necessary that sample size had enough in the data (saunders et al., 2003). hence, actual sample size was estimated using equation 2 from saunders et al., 2003: na = nr re% x 100 …………………………..………… (2) where, na is the actual sample size, nr is the representative sample size of total farm households in the study area (obtained by using equation 1) and re% is estimated journal of applied economics and business studies, volume. 5, issue 3 (2021) 19-42 https://doi.org/10.34260/jaebs.532 23 response rate expressed as a percentage. babbie (1990) suggested a 60 percent of response rate as ‘good’ and 70 percent response as ‘very good’. applying equation 2 with representative sample size nr 244 and response rate of 70 percent, the actual sample size na obtained was 349 farm households (table 1). additionally, proportional allocation technique was also applied for achieving the requisite sample size by village. 2.3. data for this study, data has been collected from primary source. primary data were directly obtained from farm households. the sampled respondents were interviewed personally by framing a well-structured pre-tested questionnaire developed in the light of study objectives. primary data were collected from the respondents on the convenient places. the data has been collected in the month of april 2019 to june 2019. the sampled farmers are promised that the data collected will be strictly confidential and will only be used for this research work. after developing rapport with the farmers, they agreed to unveil their real income. table 1. distribution of farm households by village villages small farm households in each village sample households shiggi bala 85 47 bhatian 80 45 jogani 78 43 angoor kala 73 41 jabba daudzai 70 39 garhi momin 71 40 mufti 105 58 tarkha 65 36 total 627 349 source: author calculation 2.4. measuring the assets the asset indicators were measured in different scale. index for each indicator was calculated in order to normalize them. the equation adopted following hahn et al., (2009); samsudin & kamaruddin (2013); gautam & andersen (2016); xu et al., (2018); yang et al., (2018); zhifei et al., (2018) and used by undp in human development indices: index ii = ii−imin imax−imin ……………………….………..………… (3) abdul hassan, shahnaz akhtar and muhammad ishaq 24 where: ii is the actual indicator value and imin and imax are the respective minimum and maximum indicator values. these minimum and maximum values are used to transform these indicators into a normalized index. after normalizing the indices, it ranges in between 0 to 1 that indicates low to high score respectively and are unit free. next to normalization of each variable, the variables of each asset were averaged for calculating the value of each asset as: ai = ∑ index ii n i=1 n ………………....………… (4) where, ai is one of the five capital, index ii represent the indicator by i that make the livelihood capital and n is the number of variables in each capital. 2.5. multiple regression model in this research work, the functional form of the relationship of the household capitals with household output (income) could be expressed as in equation 5, assuming ceteris paribus on all other factors. the null hypothesis will be that the capitals of the household had no effect on household income. e𝑖 is a random disturbance term on observations made on household i. the relationship can be described as: 𝑦𝑖 = α0 + 𝛽1hc𝑖 + 𝛽2nc𝑖 + 𝛽3pc𝑖 + 𝛽4fc𝑖 + 𝛽5sc𝑖 + e𝑖…………………… (5) where; α0 is the intercept and βi are the respective parameters of the explanatory variables, i.e., livelihood capitals and 𝑦 is livelihood output, hc : household capital, nc: natural capital, pc : physical capital, fc: financial capital, sc: social capital are the explanatory variables of household i. 2.5.1. selection of functional form of multiple regression model in linear regression model, there is a linear relationship between dependent variable and independent variables. none of the variables are transformed in linear regression model. however, in log-linear model, the dependent variable is transformed by natural logarithm while the independent variables are not transformed. in log-linear models, the dependent variable need to be greater than zero. similarly, in log-log model, both the dependent and independent variables are transformed by natural logarithm. in log-log model, both dependent and independent variables must be greater than zero, as the logarithm of positive numbers is only defined. likewise, in the linear-log model the dependent variable is not transformed while the independent variables are transformed journal of applied economics and business studies, volume. 5, issue 3 (2021) 19-42 https://doi.org/10.34260/jaebs.532 25 by natural logarithm. in this form of models, the independent variables need to be greater than zero (hill et al., 2011). keeping in view the requirements of the data for different functional forms of the multiple regression model, only two types of functional forms i.e., linear regression and log-linear functional forms are applicable in this study. the functional forms of log-log model and linear-log model are not applicable to the data of this study, because both this type of functional forms have a precondition to have values of independent variables greater than zero, however, the independent variables in this study has zero values. hence two models i.e., linear regression model and log-linear regression model are used and the best fit model is then selected for further interpretations. moreover, after selecting the best fit model, the model was run for both livelihood capitals and for capital variables. similarly, equation 6 gives the relationship of household capitals with household output (income). in order to identify the influence of each variables of livelihood capitals on livelihood output, by utilizing the variable of each capital, the model could be econometrically specified as: yi = α0 + β1hlc + β2hhs + β3hel + φ1nfl + φ2nsq + γ1phc + γ2pps + θ1flk + θ2ffc + δ1sim + δ2ssp + εi…………………..… (6) where; α0 is the intercept, βi, φi, γi, θi, δi; are the respective parameters of the explanatory variables and εi is a random disturbance term on observations made on household i and yi : dependent variable i.e., the total household income of household i hlc : labour capacity hhs : health status hel : education of labor force (mean years of schooling calculated by adopting the formula from unesco institute for statistics (uis, 2013). nfl : family own land nsq : soil quality phc : housing condition pps : public services flk : livestock (unit coefficients from fao (2011) was used for calculating equivalent units except for donkeys that is presumed though discussion with livestock experts ffc : access to formal financial credit sim : institutional membership ssp : access to service providers are the explanatory variables of household i. abdul hassan, shahnaz akhtar and muhammad ishaq 26 diagnostic tests are also employed to make bold statements on the validity of models’ empirical findings. 2.6. multiple case deletion data has been checked for abnormal observations. same like xu et al., (2015) the invalid questionnaires based on nonsensical responses were excluded from analysis. additionally, in a set of data, an outlier is the observations (or subset of observations) that appear inconsistent with rest of that set of data (barnett & lewis, 1994). the presence of outliers in the data set misleads the results (zakaria et al., 2014). cook’s distance is the most important measure that detect the influence of individual or subsets of observations in linear regression for cross-sectional data (zhu et al., 2012). a rough rule of thumb is if the value of cook's distances for the outliers are > 1, then these outliers are the influential points (dhakal, 2017) and have abnormal effects (abdallah, 2013). the cook’s distance was used in this study to remove the outlier observations from the data. after cleaning the data by removing the nonsensical responses and influential outliers, the total number of observations used for data analysis were 307. 3. results and discussion 3.1. characteristics of rural farm livelihood assets 3.1.1. human capital the study results revealed that 48 percent of the sampled household’s members were in the household labour category, the rest 52 percent of the sampled household members were not fall in the labour category. moreover, more than four-fifth (82 percent) of the sampled households had no member with health issues. additionally, the overall literacy rate in the study area was very low. the average schooling of the household labour force was 2.85 years (table 2). the educational level of farmers is a good measure of the flexibility of farmers in adopting better cropping practices. 3.1.2. natural capital land is a finite resource in the world (kenea, 2008; nuru & seebens, 2008) and is an important resource due to its provision of food and other necessities that requires proper care (qasim et al., 2014). size of land holdings greatly determines the economic practice and capacity of the farm households. according to jafri (undated) consideration, farm size of 5 hectares to be the minimum farm size for sustaining a farm family. soil quality is the overall condition of soil with respect to its intended use. a good quality soil is productive and sustainable over generations for farmers. an unhealthy soil can be cherished back to health with improved management (wolkowski, 2005). the family own land was 1.59 acres. the sampled farmers have had shared-in, journal of applied economics and business studies, volume. 5, issue 3 (2021) 19-42 https://doi.org/10.34260/jaebs.532 27 shared-out and rented-in land. about three quarter of the sampled households were satisfied with quality of soil in the study area (table 2). 3.1.3. physical capital the results regarding house space per capita of the sampled households in the study area revealed that on average sampled households had an area of 2.26 marlas (1 marla= 0.00625 acre) of their houses. moreover, the average distance in kilometers to public services was 10.75 km (table 2). easy access to public services like headquarters, education, health, agricultural services, banks and markets might enhance the quality of rural livelihood. 3.1.4. financial capital livestock is vital for the livelihood to secure food and cash income to the rural poor in pakistan. it is a subsistence sector mainly possessed by small holders. the overall production of livestock is low in khyber pakhtunkhwa (hassan et al., 2014). low productivity is due to poor nutrition, weak infrastructure and lack of financial facilities along with other constraints that hinders the livestock productivity of the farmers (ishaq et al., 2016). the sampled farmers had on average 0.18 livestock equivalent units that includes cattle, buffalo, goat, sheep, poultry, donkey and horses (table 2). there are two sources of rural credit in pakistan i.e., (1) informal and (2) formal. sources of informal credit are friends/relatives and landlords while zarai taraqiati bank limited (ztbl) is the main formal source of credit for farmers (amjad & hasnu, 2007). the study results revealed that majority (66 percent) of the sampled households viewed that they cannot availed the opportunity to obtain loan from formal sources (table 2). non-availability, high interest rate and the financial situation of the farmers that they were not be able to return on time were the reasons for non-accessibility to formal institutions for obtaining loan. 3.1.5. social capital model farm services centers (mfscs) after being named the farm services centers (fscs) in 2005 was established in khyber pakhtunkhwa of pakistan for increasing the farmer’s access to the quality inputs, technical advice and experience sharing. mfscs offer one window services to the farmers by keeping all the allied representative of agriculture sectorial departments (ullah et al., 2015). the study results revealed that more than half (58 percent) of the sampled households had no membership in mfscs while remaining 42 percent had membership in mfscs (table 2). positive role of mfscs in agriculture sector can increase the agricultural yield of the farmers. abdul hassan, shahnaz akhtar and muhammad ishaq 28 agricultural service providers (asps) being a private sector already exists in the area where they provide services to the farmers. according to salunkhe & movaliya, asps can be considered an agency that provides their services with various agricultural related inputs in sufficient quantity to the farmers at their places and at right time with affordable prices and quality. the farmers discuss field level problems that they faced with them and get appropriate suggestions that influence the farmers’ decision making process (2016). hassan et al., (2017) in district faisalabad of pakistan showed a significant positive influence of asps on crop productivity, their livelihood and income. the study results revealed that 66 percent of the sampled households had not accessed agricultural service providers while remaining 34 percent had access to agricultural service providers (table 2). it is noted that due to economic conditions of the sampled farmers, there is a lack of trust between them. the asps might not help those farmers in cash or in kind who had already taken loan in cash/kind or not trust worthy. table 2. characteristics of rural farm livelihood assets in the study area assets variables unit average / no std. deviation / percent human household labour percent 47.63 20.43 household with health issues no 253 82 yes 54 18 education of labour force years 2.85 2.70 natural family land acres 1.59 2.42 soil quality no 76 25 yes 231 75 physical house space per capita marla per capita 2.26 3.11 distance to the public services kilometer 10.75 3.37 financial equivalent units of livestock number 0.18 0.29 household with access to formal institutes to obtain loan no 203 66 yes 104 34 social membership in model farm services centers (mfscs) no 177 58 yes 130 42 access to service providers no 202 66 yes 105 34 note: dummy variables are in number and percentage source: survey data 2019 3.2. rural farm livelihood assets of sampled households in the study area the overall livelihood asset composite index score in the study area was 0.297. same like liu et al., (2018) the area farmers are lack of livelihood resources along with low level of living standards as well as economic development in the area. the composite index scores of the livelihood capitals revealed that natural capital dominated journal of applied economics and business studies, volume. 5, issue 3 (2021) 19-42 https://doi.org/10.34260/jaebs.532 29 the other four capitals by achieving highest value (0.429) and social capital that possessed a relatively high value (0.383) followed by human capital (0.271) and financial capital (0.202). however, physical capital possessed lowest value (0.201) among the five livelihood capitals (table 3). the overall result showed that rural farmers are lacked in livelihood asset but still natural and social capital affected livelihood of the farmers. the index of capacity structure of sustainable livelihood capitals revealed that there is a huge deficiency in all other four capitals while natural capital is also deficient among farming households in the study area. the magnitude of the natural capital index indicates that the sampled farm households can survive with the key environmental resources and services along with food production from the natural capital (nagesha et al., 2006; udoh et al., 2017). large farm households would have good economic condition and probably have more sustainable livelihoods through encompassing suitable combinations of farm enterprises (nagesha et al., 2006; udoh et al., 2017) however, small farm holders and tenants have to be less sustainable livelihood. social capital makes an important contribution to sense of dwellers of well-being by identity, honour and belonging. this capital has a direct impact on other types of household capital on the way through efficiency of economic relations (financial capital), facilitation of sharing views and experiences with other members of the organization and development along with sharing of knowledge (human capital) (udoh et al., 2017). trust is one the main factor that restrained the farm households to be benefited from the institutions. the financial status of the farm households also matters as poor farmers were generally not given with loaned inputs due to their capability of not returning the loan on time. human capital is essential for use of other forms of asset capitals (udoh et al., 2017). the results revealed that the human capital is not sufficient among sampled farm households for utilization of other capitals more efficiently. household with high dependency ratio and quality of household members’ mainly female members affected the human capital in the study area. the absence of human capital would force the sampled household members to work on low wages and would be pushed towards traditional agriculture production. financial capital is needed for adopting the innovations, new technologies and participation in non-farm sector. the magnitude of financial capital in the study area is low. the access to formal institutions by farm households are difficult due to fulfilling the formalities required for borrowing loan from these institutions. due to overall low financial conditions of the area farm households, it is difficult for one friend and / or relatives to give loan to another. in the absence of financial capital, the sampled farm households were facing the poverty and to have abundant labour force in agriculture sector. the ownership or control over physical capital in the study area was also low. in abdul hassan, shahnaz akhtar and muhammad ishaq 30 the absence of physical capital, the sampled respondents had low produce that needs to be developed according to socioeconomic and geographical characteristics of the sampled respondents in the context of global and cultural conditions of the area. it is observed during discussion with farmers throughout the survey that the rural farmers are experiencing persistent poverty. the people lost faith on the system of both formal and informal institutions. they sensed that they have no means and due to their powerlessness, there is no way-out for changing their life. the area farmers lack hope and think that working hard have no assurance that their livelihood will become better. the hopelessness expresses that small farmers have no idea and longer plans for improving their livelihood. they perceived that the life they have is well. there is need for provision of hope to the area farmers for changing their conditions. table 3. livelihood assets of the sampled households in the study area asset mean minimum maximum natural 0.429 0.000 1.000 human 0.271 0.030 0.800 financial 0.202 0.000 0.640 physical 0.201 0.010 0.800 social 0.383 0.000 1.000 livelihood 0.297 0.030 0.620 source: survey data 20190 3.3. multiple regression model 3.3.1. diagnostic tests randomness: both total household income and log-total household income follows randomness (annexure). normality: kolmogorov-smirnov test revealed that total household income as dependent variable was not normally distributed while log-total household income as dependent variable showed normal distribution at 3 percent significance level (annexure). multicollinearity: the minimum tolerance value for testing the multicollinearity was 0.536 and the maximum vif value was 1.866 for linear variables i.e., linear model and log-linear model. the result showed the rejection of multicollinearity in the explanatory variables used for linear model and log-linear model (annexure). heteroscedasticity: the current study used breusch-pagan /cook-weisberg test for heteroscedasticity. the test results cannot reject the null hypothesis for log-linear model i.e., homoscedasticity and found constant variances across all levels of independent variables. however, the results for linear model reject the null hypothesis i.e., heteroscedasticity and cannot found constant variances across all levels of independent variables (annexure). journal of applied economics and business studies, volume. 5, issue 3 (2021) 19-42 https://doi.org/10.34260/jaebs.532 31 summary of selection of model functional form the study results revealed that the log-linear model fulfilled the normality assumption at 3 percent significant level, as according to hill et al., (2011) that income as an economic variable has skewed distributions and its logarithmic form in model is common. moreover, both the model fulfills the assumption of multicollinearity and only log-linear model fulfill the assumption of homoscedasticity. based on these findings, the log-linear model was selected for further analysis and interpretations. 3.3.2. empirical results of the study the value of coefficient of determination (r2) of the asset model showed that 18 percent of the variations in the dependent variable were explained by the explanatory variables (table 4). similarly, the value of coefficient of determination (r2) of the capital model showed that about 36 percent of the variations in the dependent variable were explained by the explanatory variables (table 5). the value of f-statistics for both the models revealed that the overall models were highly significant. the study results regarding human capital of livelihood asset in the regression model revealed that human capital had significant positive effect on the livelihood output (table 4). the variables of human capital like household active labor and education of active labor had positive significant effect on the livelihood output, while family with health issues had non-significant effect on livelihood output (table 5). education in non-rural households designates high quality labor force which enabled them to be engaged in non-agricultural occupations and increased non-agricultural revenue (zhifei et al., 2018). the statistically non-significant effect of health issues is contrary and not in accordance with the expected outcome. the disease affects livelihood asset and the resource base of affected households via asset disposal along with education, health, income, family relations and social roles (maaka, 2012). this might be due to that the government had provided free of cost health care facilities to few chronic diseases. household with health issues is a perceived aggregate of chronic diseases whether any household member had chronic disease. further categorization of households with different disease and availing free health opportunities should be made that might have been different statistical result. the empirical result showed that livelihood with a predicted household output of rs. 100, the estimated increase in household output for an additional unit of human capital was 91.7 (table 4). moreover, livelihood with a predicted household output of rs. 100, the estimated increase in household output for an additional household active labor was 0.4. similarly, the livelihood with a predicted household output of rs. 100, abdul hassan, shahnaz akhtar and muhammad ishaq 32 the estimated increase in household output for an additional household education of labour of one year was 9.8 (table 5). the study findings of the regression model revealed that natural capital of livelihood asset had significant positive effect on the livelihood output (table 4). the variables of natural capital like family own land had significant positive effect on the livelihood output while soil quality had non-significant effect on the livelihood output (table 5). bazezew et al., (2013) found a non-significant correlation of farm size with annual income of households. family own land is one of the important asset as the orchard, economic trees and its produce are the possession of the land owner and there is no allocation made to the tenant in the study area. falco & zoupanidou, (2017) found that soil fertility (index) has significant positive influence on the revenues of farmers while in this study soil quality had non-significant effect on the livelihood output. the empirical result showed that livelihood with a predicted household output of rs. 100, the estimated increase in household output for an additional unit of natural capital was 40.3 (table 4). moreover, livelihood with a predicted household output of rs. 100, the estimated increase in household output for an additional acre of family own land was 4.5 (table 5). the study findings revealed that physical capital of livelihood asset had significant positive effect on the livelihood output (table 4). the variables of physical capital like access to public services had significant positive effect on livelihood output while house space per capita had non-significant effect on livelihood output (table 5). the empirical result showed that livelihood with a predicted household output of rs. 100, the estimated increase in household output for an additional unit of physical capital was 78.1 (table 4). moreover, livelihood with a predicted household output of rs. 100, the estimated increase in household output for an additional access to public services was 2.2 (table 5). the study results indicated that financial capital of livelihood asset had significant positive effect on livelihood output (table 4). additionally, both the factors, livestock and formal financial support of financial capital had significant positive effect on livelihood output (table 5). the study findings are consistent with the results found in literature, that the possession of livestock had significant and positive influence on household income (deressa, 2010; sharp et al., 2003; million, 2010; tefera, 2009; bazezew et al., 2013). similarly, access to credit has significant and positive correlation with annual income journal of applied economics and business studies, volume. 5, issue 3 (2021) 19-42 https://doi.org/10.34260/jaebs.532 33 of households (beyene, 2008; bazezew et al., 2013). however, farmers tend for interest free loans to sustain their livelihood (su & shang, 2012). the empirical result showed that livelihood with a predicted household output of rs. 100, the estimated increase in household output for an additional unit of financial capital was 52.5 (table 4). moreover, livelihood with a predicted household output of rs. 100, the estimated increase in household output for an additional livestock unit was 63.8. additionally, the empirical result showed that livelihood with a predicted access to formal financial support of rs. 100, the estimated increase in household output having access to formal financial support was 17.1 (table 5). the empirical finding of the study revealed that social capital of livelihood asset had significant positive effect on the livelihood output (table 4). both the variables of social capital like institutional membership and access to service providers used in the model had significant positive effect on livelihood output (table 5). hassan et al. (2017) reported a significant positive influence of agriculture service providers on crop productivity, their livelihood and income. the empirical result showed that livelihood with a predicted household output of rs. 100, the estimated increase in household output for an additional unit of social capital was 33.3 (table 4). moreover, livelihood with a predicted household output of rs. 100, the estimated increase in household output having institutional membership was 17.7. additionally, livelihood with a predicted household output of rs. 100, the estimated increase in household output having access to service providers was 23.1 (table 5). table 4. coefficients of log-linear regression model using livelihood capitals model unstandardized coefficients t sig. b std. error (constant) 11.989 0.115 104.18 0.000 human capital 0.917 0.266 3.452 0.001 natural capital 0.403 0.169 2.378 0.018 physical capital 0.781 0.341 2.29 0.023 financial capital 0.525 0.175 3.006 0.003 social capital 0.333 0.124 2.694 0.007 dependent variable: log of household income (pkr/year) number of observations: 307 r square: 0.180 f-statistics: 13.180*** highly significant***(000) source: survey data 2019 abdul hassan, shahnaz akhtar and muhammad ishaq 34 table 5. coefficients of log-linear regression model using capital variables unstandardized coefficients t sig. b std. error (constant) 11.643 0.170 68.570 0.000 hh active labor (percent) 0.004 0.002 1.956 0.051 education of labor (years) 0.098 0.015 6.784 0.000 family with health issues -0.110 0.109 -1.009 0.314 family own land (acre) 0.045 0.017 2.629 0.009 soil quality 0.079 0.089 0.888 0.376 livestock (equivalent units) 0.638 0.131 4.859 0.000 access to formal financial credit 0.171 0.081 2.102 0.036 house space per capita 0.007 0.013 0.541 0.589 distance to public services (km) 0.022 0.012 1.796 0.073 mfsc membership 0.177 0.081 2.184 0.030 access to service providers 0.231 0.079 2.907 0.004 dependent variable: log of household income (pkr/year) number of observations: 307 r square: 0.358 f-statistics: 14.925*** highly significant***(000) source: survey data 2019 4. conclusions on the basis of firsthand data collected for this study, the rural small farmers were lacked in livelihood asset. this specific asset status of the small farmers’ uncovered the living standard of small farmers having limited livelihood resources. the effective use of family land could increase the livelihood output. household labour force and education of household labor earners were main factors that could increase livelihood output. family size with more number of dependent household members may limit farm households in increasing their livelihood output. moreover, physical and financial capital have the capability to increase livelihood output by its effective usage. additionally, access to service providers of social asset, due to their timely provision of advice and inputs either on loan or in cash to farmers increases the livelihood of rural farmers. limited membership with institutions is due to its poor distrustment that ultimately could not affected the livelihood of rural farmers. based on the study results, following policy suggestion are made: • sustainable use of natural capital in terms of efficient utilization of land through the application of agricultural smart techniques and technologies. the agricultural departments and other allied stakeholders are required to change their approach and vision for agricultural development in view of educating the rural farmers especially youth about agricultural smart technologies and latest innovations applicable journal of applied economics and business studies, volume. 5, issue 3 (2021) 19-42 https://doi.org/10.34260/jaebs.532 35 according to socioeconomic conditions. further synchronizing it with the climate change and emerging market trends. • the conventional household structure in terms of female empowerment and dependency needs to be studied in the context of sustainable livelihood paradigm. • still there is a huge potential for livestock rearing in the study area and the farmers should be motivated to run the livestock business commercially. • organizations need to be developed in the area to develop social capitals. • local government system needs to be studied in light of sustainable livelihood paradigm and to develop favorable policies in terms of formal financial support and others to boost livelihood of the rural people. limitations of the study although the study provides empirical evidence to the direction of future policy, still numerous limitations apprehend the pragmatic results of this study. constraints of resource and time are not allowing to cover the whole khyber pakhtunkhwa inclusively. the study focused on small farm rural households. there is not enough clarity on the sufficiency of the selected indicators of livelihood asset to conclude that these represent each livelihood capital that effect livelihood output. the income data is based on cash incomes for study year that may be low and high in comparison to last or next year. the data collected is based that is reported by respondents. references abdallah, h. y. (2013). indices of distances: characteristics and detection of abnormal points. international journal of mathematics and computer science, 8(2), 55–68 adams, r.h. 1993. non-farm income and inequality in rural pakistan. the pak. dev. rev. 32(4): 1187-1198 ali, h., m. m. shafi and m. siraj. 2014. determinants of off-farm employment among small farm holders in rural areas of district mardan. sarhad j. agric. 30(1): 145-150. amjad, s., & hasnu, s. (2007). smallholders’ access to rural credit: evidence from pakistan. the lahore journal of economics, 12(2), 1-25 babbie, e. (1990). survey research methods 2nd ed. belmont, ca, wadsworth. in: sivo, s. a., saunders, c., chang, q., & jiang, j. j. (2006). how low should you go? low response rates and the validity of inference in is questionnaire research. journal of the association for information systems. 7(6), 351-414 barnett, v., & lewis, t. (1994). outliers in statistical data (3rd ed.). new york, ny: john wiley and sons. in: zakaria, a., howard, n. k., & b. k. nkansah. 2014. on the detection of influential outliers in linear regression analysis. american journal of theoretical and applied statistics. 3(4), 100-106. doi: 10.11648/j.ajtas.20140304.14 bazezew, a., bewket, w. & nicolau, m. (2013). rural households’ livelihood assets, abdul hassan, shahnaz akhtar and muhammad ishaq 36 strategies and outcomes in drought-prone areas of the amhara region, ethiopia: case study in lay gaint district. african journal of agricultural research, 8(46), 5716-5727 million, t. (2010). essays on contracts, risk coping and technology adoption in ethiopia. doctoral dissertation, norwegian university of life sciences, norway. in: bazezew, a., bewket, w. & nicolau, m. (2013). rural households’ livelihood assets, strategies and outcomes in drought-prone areas of the amhara region, ethiopia: case study in lay gaint district. african journal of agricultural research, 8(46), 5716-5727 beyene, a. d. (2008). determinants of off-farm participation decision of farm households in ethiopia. agrekon, vol 47, no 1 deressa, t. t. (2010). assessment of the vulnerability of ethiopian agriculture to climate change and farmers’ adaptation strategies. ph d. thesis. department of agricultural economics, extension and rural development, faculty of natural and agricultural sciences, university of pretoria dfid. (1999a). introduction, key sheets for sustainable livelihoods. department for international development, natural resources policy and advisory department, overseas development institute dfid. (1999b). overview, sustainable livelihoods guidance sheet. department for international development, natural resources policy and advisory department, overseas development institute dfid. (1999c). framework, sustainable livelihoods guidance sheet. department for international development, natural resources policy and advisory department, overseas development institute dhakal, c. p. (2017). dealing with outliers and influential points while fitting regression. journal of institute of science and technology. 22(1) issn: 24699062 (print), 2467-9240 (e) dominique van de walle. (2005a). are returns to investment lower for the poor? human and physical capital interactions in rural vietnam. public economics, development research group, the world bank dominique van de walle. (2005b). choosing rural road investments to help reduce poverty. public economics, development research group, the world bank falco, s. d., & zoupanidou, e. (2017). soil fertility, crop biodiversity, and farmers’ revenues: evidence from italy. ambio, 46, 162–172 fao. (2011). guidelines for the preparation of livestock sector reviews. animal production and health guidelines. no. 5. rome farooq, s. 2014. the rural non-farm economy, livelihood strategies and household welfare in rural pakistan, technical paper for 2nd adb-asian think tank development forum, to be held on november 20-21, 2014, seoul, republic of korea. available at http://www.adb-asianthinktanks.org/ sites/ all/ libraries/ researchpapers/theruralnon-farm economy livelihood strategies household http://www.adb-asianthinktanks.org/%20sites/%20all/%20libraries/%20researchpapers/theruralnon-farm%20economy%20livelihood%20strategies%20household%20welfare%20in%20ruralpakistanfarooq.pdf http://www.adb-asianthinktanks.org/%20sites/%20all/%20libraries/%20researchpapers/theruralnon-farm%20economy%20livelihood%20strategies%20household%20welfare%20in%20ruralpakistanfarooq.pdf journal of applied economics and business studies, volume. 5, issue 3 (2021) 19-42 https://doi.org/10.34260/jaebs.532 37 welfare in ruralpakistanfarooq.pdf (accessed august 01, 2018). galab, s., fenn, b., jones, n., raju, d. s. r., wilson, i., & reddy, m. g. (2006). livelihood diversification in rural andhra pradesh: household asset portfolios and implications for poverty reduction. working paper no. 34, young lives, an international study of childhood poverty, save the children uk, 1 st john's lane, london ec1m 4ar gokp. 2015. agriculture policy khyber pakhtunkhwa. a ten year perspective (20152025). available for download from www.fao.org.pk gokp. 2016. khyber pakhtunkhwa climate change policy. final draft. environmental protection agency. forestry, environment & wildlife department. government of khyber pakhtunkhwa gop. 2017. population census 2017, available at http://www.pbs.gov.pk/ sites/default/ files/ district_wise_census_results_census_2017.pdf accessed 17-09-2019 gop. 2012. agricultural census (2010). pakistan report statistics division. agricultural census organization. government of pakistan.gautam, y. & andersen, p. (2016). rural diversification and household well-being: insights from humla, nepal. journal of rural studies, 44, 239-249 hahn, m. b., riederer, a. m., & foster, s. o. (2009). the livelihood vulnerability index: a pragmatic approach to assessing risks from climate variability and change—a case study in mozambique. global environmental change, 19, 74– 88 hassan, a., ali, a., hussain, m., iqbal, a., moin, s., & iqbal, m. f. (2017). role of agricultural services providers (asps) in enhancing the productivity of crops in district faisalabad. international journal of advanced multidisciplinary research, 4(12), 13-18. doi: http://dx.doi.org/10.22192/ijamr.2017.04.12.003 hassan, a., ishaq, m., shah, n. a., & farooq, a. (2014). milk production potential in khyber pakhtunkhwa. pakistan j. agric. res. 27(1), 30-40 hill, r. c., griffiths, w. e., & lim, g. c. (2011). principles of econometrics. fourth edition. john wiley & sons, inc. ikram, h. 2016. income diversification in central punjab. int. j. agric. appl. sci. 8(1): 21-24. ishaq, m., hassan, a., farooq, a., & xiangsen, l. (2016). survey of livestock production and management in khyber pakhtunkhwa province of pakistan. pak. j. agri. sci. 53(2), 473-481 israr, m. and h. khan. 2010. an analysis of livelihood sources in hilly areas northern pakistan. sarhad j. agric. 26(4): 665-672. israr, m., h. khan, d. jan and n. ahmad. 2014. livelihood diversification: a strategy for rural income enhancement. j. finan. econ. 2(5): 194-198. jafri, s. m. y. (undated). land resources, their management and use in pakistan kenea, w. (2008). rural livelihood, land management and biodiversity, community http://www.adb-asianthinktanks.org/%20sites/%20all/%20libraries/%20researchpapers/theruralnon-farm%20economy%20livelihood%20strategies%20household%20welfare%20in%20ruralpakistanfarooq.pdf http://www.pbs.gov.pk/%20sites/default/%20files/%20district_wise_census_results_census_2017.pdf http://www.pbs.gov.pk/%20sites/default/%20files/%20district_wise_census_results_census_2017.pdf http://dx.doi.org/10.22192/ijamr.2017.04.12.003 abdul hassan, shahnaz akhtar and muhammad ishaq 38 participation for biodiversity conservation, the case of kakamega forest in western kenya. ph. d. thesis. faculty of spatial planning, university of dortmund kanwal, n., m. a. khan and z. zheng. 2016. analyzing the determinants of non-farm income diversification of farm households in peshawar district of pakistan. timisoara j. econ. buss. 9(1): 33-48. liu, z., chen, q. & xie, h. (2018). comprehensive evaluation of farm household livelihood assets in a western mountainous area of china: a case study in zunyi city. journal of resources and ecology, 9(2), 154-163 meinzen-dick, r., johnson, n., quisumbing, a., njuki, j., behrman, j., rubin, d., peterman, a. & waitanji. e. (2011). gender, assets, and agricultural development programs: a conceptual framework. capri working paper no. 99. international food policy research institute: washington, dc. http://dx.doi.org/10.2499/capriwp99 nuru, s., & seebens, h. (2008). the impact of location on crop choice and rural livelihood: evidences from villages in northern ethiopia, zef – discussion papers on development policy no. 123, center for development res., bonn, july 2008, pp. 27 ping, q., m. a. iqbal, m. abid, u. i. ahmed, a. nazir and a. rehman. 2016. adoption of off-farm diversification income sources in managing agricultural risks among cotton farmers in punjab pakistan. j. appl. environ. biol. sci. 6(8): 4753. qasim, s., khan, a. n., shrestha, r. p., & qasim, m. (2014). diffusion of agricultural innovation: farmers’ opinion on land conservation measures in pishin, pakistan, a scientific journal of comsats – science vision, 20(1) salunkhe, s. r., & movaliya, j. k. (2016). role of agro-service providers in agricultural development. review paper. advances in life sciences 5(11), issn 2278-3849, 4347-4351 samsudin, s. & kamaruddin, r. (2013). distribution of the livelihood assets among the hardcore poor: evidence from kedah, malaysia. world applied sciences journal 28 (economic, finance and management outlooks): 38-42, issn 18184952 saunders, m., lewis, p., & thornhill, a. (2003). research methods for business students. third edition. dorling kindersley (india) pvt. ltd. pearson education in south asia sharp, k., devereux, s. & amare, y. (2003). destitution in ethiopia’s northeastern highlands (amhara national regional state), final report. a policy research project funded by the uk department for international development (dfid) siegel, p. b. (2005). using an asset-based approach to identify drivers of sustainable rural growth and poverty reduction in central america: a conceptual framework. policy research working paper (3475), world bank journal of applied economics and business studies, volume. 5, issue 3 (2021) 19-42 https://doi.org/10.34260/jaebs.532 39 ifad. (2001). rural poverty report 2001: the challenge of ending rural poverty. rome. in: siegel, p. b. (2005). using an asset-based approach to identify drivers of sustainable rural growth and poverty reduction in central america: a conceptual framework. policy research working paper (3475), world bank su, f., & shang, h. y. (2012). relationship analysis between livelihood assets and livelihood strategies: a heihe river basin example. sciences in cold and arid regions, 4(3), 0265–0274 tefera, m. m. (2009). causes of rural household food insecurity: a case from kuyu district, central ethiopia. journal of sustainable development in africa. 11(4), issn: 1520-5509 udoh, e. j., akpan, s. b. & uko, e. f. (2017). assessment of sustainable livelihood assets of farming households in akwa ibom state, nigeria. journal of sustainable development; vol. 10, no. 4, issn 1913-9063 e-issn 1913-9071 nagesha, g., & gangadharappa, n. b. (2006). adoption of agroforestry systems in north eastern districts of karna-taka. my forest, 42(4): 337-347. in: udoh, e. j., akpan, s. b. & uko, e. f. (2017). assessment of sustainable livelihood assets of farming households in akwa ibom state, nigeria. journal of sustainable development; vol. 10, no. 4, issn 1913-9063 e-issn 1913-9071 shahbaz p., i. boz and s. u. haq. 2017. determinants of crop diversification in mixed cropping zone of punjab pakistan, direct res. j. agric. food sci. 5(11): 360366. uis. (2013). uis methodology for estimation of mean years of schooling. united nations educational, scientific and cultural organization. available at http://uis.unesco.org/sites/default/files/documents/uis-methodology-forestimation-of-mean-years-of-schooling-2013-en_0.pdf ullah, r., khan, m. z., ullah, k., & butt, t. m. (2015). model farm services center approach: an implication to boost farmer’s yield. agricultural sciences, 6, 953-960. http://dx.doi.org/10.4236/as.2015.69092 undp. (2019). human development report 2019. beyond income, beyond averages, beyond today: inequalities in human development in the 21st century. united nations development programme 1 un plaza, new york, ny 10017 usa urrehman, m., jehanzeb and m. f. rana. 2008. livelihood strategies of different categories of households in rural areas of abbottabad, pakistan. sarhad j. agric. 24(4): 685-692.wolkowski, d., (2005). soil quality and crop production systems. new horizons in soil science. issue #2 xu, d., deng, x., guo, s., & liu, s. (2018). sensitivity of livelihood strategy to livelihood capital: an empirical investigation using nationally representative survey data from rural china. social indicators research. https://doi.org/10.1007/s11205-018-2037-6 xu, d., zhang, j., rasul, g., liu, s., xie, f., cao, m., & liu, e. (2015). household livelihood strategies and dependence on agriculture in the mountainous http://uis.unesco.org/sites/default/files/documents/uis-methodology-for-estimation-of-mean-years-of-schooling-2013-en_0.pdf http://uis.unesco.org/sites/default/files/documents/uis-methodology-for-estimation-of-mean-years-of-schooling-2013-en_0.pdf abdul hassan, shahnaz akhtar and muhammad ishaq 40 settlements in the three gorges reservoir area, china. sustainability, 7, 48504869. doi:10.3390/su7054850 yang, l., liu, m., lun, f., min, q., zhang, c. & li, h. (2018). livelihood assets and strategies among rural households: comparative analysis of rice and dryland terrace systems in china. sustainability, 10, 2525; doi:10.3390/su10072525 zakaria, a., howard, n. k., & nkansah, b. k. (2014). on the detection of influential outliers in linear regression analysis. american journal of theoretical and applied statistics, 3(4), 100-106. doi: 10.11648/j.ajtas.20140304.14 zhifei, l., qianru, q. & hualin, x. (2018). comprehensive evaluation of farm household livelihood assets in a western mountainous area of china: a case study in zunyi city. journal of resources and ecology. 9(2) 154-163 zhu, h., ibrahim, j. g., & cho, h. (2012). perturbation and scaled cook’s distance. the annals of statistics 2012, vol. 40, no. 2, 785–811. an electronic reprint of the original article published by the institute of mathematical statistics in the annals of statistics, 40(2), 785–811 journal of applied economics and business studies, volume. 5, issue 3 (2021) 19-42 https://doi.org/10.34260/jaebs.532 41 annexure randomness runs test total hh income (rs./year) ln total hh income (rs./year) test value 367200.00 12.81 cases < test value 153 153 cases >= test value 154 154 total cases 307 307 number of runs 158 158 z 0.400 0.400 asymp. sig. (2tailed) 0.689 0.689 a. median tests of normality— kolmogorov-smirnov dependent variable statistic df sig. total hh income (rs./year) 0.133 307 0.000 ln total hh income (rs./year) 0.054 307 0.034 breusch-pagan / cook-weisberg test for heteroskedasticity chi2(1) with p-value = prob > chi2 linear model 42.50 0.0000 log-linear 1.35 0.2457 ho: error term is homoscedastic i.e. having constant variance h1: error term is not homoscedastic i.e. heteroscedastic abdul hassan, shahnaz akhtar and muhammad ishaq 42 multicollinearity tests variables linear model log-linear tolerance vif tolerance vif (constant) family land (ac) 0.744 1.344 0.744 1.344 farm land cultivated (ac) 0.672 1.488 0.672 1.488 soil quality (dummy) 0.792 1.263 0.792 1.263 family size (#) 0.635 1.575 0.635 1.575 hh off-farm earners (#) 0.536 1.866 0.536 1.866 education of labor earners (years) 0.581 1.722 0.581 1.722 livestock (equivalent units) 0.809 1.236 0.809 1.236 formal financial support (dummy) 0.908 1.101 0.908 1.101 hh asset-farm asset (#) 0.738 1.354 0.738 1.354 access to public services (km) 0.854 1.171 0.854 1.171 institutional membership 0.764 1.309 0.764 1.309 service provider 0.795 1.257 0.795 1.257 journal of applied economics and business studies, volume. 4, issue 4 (2020) 75-96 https://doi.org/10.34260/jaebs.444 75 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x the socio-economic determinants of child educational attainment in pakistan amjid iqbal 1, ghulam mustafa sajid 2 and sami ullah khan3* 1 statistical assistant at pakistan institute of medical sciences (pims), islamabad 2 assistant professor at international islamic university, islamabad 3 lecturer in economics at university of swabi and phd scholar at institute of management sciences, peshawar abstract the main purpose of conducting this study was to explore the socioeconomic determinants of child educational attainment in pakistan. data on selected variables is extracted from nationally representative survey of pakistan social and living standards measurement (pslm) 2013-14. the study considers the children of age 5-18 years who ever attended any or currently attending any educational institution. censored ordered probit model is employed. the results at aggregate level reveal that child’s gender, child’s age, age of household’s head, parental education, household assets and availability of private schools positively affect child educational attainment in pakistan. children belong to baluchistan province and urban region of pakistan also get more education as compare to reference categories. gender of household head and distance to school has negative effect on child educational attainment. to explore the inconsistency in effect of these factors across region and across gender, separate analysis is conducted for urban, rural, male and female sample. the results of gender specific analysis indicate that income of the household, region of residence (urban) and annual school fee are the main factors causing gender disparity whereas estimates at regional level analysis show that parental education, agricultural land ownership and, distance to school are the factors responsible for regional disparity in child educational attainment. keywords socioeconomic determinants, child educational attainment, censored ordered probit model jel classification c24, i21, i24 * corresponding author: samiullahkhan@uoswabi.edu.pk mailto:samiullahkhan@uoswabi.edu.pk amjid iqbal, ghulam mustafa sajid and sami ullah khan 76 1. introduction the economic development of a nation mainly depends on two factors: human capital and physical capital stock. human capital makes possible the production activities by using worker skills, knowledge, technology, and available capital stock. so, we must invest in human capital to develop them. according to babalola (2003) the rationality of investment in human capital is based on three arguments. first appropriate knowledge must be transferred from accumulated knowledge of previous generation to new generation. second new generation should think about that how existing knowledge can be utilized in better way. third the development of entirely new ideas, products, processes and method through creative approaches should be encouraged. three types of training or education such as education at school, training at workplace and other knowledge are important for human capital development (dubra, 2004). education is not only a consumption activity but also considered as an investment in formation of human capital. the human capital theory proposes that economic growth of a country is closely related to investment in education, job training, organized research, health and internal migration, and without investment in human being it is not possible to have capitalist development (schultz, 1971). according to lorey (1995) educated human capital is the most important source of growth and development for a country. by signing the united nations millennium declaration in september 2000, leaders from 189 countries agreed to the eight millennium development goals (mdgs), of which two are related to the education that are to be achieved by 2015, first every child should complete primary education (universal primary education for both male and female), and second is to remove gender disparities at all educational levels. this consensus reflects the view of most international development agencies and economists that education promotes economic growth and social development (glewwe and kremer, 2006). according to mdgs, pakistan was supposed to achieve 100 percent primary school enrollment and completion (up to grade five) and 88% literacy rate by 2015, but according to economic survey of pakistan 2015-16, the gross enrollment rates (ger) and net enrollment rates (ner) at the primary level was 89 percent and 57 percent respectively at the national level. the ger was 97 percent for male and 81 percent for female, whereas ner was 60 percent and 53 percent for male and female, respectively. the ger in punjab, khyber pakhtunkhwa, sindh, and baluchistan was 97%, 90%, 79% and 71% respectively. the ner was 61% in punjab, 56% in khyber pakhtunkhwa, 51% in sindh and 46% in baluchistan. these statistics show that regional and gender disparities are prevalent in pakistan and mdgs could not be achieved. the question is why pakistan could not achieve mdgs? there may be number of reasons of lower education level and not achieving mdgs which are related to education sectors. so, the main purpose of this study is to explore the factors which affect child education in pakistan. journal of applied economics and business studies, volume. 4, issue 4 (2020) 75-96 https://doi.org/10.34260/jaebs.444 77 education is the fundamental right of every child, both male and female, in all societies. according to the definition of the united nations convention on the rights of the child, “a child is a human being below the age of 18 years unless under the law applicable to the child, majority is attained earlier”. this is approved by 192 of 194 member countries. as today child is the part of future human capital, so we must focus on child education, because according to dubra (2004) education is the most important tool used for human capital development. according to previous studies child education is dependent on several factors i.e. age and gender of the child, school type, household, and community backgrounds etc. this study focuses on to explore all those socio-economic factors which determine child education in pakistan. for this purpose, the study classified all factors into four categories per their characteristics. first category is about child characteristics which include age and gender of the child. second category consists of household characteristics; gender and age of household head, father’s and mother’s education, dependency ratio, total assets, income, and agricultural land ownership. third is community level characteristics including region (urban/rural and province) and distance to school. final is the school characteristics category in which school type (public/ private) and annual school fee (admission/ tuition fee) are included. in pakistan, a limited amount of research work has been done about attainment of child education. most of the previous studies are conducted about the determinants of child enrollment in pakistan. for example, baluch and shahid (2008); pervaiz (2012); sajid and khan (2016). one study about attainment of child education in pakistan by holmes (2003) is outdated and its findings and conclusions may not be applicable to current education condition/situation in pakistan. another issue is that existing literature observes mixed results about the impacts of age and gender of the child, and school type on child education. the results of khan and khan (2016) and ngware et al. (2011) show that the effect of age and gender of the child, and school type on child education is positive while results of conlisk (1969) and kelley (1995) show that the effect of these variables on child education is negative. according to our knowledge no study in pakistan has been conducted which focused on both gender and regional (rural/urban) disparities in child education. keeping in view the above gap in the literature, the focus of the study is to find the socio-economic determinants of child education in pakistan. the child education means educational attainment of a child which is measured by years of education completed. first, all socio-economic determinants are classified into four categories per their characteristics, child characteristics, household characteristics, community characteristics, and school characteristics. second, data censoring analysis is used to find the latent desire level of education of currently enrolled children. third, in order to capture best picture of current educational conditions, the latest available data from pakistan social and living standards measurement (pslm) survey 2013-14 is used. in amjid iqbal, ghulam mustafa sajid and sami ullah khan 78 last, the analysis is made for whole pakistan, separately for rural and urbans areas and gender-based analysis. 2. literature review this section presents the review of relevant studies made by different authors and researchers. the focus is to review studies on the socio-economic determinants of child education. once we provide the review of available literature, it helps us in finding literature gap which we try to fill. the determinants of child education highlighted in the literature are based on child characteristics, household characteristics, community characteristics and school characteristics. at the end of section conclusions are made on the basis of literature review. conlisk (1969) finds that child age negatively affects child education because at initial ages of child, education is free and compulsory while after matriculation education is neither compulsory nor free in the united states. further he also finds that in the united states girls are attaining more education than boys. liu (1998) uses age in quadratic form and finds that probability of child educational attainment increases with age in vietnam. maitra (2001), using censored ordered probit model for bangladesh, concludes that the level of child education is higher for girls than that of boys. khan and ali (2005) find that age of the child positively affects child education of both boys and girls but the impact of age is more pronounced for boy’s education than that of girl in pakistan. they also find that the education level of girls is low as compared to boys. khan and khan (2016) conclude that the impact of child age and square of the child age is positive and negative respectively on child educational attainment in pakistan. they also find that male children get more level of education than that of female. sajid and khan (2016), by using logistic model, evidence that age and gender of the child positively affect child education in pakistan. conlisk (1969) results show that parental education and income has positive and significant effect on child educational attainment in the united state of america. ilon and moock (1991) are of the opinion that mother’s education level is an important determinant of child education, especially for poor households in peru. parker (2000) concludes that mother’s and father’s education is important for both boys and girls, but father’s education is more important for a child belong to poor family in mexico whereas maitra (2001) for bangladesh and zhao and glewwe (2010) for china, employing censored ordered probit regression, endorse that mother’s education has strong effects on child educational attainment than that of father. patrinos and psacharopoulos (1997) find a negative impact of income on child educational attainment in peru while zhao and glewwe (2010) evidence that in china household income has positive effect on years of schooling completed. kafle et al. (2017) find that durable assets have positive impact on child educational attainment while the effect of agriculture land ownership on child education is negative especially for children who journal of applied economics and business studies, volume. 4, issue 4 (2020) 75-96 https://doi.org/10.34260/jaebs.444 79 belong to rural household. olaniyan (2011) show that parents’ education positively affects child educational attainment. lloyd and blanc (1996) find that children of age 10-14 years who belong to female headed household have higher education than that of male headed household in sub saharan africa. they further find that the impact of female head is more for girl’s education than boy’s education. behrman et al (1997) provide evidence that the impact of household income on child educational attainment is positive for rural areas in pakistan. they also find a significant positive effect of father’s education on both girls’ and boys’ education and mother’s education positively affect only girl’s educational attainment. khan and ali (2003) and khan and ali (2005) find that parent’s education, per capita income and household assets positively affect child educational attainment in pakistan. they also find that household size has negative effect on child educational attainment. holmes (2003) concludes that parental education, household wealth, land ownership and other assets have positive effect on child educational attainment and the wealth influence is greater for females. hashmi et al. (2008) find that both father and mother’s education positively affect girl education in punjab. they also find that girls who belong to household having agricultural land have high level of education than that of household with no agricultural land. khan and khan (2016) estimate that the impact of mother’s education on female education is little more in rural region than that of urban areas. sajid and khan (2016) conclude that child education is positively affected by parental education and household assets while income has insignificant role in child education. honsi (1997) finds that moroccan urban children perform better than that of rural counterpart. cooksey et al. (2001) estimate that there is difference of 17%in net enrollment rates between rural and urban schools. holmes (2003) concludes that male child belongs to khyber pakhtunkhwa and baluchistan province get high level of education relative to male child who belongs to punjab. sajid and khan (2016) find that child who belongs to urban areas get more education. hamid and siddique (2001) find that distance to school has positive impact on child educational attainment while most of the studies such hashmi et al. (2008), holmes (2003), zhao and glewwe (2010), kondylis and manacorda (2010), and sajid and khan (2016) found that distance to school negatively affects child educational attainment. coleman et al. (1982), and coleman and hoffer (1987) confirm that in usa, private school positively affect child educational attainment more than that of public school. similar results are found by gannicott (1997) and long et al. (1999) and buckingham (2000) for australia. ngware et al. (2011), comparing private and public schools in nairobi, find that public schools are better than private schools and have positive impact on child educational attainment. frenette and chan (2015) and kelley (1995) confirm that private schools positively affect child educational attainment. astudy for india by amjid iqbal, ghulam mustafa sajid and sami ullah khan 80 kingdon (1996) finds that child educational attainment is positively affected by school tuition fees. from the above reviewed past literature, it is concluded that impact of some determinants/ variables i.e. age and gender of the child, and school type on child education are conflicting. some studies conclude that age and gender of the child, and school type positively affect child education (khan and khan, 2016; ngware et al., 2011) whereas some other studies come to the evidence of negative impact on child education (conlisk, 1969; kelley, 1995). so, there is need of proper consideration of the issue. to the best of our knowledge, no comprehensive study is available on child educational attainment for pakistan which captures all four types of characteristics the study of holmes (2003) is an exceptional study but it seems to be outdated and hence its findings and conclusions may not be applicable to current education situation/condition in pakistan. however, in order to have a comprehensive view of the current situation of child education there is a need to use latest available data. the study attempts to explore the socio-economic determinants of child educational attainment in pakistan. this study uses latest available data from pakistan social and living standards measurement (pslm) survey 2013-14. 3. material and methods 3.1 theoretical background as the focus of this study is to explore the socio-economic determinants of child education in pakistan. child education means the educational attainment level of a child. for this purpose, we need the support of theoretical background and model. first, educational attainment model was provided by becker (1964) in their human capital theory. according to this approach, education is an investment activity and people invest their resources on education. the decision regarding the investment on optimal level of education depends on cost and returns of additional level of schooling. on the one side, education is costly because of direct cost of books and tuition fees etc., and indirect cost in the form of opportunity cost i.e. the forgone earning during the period of education taken. on the other side education is beneficial as it is assumed that education increases one’s productivity in the form of future earnings. so, an optimizing individual will choose the level of education that maximizes his/her net return and will continue his/her investment on education up to the point where marginal cost of additional investment become equal to marginal return. according to this theory, human capital investment varies from individual to individual because of differences in cost and returns conditions. this theory was criticized on the ground that it considers education completely as a monetary phenomenon, and neglected consumption aspect of education. so, the model was extended by adding consumption motive. if education can increase the efficiency level of leisure allocation. heckman (1976) includes consumption motive into the journal of applied economics and business studies, volume. 4, issue 4 (2020) 75-96 https://doi.org/10.34260/jaebs.444 81 human capital model. kodde (1988) shows that demand for education becomes higher by integrating consumption motive in the model. becker (1965) extends the human capital model of child educational attainment into the household production model and argues that the process of the educational attainment is an aspect of household behaviour rather than individual behaviours. in this approach, child educational attainment is considered as a commodity in the household utility function, and the household is like a production unit generating some utility for its members by using some household inputs like time and family characteristics and market inputs like school quality. adults specially parents in the household make decisions about how to generate and how to use household resources. the children outcomes may be affected directly by parent’s decisions through the amount and nature of resource allocated or by their timing and may be indirectly through the decisions regarding family structure or location where children are growing up. the household production model has been improved continuously. engle (1980) hypotheses that the income of the mother is more relevant for child education as compared to other household income. muller (1990) states, as students grow older, they will increasingly be able to make their own choices and will be less dependent on their family background. hanushek (1992) comes to the fact that there is trade-off between quality and number of children for parents, because parents want to maximize household utility which is the function of children’s quantity and quality, subject to budget and time constraint. wilson (2001) integrates both human capital and production function models into a model of educational attainment and concludes that the factors that influence individual demand of education also affect educational attainment. 3.2 econometric model in order to explore the socio-economic determinants of child educational attainment, the following model is used. this model is based on study conducted by zhao and glewwe (2010). 𝐸𝐷𝑈𝑖 = 𝛼0 + 𝛽𝑖 𝐼𝑁𝐷𝐶𝑖 + 𝛾𝑖 𝐻𝐻𝐶𝑖 + 𝛿𝑖 𝐶𝐶𝑖 + 𝜃𝑖 𝑆𝐶𝑖 + 𝑒𝑖 (1) where, 𝐸𝐷𝑈𝑖 measure education level attained byi th child; 𝐼𝑁𝐷𝐶𝑖 is the vector of child’s characteristics; 𝐻𝐻𝐶𝑖 represents vector of household characteristics; 𝐶𝐶𝑖 shows vector of community level characteristics; 𝑆𝐶𝑖 is the vector of school characteristics and 𝑒𝑖 stands for error term. extended form of the model is given as; 𝐸𝐷𝑈𝑖 = 𝛼0 + 𝛽1𝐺𝑒𝑛𝑑𝑒𝑟𝑖 + 𝛽2𝐴𝑔𝑒𝑖 + 𝛽3𝐴𝑔𝑒𝑖 2 + 𝛾1𝐺𝑒𝑛𝑑𝑒𝑟𝐻𝑖 + 𝛾2𝐴𝑔𝑒𝐻𝑖 + 𝛾3𝐸𝑑𝑢𝐹𝑖 + 𝛾4𝐸𝑑𝑢𝑀𝑖 + 𝛾5𝐷𝑅𝑖 + 𝛾6𝐴𝑔𝑟𝐿𝑖 + 𝛾7𝐴𝑠𝑠𝑒𝑡𝑠𝑖 + 𝛾8𝐼𝑛𝑐𝑜𝑚𝑒𝑖 + 𝛿1𝑅𝑒𝑔𝑖𝑜𝑛𝑖 + 𝛿2𝐷𝐾𝑃𝑖 + 𝛿3𝐷𝑆𝑖𝑛𝑑ℎ𝑖 + 𝛿4𝐷𝐵𝑎𝑙𝑢𝑐ℎ𝑖 + 𝛿5𝐷1𝑆𝑖 + 𝛿6𝐷2𝑆𝑖 + 𝛿7𝐷3𝑆𝑖 + 𝜃1𝑇𝑦𝑝𝑒𝑆𝑖 + 𝜃2𝐹𝑒𝑒𝑆𝑖 + 𝑒𝑖 (2) amjid iqbal, ghulam mustafa sajid and sami ullah khan 82 definition of all variables in equation (2) is given in table-3.1. variables which are specific to child characteristics are. child age, age-square, and gender of the child. theoretically, the effect of child age estimate is not explicit, but for some countries this effect is positive and quadratic. bhalortra (2003) cited that as the age of male child increases the opportunity cost of education is expected to be increases because the labour productivity increases with age and become less harmful and socially acceptable. alternatively, for girls both school and market-based work became less acceptable and decline in favour of home-based work with increase in age. in order to control for gender effect, gender dummy has been used and the results are mixed (mansuri, 2006; mckenzie and repoport, 2007; hanson and woodruff, 2003; sharpa, 2011). the important household level variables that might affect schooling outcomes for children, are dependency ratio, education level of mother, gender, and age of the household head. gender of the household head is also an important control variable. bhalotra (2003) mention that it is supposed that household with female headship has less economic resources. there is also wage discrimination faced by female in labour market. the other reason on including gender of head is that to capture the significant role of women decision making regarding their children education. household head age is used as indicator for household lifecycle stage. the attitude to child education of aged individual might be different from younger individual. parental education levels are expected to have positive effect on child schooling because of higher desire for education of educated parents, give more importance to education, and more knowledge regarding returns on education (bhalotra, 2003; miluka and dabalen, 2008). other reason is that more educated heads or parents have more income and thus in batter position to devote more resources to child education. the dependency ratio is uses as control variable because it is expected that large households have more individuals among which the resources are divided. therefore, larger size households are associated with lower child schooling (sherpa, 2011). monthly per capita consumption expenditure, agriculture land ownership, and assets have been included to control wealth effect. in most of developing countries, education level is lower in rural areas as compare to urban areas. the main reasons for this are relative under development of market, social, economic, and school infrastructure (bhalotra, 2003). in order to capture regional disparities, regional dummy and provincial dummies are included. journal of applied economics and business studies, volume. 4, issue 4 (2020) 75-96 https://doi.org/10.34260/jaebs.444 83 table 3.1: definition of variables variable name definition child education 𝐸𝐷𝑈𝑖 child education measures education level attained by child. it is a categorical variable having five different categories i.e. no education, primary education, middle education, secondary education, others. gender of child 𝐺𝑒𝑛𝑑𝑒𝑟𝑖 gender of child is a dummy variable having value one (1) if child is male, zero (0) otherwise age of child 𝐴𝑔𝑒𝑖 it measures the age of child in completed years. age square of child 𝐴𝑔𝑒𝑖 2 to capture the non-linear effect of child, square of child age is taken. gender of household head 𝐺𝑒𝑛𝑑𝑒𝑟𝐻𝑖 dummy variable taking value one (1) if household is headed by male, zero (0) otherwise age of household head 𝐴𝑔𝑒𝐻𝑖 it measures the age of child in completed years. education level of father 𝐸𝑑𝑢𝐹𝑖 education level of child’s father is also divided into five categories taking value from 0 to 4 i.e. 0 = no education,1 = primary, 2 = high school, 3 = higher secondary and 4 = higher education and others. education level of mother 𝐸𝑑𝑢𝑀𝑖 mother education level is also divided into five categories taking value from 0 to 4 i.e. 0 = no education,1 = primary, 2 = high school, 3 = higher secondary and 4 = higher education and others. dependency ratio 𝐷𝑅𝑖 dependency ratio= (sum of household’s members younger than 15 and older than 64) divided by sum of household’s members of age 15-64. land ownership 𝐴𝑔𝑟𝐿𝑖 the dummy variable taking value one (1) if household own any agricultural land, zero (0) otherwise. total assets 𝐴𝑠𝑠𝑒𝑡𝑠𝑖 sum of the market value of all assets (financial + durable goods) income 𝐼𝑛𝑐𝑜𝑚𝑒𝑖 annual income of the household. region 𝑅𝑒𝑔𝑖𝑜𝑛𝑖 it is a dummy variable taking value 1 if household belongs to urban region zero otherwise (1= urban and 0= rural). provincial dummies three dummy variables are used in the study for four provinces. 𝐷𝐾𝑃𝑖 =1 if khyber pakhtunkhwa, 0 otherwise; 𝐷𝑆𝑖𝑛𝑑ℎ𝑖 =1 if sindh, 0 otherwise; 𝐷𝐵𝑎𝑙𝑢𝑐ℎ𝑖 =1 if baluchistan, 0 otherwise. note: punjab is used as reference category. distance to school to measure the distance to school, three dummies are used; 𝐷1𝑆𝑖 = 1 if distance is up to 5 km, 0 otherwise; 𝐷2𝑆𝑖 = 1 if distance is 6 to 10 km, 0 otherwise; 𝐷3𝑆𝑖 = 1 if distance is above 10 km, 0 otherwise. note: “hostel and other” is used as reference category. annual school fee 𝐹𝑒𝑒𝑆𝑖 it is a continuous variable. it includes admission/ tuition fee (in pkrs.) of the school school type 𝑇𝑦𝑝𝑒𝑆𝑖 it is a dummy variable taking value 1 if school is private, otherwise zero amjid iqbal, ghulam mustafa sajid and sami ullah khan 84 3.3 estimation technique in order to estimate equation 2, ordinary least square (ols) technique can be utilized, but there are some problems with using ols technique, and needs appropriate attention. first, we need to know about the completed final year of education of the children. therefore, data censoring analysis is necessary for currently enrolled children. ols did not consider the censoring and treat identically both currently enrolled children and those children who completed their education in past and will give biased results. second important issue is that education attainment is a series of discrete ordered choices. in such case ols cannot be used because it assumes continuous distribution. therefore, due to the above-mentioned problems with ols, king and lillared (1987) and zhao and glewwe (2010) used extended form of ordered probit model called censored ordered probit model. to estimate equation (2), we need to know about the child educational attainment. the level of education of children who have completed their education is directly observable whereas the level of education is not directly observable for currently enrolled children. therefore, there is need to censor the data to find the latent desired level of education for currently enrolled children. we can get the latent desire level of education for currently enrolled children by following formulas: 𝐸𝐷𝑈𝑖 = 0 𝑖𝑓 𝐸𝐷𝑈 ∗ ≤ 𝜋0 (3) 𝐸𝐷𝑈𝑖 = 1 𝑖𝑓 𝜋1 ≤ 𝐸𝐷𝑈 ∗ ≤ 𝜋5 (4) 𝐸𝐷𝑈𝑖 = 2 𝑖𝑓 𝜋6 ≤ 𝐸𝐷𝑈 ∗ ≤ 𝜋8 (5) 𝐸𝐷𝑈𝑖 = 3 𝑖𝑓 𝜋9 ≤ 𝐸𝐷𝑈 ∗ ≤ 𝜋10 (6) 𝐸𝐷𝑈𝑖 = 4 𝑖𝑓 𝐸𝐷𝑈 ∗ ≥ 𝜋4 (7) in the above system of equations 𝜋𝑖 ’s are the upper and lower limits of any education level and showing the switching from one lower education level to other higher education level i.e., from primary to middle. for those individuals who have never attended school the value of 𝐸𝐷𝑈 will be zero. for those individuals who have completed their education, we observe that discrete value of edu which falls between two cut-off points. for currently enrolled individuals the data is rightly censored with latent desired level of education. we did not know the desired education level for those individuals but we know the current level of education. therefore, it is assumed that they will complete at least that education level in which they are currently enrolled and therefore 𝐸𝐷𝑈∗ ≥ 𝜋𝑚𝑎𝑥 . the probability that the value of latent desired level of child’s education fall within certain threshold can be written as: 𝑃(𝐸𝐷𝑈 = 0) = 𝜃(𝜋0 − 𝛼𝑋) (8) 𝑃(𝐸𝐷𝑈 = 1) = 𝜃(𝜋1 − 𝛼𝑋) − 𝜃(𝜋0 − 𝛼𝑋) (9) 𝑃(𝐸𝐷𝑈 = 2) = 𝜃(𝜋2 − 𝛼𝑋) − 𝜃(𝜋1 − 𝛼𝑋) (10) 𝑃(𝐸𝐷𝑈 = 3) = 𝜃(𝜋3 − 𝛼𝑋) − 𝜃(𝜋2 − 𝛼𝑋) (11) 𝑃(𝐸𝐷𝑈 = 4) = 1 − 𝜃(𝜋4−1 − 𝛼𝑋) (12) journal of applied economics and business studies, volume. 4, issue 4 (2020) 75-96 https://doi.org/10.34260/jaebs.444 85 here 𝜋𝑖 s are the cut off points and 𝜃 represent the cumulative density function (cdf) of 𝑒𝑖. the cdf of 𝑒𝑖 in our model is standardized normal because we have used censored ordered probit model which is just extension of the probit model. x is the vector of all explanatory and control variables used in the study. educational attainment is a series of discrete ordered choice. to attain next high level of education to attain an extra year of schooling within certain level of education are two completely different decisions/choices and should be treated differently (khan and khan 2016). therefore, king and lillared (1987), holmes (2003) and zhao and glewwe (2010) proposed extended form of ordered probit model called censored ordered probit model. as child educational attainment is an ordered variable, therefore, censored ordered probit model has been implemented (miluka and dabalen, 2008; zhao and glewwe, 2009). 3.4 data to investigate the socio-economic determinants of child educational attainment in pakistan, data from pakistan social and living standards measurement (pslm) survey 2013-14 is used. pslm 2013-14 covers a large sample of about 17988 households from 1307 primary sampling units (psus) across four provinces of pakistan. 30513 children aged 5-18 has been selected of which 39% belongs to urban areas and remaining belongs to rural areas of pakistan. 43%, 24%, 23% and 10% children belong to punjab, khyber pakhtunkhwa, sindh and baluchistan respectively. 85% of the children are currently enrolled at different educational institutions. 4. results and discussion 4.1 descriptive statistics the summary statistics (minimum, maximum, mean and standard deviation) of main variables used in the study are reported in table 4.1. child enrollment is a dummy variable and taking value one if child is enrolled. the mean value is 0.85 means that 85 percent children are currently enrolled. school attainment is a categorical variable having five different categories taking value from zero (no education) to 4 (higher secondary and above). the median of child education is 1.00 which means that most of the children are enrolled in primary level. the gender of child is a dummy variable having value one if the child is male. the mean value of gender is 0.57 means that about 57 percent children are male and the deviation from the mean value is 0.496. we consider children of aged 5 to 18 years in the study. the average age of child is more than 11 years. the mean variation of child age from their mean value is 3.923 years. amjid iqbal, ghulam mustafa sajid and sami ullah khan 86 table 4.1: descriptive statistics variable minimum maximum mean st. deviation child enrollment (1 = if enrolled) 0 1 0.85 0.357 child education 0 4 1.00* 0.988 gender (1= if male) 0 1 0.57 0.496 age 5 18 11.37 3.923 gender of head (1 = if male) 0 1 0.90 0.298 age of head 15 99 46.54 11.742 father education 0 4 1.00* 1.310 mother education 0 4 0.00* 0.337 dependency ratio 0 9 1.30 1.017 agriculture land (1 = if yes) 0 1 0.08 0.269 total assets 0 9800000 24690.77 197863.81 income 0 101880000 606948.49 1305324.07 region (1 = if urban) 0 1 0.39 0.487 punjab (1 = if punjab) 0 1 0.43 0.495 kpk (1 = if kpk) 0 1 0.24 0.427 sindh (1 = if sindh) 0 1 0.23 0.420 baluch (1 = if baluchistan) 0 1 0.10 0.300 distance to school 0 4 0.93 0.565 school fee 0 700000 7034.14 15817.922 school type (1 = if non-government) 0 1 0.33 0.469 sample size 30513 * for education level, we have reported the median value. gender of head is also a dummy variable. the average value is 0.90 which means that 90 percent of the households are headed by male individuals. the range of head’s age is from 15 years to 99 years and the mean age of head is more than 46 years. for parental education, we have used ordinal variable taking values from zero (no education) to four (higher education). the median value of father’s education is 1.00 which means that mostly education level of the fathers is primary. the median education level of mother is zero means that almost mothers are illiterate. the minimum value of dependency ratio is zero and maximum is nine. the mean value of dependency ration is 1.30 means that on the average there are more than 1 dependent individual across households. agriculture land is a dummy variable, and its mean value is 0.08 which shows that only eight percent of household owned agriculture land. total assets and total income both are continuous variables. the minimum value of total assets is zero and maximum is rs. 980000/-. the mean value of total assets is about rs. 24690/-. income is ranged from zero to rs. 101880000/per year. the mean value of household journal of applied economics and business studies, volume. 4, issue 4 (2020) 75-96 https://doi.org/10.34260/jaebs.444 87 income is rs. 606946.49/per year and the deviation of household income from its mean value is rs. 1305324.07/-. the mean value of regional dummy is 0.39 which shows that 39 percent of the sample is belong to urban region and the remaining 61 percent are from rural area. we have used four dummies for provinces and the average values show that 43 percent of the sample belongs to punjab, 24 percent belongs to khyber pakhtunkhwa, 23 percent belongs to sindh and the remaining 10 percent is from baluchistan. distance to school is also categorical variable and ranged from 0 to 4 and the mean value is 0.93 which means that the average school distance is 1 to 5 kilometres (value=1). the minimum value of school tuition fee is zero and maximum is rs, 700000/-. the average annual school tuition fee paid by household is rs. 7034.14/-. school type is a dummy variable taking value one if school is non-government and its average value is 0.33 which shows that 33 percent children are enrolled in non-government schools. 4.2 results of censored ordered probit model at aggregate level the estimates of censored ordered probit model are reported in table 4.2. the coefficient of child’s gender (gender) is significantly positive. it means that male children are favoured more in providing education than female in pakistan. the main reason may be that our society is more biased toward male education and they want to educate male children on the cost of female education. age and age-square of the child both show significantly positive impact on educational attainment but the coefficient of age-square is less than age coefficient which means as the child’s age increases, that education level also increases but after some specific age point the positive impact start to decline. the main reason of diminishing the impact is that along with increase in age of the child, he opportunity cost of getting education also increases. khan and khan (2016) and liu (1998) also concluded the same. the effect of household head on educational attainment is negative and significant. it indicates that if child belong to the household for which the head is male person, the probability of getting higher education reduces. alternatively speaking, children belong to those households headed by female are more likely to attain higher education. age of household head, education level of both parents has significantly positive effect on child educational attainment. mother’s education has more positive effect on children educational attainment than that of father. the coefficient of household’s total assets on child educational attainment is positive and statistically significant. our results are consistent with khan and ali (2003), olaniyan (2011), khan and khan (2016) and sajid and khan (2016). coefficient of region shows that children living in urban constituency are more likely to attain higher education than their rural counterpart. to capture the role of province, the study used three dummies for four provinces and punjab province is used amjid iqbal, ghulam mustafa sajid and sami ullah khan 88 as reference category. the coefficients of province dummies show that if child belongs to khyber pakhtunkhwa, the chances of getting higher education and transition from lower education level to higher education level is low but in case of baluchistan, the results are contrary. these findings support the results of honsi (1997), cooksey et al. (2001), sajid and khan (2016) and holmes (2003). table 4.2: estimates of censored ordered probit model at aggregate level variable coefficient p-value constant -20.9900 0.000 genderi 0.0971 0.000 agei 0.2501 0.000 agei 2 0.0088 0.000 genderhi -0.0841 0.034 agehi 0.0043 0.000 edufi 0.0882 0.000 edumi 0.1412 0.000 dri -0.0174 0.122 agrli -0.0231 0.491 assestsi 0.0059 0.005 incomei -0.0021 0.516 regioni 0.0974 0.001 dkpi -0.0695 0.005 dsindhi 0.0204 0.475 dbaluchi 0.4026 0.000 d1si -0.7867 0.000 d2si -0.4693 0.000 d3si -0.1914 0.111 feesi 0.0051 0.782 typesi 0.0989 0.029 sample size 30513 to see the impact of distance to school on child education attainment, three dummies are used for distance. the results in table 4.1 enable us to conclude that distance to school decreases the chances of getting education for child when the distance is less or equals to 10 kilometres. if the distance is more than 10 km, the negative impact becomes insignificant. type of school, in which the child is enrolled, is used as proxy for capturing the quality of education. the coefficient of school type is significantly positive indicating that if child is going to private school, the likelihood to attain higher education is more. the results support the findings of previous studies of kondylis and manacorda (2010), buckingham (2000), gannicott (1997), long et al. (1999), and coleman et al. (1982). journal of applied economics and business studies, volume. 4, issue 4 (2020) 75-96 https://doi.org/10.34260/jaebs.444 89 4.3 results of gender based analysis the estimates of gender-based analysis are reported in table 4.3. it would help us to observe the disparities in effects of these socioeconomic factors on educational attainment for a child across gender. the coefficient of age variable is significantly positive, but the impact is more for male children as compared to his female counterpart. it means that chances of attaining higher education in older ages are more for males than for females. the studies of khan and ali (2005) and sajid and khan (2016) have also similar findings. the coefficient of gender of household head is significantly negative only for male sample. it means that in male headed households the likelihood of getting more education is lower for male child. the coefficient of household head age indicates that the positive impact is significantly almost same for both genders. the effect of parental education level is same for both samples. these findings are in line with parker (2000), hashmi et al. (2008), lloyd and blanc (1996) and khan and khan (2016). the coefficient of dependency ratio and land ownership is insignificant for each gender means that these two variables have no significant role in educational gender disparities. the impact of total assets on educational attainment of child is positive for both genders but significant only for male which shows that if assets of the household’s increase, it will increase the chances of attaining higher education only for male child. in case of male sample, the coefficient of income is positive but insignificant while in case of female, it is significantly negative. it means that as income level increases, the probability of attaining education decreases for female child. the justification for the negative impact of household’s income on female education level is that rich people of rural areas are more inclined toward socio-economic aspects (see psacharopoulos, 1997; khan and ali, 2003). the impact of region is significantly positive and almost same for both genders but slightly higher for male means that male children in urban region have slightly more chances to get education than female. the coefficient of khyber pakhtunkhwa dummy is negative in both gender but only significant in female sample. it indicates that in khyber pakhtunkhwa gender disparity in education is very high and females’ children of khyber pakhtunkhwa have fewer chances to get higher education level. in khyber pakhtunkhwa, the educational attainment is lower for female because of law and order situation, social norms and people attitude toward female education. in baluchistan province, the chances of higher level of education is little more for female than male. in baluchistan, the data is taken only from economically developed areas and ignored the remote areas. therefore, the results are biased toward female education. the results of holmes (2003) and sajid and khan (2016) are supporting our findings. amjid iqbal, ghulam mustafa sajid and sami ullah khan 90 table 4.3: estimates of censored ordered probit model at gender base variable male female coefficient p-value coefficient p-value constant agei agei 2 genderhi agehi edufi edumi dri agrli assetsi incomei regioni dkpi dsindhi dbaluchi d1si d2si d3si feesi typesi -1.9402 0.000 -2.4232 0.2782 0.0071 -0.0263 0.0024 0.0900 0.1277 -0.0198 0.0152 0.0045 -0.0119 0.0858 -0.1265 -0.0292 0.6274 -0.9147 -0.5799 -0.3242 -0.0163 0.2091 0.000 0.2218 0.000 0.000 0.0107 0.000 0.000 -0.1256 0.012 0.649 0.0057 0.000 0.070 0.0860 0.000 0.000 0.1554 0.000 0.000 -0.0173 0.195 0.213 -0.0485 0.217 0.776 0.0066 0.012 0.139 0.0054 0.179 0.021 0.1093 0.001 0.018 -0.0291 0.390 0.000 0.0550 0.101 0.454 0.2734 0.000 0.000 -0.7290 0.000 0.000 -0.4221 0.001 0.004 -0.1141 0.400 0.122 0.0404 0.043 0.495 -0.0287 0.565 0.000 sample size 17,289 13,224 distance to school almost has same negative effect on education level for both male and female. the impact of distance is negative because the availability of school facility played an important role in child educational attainment. the impact of school fee is significantly positive only for male child while the effect of school type is only positive for female significantly. these findings are consistent with the findings of kelley (1995), hamid and siddique (2001), holmes (2003), kingdon (1996), frenette and chan (2015) and colemon and hoffer (1987). 4.4 estimates of region based analysis in order to investigate the factors responsible for regional disparities in child educational attainment, we conducted separate analyses for both regions (see table 4.4). individual level characteristics of child, i.e. gender, age and age square has almost same impact on education level in both regions, but the impact of age is slightly higher for urban sample. our findings support the results of liu (1998) and khan and khan (2016). the coefficient of household head gender is significantly negative for rural areas only which means that male headed household is an obstacle for child to attain educational in rural areas. the age of household head has more or less the same significant positive impact on educational attainment of children in both regions. the effect parental education is positive and highly significant in both regions, but the journal of applied economics and business studies, volume. 4, issue 4 (2020) 75-96 https://doi.org/10.34260/jaebs.444 91 impact of father education is more pronounced in urban region while the impact of mother education is higher in rural region. it means that in urban areas, educated father increases the chances of their children’s education while in rural areas, children of educated mother have more chances than their urban counterparts to get higher education. these findings are comparable with that of ilon and moock (1991), maitra (2001), conlisk (1969), khan and khan (2016) and olaniyan (2011). the coefficient of ownership of agriculture land by household is significantly positive in urban areas while significantly negative for rural sample which means that in urban area, if the household owned any land, the probability of getting education will be more for children. if rural household owned any agriculture land, the probability of their children’s educational attainment will be lower because rural household may involve their children in agricultural activities. the impact of total assets is positive in both areas, but it is significant only for rural areas means that the possibility of attaining higher education is higher for a child who belongs to rural household having assets. the income of household has statistically insignificant impact on child educational attainment in both regions but negative for urban and positive for rural sample. our results are in line with that of holmes (2003), kafle et al. (2017), sajid and khan (2016), hashmi et al. (2008). table 4.4: estimates of censored ordered probit model at regional level variable urban rural coefficient p-value coefficient p-value constant 𝐆𝐞𝐧𝐝𝐞𝐫𝐢 𝐀𝐠𝐞𝐢 𝐀𝐠𝐞𝐢 𝟐 𝐆𝐞𝐧𝐝𝐞𝐫𝐇𝐢 𝐀𝐠𝐞𝐇𝐢 𝐄𝐝𝐮𝐅𝐢 𝐄𝐝𝐮𝐌𝐢 𝐃𝐑𝐢 𝐀𝐠𝐫𝐋𝐢 𝐀𝐬𝐬𝐞𝐭𝐬𝐢 𝐈𝐧𝐜𝐨𝐦𝐞𝐢 𝐃𝐊𝐏𝐢 𝐃𝐒𝐢𝐧𝐝𝐡𝐢 𝐃𝐁𝐚𝐥𝐮𝐜𝐡𝐢 𝐃𝟏𝐒𝐢 𝐃𝟐𝐒𝐢 𝐃𝟑𝐒𝐢 𝐅𝐞𝐞𝐒𝐢 𝐓𝐲𝐩𝐞𝐒𝐢 -4.8027 0.0945 0.2790 0.0093 -0.0830 0.0056 0.1050 0.1277 -0.0188 0.1211 0.0046 -0.0082 -0.1173 0.0233 0.3827 -0.8850 -0.6415 -0.2224 0.0374 -0.0347 0.114 0.000 0.000 0.000 0.241 0.010 0.000 0.000 0.404 0.050 0.139 0.220 0.002 0.571 0.000 0.000 0.001 0.237 0.215 0.608 0.5841 0.0873 0.2357 0.0083 -0.0820 0.0036 0.0772 0.1609 -0.0186 -0.0741 0.0066 0.0007 -0.0241 0.0313 0.4168 -0.7981 -0.4320 -0.1738 -0.0089 0.2109 0.009 0.004 0.000 0.000 0.084 0.002 0.000 0.000 0.156 0.060 0.016 0.850 0.464 0.428 0.000 0.000 0.003 0.249 0.701 0.000 sample size 11782 18731 amjid iqbal, ghulam mustafa sajid and sami ullah khan 92 the negative impact khyber pakhtunkhwa dummy is significantly negative only for urban sample means that the probability of switching from lower education level to higher education level is lower for child belongs to urban region of khyber pakhtunkhwa. the coefficient of baluchistan dummy is significantly positive and almost same in both regions means child who belongs to baluchistan possibly get more level of education irrespective of region to which he or she belongs. the results of holmes (2003) and sajid and khan (2016) are supporting our findings. the distance from home to school has significantly negative effect on child educational attainment up to 10 kilometres, after that distance to school becomes less effective in determining child educational attainment for both samples but the effect is little more for rural areas. it means that child educational attainment in rural areas is more adversely affected by distance to school than that of urban areas. the coefficient of the school type is negative and insignificant for urban areas but positive and significant for rural areas means child enrolled in rural private school may get more levels of education. these results are similar to the results of zhao and glewwe (2010), hashmi et al. (2008), sajid and khan (2016), buckingham (2000) and coleman and hoffer (1987). 5. conclusion and recommendations this study investigates the socioeconomic determinants of child educational attainment in pakistan. to do so, data from multidimensional nationally representative survey i.e. pslm 2013-14 is used. children of age 5-18 years who ever attended the school or currently attending school are considered and censored ordered probit model is implemented for analysis of data. in order to see disparities in effect of factors on educational attainment across region and across gender, separate analysis is carried out for both genders (male and female) and regions (urban and rural). the results of the overall model reveal that gender and age of child, household head gender, parental education, assets of household, region, dummy for baluchistan and school fee have significant positive effect on child educational attainment. household income, dependency ratio and land ownership have no significant impact on child educational attainment. the distance to school is a hurdle in attaining education while private schools are contributing to child educational attainment in pakistan. gender specific analysis shows that age and gender of household head, assets and school fee are the important determinants of educational attainment for male children only while the effect of household’s income and type of school is significant only in female sample. region wise separate analysis reveal that gender and age of child and household head, parental education, land ownership, distance to school are the important determinant of educational attainment in both regions. school type is only significant in journal of applied economics and business studies, volume. 4, issue 4 (2020) 75-96 https://doi.org/10.34260/jaebs.444 93 rural sample only while the effect of agriculture land ownership is positive for urban sample and negative for rural sample. the study findings recommended that to increase child education in pakistan, in order to increase child education in pakistan, there is a need of reducing gender differences in education by providing educational facilities to female and by increasing the value and importance of female education in society. there is a need of increasing higher education specially for female children by taking different steps i.e. increasing girl’s schools and educational institutions, increasing awareness about female education in society etc. our findings also suggest that by increasing the mother education level, the child education level could also increase especially female education in pakistan. the study also suggests that distance to school is an important determinant of child education. therefore, to increase child education level, we should reduce the distance to school by building new schools. quality of school is an important determinant of child education in pakistan. school attainment in non-government schools is higher. there is a need of improving quality of government school to increase child education level in pakistan. general awareness about the value of education should increase so that people give more attention to educating their children. furthermore, there is need of special focus on rural region to improve the educational attainment of children living in rural region. the future direction of the study is that there is a need of provincial wise study at disaggregate level to explore the differences in child education among provinces. further there is a need of national wise study which also includes observation from fata, gilgit baltistan and islamabad capital territory. the future direction of the study can be covered by continuing the pslm and by including more primary and secondary sampling units to make it wide. references babalola, j. b. (2003). budget preparation and expenditure control in education. basic text in educational planning. ibadan awemark industrial printers. baluch, m. u. h., & shahid, s. (2008). determinants of enrollment in primary education: a case study of district lahore. pakistan economic and social review, 46(2), 161-200. becker, g. s. (1964). human capital”, national bureau of economic research, newyork;11 becker, g. s. (1965). a theory of the allocation of time. the economic journal, 75(299) 493517. behrman, j. r., khan, s., ross, d., & sabot, r. (1997). school quality and cognitive achievement production: a case study for rural pakistan. economics of education review, 16(2), 127-142. bhalotra, s. (2003). child labour in asia and africa. bristol, background research paper for the efa monitoring report, mimeo, university of bristol. amjid iqbal, ghulam mustafa sajid and sami ullah khan 94 buckingham, j. (2000). the truth about private schools in australia. centre for independent studies coleman, j., hoffer, t., & kilgore, s. (1982). cognitive outcomes in public and private schools. sociology of education, 55(2-3) 65-76 coleman, j. s., & hoffer, t. (1987). public and private high school. the impact of communities”, (vol. 41). new york: basic books. conlisk, j. (1969). determinants of school enrollment and school performance. the journal of human resources, 140-157. cooksey, b., levey, l., & mkude, d. (2001). the partnership for higher education in africa: higher education in tanzania: a case study. new york, carnegie corporation dubra, i. (2004). investīcijas cilvēkkapitālā: teorētiskie aspekti, novērtēšanas metodes un pieejas pētniecībai investment in human capital: theory aspects, evaluation methods and research approaches. ekonomika. vadības zinātne, 7(4), 34. frenette, m., & chan, p. c. w. (2015). academic outcomes of public and private high school students: what lies behind the differences? statistics canada gannicott, k. (1997). charter schools: a new paradigm for public education policy: a journal of public policy and ideas, 13(2), 3. government of pakistan (2016). pakistan economic survey 2015-16. finance division, economic advisory wing, islamabad, government of pakistan hamid, s., & siddiqui, r. (2001). gender differences in demand for schooling. the pakistan development review, 1077-1092 hanson, g. h., & woodruff, c. (2003). emigration and educational attainment in mexico: mimeo., university of california at san diego. hanushek, e. a. (1992). the trade-off between child quantity and quality. journal of political economy, 100(1), 84-117 hashmi, n., zafar, m. i., & ahmad, m. (2008), cultural determinants of female educational attainment in rural jhang, punjab, pakistan. pakistan journal of agricultural sciences, 45(1) haveman, r., & wolfe, b. (1993). children's prospects and children's policy. the journal of economic perspectives, 7(4), 153-174 heckman, j. j. (1976) a life-cycle model of earnings, learning, and consumption. journal of political economy, 84(4, part 2), s9-s44 holmes, j. (2003). measuring the determinants of school completion in pakistan: analysis of censoring and selection bias. economics of education review, 22(3), 249-264. hosni, d. (1997). schooling and cognitive achievements of children in morocco: can the government improve outcomes? world bank discussion papers no. 264. washington, dc: the world bank, 1994. pp. ix+ 68. price: us $7.95. economics of education review, 16(1), 97-98. journal of applied economics and business studies, volume. 4, issue 4 (2020) 75-96 https://doi.org/10.34260/jaebs.444 95 ilon, l., & moock, p. (1991). school attributes, household characteristics, and demand for schooling: a case study of rural peru. international review of education, 37(4) p: 429451. kafle, k., jolliffe, d., & winter-nelson, a. (2017). do different types of assets have differential effects on child education? evidence from tanzania.. policy research working paper; no. 807,. world bank, washington, dc kelley, j. (1995). catholic schools and educational success in australia” worldwide attitudes.19950731 (31 july 1995) 1-8. 53. khan, r. e. a., & ali, k. (2003). determinants of schooling in rural areas of pakistan. the lahore journal of economics, vol.8, no.2 khan, r. e. a., & ali, k. (2005). bargaining over sons and daughters’ schooling: probit analysis of household behaviour in pakistan. working paper 01-05, department of economics, islamia university, bahawalpur, pakistan khan, s. u., & khan, m. j. (2016). the impact of remittances on child education in pakistan. the lahore journal of economics, 21(1), 69 king, e. m., & lillard, l. a. (1987). education policy and schooling attainment in malaysia and the philippines. economics of education review, 6(2), 167-181. kingdon, g. (1996). the quality and efficiency of private and public education: a case‐study of urban india. oxford bulletin of economics and statistics, 58(1), 57-82. kodde, d. a. (1988). unemployment expectations and human capital formation”, the european economic review, 32(8), 1645-1660. kondylis, f. & manacorda, m. (2010). school proximity and child labor: evidence from rural tanzania. cep discussion paper no 1035. liu, a. y. (1998). school children’s participation behaviour in vietnam: an empirical analysis” in twelfth annual conference of the european society for population economics, amsterdam, netherlands. lloyd, c. and a.k. blanc, (1996). schooling in sub-saharan africa, the role of fathers, mothers and others. population and development review, 22(2):265-98. long, m., carpenter, p., & hayden, m. (1999). participation in education and training, 19801994. longitudinal surveys of australian youth. research report. acer customer service, private bag 55, camberwell, victoria 3124 australia (code: a113lsa; $37.50 australian). lorey, d. e. (1995). education and the challenges of mexican development”, challenge, 38(2), 51-55 maitra, p. (2001). schooling and educational attainment: evidence from bangladesh. education economics, 11(2), 129-153. mckenzie, d., & rapoport, h. (2006). can migration reduce educational attainment? evidence from mexico. evidence from mexico (june 1, 2006). world bank policy research working paper (3952). amjid iqbal, ghulam mustafa sajid and sami ullah khan 96 miluka, j., & dabalen, a. (2008). exploring the role of albanian international migration on education. in albania-world bank conference müller, w. (1990). does education matter? evidence from cross-national comparisons. university of mannheim, manuscript. ngware, m. w., oketch, m., & ezeh, a. c. (2011). quality of primary education inputs in urban schools: evidence from nairobi. education and urban society, 43(1), 91-116. olaniyan, o. (2011). the determinants of child schooling in nigeria. (no rp_217), african economic research consortium parker, s. (2000). school and work in rural marginated communities of mexico: evidence from progresa. population association of america annual meetings, los angeles, ca. patrinos, h. a., & psacharopoulos, g. (1997). family size, schooling, and child labor in peru– an empirical analysis. journal of population economics, 10(4), 387-405. pervaiz, z. (2012). what factors determine primary gross enrollments in schools across provinces in pakistan? a comparison of the four provinces. rammohan, a., & dancer, d. (2008). gender differences in intra-household schooling outcomes: the role of sibling characteristics and birth‐order effects. education economics, 16(2), 111-126. sajid, g. m., & khan, m. (2016). determinants of primary school enrollment in pakistan: is poverty a hurdle in the way of educational attainment? the pakistan journal of social issues, volume 7. 1-18 schultz, t. w. (1971). investment in human capital. the role of education and of research. sherpa, m. (2011). essays on determinants of human capital accumulation. wilson, k. (2001). the determinants of educational attainment: modeling and estimating the human capital model and education production functions. southern economic journal, 67(3), 518-551. wolfe, b. l., & behrman, j. r. (1984). who is schooled in developing countries? the roles of income, parental schooling, sex, residence, and family size. economics of education review, 3(3), 231-245 zhao, m., & glewwe, p. (2010). what determines basic school attainment in developing countries? evidence from rural china. economics of education review, 29(3), 451-460. journal of applied economics and business studies, volume. 4, issue 2 (2020) 1-22 https://doi.org/10.34260/jaebs.421 1 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x is there any threshold in the causal association of mother’s education with child’s health? a case study of pakistan muhammad umar farooq1*, syed asif ali naqvi2 and shaoan huang3 1assistant professor, department of economics, government college university faisalabad, pakistan. 2assistant professor, department of economics, government college university faisalabad, pakistan. 3dean, center for economic research, shandong university, 27 shanda, nanlu, jinan, p.r. china. abstract past studies failed to solve the problem of nonlinearity and threshold in the association of mother education with infants’ health. this research diagnoses and investigates the existence of threshold. successive eight educational plans during 1955-1998 in pakistan record a large variation in the educational policies and investment and a great focus on primary schooling. it motivates us to select pakistan and construct the instrument for purging the coefficients of mother education from endogeneity bias along with the threshold impact. iv regression outcomes disclose that only 11-16 years of mother education has significant impact on infants’ health outcomes, implying a threshold level at 6-10 years schooling. thus, it is argued that low cognitive capability via lower level of education, poor educational quality, ineffective education for health even in the text books of 6-10 years of education leads to a threshold in the relationship of mother education with child nutrition. keywords haz; threshold; iv regression; pakistan jel classification h51; h75; i18 1. introduction if deserved nourishment is not provided to the kid during his or hers initial 2 years this could lead to some grotesque effects on the overall health, mental health and physical growth of a child. this in return give goading to bad grades in school and could well result in feeble output from the student and give halt to his growth which in result stop the churning wheel of the economy as well (glewwe & miguel, 2007). malnutrition is a bane for the economy as it causes parents to be * umarfarooqgcuf@yahoo.com mailto:umarfarooqgcuf@yahoo.com muhammad umar farooq, syed asif ali naqvi and shaoan huang 2 spendthrift on the health of the kid. it results in the physical damages of the nerves tissues and if not dealt properly and on time causes pecuniary loss to the parents. where many countries have been at cursing end as their children are underweight and to our dismay pakistan is not far behind. pakistan is well in the race to be in the top echelons of being the world’s most malnutrition country. as per the study children under age 5 are suffering malnutrition and the numbers are 44 percent and the number of children underweight is 31 percent which makes the situation worth mooting over (national institute of population studies, 2013). though these issues that has been discussed above make us lose sleep children’s being underweight and unhealthy. this study aims to bring about some solutions to this scorching task at hand. to get into the nitty-gritty of this study it was imperative for us to look into the behavior of the mothers. because it goes without saying the behavior of the mother will rub off of the child as well. this study is pioneering in this aspect in which we check the educational level and the awareness of the mother which is directly related to the health of the child. to pull through this study we made certain groups in which we divided the mothers in different groups based on their educational background such as (1-5, 6-10 and 11-16 years schooling) to check how does this effect of child’s health. infants look out to their parents in every dire need. father is responsible to bring home the bacon and to keep the roof over heads. mothers are all together are different ball games. in countries like pakistan mothers are revered as does god. they look out for every minute demands of their child. it is incumbent upon mothers to make the child nourished with all the nutrition’s one could have had asked for. the happiness and mental growth of the child depends upon this aspect of the study. in pakistan which is known to be a third world country mother’s education makes a lot of difference as it does in developed economies. if a mother is not educated which makes her less knowledgeable about certain issues? how does a mother who herself has no idea what health issues are supposed to make the demands of the child meet? education has always had an importance. developed economies have got there where they are through education and innovations. women’s do need to get educated. they are the ones who are going to keep our next generation to the next levels. in pakistan there is an adage that goes by mothers are the first teachers to their kids (novak, 2016). when on takes into account the health of a child his mother’s health also takes the precedence. as the famous maxim goes by healthy mind makes for the healthy body. if god forbid there is some issue with the mother the health of the child will be affected as well. they are the cog of the same machine (nafiu & hamidu, 2017). as per the study in america which showed that 54 percent of the deaths can be curbed by educating the mothers (polakow, butler, deprez, & kahn, 2012). this makes it indispensible tool to curb this menace of child death. ergo it is important to assume the link which has been developed here that education makes for a healthy nation. a healthy nation will turn journal of applied economics and business studies, volume. 4, issue 2 (2020) 1-22 https://doi.org/10.34260/jaebs.421 3 around the fate of the economy with adept mind and nimble bodies in place. dexterous mind is all today’s world ask for (cutler & lleras-muney, 2006). erstwhile studies have showed that there is a positive relation between mother’s education and a child health. this also has a strong impact on the income of the household. if a child is not adept the big chunk of the income of the household will go into the kitty of doctors and teachers which will cash in on this. this issue inadvertently makes up for a larger chunk in the country. economic growth of any country depends on how nimble and agile their children’s are as they are the ones who are going to reign in the workings of the country (borhan, ahmed, & hitam, 2018). how does this issue play out in the economy of any country regardless of it being developed our under developed? it is known to all and sundry that an education is an indispensible ingredient in the dish of the economy. taking a further plunge into the studies it is being noted mother’s education takes precedence over anything else in here. in pakistan mothers are known to take the bull by the horns inasmuch as going to any length to save their kid (ramos, dumith, & césar, 2015). hence, the quality education foremothers has a great share in every part of economy (wang & wong, 2011). 1.1 theoretical framework about the association between education and health the related theories used in this paper have been introduced by becker (becker & becker, 2009) and discussed by (chou, liu, grossman, & joyce, 2010). becker’s theory was associated with the allocation of time for both non-market and market activities by households. the foremost assumption is that they produce non-market commodities (e.g. health) consuming market commodities and time. becker & becker (2009) empirically examined the households’ choices about the number of babies and their health. according to his model, households get utility from formal commodities, as well as from the number of babies and their health quality (informal commodities), calculated by expenses per baby and babies’ health. this study follows becker’s model, with child health as an argument in the utility function for household. chou, liu, grossman, & joyce (2010)and grossman (1972)have estimated the impacts of education on non-market goods, i.e., the causal impacts of education and the mechanisms through which it impacts non-market goods. he has introduced a background in which education creates allocate efficiency impacts and productive efficiency impacts. productive efficiency impacts indicate that more education results in more efficiency in production process of non-market-sector; for a given number of inputs, more educated persons obtain better health outcomes than less educated ones. theoretical models mostly show that the economists are agreed upon the positive and significant parents’ education child health connection (grantham-mcgregor et al., 2007; behrman & rosenzweig, 2004; kovsted, pörtner, & tarp, 2002; strauss & thomas, 1995). some studies have marked little or no causal relationship (desai & alva, 1998; glewwe & miguel, 2007) muhammad umar farooq, syed asif ali naqvi and shaoan huang 4 while few studies have observed negative association (alderman & lavy, 1996). the related literature detects some nonlinearities in this association (iltaf, shahid, & khan, 2017; pal, pari, sinha, & dhara, 2017). desai & alva (1998) also concludes the presence of non-linearity in this association; arise from the greater magnitude of the estimate of higher education relative to the primary education’s estimate. barnet (1995) has reported that health based education is also more effective and comprehensive in secondary grade than the primary grade in pakistan. although this nonlinearity has substantial implications on policy measures, no serious effort to highlight this issue has been made in the literature. now the important question is how a threshold or non-convexity may arise in this relationship. it is argued that low cognitive capability via poor education, lower educational quality, ineffective education for health in prospects leads to a permanent cost, and hence to the threshold in the relationship of mother education with child nutrition. technically speaking, cognitive talent comprises a large range of characteristics, such as memory, perception, attention, action, mental imagery and problem solving. it is proven that mother’s cognitive talent should approach to a threshold so as to be effective in the production of child health. some past studies evidence that early child birth, and the corresponding diseases of mother and babies are an outcome of low cognitive talent. the mothers with low cognitive talent have three times higher chances to have two births before age 20 than those with the higher cognitive talent in usa. moreover, early child birth is also associated with lbw and miserable child health (shearer et al., 2002). thus, education quality has great importance in this association because it is essential in the development of health knowledge and numeracy skill, and its utilization. it is quite imaginable that a student with 5 years schooling may not have ability to write and read at all in developing nations. in this scenario, nonlinearity or a threshold in this association is not surprising. we can find the chances of threshold in the descriptive analysis of mother literacy and her education in table 8 given in appendix, using the filtered data set. this table 8 shows that 87.5%, 49%, 31.2%,18.7% and 10.7% mothers cannot read a single word at 1-5 years of schooling, respectively while 18.4%, 32.5%,38.7%, 16.9%, 16.7%, and 9.3% mothers can read a sentence but by parts. thus, this lower quality of 1-5 years of schooling may suggest a threshold in this relationship. this study contributes to the current literature by searching out the threshold level in mother’s education child nutrition gradient for pakistan. this work is essential for two reasons. firstly, the existence of a threshold may result in a poverty trap. inappropriate public expenditure on education may produce worsening health of the succeeding generations which results in low level of efficiency and income. this may deteriorate the income inequality condition in a country. secondly, it has imperative implications on policy problems. mostly, the governments of journal of applied economics and business studies, volume. 4, issue 2 (2020) 1-22 https://doi.org/10.34260/jaebs.421 5 developing nations offer subsidy to primary education for females due to the fact that the nonmarket benefits, like, children’s health, fertility, etc. are reasonably important, while its effect on female contribution into labor force is very little. the institutions’ statistics show a successive rise in the construction of primary institutions in pakistan along with a poor health status of children among primary passed mothers (national institute of population studies, islamabad, 2013). therefore, if the education for mother has no or negative effect on child health for a particular lower (primary) level of schooling (in other words, if a threshold level is present), policy makers need to call for steps to erase the possible reasons of threshold or reconsider about the size of the subsidy program. firstly, non-parametric methods are employed to explore the bivariate relationship of mother education with child health in order to determine the presence of threshold effects. if this association is nonlinear, dummies of different level of mother schooling along with zero schooling as reference category are added in the regression of child health to explore the effective level of education of mothers to impact child health. thus, this study will extends the literature on mother education child health gradient to search out any nonlinearity or threshold using pdhs, 2012-13) data. where threshold level implies the difference in the cognitive talent among mothers having different level of formal education, it also refers to the unseen able factors which may impact both mother education and child health, simultaneously. now an important question is how to deal with mother education in the econometric regression. in a different channel, given that school enrollment for female is lower; mothers who attain education are inherently more talented and motivated. their motivation and talent are unseen able factors for the researcher and missing from the regression. therefore, estimates of education endogeneity bias. mostly, past studies regarding the relationship of mother education with child health have tried to employ instrumental factors (factors which are correlated with mother education but are uncorrelated with unseen able factors that directly impact child health) to get rid of the biasness arisen from the endogeneity problem. since it is very difficult to find the instruments that obviously satisfy this condition, they do not promise to solve the endogeneity issue. the instruments may consist of such different variables that may explain mothers’ access to different kind of institutions over the periods when she was schooling going age (6-10 years) like some studies have used a dummy of mother born place as an instrument, showing that whether educational institutions were easily available in her living place over the years of her school age (6-10) or not and some studies have assumed mother’s birth cohort as instrumental factor indicating the higher intensity of institutions’ construction followed by the policies on education muhammad umar farooq, syed asif ali naqvi and shaoan huang 6 over the periods of her birth cohort(currie & almond, 2011). the logic behind this instrument is to reveal mother’s accessibility to the institutions in her resident place and her birth cluster. this study also makes a contribution in term of introducing instrumental factors for mother education. we find a great variation in health policies, reforms and higher public expenditures on education over 1955-1998 in pakistan. the mothers born over the period of 1963-1996 are highly affected by these reforms and plans because there is a wide difference in the intensity of institutions’ construction in this period. furthermore, we also see a clear difference in the average education within different birth cohorts. thus, mother education is instrumented with her birth cohort because where it reflects a differential intensity of institutions’ construction; it also captures a great variation of the relevant institutions’ supply in the period of enrollment (6-10 years), which presents her differential access to them. this factor is correlated with the education of mothers and uncorrelated with the unseen able factors that could affect child health. due to missing the data on her birth place, only birth cohort is focused. our study also takes mother nutritional level as instrument for mother education because it impacts her education and child health, simultaneously (ahmed & iqbal, 2007). secondly, a large variation in the educational reforms and plans, and the public expenditure in the education sector during 1955-1998 provoke the selection of this country and the construction of this instrument for mother education. up to my knowledge, none study has used a variation in education policies to get information about her education and child health associations in pakistan. however, some past studies have done this kind of work for nigeria, taiwan, indonesia, and the u.s. a list of health inputs included to produce informal goods (infants’ health) may also be endogenous (e.g. food intake, mother decision to vaccinate babies, her health practices and sanitary). the household economic theory states that both child health and his nutrition inputs (e.g. calorie intake, duration of breastfeeding) are jointly determined by households (becker &becker, 2009). however, the latest data-set pdhs 2012-13 on these inputs are missing. the number of living children and their health quality are also simultaneously determined as in the becker and lewis’ model (1973). likewise, mothers’ participation decision in the economic activities and their infants’ health are expected to be jointly determined. hence, family income level may be endogenous factor because infants’ illness may negatively influence parents’ working hours (labor supply) and earning(chen & li, 2009). in our analysis, mother education, her knowledge about health, wealth index and access to health services are assumed endogenous health inputs due to any of these possible biases. given the possible biases, 2sls-iv regression is employed to estimate the reduced form demand function for child health and the endogenous health inputs to get rid of the endogeneity issue. journal of applied economics and business studies, volume. 4, issue 2 (2020) 1-22 https://doi.org/10.34260/jaebs.421 7 2. data and methods successive 8 educational plans in pakistan over the period 1955-1998 record large variations in the educational policies and investment in education sector, just showing a great focus on primary schooling. it motivates us to select pakistan and construct the policy instrument for purging the coefficients of mother education from endogeneity bias along with finding the threshold in the relationship of mother education with child health. thus, this case study for pakistan is primarily chosen due to the choice of instrument. mother education in the production function for child health is endogenous as mother’s unobserved characteristics for instance, her ability, determination and motivation etc. impact her educational achievement and ultimately, affects her children’s health. in this study, instrument for mother education is constructed on the basis of a large variation in educational policies, programs and the government investment in this sector(central bureau of education and federal bureau of statistics, 1998). relevant data are filtered from the pdhs 201213 data-set. a sample of 3404 women aged 15-49 who could produce birth cards containing child birth height and weight, were selected from the data set. these children were under 5. height-for-age z score (haz) and weight-for-age z score are taken as the indicators of long (stunting) and short run (underweight) health outcomes for under 5 children. since one of key objective of this research is to see the existence of a threshold that may produce a poverty trap, long term child health indicator is much closer to our analysis in the scenario of achieving 4th mdg (reducing under 5 infants’ mortality rate by 2/3 during 1990-2015). three dummies for mother education are created: 1-5=1, 6-10=2 and 11-16=3, with zeros schooling as the reference category. we could not produce dummies for each category of education because the observations for the grades of 1, 2, 3, 6, 7 and 9 are not enough to accurately estimate the marginal impact. mother knowledge about health is also a part of key determinants of child health. mother knowledge about hepatitis disease is taken as the proxy for it. it is in the binary form (if she doesn’t know about hepatitis =0, otherwise=1). it is also assumed an endogenous factor because maternal knowledge about health is mostly judged via her ability to understand, evaluate, and apply health knowledge with the usage of health promotion resources. therefore, this endogenous variable is instrumented by the information containing watching tv, reading newspapers, and listening radio(strauss & thomas, 1995) to get rid of possible omitted factor bias. all instruments are also in the binary form (no=0 and yes=1). lifetime permanent household income is a significant predictor of the long term child health status, thus it should be a part of the health regression controlling for the income impact. as the data on permanent household wealth are hardly available for the research work, present income or present expenditures are mostly taken as a proxy. when current income obviously leads to measurement error problem, it is likely to be endogenous to the decisions of household health. to keep away from this kind of bias, wealth statistics of the household and non-labor income are commonly taken as a proxy of permanent income. unfortunately, pdhs 2012-13 misses the data muhammad umar farooq, syed asif ali naqvi and shaoan huang 8 collection on expenditure or any income. however, but it successfully collected the information about various kind of household holdings for instance, ownership of a radio, a television, a scooter, a bicycle etc. as well as residence characteristics like drinking water, sanitation facilities and the types of floor and roof material. a wealth index is built by pdhs using the information about all possible assets, and principle component analysis. different weights are assigned to the households’ holdings for instance, television, radio, etc. and to the living circumstances (for example, type of roof materials, etc.). the wealth index is also used to minimize the endogeneity issue, because included assets in this index are accumulated by a household over time and are less likely to be jointly determined through child health (linnemayr, alderman, & ka, 2008). thus, wealth index is accepted as a proxy of permanent level of income and the households’ living standard. thus, wi scores are categorized into poor=0, middle class=1 and rich=2. however, due to lack of guarantee that wi is exogenous, durbin chi-square and wu-hausman f-statistics for endogeneity test are conducted in the following sections. unluckily, pdhs 2012-13 also failed to collect information about health facilities, the availability of health workers, or any factor of community living environment. however, the survey contained the questions like whether the prenatal care is received by the mother or not, whether family planning employee visited her in the last 12 months, and whether the piped water facility was available for the household. all of these variables are binary. this research used medical assistance from nurses as a proxy for the community health services in the binary form. some studies have also assumed it as endogenous(becker & becker, 2009).residence factor captures urban-rural variation in infants’ health outcomes. 2.1 methods mukabutera et al. (2016) and strauss & thomas (1995) have introduced a theoretical framework of households’ decision making for the derivation of the reduced form demand function for children’s health. their specification is followed to estimate the regression equation as given below: 𝑯𝒊 = 𝜶 + 𝜷𝟏 𝑴𝒆𝒅𝒖𝒊 + 𝜷𝟐𝑿𝒊 + 𝒆𝒊 (1) where, hi is health status of the child i in a household, x is the vector of controlled factors such as child’s age, its square, parents’ age, region, access to health services dummy, mother knowledge dummy and scores of wealth index, med is a vector of mother education dummies, and e is the residual. 2.2 causal association of mother education with children’s health the modeling scheme will incorporate netting out the influence of the unobserved elements in the estimation procedure, which may impact both mother education and child health association, simultaneously. to consider this possible endogeneity, we use two-stage least squares instrumental variables methods (2slsiv).partially, duflo (2001)has helped to erase the issue of possible journal of applied economics and business studies, volume. 4, issue 2 (2020) 1-22 https://doi.org/10.34260/jaebs.421 9 endogeneity by utilizing mothers ‘birth cohort as instrument variable and has reported that indonesian government made a huge investment in schooling construction programs between 1973-78 and introduced the interaction between the total schools built in the mothers’ birth region and her birth cohort as instrument for her education. beginning with the national conference on education held in 1947, seven national policies and eight 5 year plans on education, and more than six other schemes were announced and launched over the period 1955-1998 in pakistan. under the national educational policy of 1992, numerous projects under nation and worldwide organizations were launched to enhance women participation in education. table 1: numbers of different type of institutions over different time spans institutions 1964-65 1974-75 1984-85 1994-95 1996-97 mother education between 1-5 821 15673 21551 39987 46691 middle schools 589 1266 1817 5582 6425 higher secondary vocational institutions 425 218 1434 3410 3736 arts and science colleges 62 96 153 251 296 professional colleges 5 8 8 9 9 universities (male and female) 6 10 21 24 25 where a wide variation in the record of newly established female institutions was a response of earlier educational plans and policies, it also implied different accessibility for females born in the corresponding time spans. a previous study assessed a stark difference of public investment in schools’ construction in rural-urban regions of nigerian order to suggest birth place of mother as instrument (moja, 2000). mothers ‘birth years are divided into four cohorts, 1963-72, 1973-79, 1980-88 and 1989-96 in the light of table 1. this will help us to investigate the impacts of different educational institutions’ construction on the average education for mothers in each cohort. table 2 depicts the average education for mothers in different birth cohorts. table 2: average education in years by mothers’ birth cohort mother birth cohorts mean n std. deviation std. error of mean 1963-72 1.67 226 3.732 0.248 1973-79 3.43 741 5.022 0.184 1980-88 4.36 1636 5.037 0.125 1989-96 3.69 801 4.548 0.161 total 3.82 3404 4.893 0.084 we find a significant variation in the average education of mothers across cohorts. the younger mothers have higher average education relative to the elder mothers regardless of their residence place in childhood because overall institutions for female education have increased over the periods of 1974-75 and 1984-85. however, the mothers in the last cohort show little decline in the average education from their elders. this may be attributed to less construction of formal muhammad umar farooq, syed asif ali naqvi and shaoan huang 10 educational institutions in this period. table 3 revels a large variation in the average education among mothers born in different cohorts. table 3: variation in average level of education across cohorts birth cohorts difference s.e unadjusted c.i at 95% (1973-79)-(1963-72) 1.757 0.368 1.035 2.479 ( 1980-88)-(1963-72) 2.683 0.344 2.009 3.357 (1989-96)-(1963-72) 2.020 0.365 1.305 2.736 (1980-88)-(1973-79) 0.927 0.215 0.506 1.347 (1989-96)-(1973-79) 0.264 0.247 -0.221 0.748 (1989-96)-(1979-80) -0.663 0.209 -1.073 -0.253 a significant variation in the average education for mothers between cohorts is observed. it implies that mothers related to different cohorts have differential access to institutions as highlighted in table 2. consequently, the variation in the public investment for education over time is resulted in differential average education for mothers. this demands to employ her birth cohort as instrument variable for mother education. moreover, mother education is also instrumented by her bmi because healthy mothers have more likelihood of attaining more education (ahmed, 2007). thus, to measure the influence of instruments on endogenous factors, the following equation will be estimated: 𝑬𝒊 = 𝜶 + 𝜷𝟏𝒁𝒊 + 𝜷𝟐𝑿𝒊 + 𝜺𝒊 (2) where e is the vector of endogenous factors, α is a constant, z is a vector of instruments, x is a vector of controlled factors, β1 and β2 are parameters, and 𝜺 is an error term. the estimation of equation (2) will give us predicted values which are not directly linked with child health. this is the first stage of a 2sls estimation to find out the existence of causal connection between maternal education and child health (hypothesis 1). the 2nd stage of 2sls equation is: 𝑯𝒊 = 𝜸 + 𝜽𝟏𝑬𝒊 + 𝜽𝟐𝑿𝒊 + 𝝁𝒊 (3) where hi is the health outcomes for child i, γ is a constant, ei is a vector of endogenous factors, x is a vector of controlled factors and μi is an error term. the predicted values from 1st stage are replaced with endogenous factors (ei) in equation (3) and later on, this equation is estimated to get unbiased estimators confirming that the predicted values are not correlated with the µiin equation (3). 3. regression results this study demonstrates the association of mother education with health outcomes of under 5’s infants. as mother education and her health knowledge are endogenous regressors, this study employs 2sls-iv regression. these two endogenous regressors are incorporated in the model after verification of over-identification and endogeneity tests. results from the sargan chi-square test journal of applied economics and business studies, volume. 4, issue 2 (2020) 1-22 https://doi.org/10.34260/jaebs.421 11 for over-identification and durbin chi-square, and wu-hausman f statistics for endogeneity are added in the bottom of table 4 and 5. the p-values of sargan statistics are 0.48 and 0.59 in haz regressions and 0.3 and 0.3 in waz regressions, indicating that “the joint h0 = the instruments are valid” cannot be rejected. thus, the instruments qualify the over-identification criteria and thus, they are jointly valid instruments and are correctly omitted from the second stage estimations. it infers that instruments are not correlated with the residual term. this statistics also implies that although some instruments may be weak, overall they are valid. the durbin chi-square and wuhausman f-statistics confirm that the regressors assumed endogenous are really endogenous. the null hypothesis of these statistics is that the regressors are exogenous. the p-values <0.05 in both specifications show that these two regressors are endogenous. however, wi and access to health services factors incorporated in the following specifications are exogenous factors due to insignificant endogeneity statistics. given that these possible endogenous factors are indeed endogenous, 2sls-iv regression method is preferred over ols for estimation as shown in table 4 and 5. table 4: impact of mother’s education on child’s height for age: specification (1) regressors ols (1) iv (2) ols (3) iv (4) mother education between 1-5 0.3 (0)*** 1.18 (0.55) 0.173 (0.12) -0.38 (0.89) 6-10 0.67 (0)*** 1.8 (0.36) 0.51 (0)*** 0.82 (0.67) 11-16 0.77 (0)*** 3.61 (0.05)** 0.6 (0)*** 3.37 (0.042)** mother health knowledge 0.15 (0.1)* 3.02 (0)*** 0.23 (0.02)** 2.99 (0)*** mother age -0.012 (0.07)* 0.02 (0.18) -0.015 (0.03)** .015 (0.36) father age -0.015 (0)*** 0.001 (0.8) -0.016 (0)*** -0.001 (0.94) child age 0.08 (0)*** -.05 (0)*** -0.08 (0)*** -0.055 (0)*** age squared 0.001 (0)*** 0.001 (0)*** 0.001 (0)*** 0.001 (0)*** residence (1) -0.17 (0.02)** -0.5 (0)*** -0.074 (0.36) -0.58 (0)*** access to health services -0.01 (0.9) -.103 (0.44) wealth index (1) 0.18 (0.07)* 0.88 (0)*** wealth index (2) 0.41 (0)*** 1.01 (0.04)** over identification test (d.f) 3.50 (4) (0.48) 2.83 (4)(0.59) wu-hausman test (d.f) 15.4 (6) (0) 17.72 (6)(0) durbin chi-squared test (d.f) 90.35 (6) (0) 103.6 (6)(0) observations 3404 3404 3404 3404 r-squared 0.47 o.48 instruments=mother birth cohort, her bmi, listening radio, and reading newspaper, p-values are in parentheses, *** significant at 1%; ** significant at 5%; * significant at 10%. firstly, we discuss columns (1) and (2) in table 4, which present ols and 2sls-iv estimates for haz specification, respectively. this specification contains mother education dummies, mother and father age, child age, child age squared and residence dummy as key regressors. as statistics rejects h0, iv will be preferred over ols, we will only focus on the iv results and will take ols results to throw light on the magnitude and direction of the biased coefficients.ols muhammad umar farooq, syed asif ali naqvi and shaoan huang 12 estimates for all grades of mother education are highly significant and interestingly, the coefficients of 6-10 and 11-16 years education get double and more than double than that of1-5.it implies that hfz improves at different rate as education level of mothers increases. this concludes that this impact may not be linear in education. although ols outcomes verify a highly significant influence of 1-5 and 6-10 years education on haz, this significance impact goes away with iv. however, iv regression indorses that only 11-16 years of education is significantly (p<0.05) impacting haz and this impact is positive because mothers with higher education are more likely to avail more economic opportunities and are more empowered in household decision making and utilizing resources that definitely has positive influence on haz. it is also confirmed that well educated mother’s truly increase the hygienic conditions for their babies, which finally keep them healthier. this outcome is also aligned with (ahmed, 2007; currie & moretti, 2003). the magnitude of its coefficient much bigger (3.61) than that of the ols estimate for two reasons implying biased ols estimates. firstly, this much bigger estimate of the higher education in iv regression explains the omitted variable baseness in the corresponding ols estimate as well as broadly shows that the measurement error in education, which may result in downward biasness in the ols coefficient, partly offsets any upward ability bias (angrist & newey, 1991; card, 1993; psacharopoulos & patrinos*, 2004; staiger & stock, 1994). furthermore, this measurement error impact is compounded if talent bias in ols coefficients is relatively small. however, this measurement error in education alone may not elucidate such big differences (card, 1993). thus, the bigger iv estimates highlights that smaller ols estimates are highly biased and show the causal association between mother education and waz, and haz in table 4 and 5. furthermore, inconsistent association of 1-5 and 6-10 years of schooling on waz and haz may be due to the fact that 1-10 years of schooling fail to equip them with necessary health knowledge to promote child health despite of the successive 8 education plans to boost up females’ primary schooling in their birth cohorts. thus, it refers to the existence of threshold at 6-10 years of schooling in this specification. this outcome is contradicted to (ahmed, 2007; maïga, 2013)who empirically find the threshold level at 5 years of schooling. the empirical outcomes underline the same scenario in the comparison of ols and iv estimates of mother knowledge. the coefficient of mother knowledge about health is highly positive and significant (p<0.01) showing that mothers with rising trend in health knowledge have higher haz by 3.02 s.d of the height for the similar sex and age group of children, holding other factors constant due the fact that health knowledge produces common awareness within society and ultimately, they can get more benefits from health oriented programs launched by public sector. according to abuya, ciera, & kimani-murage (2012), the risk of contagious diseases can be curtailed via health knowledge and easy approach to latest medicines. the columns 3 and 4 approximately presents the same outcomes for the estimates of mother education and her journal of applied economics and business studies, volume. 4, issue 2 (2020) 1-22 https://doi.org/10.34260/jaebs.421 13 knowledge in term of comparison between ols and iv regressions except the coefficient of primary education in ols estimation becomes insignificant. a convex connection between age and haz is measured in all columns of this specification. the estimates of both child age and its squared are significantly (p<0.01) negative and positive, respectively, demonstrating that haz decreases with age but with diminishing slope. it means that elder infants are worse in haz than younger ones. this outcomes is also supported by (ahmed, 2007; krueger & malečková, 2003).however, the estimates of mother and father’s age are inconsistent in iv regression while they are highly significant in ols estimation with a very small magnitude. rural mothers have a disadvantage of 0.50and 0.58 s.d (p<0.01) in terms of height comparing with the urban ones in columns 2 & 4 because they mostly lack basic health services and hold poor infrastructure and environmental settings. this result is also favored by(fakir & khan, 2015; khattak, iqbal, & ghazanfar, 2017). the estimates of residence are also improved in iv regression, implying the better choice of educated mother in residence for better health their babies. wealth index and access to health services are added in the 3 and 4 columns to understand how the magnitude of mother education’s estimates is altered with the inclusion of these variables. by doing so, the coefficient of mother education stays on significant (p<0.05), but it reduces marginally, indicating that the key impact of higher education on haz is not via income. this results is supported by (maiga, 2015).this outcome shows that the rising wealth status of middle and rich class families gives a rise to haz by 0.88 and 1.01 s.d respectively, holding all else constant because rich parents can meet the expense of more nutritious food and better medication, and can afford a healthy environment setting for their children so that they remain healthier (chai et al., 2016; farooq, 2012).however, an access to health services has inconsistent connection with haz in columns 2 and 4. summary statistics of variables of interest presented in table 7 (in appendix) show that about 54% of mothers have zero education, 15% have 1-5 years of schooling, 19% have 6-10 years of schooling and only 11% have 11-16 years of education. the mean of haz at -1.97 is very closed to the who declared moderate stunting (-3 < haz < -2) criteria while average waz (-1.34) is far from moderate underweight condition (-3 < haz < -2)(wolde, berhan, & chala, 2015). 42% mothers grew up in urban areas while 58% are resided in rural regions. the average mbmi shows that overall; mothers are healthy in the sample. on average, 39% households belong to rich and middle class. muhammad umar farooq, syed asif ali naqvi and shaoan huang 14 table 5: impact of mother’s education on child’s weight for age : specification (2) regressors ols (1) iv (2) ols (3) iv (4) mother education between 1-5 0.05 (0.52) 1.1 (0.53) 0.04 (0.6) -1.42 (0.55) 6-10 0.34 (0)*** -1.66 (0.4) 0.3 (0)*** -1.7 (0.34) 11-16 0.51 (0)*** 4.39 (0)*** 0.48 (0)*** 3.8 (0)*** mother health knowledge 0.34 (0)*** 1.95 (0)*** 0.34 (0)*** 1.9 (0)*** mother age -0.003 (0)*** -0.003 (0.84) -0.014 (0)*** -0.008 (0.62) father age -0.002 (0.48) 0.009 (0.12) -0.002 (0.5) 0.008 (0.13) child age -0.03 (0)*** -.007 (0.44) -0.03 (0)*** -0.008 (0.38) age squared 0.0001 (0)*** 0.00001 (0.9) 0.0003 (0)*** 0.00001(0.9) residence (1) 0.28 (0)*** -0.12 (0.25) -0.25 (0)*** -0.4 (0)*** access to health services -0.062 (0.33) -0.15 (0.15) wealth index (1) -0.038 (0.6) 0.72 (0.02)** wealth index (2) 0.075 (0.4) 1.1 (0)*** over identification test (d.f) 2.4 (2) (0.3) 2.35 (2)(0.3) wu-hausman test (d.f) 20.55 (6) (0) 21 (6)(0) durbin chi-square test (d.f) 119.5 (6) (0) 122 (6)(0) observations 3404 3404 3404 3404 r-squared 0.4 0.4 instruments=mother birth cohort,listening radio, and reading newspaper, p-values are in parentheses, *** significant at 1%; ** significant at 5%; * significant at 10%. approximately, the similar kind of outcomes for mother education and her knowledge in term of comparison between ols and iv regressions in all columns of table 5 are measured except the primary education in ols columns becomes insignificant. similarly, the bigger coefficients of 1116 years of education in iv columns validate that the corresponding ols coefficients are highly biased, showing the causal connection between mother education and waz. iv columns reveal that mothers with an increasing rate of higher education raises waz by 4.39 s.d and 3.8 s.d respectively, controlling other factors. the signs of the other controlled factors are reasonable in ols columns. the negative estimate of child age and positive estimate of its square in ols point out that child’s waz firstly reduces with age and improves at the later time. however, these age variables are inconsistent in iv regression. the magnitude of residence factor in iv gets little better, implying that that rural child also suffers a disadvantage of waz by 0.874 s.d due to lack of basic health facilities. in this specification, wealth index and access to health services are also added in columns 3 and 4 because we want to see how the magnitude of estimates is changed with the inclusion of these variables. by doing so, access to health services remains inconsistent, but the estimate of the higher education is significantly reduced while controlling for other factors, indicating that the influence of the higher education on waz is partially through the channel of wealth status. this outcome is aligned with (chen & li, 2009; chou et al., 2010). however, it is still significantly journal of applied economics and business studies, volume. 4, issue 2 (2020) 1-22 https://doi.org/10.34260/jaebs.421 15 much bigger than that of ols. iv estimates show that an increasing trend of middle and rich families’ wealth significantly scales up waz by 0.72 and 1.1 s.d (p<0.05 and p<0.01) respectively. to sum up, it is reliably found that the 11-16 years of education has highly significant and positive influence on waz and haz while ols estimates of mother education are also highly significant, but they are seriously a victim of biasness as shown from their magnitudes. this outcomes is also aligned with (ahmed, 2007; currie & moretti, 2003). iv estimations for both cases indicate threshold at 6-10 years of mother education. thus, mother education impactswaz and haz only at 11-16 years of education. 3.1 semi-parametric regressions results to examine the threshold effects, the bivariate maternal education infants’ health outcomes’ connections are measured. a local polynomial smoothing of the semi-parametric relationships between haz, waz and mother education at 95% confidence interval curves, are shown in figure 1 and 2. a 2ndorder polynomial degree is taken. the graph predicts the non-linear relationship. given that overall mother education in pakistan is very low, the confidence intervals are appeared much wider after 14 and 15 years education for haz and waz, respectively because few mothers are highly educated. therefore, the coefficients for those grades of education are fewer precise. -1 0 -5 0 5 m o th e r y e a rs o f e d u c a ti o n 0 5 10 15 haz 95% ci wfa lpoly smooth kernel = rectangle, degree = 2, bandwidth = 3.35, pwidth = 5.02 figure 2: semi-parametric regression of waz on mother’s years of education muhammad umar farooq, syed asif ali naqvi and shaoan huang 16 3.2. threshold effects in the preceding section, the bivariate associations between mother education, waz, and haz are non-linear. therefore, haz and waz are regressed on the dummies for all grades of mother education in the health production function, along with zero education as reference category to verify the threshold impacts. the table 6 shows that the estimates of 16 years are 1.73 (p<0.01) and 0.88 (p<0.01) for haz and waz specifications, respectively. these are the biggest and positive estimates, implying that 16 years education will raise haz by 1.73 s.d relative to 0.51 s.d for 5 years of schooling and 16 years education will also scale up waz by 0.9 s.d relative to 0.52 s.d for 10 years of education. where the pattern of coefficients at different education dummies provides justification for non-linear relationship between mother education, waz, and haz, it also confirms that the influence of mother education on waz and haz is much strong only at 16 years education, implying the presence of threshold at secondary education. the negative estimates may reveal that students are very bad at that level of education. table 6: threshold effects regression results education dummies haz(ols) waz (ols) 1 0.39 (0.53) 0.061(0.41) 2 0.31 (0.30) -0.08 (0.24) 3 0.38 (0.24) ** 0.419 (0.19) 4 0.42 (0.25) * 0.122 (0.19) 5 0.510 (0.13) *** 0.16(0.1) 6 0.82 (0.31) *** 0.34 (0.24) 7 1.117 (0.32) *** 0.8 (0.25) *** 8 0.85 (0.164) *** 0.43 (0.13) *** -5 0 5 m ot he r ye ar s of e du ca tio n 0 5 10 15 haz 95% ci hfa lpoly smooth kernel = rectangle, degree = 2, bandwidth = 2.65, pwidth = 3.97 figure 1: semi-parametric regression of haz on mother’s years of education journal of applied economics and business studies, volume. 4, issue 2 (2020) 1-22 https://doi.org/10.34260/jaebs.421 17 9 0.57 (0.34) * 0.33 (0.26) 10 0.9 (0.12) *** 0.52 (0.09) *** 11 -0.83 (.64) -0.41 (0.5) 12 1.003 (0.17) *** 0.735 (0.13) *** 13 1.187 (1.48) 1.42(1.15) 14 0.59 (0.2) *** 0.59(0.16) *** 15 0.24 (1.48) 0.38 (1.15) 16 1.73 (0.237) *** 0.88(0.18) *** constant -2.32 (0.048) *** -1.54(0.04)*** robust standard errors in parentheses *** p<0.01, ** p<0.05, * p 4. conclusion and recommendations this paper elaborates the mother education child health gradient in pakistan for under 5’s children. the summary statistics declare that the mean value of haz at -1.97 and waz at 1.34.though the past studies point out the presence of nonlinearity and threshold in this association, no effort has been made to investigate this issue, despite its substantial implications on public health programs. a change in the educational policies captured by mother birth cluster is taken to struggle for purging the coefficients of mother education from endogeneity bias. as hausman statistics discards h0, iv is preferred over ols, only iv results are focused and ols results are taken to highlight the magnitude of the biased coefficients. though ols estimates corroborate a highly significant and positive influence of 1-5 and 6-10 years education on haz and similar effect of 6-10 years of schooling on waz, these significance impacts go away with iv in both specifications. however, iv regressions indorse that only 11-16 years education of mother is significantly impacting haz and waz and the bigger iv estimates of 11-16 years education in both specifications highlight that smaller ols estimates are highly biased and show the causal association between mother education and child health due to endogeneity bias. thus, iv estimations in both cases indicate threshold at 6-10 years of mother education, signifying that mother education impacts waz and haz only at 11-16 years of education. similarly, where the pattern of coefficients at different education dummies provides justification for non-linear relationship between mothers’ education and waz, and haz, it also validates that the influence of mothers’ education on waz and haz is much strong only at 16 years education implying the existence of threshold at secondary education. where an inconsistent association of 1-5 and 6-10 years of schooling with waz and haz in iv regressions defines the limit of threshold, it also highlight that 1-10 years of schooling fail to equip them with necessary health knowledge to give arise child nutrition in pakistan. the empirical outcomes underline the same scenario of endogeneity biasing the comparison of ols and iv estimates of mother knowledge for health. as muhammad umar farooq, syed asif ali naqvi and shaoan huang 18 wealth index and access to health services are added in both specifications, the coefficients of mother 11-16 years education stays on significant, but it reduces marginally in haz specification, indicating that the key impact of higher education on haz is not via wealth status while this estimate in waz specification is significantly reduced, signifying that the influence of the higher education on waz partially works through the channel of wealth status. however, access to health services remains inconsistent in both specifications. a local polynomial smoothing graph also predicts the non-linear relationship between mother education, haz, and waz. it is being vehemently argued here that the nutrition initial 2 years of a child is pivotal for both mental and physical growth. it has been proven both rationally and by putting past data sets together that the deficiency of nutrients can have long lasting and gruesome effects not on the child alone but on the whole nine yards of the society as well. investment in female education can bring about a change for good in labor productivity. the threshold impact identifies that the biggest impact of education on haz and waz is occurred at 11-16 years of education. therefore, educational policies designed for girls should emphasize on trying to attach them with schools and colleges after 10 year of schooling to equip them with necessary health knowledge to give rise to child nutrition in pakistan, instead of plans to promote only primary schooling. this will only be possible at higher short term financial cost on the behalf of other productive sectors, but it will raise long term future labor productivity and economic growth in pakistan. health knowledge effect suggests that health care information programs designed for girls can be helpful to reduce haz and waz in pakistan. the wealth index influence in wfa and haz specifications direct that educated women should be facilitated to join income-generating activities in order to share household expenditures and finance their infants’ nutrition. references abuya, b. a., ciera, j., & kimani-murage, e. 2012. effect of mother’s education on child’s nutritional status in the slums of nairobi. bmc pediatrics, 12(1), 80. ahmed, m. 2007. is there any threshold in mother’s education and child health relationship, (august). ahmed, m., & iqbal, k. 2007. is there any threshold in mother’s education and child health relationship? evidence from nigeria. carleton college department of economics. working series. alderman, h., & lavy, v. 1996. household responses to public health services: cost and quality tradeoffs. the world bank research observer, 11(1), 3–22. angrist, j. d., & newey, w. k. 1991. over-identification tests in earnings functions with fixed journal of applied economics and business studies, volume. 4, issue 2 (2020) 1-22 https://doi.org/10.34260/jaebs.421 19 effects. journal of business & economic statistics, 9(3), 317–323. balint, a. (2018). love for the mother and mother love 1. in primary love and psychoanalytic technique (pp. 109–127). routledge. barnet, e. (1995). health and hiv/aids education in primary and secondary schools in africa and asia-research paper no. 14. liverpool: school of tropical medicine. becker, g. s., & becker, g. s. (2009). a treatise on the family. harvard university press. behrman, j. r., & rosenzweig, m. r. (2004). returns to birthweight. review of economics and statistics, 86(2), 586–601. borhan, h., ahmed, e. m., & hitam, m. (2018). co2, quality of life and economic growth in asean 8. journal of asian behavioural studies, 3(6), 55–63. card, d. (1993). using geographic variation in college proximity to estimate the return to schooling. national bureau of economic research. central bureau of education and federal bureau of statistics, p. (1998). education s. retrieved from http://www.pbs.gov.pk/ chai, j., fink, g., kaaya, s., danaei, g., fawzi, w., ezzati, m., … fawzi, m. c. s. (2016). association between intimate partner violence and poor child growth: results from 42 demographic and health surveys. bulletin of the world health organization, 94(5), 331. chen, y., & li, h. (2009). mother’s education and child health: is there a nurturing effect? journal of health economics, 28(2), 413–426. chou, s.-y., liu, j.-t., grossman, m., & joyce, t. 2010. parental education and child health: evidence from a natural experiment in taiwan. american economic journal: applied economics, 2(1), 33–61. currie, j., & almond, d. 2011. human capital development before age five. in handbook of labor economics (vol. 4, pp. 1315–1486). elsevier. currie, j., & moretti, e. 2003. mother’s education and the intergenerational transmission of human capital: evidence from college openings. the quarterly journal of economics, 118(4), 1495–1532. cutler, d. m., & lleras-muney, a. 2006. education and health: evaluating theories and evidence. national bureau of economic research. desai, s., & alva, s. 1998. maternal education and child health: is there a strong causal relationship? demography, 35(1), 71. https://doi.org/10.2307/3004028 duflo, e. 2001. schooling and labor market consequences of school construction in indonesia: muhammad umar farooq, syed asif ali naqvi and shaoan huang 20 evidence from an unusual policy experiment. american economic review, 91(4), 795– 813. fakir, a. m. s., & khan, m. w. r. 2015. determinants of malnutrition among urban slum children in bangladesh. health economics review, 5(1), 1–11. https://doi.org/10.1186/s13561015-0059-1 farooq, m. 2012. maternal and child health journal mother education and child health status : a case study of pakistan muhammad umar farooq center for economic research , shandong university , jinan address : center for economic research , shandong university , 27 shanda . glewwe, p., & miguel, e. a. 2007. the impact of child health and nutrition on education in less developed countries. handbook of development economics, 4, 3561–3606. grossman, m. 1972. on the concept of health capital and the demand for health. journal of political economy, 80(2), 223–255. iltaf, g., shahid, b., & khan, m. i. 2017. incidence and associated risk factors of low birth weight babies born in shaikh khalifa bin zayad al-nayan hospital muzaffarabad, azad jammu and kashmir. pakistan journal of medical sciences, 33(3), 626. khattak, u. k., iqbal, s. p., & ghazanfar, h. 2017. the role of parents ’ literacy in malnutrition of children under the age of five years in a semi-urban community of pakistan : a casecontrol study, 9(6). https://doi.org/10.7759/cureus.1316 krueger, a. b., & malečková, j. 2003. education, poverty and terrorism: is there a causal connection? journal of economic perspectives, 17(4), 119–144. linnemayr, s., alderman, h., & ka, a. 2008. determinants of malnutrition in senegal: individual, household, community variables, and their interaction. economics & human biology, 6(2), 252–263. maiga, e. 2015. agrodep working paper 0019 mother ’ s education and children ’ s nutrition outcomes in burkina faso : is there a strong causal relationship ?, (november). maïga, e. w. h. 2013. the impact of mother’s education on child health and nutrition in developing countries: evidence from a natural experiment in burkina faso. in african economic conference. moja, t. 2000. nigeria education sector analysis: an analytical synthesis of performance and main issues. world bank report, 3, 46–56. mukabutera, a., thomson, d. r., hedt-gauthier, b. l., basinga, p., nyirazinyoye, l., & murray, m. 2016. risk factors associated with underweight status in children under five: an journal of applied economics and business studies, volume. 4, issue 2 (2020) 1-22 https://doi.org/10.34260/jaebs.421 21 analysis of the 2010 rwanda demographic health survey (rdhs). bmc nutrition, 2(1), 40. https://doi.org/10.1186/s40795-016-0078-2 nafiu, l. a., & hamidu, u. w. 2017. prevalence of five-child-killer diseases and under-five mortality in adamawa state, nigeria. kiu journal of social sciences, 3(1), 13–20. national institute of population studies, islamabad, p. 2013. pakistan demographic and health survey 2012-13. islamabad, pakistan. novak, l. 2016. persistent norms and tipping points: female genital cutting in burkina faso. working paper. http://www. lindseyknovak. com/research. pal, a., pari, a. k., sinha, a., & dhara, p. c. 2017. prevalence of undernutrition and associated factors: a cross-sectional study among rural adolescents in west bengal, india. international journal of pediatrics and adolescent medicine, 4(1), 9–18. https://doi.org/https://doi.org/10.1016/j.ijpam.2016.08.009 polakow, v., butler, s. s., deprez, l. s., & kahn, p. 2012. shut out: low income mothers and higher education in post-welfare america. suny press. psacharopoulos, g., & patrinos, h. a. 2004. returns to investment in education: a further update. education economics, 12(2), 111–134. ramos, c. v, dumith, s. c., & césar, j. a. 2015. prevalence and factors associated with stunting and excess weight in children aged 0‐5 years from the brazilian semi‐arid region. jornal de pediatria (versão em português), 91(2), 175–182. shearer, d. l., mulvihill, b. a., klerman, l. v, wallander, j. l., hovinga, m. e., & redden, d. t. 2002. association of early childbearing and low cognitive ability. perspectives on sexual and reproductive health, 236–243. staiger, d. o., & stock, j. h. 1994. instrumental variables regression with weak instruments. national bureau of economic research cambridge, mass., usa. strauss, j., & thomas, d. 1995. human resources: empirical modeling of household and family decisions. handbook of development economics, 3, 1883–2023. wang, m., & wong, m. c. s. 2011. fdi, education, and economic growth: quality matters. atlantic economic journal, 39(2), 103–115. wolde, m., berhan, y., & chala, a. 2015. determinants of underweight, stunting and wasting among schoolchildren. bmc public health, 15(1), 8. https://doi.org/10.1186/s12889-0141337-2 muhammad umar farooq, syed asif ali naqvi and shaoan huang 22 index table 7: summary statistics variables mean standard deviation child age 29.9 17.26 hfa -1.97 2.14 wfa -1.34 1.64 mbmi 23.64 4.86 mother age 29.37 6.15 husband age 35 9 mother education 0 0.54 17.26 1-5 0.15 2.14 6-10 0.19 1.64 11-16 0.1 4.86 urban 0.42 0.49 rural 0.57 0.49 health knowledge she did not heard about hepatitis 0.16 0.36 she heard about hepatitis 0.84 0.36 wealth index scores poor household 0.22 0.41 middle household 0.38 0.48 rich household 0.39 0.48 mother birth cohort 1963-72 0.07 0.25 1973-79 0.22 0.41 1980-88 0.48 0.5 1989-96 0.23 0.42 table 8: fraction of mothers who cannot read and can read a sentence by parts at different levels of schooling years of schooling 1 2 3 4 5 6 7 8 mothers who cannot read 87.5% 49% 31.2% 18.7% 10.7% 0% 2.3% 1.1% mothers who can read a sentence by parts 6.3% 18.4% 32.5% 38.7% 16.9% 16.7% 9.3% 3.9% journal of applied economics and business studies, volume. 4, issue 1 (2020) 143-167 https://doi.org/10.34260/jaebs.417 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x the role of monetary policy in transmission of asset prices into goods prices sidra mariyam1 & wasim shahid malik2 * 1 phd scholar, school of economics, quaid-i-azam university, islamabad. 2 associate professor, school of economics, quaid-i-azam university, islamabad and professor, sbp chair, university of peshawar abstract monetary policy in the contemporary world reacts, through short term interest rate, to deviations of inflation rate and output from their respective targets, while asset prices are responded to the extent they contribute to these deviations. this practice significantly affects transmission of asset prices into goods prices, which has serious implications for income distribution. this paper sets the objectives of estimating transmission of asset prices into goods prices and the role of monetary policy in influencing this transmission. in this regard, the paper hypothesizes that inflation rate positively responds to asset prices and this response weakens if interest rate leans against the winds of inflation, output and asset prices. to test these hypotheses, we have estimated different specifications of vector autoregressive (var) model and impulse response functions have been found after identifying structural shocks. data of pakistan’s economy on inflation rate, large scale manufacturing index, interest rate and asset price index – comprising house prices, stock prices and exchange rate – are used for the time period 2000m01 to 2019m06. we find evidence in support of both hypotheses; asset price inflation positively transmits into goods price inflation and this transmission intensifies if interest rate does not respond to other variables in the model. moreover, transmission of asset prices into inflation rate, as compared to output, is influenced more by monetary policy. finally, we find that the transmission of exchange rate and house prices to inflation rate are very much affected by monetary policy while in case of stock prices the influence of policy is moderate. keywords asset prices, inflation rate, structural vector autoregressive (svar) model jel classification e31, e44 1. introduction the quantitative strength of asset price channel in the transmission of monetary policy depends on how much policy affects asset prices and the contribution these prices make in shaping the target variables – inflation rate and output. three of the asset prices – stock prices, real estate prices and exchange rate – are most affected by changes in monetary * corresponding author: wsmalick@gmail.com 143 sidra mariyam & wasim shahid malik policy instrument and in turn these prices influence the paths of inflation rate and output. asset prices affect target variables via their interim effect on a number of variables such as, tobin’s q, cost of credit, net worth of firms and banks, probability of households being financially distressed, household wealth, and competitive position of a country in international trade (modigliani, 1971; bosworth, 1975, hayashi, 1982, bernanke and gertler, 1995; steindel, lettau and ludvigson, 2001; mishkin, 2001). the debate on whether or not central banks should take asset prices as part of their policy reaction function, when loss function is defined over deviation of inflation rate from its target and that of output from its potential level with the knowledge that asset prices affect these deviations, occupies significant place in the literature. for example, cecchetti, et al. (2000) warned against asset price bubbles, which when collapse are damaging to real economic activity, and argue for active role of monetary policy to contain bubbles from building. however, this is possible only if central bankers have informational advantage, asset prices are predominantly driven by monetary policy and central banks have mandate to do that (mishkin, 2001). most of the studies like goodhart and hofmann (1998), bernanke and gertler (1999), filardo (2001), issing (2003), and svensson (2011) argue for indirect response of monetary policy to asset prices; central banks react to changes in asset prices when these are reflected in goods price inflation (or inflation forecast), which is a target variable in policy reaction function. this recommendation is practiced by most of the central banks. existing literature considers asset prices as mediating variable in the transmission of monetary policy to target variables. however, variation in asset prices can be driven by many factors and monetary policy instrument is one of them. for instance, stock prices fluctuate in response to drivers of economic growth, all sort of shocks that change the investors’ willingness to take risk, and shocks that change income distribution between workers and shareholders (greenwald, lettau, and ludvigson, 2014). hence, it is needed to estimate the responses of goods prices and economic activity to shocks in asset prices, no matter what are the sources of asset price fluctuations. this is important because changes in monetary policy instrument interact with other sources of asset price fluctuations while affecting asset prices and in turn inflation and output. moreover, if there is change in asset prices, which may be because of a factor independent of monetary policy, then economic activity and inflation tend to change. this calls for a change in monetary policy instrument but recommendation of such intervention crucially depends on the knowledge of whether or not systematic monetary policy remains successful in weakening the effect of asset prices on inflation and output; this necessitates estimation of the moderating role of monetary 144 journal of applied economics and business studies, volume. 4, issue 1 (2020) 143-167 https://doi.org/10.34260/jaebs.417 policy in the transmission of asset prices to goods prices and economic activity 2 . furthermore, knowledge of moderating role of monetary policy is essential for future work in the area of monetary policy and income distribution. goods prices are reward to producers in the production process while the same are deflator of assets or wealth and wages of workers; thus increase in goods prices may increase producers’ profits but deflate assets, especially the ones that are unrelated to producers’ profits. in this paper, we state that changes in asset prices are transmitted into goods prices and monetary policy, the way it is practiced, can affect this transmission. when there is increase in asset prices then goods prices tend to increase via wealth channel, tobin’s q, or balancesheet channel. monetary policy then controls inflation through discouraging aggregate demand, damaging some economic activity. therefore, when increase in asset prices trigger aggregate demand through tobin’s q or wealth and balance sheet channels, central banks create a parallel opposing force through increasing interest rate. this opposing force makes transmission of asset prices to goods prices slow and incomplete. therefore, in a model, which allows reaction of monetary policy instrument to target variables, response of goods prices to asset prices is found weaker compared to the case when policy variable is super exogenous 3 . the objectives of this paper are twofold. firstly, the paper estimates transmission of asset prices into goods prices for pakistan’s economy. the existing literature of pakistan economy mostly focuses on finding whether or not stock prices provide hedge against goods price inflation 4 or estimating the pass-through effect of exchange rate into goods prices 5 ; the real estate sector is largely ignored 6 and there is dearth of research on transmission of stock and house prices into goods prices. the second objective is to estimate the extent to which monetary policy remains successful in diluting this transmission. the existing literature of pakistan economy estimates the mediating role of asset prices in the transmission of monetary policy into target variables 7 but does not consider the moderating role of policy in transmission of asset prices into goods prices. for these two objectives we use structural 2 shiratsuka (1999) states that japanese economy experienced large fluctuations in asset prices and economic activity, but inflation remained stable. this points to moderating role of monetary policy, though the study does not estimate such role. 3 super exogeneity entails absence of reaction of policy instrument to target variables at all lags, including zero. 4 (see, tiwari, et al. 2015; shahbaz, islam and rehman, 2016 and ali and wahid, 2019) 5 (see, choudhri, et al. 2005; hyder and shah, 2005; jaffri, 2010; younus and yucel, 2020) 6 shafiq and malik (2018) is the only exception. 7 see for instance, agha et al. (2005) 145 sidra mariyam & wasim shahid malik vector autoregressive (svar) 8 model for pakistan’s economy and estimate impulse response functions. we claim two contributions of our paper in the existing literature with regards to pakistan’s economy. first, we estimate the role of monetary policy in weakening the transmission of asset prices into goods prices. for this, two var models have been estimated; one in which monetary policy instrument does not respond to other variables in the model while in the second it does. we find significant role of monetary policy in fading the transmission. second, we take prices of three assets, namely stocks, housing, and foreign exchange 9 . we estimate the responses of inflation rate and economic activity to changes in these three asset prices, while previous literature focuses only on pass-through effect of exchange rate into inflation. it is noteworthy that scope of this paper is delimited to estimation of the transmission of asset prices to goods prices and economic activity and the moderating role of monetary policy in this transmission. the paper is not focused on identifying any specific channel through which this transmission occurs. in the same way, the paper does not estimate the effects of monetary policy on income distribution that occur due to moderating role of policy, though the paper offers an opportunity for future research to look into this issue. rest of the paper proceeds as follows: section 2 discusses some theoretical linkages from existing literature that support our empirical estimation; section 3 elaborates methodological issues, construction of variables and data sources; section 4 explains empirical findings; and section 5 concludes the paper. 2. theoretical framework theoretically, there are some channels through which changes in asset prices are transmitted into goods prices or the latter responds to changes in the former. in this regard, transmission channels of equity prices and house prices are different to those of exchange rate. equity and house prices affect goods prices in three important ways. first, equity and housing occupy significant share of consumers’ wealth. therefore, an increase in prices of these assets translates into higher wealth, which encourages aggregate demand to rise thereby causing an increase in goods prices. the quantitative strength of this wealth channel hinges on proportion of equity and housing in total wealth and consumers’ propensity to consume out of these two assets (see for instance, friedman, 1953; ando and modigliani, 1963; bover and muellbauer, 1989; girouard and blondal, 2001). second, increase in value of equity or housing that markets attach to these assets may cause this value to be higher 8 it must be mentioned at the outset that after estimating reduced form var we have used choleski decomposition to identify structural shocks and estimate impulse response functions 9 we follow shafiq and malik (2018) for constructing series on house price index and asset price index 146 journal of applied economics and business studies, volume. 4, issue 1 (2020) 143-167 https://doi.org/10.34260/jaebs.417 than replacement cost of capital or construction cost of housing. this makes attractive for firms to raise capital from market and invest in these assets. this tobin’s q channel is related to investment (tobin, 1969) as opposed to wealth channel that works through consumers’ expenditure. third, changes in asset prices may transmit into goods prices through credit channel. asymmetric information between borrowers and lenders lead to adverse selection and moral hazard problems and limit the credit expansion. increase in asset prices improves balance sheet of households/firms causing credit expansion through increase in the value of collateral. extended credit fuels into higher aggregate demand and in turn into higher goods prices (bernanke and gertler, 1995). exchange rate also affects goods prices in four different ways. first, higher exchange rate leads to lesser value of domestic currency as compared to foreign currency and therefore makes locally manufactured finished goods more attractive for both domestic consumers and foreigners while discourages imports. higher net exports then expand aggregate demand for locally manufactured finished goods and therefore result in higher domestic prices of goods and services (see for instance, dornbusch, 1976). second, higher exchange rate increases prices of imported raw material or intermediate products, which result in higher cost of production leading to lesser supply of locally manufactured finished goods if demand is elastic or in higher prices of finished goods if demand is inelastic. third, if imports and exports are less elastic, especially in the short run, then higher exchange rate results in higher value of imported goods thereby causing an increase in the cost of consumers’ basket. this induces labor to demand higher wages, which, if actually increase, results in higher cost of production (obstfeld and rogoff, 2001; engel, 2002). fourth, balance sheet channel, which is especially relevant for emerging economies, states that higher exchange rate leads to higher liabilities, denominated in foreign currency, of domestic firms and therefore decreases the net worth if assets are denominated in domestic currency. this limits borrowing capacity of domestic firms and discourages economic activity thereby putting lesser pressure on goods prices to increase. the transmission of changes in asset prices into goods prices can be significantly influenced by monetary policy. most of the literature in this field focuses only on whether or not monetary policy should target asset prices. for example, cecchetti, et al. (2000) suggest active role of monetary policy in containing asset prices as asset price bubbles, when bursts, may cause subsequent disruption in real economic activity 10 . this suggestion hinges on informational advantage of central banks, their mandate to contain asset prices, and significance of the role that monetary policy can play in determining asset prices (mishkin, 10 asset price bubbles, if they are left unattended, cause slowdown in economic activity when burst. this hypothesis gained much importance after global financial crisis, when minsky’s (1986) theory was revitalized. 147 sidra mariyam & wasim shahid malik 2001). kontonikas and ioannidis (2005) conduct stochastic simulations using inflation forecast targeting and standard taylor rule; they find that macroeconomic performance can be improved by making interest rate responsive to asset prices. goodhart and hofmann (1998), bernanke and gertler (1999), filardo (2001), issing (2003), and svensson (2011), on the other hand, advocate indirect response of monetary policy to asset prices; central banks react to changes in asset prices only when these are reflected in goods price inflation – a target variable in the policy reaction function. similarly, reinhart (1998) states that targeting asset prices by making it part of reaction function is not desirable as relationship between asset prices and inflation rate may not remain stable if monetary policy tries to exploit it. papademos and stark (2010) raise some practical concerns regarding leaning against the wind of asset price surges. it is difficult to differentiate between surges of asset prices that are driven by fundamentals and those that are speculative. however, monetary aggregates and credit expansion can be used as early warning alarms for asset price bubbles as historical data show significant correlation among money, credit and asset prices. in this paper, we focus on the moderating role of monetary policy on the transmission of asset prices into goods prices and economic activity. when central banks target asset prices then they can keep bubbles from building thereby saving economic activity to be hurt by bursting of asset price bubbles 11 . for instance, kent and lowe (1998) conclude that monetary policy should take contractionary measures in response to an emerging asset price bubble. this policy action aims at bursting the bubble before it becomes large and cause disturbance in the economic activity. they call this action an optimal response to asset price bubble even if it causes inflationary expectations to be lower than the target. on the other hand, when asset prices do not directly appear in the policy reaction function and central banks respond only to deviations of goods price inflation and economic activity from their respective targets then policy response to asset price bubble is delayed until it is reflected in inflation or at least in its forecast. this policy makes asset prices and economic activity much more volatile as compared to goods price inflation 12 . hence, monetary policy allows asset prices to increase in response to positive shock in asset market but does not allow inflation rate to gain momentum, thereby weakening the transmission of asset prices into goods prices. in this case, in a model with systematic response of monetary policy instrument to target variables, the effects of asset prices on goods prices are found weak. this provides rationale for estimating response of goods prices and economic activity to asset prices in a model in which policy instrument does not systematically respond to any 11 however, targeting asset prices may itself render economic activity unstable (bernanke, laubach, mishkin and posen, 1999). 12 shiratsuka (1999) mentions this regarding japanese economy in 1980s. 148 journal of applied economics and business studies, volume. 4, issue 1 (2020) 143-167 https://doi.org/10.34260/jaebs.417 of the target variables. theoretically, such response would be higher in magnitude as compared to the case in which monetary policy is actively used as state contingent rule. 3. econometric methodology in line with the objectives set in this paper, we need to identify structural shocks in asset prices that are orthogonal to other variables in the model. there are two methods to identify these shocks; finding residuals from single equation in which concerned variable is regressed on relevant variables along with their lags or finding structural errors from svar model using certain identifying restrictions. first approach requires valid instruments for endogenous variables to find consistent estimates of parameters, while the second approach needs identifying restrictions to be placed on structural parameters. while both methods make some arbitrary assumptions regarding valid instruments and sequencing of variables in a dynamic setup, the second method is mostly used in the literature due to its appealing features. var models treat all variables as endogenous and then sequencing of variables in dynamic setup can be imposed to identify structural shocks. there are other considerations that need to be discussed for the choice of appropriate model. for instance, a priori knowledge about exogeneity of the regressor of interest is helpful in identifying the response of dependent variable to that regressor in an autoregressive distributed lag (ardl) framework. however, if exogeneity is not known or the concerned regressor is contemporaneously responsive to other variables in the model, then svar model is the suitable option. another issue is regarding the order of integration. if scope of the study is to find level relationship in the long run and order of integration of variables is not known with certainty then ardl model provides suitable way to find cointegration among variables and estimate legitimate long run relationship. however, if interest is not in estimating values of parameters and there is uncertainty regarding exogeneity of variable of interest then svar is more relevant model to use. in this paper, the objective is to find structural shocks in asset prices and then estimate the dynamic response of inflation and output to these shocks. consistent with the literature, we assume asset prices to respond contemporaneously to other variables (interest rate and output) and therefore do not treat asset price index as weakly exogenous. moreover, we have three variables measured as percentage growth rates, which are less likely to be integrated of order greater than zero. therefore, we use svar model to satisfy objectives set in the paper. to estimate the effects of asset prices on inflation rate (and output), we have used structural var model containing four variables: bxt = b0 + ci xt −i + t ----------------------------(1) 149 sidra mariyam & wasim shahid malik here, b is a matrix of contemporaneous response coefficients, is a vector of endogenous variables, (inflation rate, output, asset prices and interest rate), c is a matrix of coefficients attached to lagged endogenous variables, and ɛ contains zero mean, constant variance and serially and contemporaneously uncorrelated structural shocks. the above system of equations can be converted into reduced form var as: xt = a0 +  ai xt + et ----------------------------(2) here, consists of one step ahead forecast errors with zero mean and constant variance. these errors are serially uncorrelated but may be contemporaneously correlated. our objective in this paper is to estimate the effects of shocks in asset prices on inflation rate (and output). for that we estimate the model in equation (2) and then use estimated parameters of these equations to identify structural parameters and recover structural shocks from the system of equations (1) by imposing appropriate restrictions on structural parameters. the moving average representation of var in equation (2) is given as: xt = b(l)et --------------------------------------------(3) b(l) denotes the matrix of coefficients polynomial in the lag operator l, defined as b(l) =∑∞ =0 ∗ . et can be written as linear combination of structural shocks as: = so vma can be written in terms of structural shocks as: xt = b(l)st ------------------------------------------(4) xt = (l)t -------------------------------------------(5) here, ∅( ) represents impulse response functions. 3.1 identifying restrictions though the forecast errors, et, are estimated, the structural shocks, ɛt, need to be recovered. for that, n(n-1)/2 restrictions are needed on s matrix to identify the system, where n is the number of variables in the var model. in our case, we need minimum six restrictions on structural parameters to identify var model. we allow sluggish response of monetary policy to all variables, which restricts three parameters to be zero. more specifically, interest rate is allowed to respond to all other variables in the var model only after lags and not contemporaneously. monetary policy is either forward-looking or backward-looking. in the case of former, monetary policy responds to expected changes in 150 journal of applied economics and business studies, volume. 4, issue 1 (2020) 143-167 https://doi.org/10.34260/jaebs.417 inflation rate and output. future forecasts of these variables depend on their own current and lagged values and on current and lagged values of other variables. in this case, interest rate contemporaneously respond to other variables in the model. however, in developing economies prompt response of monetary policy to future forecasts is lacking due to time lags involved in data availability, less than complete operational autonomy of the central banks and economies being hit by frequent shocks. this inertia in policy decisions prevails even if considerable weight is given to inflation forecast in monetary policy committee’s meetings. sluggish response of monetary policy to target variables is consistent with iacoviello (2005) and sims and zha (2006). inflation rate, on the other hand, responds contemporaneously to all other variables. this is somewhat unusual assumption if goods prices are sluggish to adjust. however, the way we constructed inflation rate (explained below), and the empirical finding that goods prices in pakistan are not so rigid (see, malik, satti and saghir, 2008; chaudhary et al, 2016, iqbal and amin, 2019) justify this identifying restriction. in between these two extremes are placed output and asset prices with the assumptions that output contemporaneously responds only to interest rate 13 while asset prices contemporaneously respond to interest rate and output but not to inflation rate 14 . though, asset prices can hedge savers against positive inflation in which case inflation rate should appear before asset prices but this channel has been found weak in pakistan (see, khan, 2004; nishat and shahieen, 2004, sohail and hussain, 2009, zafar and rafique, 2013, hussain, zaman and ahmad, 2015; qamri, haq and akram, 2015; saleem, et al., 2015; and muzammil et al. 2019). placing output before asset prices and inflation reflects inertia in economic activity and is consistent with the empirical literature (see for instance, christiano, eichenbaum and evans, 1999). economic shocks hitting the economy are first absorbed by asset prices and inflationary expectations and then decisions about investment and consumption follow. finally, placing asset prices after policy instrument is consistent with choudhri, jan and malik (2015). that study however places goods market variables before interest rate; therefore, for sensitivity analysis, we change identifying restrictions and place output before interest rate – results largely remain same. based on these assumptions, we can write the system of equations (4) with zero restrictions on coefficients as: 13 this restricts two structural parameters to be zero. 14 this restricts one structural parameter to be zero. 151 sidra mariyam & wasim shahid malik  it   yt    = api t     t   s11 0 0 0    i t  s 21 s 22 0 0    y t  ------------(6) b(l)     s 31 s 32 s 33 0   api t   s s s s     t   41 42 43 44      to estimate the effect of monetary policy on transmission of asset prices into goods prices, we modify the above baseline var model. interest rate – monetary policy instrument – is taken as exogenous variable in the var model by disallowing it to respond to any other variable, contemporaneously as well as with lags. however, interest rate appears in equations of all other variables. in this specification, interest rate does not react to any of the target variables and policy is passive. the gap between impulse response functions of this counterfactual model and that of baseline var model is a measure of the role monetary policy plays in diluting the transmission of asset prices into goods prices. a point of caution in the use of the term svar is in order here. as explained above, after estimating reduced form var we have used choleski decomposition to identify structural shocks and estimate impulse response functions and variance decomposition. some researchers use the term svar only if structural shocks are identified through simsbernanke or blanchard-quah methodologies. however, svar refers to a model in which dynamics of all variables are determined by their own lags and current and lagged values of other variables. moreover, error terms in all equations are structural shocks that are not correlated across equations. therefore, the term svar can be used if structural parameters and structural shocks are recovered from reduced form var using any of sims-bernanke, blanchard-quah or choleski identifying restrictions. 3.2 data and construction of variables we measure output by large scale manufacturing index (lsm), which is available at monthly frequency and portrays fluctuations in aggregate economic activity. in the empirical model, annualized growth rate of this variable is used. karachi interbank offered rate (kibor) is used for short term interest rate. in pakistan, money market rate is targeted by state bank of pakistan (sbp) through interest rate corridor system. this follows most of the studies in the literature, which use money market rate as proxy of policy variable (see taylor, 1993; thorbecke, 1997; svensson, 1999; bernanke and mihov, 1998; gali and gertler, 1999, among others). some studies (notably, romer and romer, 2004) use discount rate as policy variable but this proxy is not appropriate. in contemporary world, monetary policy committee sets an operating target for money market rate, which is contained by repo rate and reverse repo rate (discount rate). central banks use different policy tools to keep money market rate close to the operating target. discount rate signals stance of monetary policy and is upper bound for money market rate but central banks also use other tools, like 152 journal of applied economics and business studies, volume. 4, issue 1 (2020) 143-167 https://doi.org/10.34260/jaebs.417 open market operation, to keep money market rate on target. in this case, discount rate does not portray complete picture of monetary policy actions. moreover, in case of pakistan, monetary policy committee meets after every two months and most of the times discount rate is not changed 15 . this makes discount rate inappropriate measure of monetary policy at monthly frequency. for instance, in our sample period, discount rate remained constant for many long episodes – july 2005 to june 2006, july 2006 to january 2008, november 2011 to august 2012, and june 2016 to december 2017. we define inflation rate as year-on-year percentage change in non-house-rent consumer price index (cpi) 16 . we then adjust this inflation rate by a common factor measured as factor score using eight macroeconomic variables 17 . the inflation rate, which we use for our analysis, is the residual estimated from a regression of inflation rate, measured from non-house-rent cpi basket, on this factor score. this residual series carries all information regarding variation in goods price but the effect of large shocks that change the paths of macroeconomic variables, especially inflation rate. we follow shafiq and malik (2018) for construction of asset price index from equity/stock prices, house prices and exchange rate, and simulating house price index using house rent index. data on house price index are available from january 2011 onwards on www.zameen.com website. following shafiq and malik (2018) we estimate two variables var model comprising house price index and house rent index for the period 2011m01 to 2019m06 and then data on house price index for the period 2000m01 to 2010m12 are backcasted. asset price index and individual asset prices are also used in terms of annualized percentage growth rates in empirical model. we have used monthly data on output, inflation rate, asset prices and interest rate for the period 2000m01 to 2019m06. the data on exchange rate and stock prices are taken from international financial statistics (ifs). the data on cpi, lsm, and kibor are taken from monthly bulletin of statistics published by sbp. the data on house rent index is taken from monthly price indices published by pakistan bureau of statistic (pbs), which is then used to construct house price index. 4. empirical results the main objective of this paper is to estimate the response of goods prices to changes in asset prices. for that we construct four variables var model and select lags using likelihood ratio statistics. we check the presence of autocorrelation in the errors of each 15 monetary policy committee sets operating target for money market rate. whenever, this target is changed, discount rate is also set accordingly. 16 house rent is excluded from cpi basket because our model contains house price index, which is constructed using house rent. 17 these variables include large scale manufacturing index, interest rate, house price index, stock prices, exchange rate, money supply, trade deficit, and budget deficit. 153 sidra mariyam & wasim shahid malik model and proceed further only if errors are found serially uncorrelated. results in table a1 (in appendix) show that errors in all of our estimated models are free from autocorrelation. as all variables, except interest rate, are measured as annualized growth rate so we use them in level form 18 . interest rate is non-stationary at level 19 so there is a choice between using this variable in level form or in first differenced form. sims (1980) and sims, stock, and watson (1990) warn against differencing in var model if variables are non-stationary. therefore, we use interest rate in level form. however, to check robustness of results we also estimate var model with first differenced interest rate and give results of both specifications for comparison. moreover, for robustness, identifying restrictions are changed by placing output before interest rate. our results are robust to three specifications as in all cases we find same responses of inflation rate and output to asset prices (figure 1). 18 growth rate of a variable is stationary if the variable is non-stationary. 19 adf test statistic for null of unit root at level is -2.94 with probability 0.16 and -17.64 with probability < 0. 01 for the null that first differenced interest rate is unit root process. 154 journal of applied economics and business studies, volume. 4, issue 2 (2020) 143-167 https://doi.org/10.34260/jaebs.417 response of inflation rate to asset prices 4 2 0 -2 -4 -6 -8 2 4 6 8 10 12 14 16 18 20 22 24 response of inflation rate to asset prices 4 2 0 -2 -4 -6 -8 2 4 6 8 10 12 14 16 18 20 22 24 response of inflation rate to asset prices 4 2 0 -2 -4 -6 -8 2 4 6 8 10 12 14 16 18 20 22 24 response of output to asset prices 12 8 4 0 -4 2 4 6 8 10 12 14 16 18 20 22 24 response of output to asset prices 10 5 0 -5 -10 2 4 6 8 10 12 14 16 18 20 22 24 response of output to asset prices 12 8 4 0 -4 2 4 6 8 10 12 14 16 18 20 22 24 figure 1: accumulated impulse response functions (± 2 s.e) (source: authors’ estimations) note: dotted lines represent ±2 standard errors band. results in upper and lower horizontal panels are from a var model with interest rate at level while the middle horizontal panel shows result of var model with interest rate in first differenced form. choleski ordering: interest rate, output, asset prices and inflation rate in upper and middle panels, while output, interest rate, asset prices and inflation rate in lower panel. 155 sidra mariyam & wasim shahid malik the response of inflation rate to asset prices is found positive and significant. this positive response coexists with the positive effect of asset prices on output. this finding makes it clear that higher asset prices expand business activity via wealth channel, tobin’s q channel or balance sheet channel which in turn puts upward pressure on goods prices because of excess demand. our results are in conformity with all theoretical channels explained in section 2, which predict positive response of inflation to asset prices, except the balance sheet channel of exchange rate. according to that channel exchange rate depreciation worsens balance sheet of non-financial firms of emerging economies as their assets are denominated in domestic currency but liabilities are denominated in foreign currency. this limits their capacity to borrow and credit constrained firms are unable to expand business activity, which results in low inflation. all other channels predict expansion in economic activity led by increased asset prices which in turn results in high inflation. our results are such as if the balance sheet channel of exchange rate is dominated by other channels in pakistan; this is likely because pakistani firms rely more on domestic credit market and their exposure to exchange rate fluctuations is low (ihsan, rashid and naz, 2018) 20 . we also find that interest rate does not respond to asset prices (figure a1 in appendix). this result is not surprising as most of the central banks do not directly respond to asset prices. monetary policy statements of pakistan also show that policy decisions are predominantly driven by state of economic activity, inflation and external account position. asset prices are responded only to the extent they influence paths of inflation and economic activity. though the responses of inflation and output to changes in asset prices are in conformity with the economic theory, the confidence intervals of impulse responses are wide. this is a problem researchers more often face while estimating impulse response functions. the dilemma lies in the fact that lesser lags make errors of the var model auto-correlated while long lags make var model over-parameterized and estimation errors attached with more parameters result in wide confidence intervals of impulse response functions. our results in figure 1 show that response of inflation to asset price shock is statistically significant only for four months and then it becomes insignificant. this may be of concern if policy implications are to be drawn from this empirical finding. however, two points make our result relevant for policy even with statistical significance for short period of time. first, inflation rate has strong inertia in it and most of the variation in this variable is explained by its own lags (variance decomposition shows that about 70 percent of the variation in inflation rate at 24 months horizon is explained by its own lags) 21 . in this case, actual effect of asset prices on inflation rate is long lasting despite an apparent short lived direct effect. 20 in that paper authors show that domestic firms have less exchange rate exposure as compared to multinational firms. 21 results can be shared on request from the authors 156 journal of applied economics and business studies, volume. 4, issue 1 (2020) 143-167 https://doi.org/10.34260/jaebs.417 second, the short lived effect of asset prices on inflation is found in baseline var model in which monetary policy moderates this effect. as shown below, the time period of significant effect increases when interest rate is not allowed to respond to other variables in the model (figure a2 in appendix). monetary policy has a role in determining inflation rate and asset prices. for instance, central banks can directly put asset prices in their reaction function and react to any undesirable movement in this variable. this action of central banks keep asset prices in check and associated inflation rate is also controlled. however, asset prices are more often treated as intermediate targets in which case they cannot be made part of policy reaction function. but there is ample evidence in the literature to augment policy reaction function by other variables like exchange rate and lagged interest rate (clardia, gali and gertler, 1998; amato and laubach, 1999; taylor, 2001; reudebusch, 2002; saghir and malik, 2017). asset prices too have justification to be treated as ultimate target and not only as intermediate target. for instance, biasedness in inflation rate as a measure of change in overall cost of living (alchian and klein, 1973) makes asset prices suitable target of monetary policy. these arguments notwithstanding, most of the central banks target only inflation rate and do not intervene in the asset market. in this case, central banks respond to asset prices only indirectly; monetary policy instrument is changed whenever asset prices make inflation forecast deviate from the target. we hypothesize in this paper that if central banks directly respond only to inflation rate (along with output) and not to asset prices then transmission of asset prices to inflation rate is diluted. in the absence of active monetary policy, response of inflation to asset prices is higher compared to the case when central banks target variables like inflation, output or asset prices. we test this hypothesis by assuming a counterfactual scenario in which interest rate does not change in response to any of the other three variables in the var model. more specifically, interest rate is taken as exogenous variable in which case it cannot change in response to any other variable in the var model at any lag. however, interest rate appears in equations of other three variables. in this setting policy is not state contingent but it has the power to change the economic state. we find support for our hypothesis as response of inflation rate to asset prices significantly changes in this counterfactual scenario (figure 2). in the absence of active monetary policy, transmission of asset prices to goods prices is higher compared to what is found in case when policy does respond to other variables in the var model. in quantitative terms, accumulated response of inflation to asset prices in baseline scenario is 83% of that in counterfactual scenario after one lag. the difference increases with higher lags as response in baseline scenario is 58% of that in counterfactual scenario after 6 lags; 44% after 12 lags; and just 5% after 18 lags. same result is found for the case of output but the difference between two impulse response functions is smaller 157 sidra mariyam & wasim shahid malik compared to what has been found for inflation; response of output in baseline var model is approximately 90% (at all lags) of that in counterfactual model. figure 2: moderating role of monetary policy (the case of asset price index) (source: authors’ estimations) we also estimate the effect of monetary policy on transmission of asset prices into inflation rate for individual assets. this analysis makes sense as different assets may have different transmission channels. we have used same methodology for individual asset prices as that in the case of asset price index. more specifically, we have estimated three var models with asset price index replaced by one of the asset prices 22 . lag length for each var model is selected using likelihood ratio statistics which is allowed to be different in different models and dummy variables are included in the model to capture structural breaks. we find some interesting results. first, for all of the asset prices the transmission to inflation rate is higher in counterfactual scenario compared to that in the baseline scenario. this shows robustness of result found in case of asset price index that monetary policy lowers the transmission of asset prices into goods prices. moreover, statistical significance of impulse response functions is also improved. for instance, in the baseline model, response of inflation rate to asset prices is statistically significant only for four months but in the counterfactual model the response becomes insignificant only after 10 months (see figure a2 in appendix). second, the highest gap between responses in the baseline and counterfactual scenarios is found for exchange rate followed by house prices while in case of stock prices 23 the 22 exchange rate depreciated sharply after a period of overvaluation twice in our sample; in 2008-09 and 2018 19. we adjusted exchange rate for these abnormal fluctuations using dummy variables. similarly, asset prices sharply declined in the year 2009 and remained volatile till the end of year 2010. dummy variable for this period is also used for this period in var model with stock prices. we test for structural breaks in the paths of structural shocks identified from svar model. we find no evidence of such breaks as shown in table a2 in the appendix, which implies that the effects of asset price shocks on output and inflation are not biased subject to structural breaks in the data. 23 we find positive response of inflation to shocks in asset prices, while mukhtar and younas (2019) find negative response. the reason behind this difference of results is the way we measure inflation rate in our 158 journal of applied economics and business studies, volume. 4, issue 1 (2020) 143-167 https://doi.org/10.34260/jaebs.417 difference is found low. in quantitative terms, after 12 lags, accumulated response of inflation to asset prices in baseline scenario is 10% of that in counterfactual scenario for exchange rate, 13% for house prices and 67% for stock prices. in pakistan, exchange rate plays major role in monetary policy decisions. interest rate is kept high whenever there is upward pressure on exchange rate. this is evident from recent contractionary stance of monetary policy in the middle of economic crisis 24 . in this case, higher exchange rate does not fully transmit into goods prices because of strict reaction of monetary policy to exchange rate depreciation and associated expected hike in inflation rate. in case of stock and housing markets sbp does not directly respond to asset price hike 25 . but there are certain reasons behind why gap is more prominent in house prices 26 . housing market has close linkages with other economic activities. therefore, expansion in housing market activities can potentially expand aggregate demand thereby putting upward pressure on goods prices. moreover, house prices are closely linked with house rent, which carries significant weight in cpi; increase in house prices are quickly reflected in cpi. finally, housing market activities are broad based and high in market value. all these characteristics make central banks respond more quickly to rising house prices compared to stock prices. this response may not be direct however, as most of the times central banks respond to asset prices indirectly through their response to inflation rate and economic activity. third, response of inflation rate to exchange rate is negative in baseline model but positive in counterfactual model. the result in baseline model looks counterintuitive as historically inflation rate in pakistan always soared after massive currency depreciation. however, there are certain reasons that justify such a result. we adjusted inflation rate for the effects of big shocks measured by factor score using macroeconomic variables. therefore, our inflation rate does not carry the effects of big shocks. three such shocks in our sample are massive capital inflow after 9/11, oil price hike in 2008, and foreign exchange crisis starting in fiscal year 2019. in all of these three episodes inflation rate gained momentum but our inflation series does not contain the unusual change of path of inflation rate. moreover, we adjusted exchange rate for massive depreciation during these time periods, so our results are not applicable for abnormal times. finally, above two issues coupled with the strict response of monetary policy to exchange rate depreciation make response of inflation rate to exchange rate negative in baseline model. chaudhry et al (2011) study. if inflation rate is not adjusted for common component of macroeconomic variables then it is negatively correlated with stock prices. 24 at the start of fiscal year 2018-19 economic activity was slowing down and exchange rate started depreciating after a long period of overvaluation. inflation rate was not that high but it was expected to gain momentum because of further currency depreciation. in this scenario sbp took tight stance of monetary policy despite criticism from academia, businesses and general public. this shows sbp’s preference to stabilize exchange rate. 25 this is evident from monetary policy statements. 26 umer et al (2019) find positive effect of inflation on house prices. though our objective is to estimate the effect of house prices on inflation, our results are consistent with umer et al (2019) in correlational context. 159 sidra mariyam & wasim shahid malik also find negative relationship between exchange rate and inflation, while shaheen (2013) finds positive association. figure 3: moderating role of monetary policy (the case of individual asset prices) (source: authors’ estimations) 160 journal of applied economics and business studies, volume. 4, issue 1 (2020) 143-167 https://doi.org/10.34260/jaebs.417 5. conclusion this paper estimates the transmission of asset prices into inflation rate and the role of monetary policy in weakening this transmission. the paper hypothesizes that inflation rate positively responds to asset prices and this response is influenced by the way monetary policy is conducted in contemporary world. to test these hypotheses different specifications of four variables var model have been estimated and impulse response functions are found. data of pakistan’s economy on non-house-rent cpi, large scale manufacturing index, asset prices and kibor are used for the time period 2000m01 to 2019m06. asset price index includes stocks prices, house price index and exchange rate. var models have been estimated for asset price index as well as for individual asset prices. we find evidence in support of both hypotheses in that asset price inflation positively transmits into goods price inflation and this transmission intensifies if interest rate does not respond to other variables in the model. moreover, transmission of asset prices to inflation rate is influenced more, as compared to output, by monetary policy. finally, we find that the transmission of exchange rate and house prices to inflation rate is affected significantly by monetary policy while in case of stock prices the influence of policy has been found moderate. according to these results, the way monetary policy is currently practiced has important implications for income distribution. if monetary policy does not respond to asset prices but remains strict on deviations of inflation and output from their respective targets then growth rates of assets’ value and nominal gdp will be different. more specifically, central banks respond to inflation and output if they are not on target; hence, asset prices are responded to the extent they affect output or inflation rate. it means asset prices are responded when they are already changed and are reflected in inflation forecast. in this process economic activity in real sector is affected through changing interest rate. hence, real sector bears the cost of contractionary monetary policy, in case of high inflation, that has been necessitated by activities in the asset markets. if this process continues then growth rate of gdp will be lesser than that of assets’ value and income distribution will be skewed towards asset markets. this discussion entails monetary policy to differentiate between inflation caused by asset market activities and the one caused by goods market activities. goods market activities should not be discouraged when inflation is induced by speculative activities in the asset markets. references agha, a.i., ahmed, n., mubarik, y.a., & shah, h. (2005). transmission mechanism of monetary policy in pakistan. sbpresearch bulletin, 1(1), ali, g., & wahid, f. (2019). does stock returns act as a hedge against inflation in pakistan? an empirical analysis. journal of research in social science, 7(1), 2305-6533. 161 sidra mariyam & wasim shahid malik alchian, a., & klein, b. (1973). on a correct measure of inflation. journal of money, credit and banking, 5(1), 173-91. amato, j.d., & laubach, t. (1999). the value of interest-rate smoothing: how the private sector helps the federal reserve. economic review, 84(3), 47-64 bernanke, b., & gertler, m. (1995). inside the black box: the credit channel of monetary policy. the journal of conomic perspective, 9(4), 27-48. bernanke, b., laubach, t. mishkin, f.s & pseon, a.s. (1999). inflation targeting: lessons from the international experience, princeton, nj: princeton university press. bernanke, b. & gertler, m. (1999). monetary policy and asset price volatility, in new challenges for monetary policy. jackson hole conference, federal reserve bank of kansas city. bernanke, b. & mihov, i. (1998). measuring monetary policy. the quarterly journal of economics, 113(3), 869-902. bosworth, b. (1975). the stock market and the economy. brookings paper on economic activity, 6(2), 257-300. bover, o., & muellbauer, j. (1989). housing, wages and uk labor markets. oxford bulletin of economics and statistics, 51(2), 96-136. cecchetti, s., genberg, h. lipsky, j. & wadhwani, s. (2000). asset prices and central bank policy. the geneva reports on the world economy no.2 icmb. chaudhry, i.s., akhtar, m.h., mahmood, k., and faridi, m.z. (2011). foreign exchange reserves and inflation in pakistan: evidence from ardl modelling approach. international journal of economics and finance, 3 choudhri, e.u., faruqee, h., & hakura, d.s. (2005). explaining the exchange rate pass-through in different prices. journal of international economics, 65(2), 349-374. choudhary, a.m., faheem, a., hanif, m.nadim., naeem, s., & pasha, f. (2016). price setting and price stickiness. a developing economy perspective. journal of macroeconomics, 48(c), 44-61. choudhary, e.u., jan, a., and malik, h. (2015). monetary policy in pakistan: effectiveness in inflation control and stabilization. international growth centre, working paper, no s37204-pak-1 christiano, l., eichenbaum, m., and evans, c. (1999). monetary policy shocks: what have we learned and to what end?. handbook of macroeconomics, in john taylor and michael woodford, 1, chapter 2, 65-148. clarida, r., gali, j., & gertler, m. (1998). monetary policy rules in practice: some international evidence. european economic review, 42(6), 1033-1067. dornbusch, r. (1976). expectations and exchange rate dynamics. journal of political economy, 84(6), 1161-1176. engel, c. (2002). the responsiveness of consumer prices to exchange rates: a synthesis of some new open economy macro models. the mancester school, 70(1), 1-15. filardo, a.j. (2001). should monetary policy respond to asset price bubbles? some experimental results, fbr of kansas city, working paper, no. 04. friedman, m. (1953). the methodology of positive economics, in essays in positive economics, chicago university press, 47-99. gali, j., & gertler, m. (1999). inflation dynamics: a structural econometric analysis. journal of monetary economics, 44(2), 195-222. 162 journal of applied economics and business studies, volume. 4, issue 1 (2020) 143-167 https://doi.org/10.34260/jaebs.417 girouard, n., & blondal, s. (2001). house prices and economic activity. oecd economic department, working paper, 279. goodhart, c., & hofmann, b. (1998). asset prices and the conduct of monetary policy. royal economic society annual conference, royal economic society. greenwald, d.l., lettau, m., and ludvigson, s.c. (2014). origins of stock market fluctuations. nber, working paper, no, 19818. hayashi, f. (1982). tobin’s marginal q and average q: a neoclassical interpretation. the econometric society, 50(1), 213-224. hussain, m., zaman, b., & ahmad, n. (2015). relationship between stock market volatility and macroeconomic variables: evidence from pakistan. pakistan business review, 16(4), 723743. hyder, z., & shah, s. (2005). exchange rate pass-through to domestic prices in pakistan. state bank of pakistan, working paper, no. 05 iacoviello, m. (2005). house prices, borrowing constraints, and monetary policy in the business cycle. the american economic review, 95(3), 739-764. ihsan, h., rashid, a. & naz, a. (2018). exchange rate exposure and firms value: an assessment of domestic versus multinational firms. lahore journal of economics, 23(1), 51-77. issing, o. (2003). monetary stability, financial stability and the business cycle. bank for international settlements, basle. jaffri, a.a. (2010). exchange rate pass-through to consumer prices in pakistan: does misalignment matter?, the pakistan development review, 49(1), 19-35. kent, c., and lowe, p. (1998). property price cycles and monetary policy. in the role of asset prices in the formation of monetary policy. basle, bank for international settlements, 239263. khan, k.n. (2004). inflation and stock market performance: a case study for pakistan. giordano dell-amore foundation, 28(1), 87-101. kontonikas, a., & ioannidis, c. (2005). should monetary policy respond to asset price misalignments?. economic modelling, 22(6), 1105-1121. mahmood, i., nazir, f., & junaid, m. (2014). stock prices and inflation: a case study of pakistan. journal of asian business strategy, 4(12), 217-223. malik w s, satti a h, saghir g (2008): price setting behavior of pakistani firms: evidence from four industrial cities of punjab. . the pakistan development review, 47(3): 247-266. minsky, h.p., (1986). stabilizing an unstable economy. yale university press, new havan and london (2008). modigliani, f. (1971). monetary policy and consumption in consumer spending and monetary policy: the linkages. federal reserve bank of boston, 9-84. modigliani, f., & ando, a. (1963). the “life cycle” hypothesis of saving: aggregate implications and tests. the american economic review, 53(1), 55-84. mishkin, f.s. (2001). the transmission mechanism and the role of asset prices in monetary policy. nber working paper series, no. 8617. mukhtar, t., & younas, m.z. (2019). monetary policy transmission mechanism of pakistan: evidence from bank lending and asset price channels. asian journal of economic modelling, 7(3). 121-139. muzammil, m., zaman, s., asim, a., & haider, s. (2019). the impact of inflation on pakistan stock exchange. european journal of business and managemnet, 11(16), 81-92. 163 sidra mariyam & wasim shahid malik nishat, m., & shaheen, r. (2004). macroeconomic factors and pakistani equity market. pakistan development review, 43(4), 619-637. obstfeld, m., & rogoff, k. (2001). the six major puzzles in international macroeconomics: is there a common cause? nber, working paper, no. 7777. papademos, l.d. and stark, j. (2010), enhancing monetary analysis, frankfurt am main: european central bank. qamri, g.m., haq, m.a., & akram, f. (2015). the impact on stock prices: evidence from pakistan. microeconomics and macroeconomics, 3(4), 83-88 reinhart, v. (1998). equity prices and monetary policy in the united states. in the role of asset prices in the formation of monetary policy. basle, bank for international settlements, 280300. romer, c.d., & romer, d.h. (2004). a new measure of monetary shocks: derivation and implications. american economic review, 94(4), 1055-1084. rudebusch, g.d. (2002). term structure evidence on interest rate smoothing and monetary policy inertia. journal of monetary economics, 49(6), 1161-1187. saleem, f., zafar, l., & rafique, b. (2013). long run relationship between inflation and stock return: evidence from pakistan. social science and humanities, 4(2), 407-415. sims, c.a. (1980a). macroeconomics and reality. econometrica, 48(1), 1-48. sims, c.a., stock, j.h. & watson, m.w. (1990). inference in linear time series models with some unit roots. econometrica, 58(1), 113-144. sims, c.a. and zha, t. (2006). were there regime switches in u.s. monetary policy? american economic review, 96, 1, 54-81. saghir, m., & malik, w.s. (2017). estimating monetary policy reaction function of state bank of pakistan. pakistan economic and social review, 55(1), 119-157. shaheen, f. (2013). fluctuations in exchange rate and its impact on macroeconomic performance of pakistan. the dialogue, 8(4), 141-418 shahbaz, m., & islam, f. & rehman, i. (2016). stock as hedge against inflation in pakistan: evidence from ardl approach. global business review, vol, 17(6), 1280-1295. shafiq, h., & malik, w.s. (2018). money demand function revisited: the role of asset prices in pakistan. pakistan economic and social review, 56(1), 67-92. shiratsuka, s. (1999). asset price fluctuation and price indices. monetary and economic studies, federal reserve bank of chicago, working paper, no, 99. sohail, n., & hussain, z. (2009). long-run and short-run relationship between macroeconomic variables and stock prices in pakistan. pakistan economic and social review, 47(2), 183198. svensson, l.e.o. (1999). inflation targeting as a monetary policy rule. journal of monetary economics, 43(3), 607-654. svensson, l.e.o. (2011). monetary policy after the crisis, speech, federal reserve bank of san francisco. steindel, c., lettau, m. & ludvigson, s.c. (2005). monetary policy transmission through the consumption –wealth channel. economic policy reveiew, 8(1), 117-133. taylor, j.b. (2001). the role of the exchange rate in monetary policy rules. the american economic review, 91(2), 263-267. thorbecke, w. (1997). on stock market returns and monetary policy. the journal of finance, 52(2), 635-654. 164 journal of applied economics and business studies, volume. 4, issue 1 (2020) 143-167 https://doi.org/10.34260/jaebs.417 tobin, j. (1969). a general equilibrium approach to monetary theory, journal of money, credit and banking, 1(1), 15-29. tiwari, a., dar, a, bhanja, n. & arouri, m. (2015). stock returns and inflation in pakistan. economic modelling, 47(c), 23-31. umer, m., akhtar, m., shafiq, m., & rao, z.r. (2019). impact of monetary policy on house prices: case of pakistan. international journal of housing markets and analysis, younus, r.a., & yucel, e. (2020). exchange rate pass through in pakistan. mpra paper, 98955. 165 sidra mariyam & wasim shahid malik appendix response of interest rate to asset prices response of output to asset prices 4 2 0 -2 -4 -6 2 4 6 8 10 12 14 16 18 20 22 24 response of asset prices to asset prices 12 8 4 0 -4 2 4 6 8 10 12 14 16 18 20 22 24 response of inflation rate to asset prices 60 50 40 30 20 10 2 4 6 8 10 12 14 16 18 20 22 24 4 2 0 -2 -4 -6 -8 2 4 6 8 10 12 14 16 18 20 22 24 figure a1: accumulated impulse response functions (± 2 s.e): baseline model (source: authors’ estimations) response of output to asset prices response of asset prices to asset prices response of inflation rate to asset prices 10 8 6 4 2 0 2 4 6 8 10 12 14 16 18 20 22 24 60 50 40 30 20 10 2 4 6 8 10 12 14 16 18 20 22 24 4 0 -4 2 4 6 8 10 12 14 16 18 20 22 24 figure a2: accumulated impulse response functions (± 2 s.e): counterfactual model (source: authors’ estimations) 166 journal of applied economics and business studies, volume. 4, issue 1 (2020) 143-167 https://doi.org/10.34260/jaebs.417 table a1: autocorrelation test in var models max lags lre stats probability model 1 14 18.02 0.32 model 2 14 10.47 0.84 model 3 14 18.02 0.32 model 4 14 12.53 0.19 model 5 14 28.77 0.05 model 6 14 11.85 0.22 model 7 14 20.82 0.19 model 8 14 15.44 0.08 model 9 14 22.02 0.14 model 10 14 16.35 0.06 source: authors’ estimations note: model 1 corresponds to impulse response functions (irf) in upper panel of figure 1, model 2 to middle panel and model 3 to lower panel. model 4 corresponds to counterfactual irf in figure 2. model 5 (6) corresponds to baseline (counterfactual) irf in upper panel, model 7 (8) in middle panel and model 9 (10) in lower panel of figure 3. table a2: testing structural breaks in structural shocks variable break test f-statistic scaled f-statistic critical value kibor 0 vs. 1 5.88 5.88 8.58 lsm 0 vs. 1 1.94 1.94 8.58 ap 0 vs. 1 1.80 1.80 8.58 inflation 0 vs. 1 3.27 3.27 8.58 source: authors’ estimations 167 journal of applied economics and business studies, volume. 4, issue 4 (2020) 187-214 https://doi.org/10.34260/jaebs.449 187 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x contribution of spatial concentration in exaggerating inequalities: across various urban regions of pakistan uzma tabassum1*, shaista alam2 and ambreen fatima3 1 phd scholar and research assistant, applied economics research centre (aerc), university of karachi 2 senior research economist, applied economics research centre (aerc), university of karachi 3senior research economist, applied economics research centre (aerc), university of karachi abstract industrial agglomeration and inequalities are much of the concerns in recent literature. pakistan, being a developing country, is also restricted by resource availability to treat all regions equally with respect to investment and development. as a result, regions with growing agglomeration experience higher income levels relative to other regions. to investigate this empirically this study employed propensity score matching (psm) across urban regions in pakistan using labour force survey data 2017-18. for agglomeration regional herfindahl indices were estimated and regions with above average index value along with having positive index growth were considered as treated or agglomerated regions. the positive and significant coefficient of regions with treatment signifies that regions with agglomeration were found to have higher relatively income. hence introducing industrial concentration in untreated regions would be effective in reducing inequalities rather tackling them by reducing agglomeration in agglomerated regions. keywords inequality, agglomeration, propensity score matching, urban regions, herfindahl index. jel classification c31; d30; d3; o14; o15; 1. introduction over the last few epochs world trends draw attention to two apparent traits in economic growth. one is rising inequalities and the other is increasing geographical concentration of economic activities across regions. various strands of the theory suggest that openness of region and removal of barriers in free mobility of factor will probably result in a higher level of concentration of industries and as a result greater specialization of the region. this might have repercussions in growing income disparities across regions. such growth is also observable in pakistan where unbalanced * uzma_tbsm@hotmail.com the work is the part of phd thesis under process. uzma tabassum, shaista alam and ambreen fatima 188 growth strategy is pervaded because of resource constraints (tabassum, 2016).therefore investments and economic activities concentrate in some regions of the country that amplifies inequalities across regions, as the exogenous characteristics of regions, skill composition of the workforce and agglomeration economies attributed to productivity differences that in turn assumed to be reflected in income distribution. given that the distributions of industries are not uniform across regions in an effort to specifically elicit the effect of agglomeration on spatial disparities it is important to know that how the concentration of industries affect the growth process of a particular region. what is the role of urbanization and localization economies in provoking spatial disparities? how these spatial disparities are connected with inequalities in the income distribution of individuals living in that region? the various regions would be different in sizes with respect to their physical and economic structure, depending on the agglomeration behaviour. the swiftness and assortment of industrial agglomeration in a region describe the economic conditions of that region. on the theoretical side, the concept of agglomeration economies define the process of clustering of firms in close proximity to one another to exploit the benefits of being together within intension to minimize per-unit cost of production or achieved economies of scale (un-habitat, 2011). the concept was first acknowledged by weber (1909) while the detailed explanation of the sources of these economies was given by marshall (1890). marshall (1890) provided three bases through which economies of scale can be achieved. according to him agglomeration enables firms to economize on their unit cost by sharing knowledge regarding efficient production technologies and management strategies. further firms may agglomerate to share supply of intermediate inputs. the demand for intermediate input is greater when there are many firms located at the same place the average cost of providing such intermediate inputs to each firm will be low as high demand enable the producer of intermediate input to achieve scale economies. lastly sharing labour pool benefit firm when the variation in firm product demand is greater than the variation in industry level demand. the labour fired from the firm that is facing low product demand can be hired by a firm that is facing high product demand. for doing so firm have to bear no cost for searching and training labour thus declining unit cost for two reasons: one by firing unwanted labours by a firm facing low product demand. and secondly by reducing search and training cost by high product demand firm. duranton & puga (2004) also describe all these as a process of learning, sharing and matching. in a nutshell productivity and income growth of region positively affect by agglomeration but these positive effects may come from urbanization economies or localization economies. in localization economies firms that belong to a single industry cluster together to exploit agglomeration benefits thus they specialize in that particular industry production. while in urbanization economies firms that belong to different industries cluster together to exploit agglomeration benefits that results diversified journal of applied economics and business studies, volume. 4, issue 4 (2020) 187-214 https://doi.org/10.34260/jaebs.449 189 production. in viewpoint of efficiency spatial inequalities may have positive and negative effects. inequalities resulting from regional specialization based on localization economies may be beneficial as it increases productivity but if these are caused by external economies or urbanization economies then they may not be optimal (kim, 2008). spatial inequalities caused by unrestrained population concentration in large primate cities may inflict many social ills in society. thus spatial inequalities possibly are undesirable for society if it provoke social inequalities across the region. this research seeks to shed some light on the process of agglomeration and its effect on the income of region that experiencing these agglomerations. the literature focusing spatial disparities between urban regions based on agglomeration are rather scarce in the context of pakistan. a recent set of empirical literature both nationally and internationally majorly investigated the determinants of income inequalities. these studies point out socio-economic, political and open economy variables to contribute in influencing inequalities. disaggregating these determinants includes individual and region specific variables such as education, health, gender etc. for the former and industrial clustering, policies and public service provision, agglomeration patterns etc. for the later (sial et al, 2018., fambon, 2017., naseer & ahmed, 2016., burki & khan, 2012., davtyan; 2014., jamal, 2006). extensive research is already available domestically on the aspects explaining prevalence of regional inequalities but it largely focuses on socio-economic determinants of inequalities while the impact of location specific industrial clustering on embellishing inequalities has mainly overlooked. this research will evaluate the contribution of spatial concentration in inflating inequalities across various regions by exploring the likelihood of prosperous region become more prosperous, as heterogeneity across industrial expansion and in labour demand frames inequalities among individuals. the empirical assessment is based on survey data conducted by the pakistan bureau of statistics (pbs). in the study large cities and other urban areas have taken as spatial units. because some urban regions were not included in past years survey so they left intentionally from recent sets of survey to avoid inconsistency in data overtime. for estimating the agglomeration impact, this study used the labour force survey (lfs) for the period of 2017-18. further to identify the urban regions that are experiencing persistent inter-temporal spatial concentration in the case of pakistan, the herfindahl indices are computed for the years 2010 and 2018. the regions that have above average level of agglomeration in a most recent period and positive growth from the base year marked as treated regions. the study used non-parametric propensity score matching (psm) techniques to overcome the problem of selection bias arise due to the selection of regions who received treatment-spatial agglomeration, particularly in nonexperimental designs. this research is a value addition to existing literature of pakistan as it is the first study that employed propensity score matching to study agglomeration linkages with spatial inequalities arises from the unbalanced growth of regions. further, it also contributes in empirical literature by estimating herfindahl index for spatial uzma tabassum, shaista alam and ambreen fatima 190 concentration at regional level over time which is used as a base for identification of treated and untreated urban regions. the rest of the paper is proceed by highlighting empirical literature done nationally and internationally, followed by the review of theoretical model, section four sheds light on the estimation technique and data employed, section five discusses the results and final section wrap up finding and expected policy implications. 2. review of literature spatial concentration of industries is placed among the highest feature of economic activity in both developed and seeking to be developed nations. a vast existing empirical literature suggest that industries are unevenly distributed across spatial units because of difference in transportation cost, immediacy to market, presence of skilled and cheap worked force and natural locational advantage etc. thus industries agglomerate in a few regions more than others that are comparatively less attractive (ellison & glaser; 1997, maurel & sedilot, 1999 and abid & mushtaq, 2011). geographic concentration and regional specialization of industries were seen by many researchers as to be two sides of the same coin. specialization of region and concentration of industries in specific location were defined using the analogous production structures as base so as to reflect the same reality which is evident from various studies in the existing literature of regional economics (aiginger, 1999). it was since 19th and 20th century that researchers were modeling regional specialization and geographical concentration rooting from trade theory and location theory respectively and conducting empirical studies on them. the literature about regional concentration probably commence with krugman (1991). krugman (1991) was the one which estimated gini coefficient to capture concentration in various regions by comparing the regional market shares for a single industry with manufacturing’s employment structure. setting a traditionally localized industry (automotive industry) as a benchmark he found that large numbers of industries were localized though most of these were not cutting-edge and that industries related to high technology sectors such as textile, were highly clustered. however estimated concentration of such industries was not solely depends on technological spillover and what he found was biased in terms of the data he got. aiginger & rossi-hansberg (2006) deliberated regional specialization and geographical concentration as the two sides of the coin and theses two could be calculated from the same matrix showing industries in rows and geographical regions in columns. for the estimation of specialization columns are to be measured while for that of concentration rows are to be considered. aiginger & davies (2004) using a mathematical model pointed out that if inequalities across columns vary so does it varies across rows. the pragmatic evidence on the relationship between agglomeration economies and regional inequalities in developed nations may have immense importance. geppert et al journal of applied economics and business studies, volume. 4, issue 4 (2020) 187-214 https://doi.org/10.34260/jaebs.449 191 (2008) explored this connection in the context of west germany using data from 1980 to 2000 and suggest that industrial agglomeration can be a considerable contributor to regional income differentiation. they find that the number of regions that are already rich with large agglomeration of industries has succeeded in further distancing themselves from the others. thus, concentration of industries has sizeable contribution to spatial income differentiation. larger agglomerated regions such as munich, hamburg and frankfurt are leading among the richer regions that are becoming richer and showing a sign of urbanization economies. though, the analysis does not ascertain the particular nature of causal mechanism that could be underlying this relationship. kim (2008) surveyed the theoretical and empirical literature about spatial and regional inequalities. the survey of literature underscores that endowment of resources, proximity to port and increasing returns attributable to concentration of human interaction increases spatial inequalities. all these are beneficial for region to enhance their productivity therefore these regions grow relatively with faster rate of growth. however congestion cost and excessive urban concentration may be harmful. thus, in equilibrium there is an optimal level of spatial disparities to combat with inefficient level of agglomeration. midelfart (2004) examined the role played by agglomeration of economic activities, as measured by density of employment, and composition of labour force skills in explaining regional income difference in the norwegian economy especially focusing on norway's manufacturing sector. using data for the sub-regional nuts 4 spatial units of norway for two time period 1989 &1999 the study concluded that high income seems to be an upshot of agglomerated economic activities in contrast to differences in skill composition of labor force. two different measures of income are considered as a proxy to measure income disparities, per hour gdp of employees and mean hourly earnings. considering gdp per hour the regional inequalities are found significant and much more distinct relative to earnings per hour. further disparities shows widening trend while considering gdp per hour in contrast to the result of average hourly earnings which shows reduction in regional disparities for same time span. regional specialization and the geographic concentration patterns of manufacturing industries as measured by dissimilarity index in bulgaria, estonia, hungary, romania and slovenia had changed during 1990-1999 (traistaru et al., 2002). they found that in bulgaria and romania regional specialization had increased with no significant changes in estonia, hungary and slovenia. further they found that large economies of scale, high technology and high wages were associated with highly concentrated industries therefore agglomeration of these industries foster the growth of these region in which they clustered. on the other hand industries with low technology and low wages were more likely to be dispersed spatially. glaeser et al (1992) and henderson et al (1995) have made decisive contributions to the empirical literature focus on urban area-city. they estimate the effect of city uzma tabassum, shaista alam and ambreen fatima 192 employment growth on specialization and diversification index. the specialization index is measured as a proportion of industry employment to the total employment of city. the diversity index computed by glaeser et al (1992) is calculate by taking the employment share of five largest industries in the city while henderson et al (1995) measured diversity index as employment share of all industry in the city aggregate employment. glaeser et al (1992) also include the inverse of plant average size as a competition index. the result of gleaeser's model shows that city growth is positively associated with diversity index and negatively by specialization and competition index. along with specialization and diversification index henderson et al (1995) also used city employment in all other industries as an additional urbanization variable and found positive impact of both specialization and overall employment on the growth of employment. focusing on the role of agglomeration of industries in specific regions as pedestal of productivity effects and developmental impulses fan & scott (2003) investigates the interconnections between economic growth and geography in the developing countries of east asia. findings indicate that agglomeration and productivity have strong association in chinese industries, especially electronics, computer and garments industries since the public policies in china follows economic liberalization. the scrutiny of empirical literature done in pakistan indicates that, the literature on agglomeration is usually focuses on the extent with which industries agglomerate in a particular region and the effect of agglomeration on the efficiency of similar and dissimilar firms in a local system. burki & khan (2011) measure agglomeration of industries using data taken from census of manufacturing industries (cmi). as per results the agglomeration is pervasive in pakistani industries, the 35.3% industries are profoundly agglomerated. while, 38.2% industries are somewhat less agglomerated and 26.5% pakistani industries are dispersed. district population, increase in road density, skilled worked force are found significant contributor in promoting spatial concentration of industries. chaudhry & haroon (2015) evaluate the economic impact of new firm in the market using punjab’s directory of industries for the year 2002, 2006 and 2010. the finding indicates that new firm entry has noteworthy effect on socioeconomic outcome. some of these outcomes are immediate while some outcomes become visible with lag. they recommended that policy makers should be familiar with nature of firm as impacts of new firm differ with the type of firms that affirms the need of specific industrial policy. azhar & adil (2019) explore the effect of agglomeration on social inclusion & firm efficiency for punjab districts. social inclusion is made of mortality rate, antenatal care, improved water & sanitation which then compressed into a single variable using factor component analysis. the results demonstrate that agglomeration is positively associated with firm efficiency and also it has significant association with social-inclusion. the districts with higher agglomeration also experience higher social inclusion. in developed nations the empirical evidence on spatial inequalities and agglomeration may journal of applied economics and business studies, volume. 4, issue 4 (2020) 187-214 https://doi.org/10.34260/jaebs.449 193 contain immense importance, the pattern of agglomeration and its impact on spatial inequalities in developing nations like pakistan is crucial to study. but despite the obvious policy concerns, there has been no empirical work found that explores the role of agglomeration in defining the income level of that region where these agglomerations take place. this study tries to fill the gap that is evident in the literature. 3. the review of selected theoretical model in the theoretical literature the most cited work is the "new economic geography (neg)" model put forward by krugman (1991). researchers have based their empirical findings on the theoretical arguments placed by new economic geography (neg) model. according to neg model spatial inequality is the net effect of two contrasting forces, one is centripetal force and the other is centrifugal force. centripetal forces determine agglomeration or concentration of economic activities that is caused by imperfect competition, increasing returns to scale and mobility of factor. on the other side centrifugal forces such as congestion, high transportation cost, restricted mobility of factor encourages dispersion of economic activities. when the cost of supplying goods at far off places would is high then firm would be dispersed around markets that are immobile. in contrast when transportation cost is low immobile market can easily get access to distinct places that encourage firms to agglomerate at particular place gaining advantages of economies of scale and externalities. the region that experiencing increasing agglomeration also come up with higher economic activities which results in spatial inequalities showing the positive inter relation (curve aa). at certain level higher agglomeration will results in increase congestion and negative externalities offsetting the positive gain of agglomeration. consequently a new relationship established between agglomeration and spatial inequality that is opposite to the previous one (curve rr). figure-1: nexus between income growth and regional inequalities source: gardiner et al. 2010. spatial agglomeration of economic activities positively contributes in the economic growth of nation through localized positive spillovers. agglomeration increase productivities and real output via labour pooling and knowledge spill over as shown uzma tabassum, shaista alam and ambreen fatima 194 from the curve ss. the equilibrium level of agglomeration, inequalities and growth comes from the intersection of two curves rr and aa. there is a tradeoff between national growth and spatial inequalities produce from agglomeration. the effective intervention of government that break up agglomeration shift rr curve to rr1 and reduces spatial inequality by the transfer of resources to poorer regions but also erase the spillover effects that results in decreasing national growth. the income of region in which economic activities clustered grow relatively at a faster rate therefore a similar trade-off became visible within regions. therefore on the basis of neg model it is expected that the regions that are experiencing agglomeration may undergo high income level relative to other regions. 4. estimation techniques first the herfindahl indices will be computed to measure the inter-temporal spatial concentration of sector. using these indices values study will identifies regions that experience higher than average initial level of concentration and positive growth of spatial concentration. 4.1 measurement of agglomeration herfindahl index of spatial concentration is calculated as the sum of the regional shares in national employment in the particular industry both by industry and regions. symbolically, 𝐒𝐂𝐑𝐣 𝐜 = ∑𝐢𝐂𝐑𝐢𝐣 𝐜𝟐 where: 𝑪𝑹𝒊𝒋 𝒄 = the concentration ratio: the weight of employment in the sector i from the city j in the overall employment of the sector 𝑪𝑹𝒊𝒋 𝒄 = 𝑬𝒊𝒋 𝑬𝒊 = 𝑬𝒊𝒋 ∑ 𝑬𝒊𝒋𝒊 the above index ranges from 0 to 1. the more a region’s herfindahl index is closer to 0 the less spatially concentrated the region is, and the more it is away from 0 or closer to 1 the more it is said to be spatially concentrated. 4.2 measuring the linkage between spatial concentration and income to estimate causal treatment effects of agglomeration in exaggerating income inequalities across various regions propensity-score matching (psm) technique is used. the psm technique was initially presented in the pioneering work of rosenbaum and robin (1983a, 1983b, 1985). propensity score matching was the first time applied in the field of macroeconomics, to find the effects of currency unions on trade flow. afterward, many economists used this technique to evaluate the causal effects in non experimental research. journal of applied economics and business studies, volume. 4, issue 4 (2020) 187-214 https://doi.org/10.34260/jaebs.449 195 propensity score matching (psm) in an experimental research randomization is assumed as there is an equal distribution of group members' characteristics for both treated and control group via random sampling and random assignments. thus randomization limits potential selection bias therefore the only interest is to estimate cause of an effect. the observational studies lack randomization as conclusions are drawn based on data that is collected without random assignment therefore the results become inconsistent and biased. in such condition propensity score matching provides the most accurate estimates of the impact. the study employed propensity score matching (psm) to analyze the effect of agglomeration of industries on the regional income in pakistan. for the analysis purpose regions that are experiencing increasing growth of agglomeration are treated as “treatment group” while the regions where agglomeration is not persistent are considered as “control group or counterfactuals.” as we have n units (subjects or regions). c denotes a treatment condition and y as a potential outcome. for each unit i (i=1.......41), ci =1 indicates that the subject unit is belongs to the treatment group with subsequent potential outcome o1i and ci =0 shows that the subject i is in counterfactual group with potential outcome o0i. to analyze causal effects in counterfactual framework the core interest is to find treatment effect for each unit i defined as ∆i= o1i o0i 1 for each unit i, y1i and y0i are not discernible at the same time because the same unit cannot simultaneously be in both the treated and untreated groups. alternatively, average treatment effects (ate) can be estimated ate = e (o1 – o0) = e (o1) –e (o0) 2 the second frequently estimated effects are average treatment on treated (att). it is the average effect of treatment on the units which are treated defined as att = (o1 o0 | c=1) = e (o1/c = 1)] e (o0/c = 1) 3 the ate is the difference between average outcomes of treated group and the average outcome of untreated group, no one of which is fully observe. the estimation of o1 requires average over the outcome of treatment for treated group (o1|c = 1) which are observed as well as average over the outcome of treatment for the untreated group (o1|c = 0), the unobserved counterfactual. in a similar way to estimate o0 average over the outcome of non-treatment for untreated group (o0|c = 0) and average over the outcome of non-treatment for treated group (o0|c = 1) required, from which (o0|c = 1) is also an unobserved counterfactual. ate= e (o1 o0) = e (o1) e (o0) 4 uzma tabassum, shaista alam and ambreen fatima 196 = [pr (c =1) e (o1|c = 1)] +[ pr (c =0) e(o1|c = 0)] [pr (c =1) e(o0|c = 1)] + [ pr (c =0) e(o0|c = 0)] 5 = pr(c=1) [e (o1|c = 1) (e (o0|c = 1)] + pr (c = 0) [e (o1|c = 0) e (o0|c = 0)] 6 here pr (c=1) is the probability that unit is treated and pr (c = 0) is the probability that unit is not treated. the first part of the equation-6 is just the proportion of units receiving the treatment multiply by the average treatment effect on treatment. it is very likely in the absence of randomization that subjects' background characteristics in the two groups are significantly different which may also influence potential outcome so the present study cannot directly assess the effect of treatment without controlling for other influential factors. the unbalanced distribution influential factor between two groups may create selection which creates biased ate. hence, to solve the selection bias problem and evaluate the treatment effect matching adopts two postulations. conditional independence assumption (cia) conditional on influential factors x, c and (o1 & o0) are independent. x is the vector of observed cofounders. to estimate att the conditional independence assumption (cia) states o0⊥ c | x or o0⊥ c | pr(x) 7 assumption states that there is no association among treatment and the potential outcome that arise without treatment, conditional on observable characteristics or the propensity score (rosenbaum and rubin; 1983). further for estimating ate the cia states that conditional on observable characteristic or propensity score, treatment is independent of both treated and untreated outcomes because for the estimation of ate treated units are required to serve as a counterfactuals for untreated units. o1, o0 c | x or o1, o0 c | pr(x) 8 common support assumption (csa) a second important assumption besides independence required in matching method is the common support or overlap condition. it precludes the occurrence of perfect predictability of c conditional on x. it ensures that units with the similar values of x possess a positive probability of being both in treated and untreated (heckman et al; 1999). furthermore, the common support assumption makes sure that these propensity scores must fall between 0 and 1. 0 <pr(c = 1 | x) < 1 9 the common support assumption in strict sense can be written as 0 < c <pr(c = 1 | x) <1-c < 1 10 journal of applied economics and business studies, volume. 4, issue 4 (2020) 187-214 https://doi.org/10.34260/jaebs.449 197 in a nutshell propensity score analysis set forth by rosenbaum and rubin (1983) ensures to reduce selection biased by balancing the distribution of observed influential characteristics (cofactors) between two groups; treated and control. therefore psm allow one to get unbiased estimates from observational analysis. psm balances the allocation of observed cofactors among treated and untreated group based on resemblance of their predicted probabilities of having treatment; propensity scores. the psm estimates mean effects without taking any arbitrary assumptions related to functional form and error distribution and not require parametric modeling relating subject to outcome. estimating the propensity score for practical implementation of the propensity score two actions are required to take, one is about the model to be used for the estimation, and the other is about selection of variables that include in the model. for dichotomous dependent variable or binary treatment both logit and probit models set forth similar bounded results as compare to linear probability model that gives predictions exterior to [0,1] bounds. while deciding the variable to be included for estimating of propensity score it must be kept in mind that propensity score balances the covariate across treatment status. thus, executing matching necessitate selection of covariates x that plausibly satisfy this condition. omitting essential covariates can critically raises bias in resulting estimates (heckman et al., 1997). only those covariates should be included that simultaneously influence the treatment decision and the outcome variable. further it should also be ensures that variables included in the model should not be affected by treatment or the anticipation of it. heckman et al (1999) further indicate that the data for both groups should originate from the same resource e.g. from same questionnaire. the too good data is also not appropriate because if propensity score is either 0 or 1 for some values of x then conditional on those x values, matching would not be possible. these characteristics either never or also receive treatment. so, the common support condition not be realized and matches cannot be carried out (kahn and tamer; 2010). randomization is needed to guarantee that unit with indistinguishable characteristics can be observed in two groups (heckman et al., 1998). selection of matching algorithm a variety of different strategies can be use to match units on their propensity scores. these method are summarize in figure-2 figure-2 ways to perform matching uzma tabassum, shaista alam and ambreen fatima 198 source: author visualization the commonly used method of propensity score matching is nearest neighbor matching (rosenbaum & rubin; 1985) that matches each unit "i" in the treatment group with a unit "j" in untreated group with closets absolute distance among their propensities expressed as ɗ (i,j) = min{|ƥ(xiƥ(xj)|} 11 alternatively, caliper matching matches each unit "i" in the treatment group with "j" in untreated group within a particular caliper band "b" (cochran & rubin, 1973), stated as ɗ (i,j) = min{|ƥ(xi-ƥ(xj)| <b} 12 rosenbaum and rubin (1985) suggested that the caliper band ought to be less than or equivalent to 0.25 of the standard deviation of propensity score. later austin (2011) suggested that the optimal value of b should be 0.20. dehejia & wahba (2002) proposed the idea of radius matching which matches each unit "i" in the treated group with multiple units in untreated group within a particular band ɗ (i,j) = min{|ƥ(xiƥ(xj)| < b} 13 in the strategy of stratification matching or interval matching the common support of propensity score is alienated into a set of interval. the impact within each intermission is measured by computing the mean difference in potential outcomes between treated and untreated (rosenbaum and rubin; 1983). under the normality the use of five intervals are often enough to remove 95% of the biasness associated with covariates (cochrane and chambers; 1965 and imbens; 2004).kernel matching and local linear matching are non-parametric matching estimators. in kernel matching weighted averages of all units in the control group are utilized to create the ten counterfactual outcomes. weights rely on the extent with which each treated unit is farther from the untreated group and the accomplice observation for which the counterfactual is estimated. different matching algorithms nearest neighbor (nn) caliper and radius stratification and interval kernel and local linear journal of applied economics and business studies, volume. 4, issue 4 (2020) 187-214 https://doi.org/10.34260/jaebs.449 199 this study employ the most commonly used matching strategy nearest neighbor (nn) matching. differences are attained after matching the entire treated unit with control units and come up with att by averaging these differences. this kind of matching can be performed with or without replacement. when the matching is done with replacement (one nearest neighbor) it curtails biasness but enlarge the variance. in contrast, matching with more nearest neighbors inflate biasness but reduces the variability. the general formula of att is att = (o1 o0 | c=1) = e (o1|c = 1)] e (o0|c = 1) 14 using nearest neighbor (nn) matching att is estimated as attnn = 1 𝑁𝐶 ∑ [𝑂𝑖 𝑜𝑏𝑠 − ∑ 𝑊𝑖𝑗𝑗∊𝐶(𝑖)𝑀 𝑂𝑖𝑗 𝑜𝑏𝑠 ]i:wi=1 15 = 1 𝑁𝐶 ∑ 𝑂𝑖 𝑜𝑏𝑠 𝑖:𝑤𝑖=1 − 1 𝑁𝑇 ∑ 𝑊𝑗𝑗∊𝐶(𝑖)𝑀 𝑂𝑖𝑗 𝑜𝑏𝑠 16 nc indicate the no. of observations in the treated group ni u indicate the no. of controls or untreated matched with treated observations i wij = 1 𝑁𝑈 if j is a control unit of i and 0 otherwise and wj = ∑iwij c (i)m indicate the set of first m matches for unit i 5. empirical findings regions experiencing positive growth of agglomeration and above-average level of agglomeration in most recent period 2017-18 are listed in table-1.the results are obtained using lfs data for the period 2010 to 2018. for this time span only eight of the 40 urban regions fulfilled both criteria necessary to qualify as a ‘treated regions. table-1: summary statistics of herfindahl-hirschman index (hhi) uzma tabassum, shaista alam and ambreen fatima 200 source: authors’ own calculation based on herfindahl index after identifying the treated and untreated regions the second step is to estimate causal treatment effects of agglomeration on regions income that responsible to exaggerate inequalities across treated and untreated regions. as stated earlier for this purpose propensity-score matching (psm) technique is used. based on the hypothesis to be tested, number of covariates appears to be critical for this research. these covariates include the demographics of the household resides in the regions which are spatially concentrated, the knowledge and education variables. in addition with other categorical variables such as migrant or non migrant household, occupation and non categorical labour market variables such as female participation in the labour market, no of work hours etc all these variables expected to influence the average income of the household. the comparison of covariates across treated and untreated subject can be done by comparing the mean or median of continuous variables and the allocation of categorical journal of applied economics and business studies, volume. 4, issue 4 (2020) 187-214 https://doi.org/10.34260/jaebs.449 201 variables among these two groups. for this purpose the standardized difference (sd) can be used. the standardized difference evaluates the variation in means in units of the pooled standard deviation (flury & riedwyl; 1986). for continuous variable the sd is measured as: 𝑆𝐷𝑐𝑜𝑛 = �̅�𝑡𝑟𝑒𝑎𝑡𝑒𝑑 − �̅�𝑐𝑜𝑛𝑡𝑟𝑜𝑙 √ 𝑆𝑡𝑟𝑒𝑎𝑡𝑒𝑑 2 +𝑆𝑐𝑜𝑛𝑡𝑟𝑜𝑙 2 2 where �̅�𝑡𝑟𝑒𝑎𝑡𝑒𝑑&�̅�𝑐𝑜𝑛𝑡𝑟𝑜𝑙 are the sample mean of continuous variables in treated and untreated groups respectively, 𝑆𝑡𝑟𝑒𝑎𝑡𝑒𝑑 2 &𝑆𝑐𝑜𝑛𝑡𝑟𝑜𝑙 2 are the sample variance in both groups. for dichotomous variables sd is measured as: 𝑆𝐷𝑑𝑖𝑐ℎ𝑜 = ƥ̂𝑡𝑟𝑒𝑎𝑡𝑒𝑑 − ƥ̂𝑐𝑜𝑛𝑡𝑟𝑜𝑙 √ ƥ̂𝑡𝑟𝑒𝑎𝑡𝑒𝑑∗(1−ƥ̂𝑡𝑟𝑒𝑎𝑡𝑒𝑑 )+ƥ̂𝑐𝑜𝑛𝑡𝑟𝑜𝑙∗(1−ƥ̂𝑐𝑜𝑛𝑡𝑟𝑜𝑙) 2 where ƥ̂𝑡𝑟𝑒𝑎𝑡𝑒𝑑&ƥ̂𝑐𝑜𝑛𝑡𝑟𝑜𝑙 are the occurrence of dichotomous variable in treated and untreated group. smd is not influenced by sample size and allow comparison of the variables measured in different units. less than 0.1 value of smd indicates aim perceptible difference in mean or prevalence of variables between two groups (normand et al; 2001). the below table shows the "standardized difference (sd)" the difference in term of standard error of some of demographic variables in treated and untreated group. the result shows that the treated and untreated differ by about -0.24 sd in dependency ratio, dependency ratio is higher in the untreated regions than the regions that are experiencing concentration. education and age of highest earner in treated and untreated differ by 0.12 sd while the age of head of household is more similar in treated and untreated as sd is about 0.08. significant difference appears in average years of schooling that is about 0.37 sd. table-2 standardized differences in selected demographic variables mean in treated mean in untreated standardised diff. dependency ratio 2.87 3.39 -0.24 average education 8.25 7.15 0.37 education of highest earner 5.9 5.45 0.12 age of highest earner 37.25 35.84 0.12 age of hoh 48.25 47.17 0.08 source: authors calculations based on lfs 2017-18 the standardized difference in selected binary variable such as gender of head of household and gender of highest earner is quite negligible but the proportion of migrant household in treated group is quite differ than comparative group as the value of sd is about 0.26. uzma tabassum, shaista alam and ambreen fatima 202 table-3 standardized differences in selected dichotomous variables proportion in treated proportion in untreated standardised diff. migration 0.20 0.11 0.26 gender of hoh (male) 0.91 0.94 -0.09 gender of highest earner (male) 0.90 0.92 -0.09 source: authors calculations based on lfs 2017-18 the standardized difference in labour market related variables shows that percentage of females participate in labour market both in treated and untreated group differ by about -0.21 sd. female in untreated group participate more in labour market than the female resides in spatially agglomerated regions. the individuals belongs to the household located in treated regions are technically more sound and the standardized difference is about 0.32 between treated and untreated group. the average weekly hours work is differ by 0.4 sd, households resides in spatially concentrated regions weekly works more hours than the comparative group. there is a negligible difference (0.06 sd) in working age population between treated and untreated regions. table-4 standardized differences in labour market variables mean in treated mean in untreated standardised diff. female participation ratio 0.14 0.21 -0.218 % of technically trained working age individuals in hh 0.26 0.17 0.322 working age population 3.84 3.71 0.062 average weekly hours works 51.26 45.81 0.401 source: authors calculations based on lfs 2017-18 the figure-3 demonstrates the outcome variable-average income of the household from the districts where spatial concentration is observed during the considered time span and the average income of the household from the districts that are spatially disbursed. it is visible from the figure that the households living in those urban regions which are experiencing agglomeration enjoys high level of per-capita income in contrast to those household which are living in regions where concentration is not high and persistent. figure-3: average income of household in treated and untreated region journal of applied economics and business studies, volume. 4, issue 4 (2020) 187-214 https://doi.org/10.34260/jaebs.449 203 source: authors visualization based on lfs 2017-18 next figure also shows significant difference in average income of household resides in spatially concentrated urban regions and the regions where such concentration is not quite visible. the largest part of the data points are either overlap the diagonal or to the left of the diagonal that shows that the values of the distribution on the y-axis are higher than those of the distribution on the x-axis means income of the household reside in agglomerated region is much higher than the income of the households that reside in the regions which does not fulfill the criteria to be called agglomerated. it can also be observed that income inequality is somewhat stronger for higher income but it is quite weak for lower income. figure -4: quantile-quantile plot of income distribution source: authors visualization based on lfs 2017-18 estimation of propensity score the estimates of probit models are used to estimates the propensity scores based on which the matching is done afterward. the probit model provides the probabilities or propensities to calculate the average treatment effect of spatial concentration on the average income of household resides in these regions. 0 10 00 00 20 00 00 30 00 00 40 00 00 m ea n of in co m e 0 1 0 50 00 00 0 1. 00 e+ 07 1. 50 e+ 07 in co m e, a gg _t = = 1 0 1000000 2000000 3000000 4000000 income, agg_t == 0 quantile-quantile plot uzma tabassum, shaista alam and ambreen fatima 204 to estimates the propensities the dichotomous treatment variable -spatial concentration run on the covariates selected however, the estimates of the probit model are just to predict the propensities and hence not provide any meaning full explanation. the study has first estimated probability of receiving treatment by regressing binary variable agglomeration equals 1 if regions spatial concentration is above average in base period and with positive growth and 0 otherwise as a function of household characteristics including average education, age & gender of head of household, age and education of highest earner, dependency ratio, female participation ratio, percentage of technically trained working age individuals in household, average weekly hours works and provincial and occupational dummies. the probability of receiving treatment is positively correlated with average schooling, age of highest earner and head of household, migrant household, male headed household, no of working age individuals in household, average weekly hours works, percentage of technically trained working age individuals in household and negatively correlated with age of head of household, education of highest earner, dependency ratio and female participation rate. almost all variables are significant, except some occupational and regional dummies. the results as stated above are annexed (see table a1). overall, the model is statistically fit on the basis of the likelihood chi-square. for the estimation of ps, 18,722 observations that fall within the common support region [.002032, .8526713] have been used (see annex table-a2). common support or overlap condition it is essential to test out the common support region for the both treated and untreated group. if the distribution does not overlap it would be problematic but using the minimum and maximum comparison the problem could be tackled. to do this all those observation that has propensity score smaller than minimum or larger than the maximum would be deleted. hence all those observations are withdrawn from the analysis that comes into view outside the common support region. table-5 shows the number of observation removed from the analysis on the basis of nearest neighbor matching. the number of removed observation is quite low in this case only 3 observations have been removed from the analysis that appeared off support. table 5: common support condition for nearest neighbor psmatch2 treatment assignment psmatch2 common off support on support total untreated 0 14,036 14,036 treated 3 4,683 4,686 total 3 18,719 18,722 source: author's estimations based on lfs 2017-18. a description of the density distribution of propensity score can be seen in figure-1. the upper part of the figure shows propensity score distribution for treated, whereas the below part of the figure refers to the treated. the symmetric distribution of the covariates after the matching indicates the overlapping of the treated and non-treated groups and hence it is concluded that the common support condition is satisfied. journal of applied economics and business studies, volume. 4, issue 4 (2020) 187-214 https://doi.org/10.34260/jaebs.449 205 figure-5: distribution of propensity score to test quality of matching among treated and untreated groups source: author's estimations based on lfs 2017-18. effects of spatial concentration on average income: this research work has empirically tested that concentration of industries at particular region has positive impact on the livelihood of household resides in these regions. the role of spatial concentration in raising livelihood by positively affect income of the household is proved valid. the empirical result of average treatment effect of the spatial concentration on income level of household is reported in table-6. statistically significant and positive coefficient indicated that the one percent increase in spatial concentration raises income level of the households that reside in these urban districts of pakistan by 346854. table-6: average treatment effects of agglomeration on income matched/unmatched treated controls difference t-stat unmatched 12.75 12.38 .374 27.07 att 12.76 12.55 .203 9.09 source: author's estimations based on lfs 2017-18. the positive and statistically significant value of average treatment effect of unmatched group (att) shows that on average, income of household increased by [(e0.374– 1)*100 =45.4 %] for 2017-18. further, the positive and statistically significant value of the average treatment effect on treated (matched att) shows that income of household increased by [(e0.20335– 1) = 22.55%]. 0 .2 .4 .6 .8 propensity score untreated treated: on support treated: off support uzma tabassum, shaista alam and ambreen fatima 206 quality of matching the study carried out t-tests to evaluate either difference between treated and nontreated groups are statistically significant or not. once the balance is attained with psm, tstatistic turns out to be statistically insignificant because there is negligible difference between treated and non-treated. in this regards, the covariates’ variance ratios of the treated over the untreated measures are also estimated. the ratio should be equal to one if there is perfect balance across samples. all these measures provide portent of whether particular covariates are balanced across treated and non-treated groups. the results indicate that before matching most of the variables have demonstrated statistically significant differences but after matching most of the covariates found balanced and no significant statistical differences have been found1.further the table-7 indicates that the bias percentage for each of the covariate is also reduced significantly after applying matching. table-7 covariates balancing test (ps-test) variable unmatched/ mean %reduction t-test v(t)/v(c1) matched treated control %bias bias t female participation rate u 0.138 0.213 -21.8 74 -30.26 0.63* m 0.258 0.239 5.7 2.58 1.03 gender of the head – male u 1.084 1.059 9.8 77.3 16.08 1.39* m 1.077 1.071 2.2 1.06 1.07* working age members in household u 3.841 3.713 6.2 -70.7 9.35 0.94* m 4.472 4.254 10.6 5.23 1.12* education of the highest earner u 5.631 4.897 19.5 93.1 23.33 1.09* m 5.731 5.781 -1.3 -0.66 1.08* age of the highest earner u 37.07 35.59 12.6 97.8 15.07 1.06* m 36.27 36.30 -0.3 -0.12 0.95 technical/vocational training u 0.253 0.166 30.9 99.4 50.06 1.32* m 0.285 0.285 0.2 0.09 1.05 average hours work u 52.061 47.177 42.8 96.2 62.69 0.95* m 49.668 49.851 -1.6 -0.86 0.79* average schooling u 8.203 6.806 48.5 99 71.9 1 m 8.263 8.278 -0.5 -0.23 1.01 migrant hoh u 1.166 1.089 33 99.3 56.51 1.68* m 1.178 1.179 -0.2 -0.1 1.05 dependency ratio u 3.189 3.448 -11.5 45.1 -15.82 0.67* m 2.091 2.234 -6.3 -4.26 0.94* age of hoh u 47.15 45.826 10.3 70.3 15.62 0.95 m 48.863 48.47 3.1 1.52 1.03 *if variance ratio outside [0.98; 1.02] for u and [0.94; 1.06] for m source: author estimation using lfs-2017-18 the absolute percentage bias for treatment and the control group before and after the matching are presented in the table-8. it can be observed from the table that the absolute mean bias has reduced from 18.8 to 1.6 percent and absolute median bias reduced from 9.2 to 1.2 percent for the year 2017-18 after performing the psm. also, 1occupational and provincial dummies were included in pstest and also found statistically insignificant journal of applied economics and business studies, volume. 4, issue 4 (2020) 187-214 https://doi.org/10.34260/jaebs.449 207 after matching pseudo r2 is reduced significantly from 0.21 to 0.003. insignificant pvalue of likelihood ratio after matching implies that estimation of the propensity score matching is efficient regarding balancing the distribution of covariates between two groups. so, on the basis of the measures presented it can be concluded the (nearest neighbor) propensity score matching is an efficient method to construct a similar control group and to estimate the average treatment effect. table-8 average percentage bias sample ps r2 lr chi2 p>chi2 mean bias med bias unmatched 0.215 3143.96 0.000 18.8 9.2 matched 0.003 43.05 0.555 1.6 1.2 source: author estimation using lfs-2017-18 the graph indicates how individual variables balance after matching. the x-axis displays the standardized bias, which is the percentage difference of the sample means in the treated and non-treated as a percentage of the square root of the average of the sample variances in the treated and non-treated groups (rosenbaum and rubin; 1985) figure-6 standardized bias across covariates source: author estimation using lfs-2017-18 it can be visualized from figure that the unmatched sample exhibit large imbalances with standardized bias being present across many of covariates but once the matching is done the standardized differences diminished significantly. further the same results can also be seen through the histogram. it can be observed that the standardized -100 -50 0 50 standardized % bias across covariates unmatched matched uzma tabassum, shaista alam and ambreen fatima 208 percentage bias has reduced significantly after the matching. comparing the upper panel which shows the percentage bias in unmatched (non-treated group) with the second panel reveals significant reduction in percentage bias after the matching. figure-7 histogram of standardized % bias source: author estimation using lfs-2017-18 6. conclusion and policy implications the purpose of this research is to explore the variation of spatial concentration /agglomeration across urban districts of pakistan and analyses the effect of this concentration on the prevalence of income inequalities. the basic view is that spatial concentration amplifies inequalities across regions, because of the differences in productivity. the empirical literatures on the spatial disparities between urban regions based on agglomeration are rather scarce. this study attempts to start bridging this gap by using survey data conducted by pakistan bureau of statistics (pbs). the empirical results and the evidences provided clearly indicate that distribution of income is significantly affected by spatial concentration supports the viewpoint of new economic geography (neg) model. as per new economic geography (neg) model the economic activities will be higher in the region of increased agglomeration which results in spatial inequalities showing the positive inter relation. to reveal results this study used two steps methodology. first the herfindahl indices have computed, using labour force survey (lfs) for the year 2010 to 2018, to measure the inter-temporal spatial concentration of industries. using the values of indices urban regions are identified that have positive and above average level of concentration. four large cities, gujranwala, sialkot, lahore, multan and four other urban areas rawalpindi, faisalabad, gujranwala 0 .0 5 .1 .1 5 d en si ty -80 -60 -40 -20 0 20 40 60 80 standardized % bias across covariates unmatched 0 .0 5 .1 .1 5 d en si ty -80 -60 -40 -20 0 20 40 60 80 standardized % bias across covariates matched journal of applied economics and business studies, volume. 4, issue 4 (2020) 187-214 https://doi.org/10.34260/jaebs.449 209 and dg khan fulfilled both criteria. after identifying the agglomerated regions the second step was to estimate causal treatment effects of agglomeration on regions income that responsible to exaggerate inequalities across agglomerated and non agglomerated regions. propensity-score matching (psm) was applied for this purpose. for the analysis eight regions that are experiencing increasing growth of agglomeration was considered as "treated group" while the regions where agglomeration appeared non persistent was considered as “control group.” the statistically significant coefficient of average treatment effect on treated (att) with positive sign indicated that the one percent increase in spatial concentration increase income level of the households that reside in these urban districts of pakistan by 346854. further no significant statistical differences have been found after matching and most of the covariates found balanced as per ps-test results. after performing psm absolute percentage bias also decrease significantly. the absolute mean bias has reduced from 18.8 to 1.6 percent and absolute median bias reduced from 9.2 to 1.2. in a nut shell, it can be concluded that agglomeration is important determinant of regional variations in income. the reported findings have significant implications for regional policy. it means that obstructing the agglomeration of industries with the intention of encouraging equality may hurts productivity. in order to lessen spatial income inequalities focusing on the development of regional agglomeration of activities, rather than dispersion of resources to every single region, seems to be the correct way to carry on. the effective policies to promote agglomeration are needed for growing urban population that has the potential to turn out to be an effective part of economic growth. pakistan fronting many challenges; key to this is the provision of employment to growing labour force and create vibrant industrial sector that can provide variety of jobs requiring skills of different intensity. in the absences of such opportunities the standard of living for those that are affected abated which may lead social unrest. there is also need to improve infrastructure in order to promote agglomeration and greater inclusion and equity. references abid a. burki& mushtaq a. khan (2011). "agglomeration economies and their effects on technical inefficiency of manufacturing firms: evidence from pakistan", international growth centre working paper aiginger, k. &rossi-hansberg, e. (2006). “specialization and concentration: a note on theory and evidence” empirica 33, pp. 255–266. aiginger, k. (1999) “do industrial structures converge? a survey on the empirical literature on specialization and concentration of industries”, wifo-working papers. aiginger, k., davies, d. (2004). “industrial specialization and geographical concentration; two sides of the same coin?” journal of applied economics 7(2), pp. 231-248. uzma tabassum, shaista alam and ambreen fatima 210 austin p.c. (2011) "a tutorial and case study in propensity score analysis: an application to estimating the effect of in-hospital smoking cessation counseling on mortality". multivariate behavioral research, 46 pp. 119–151. azhar, a. & adil, s. (2016). " effect of agglomeration on socio-economic outcomes: a district level panel study of punjab" the pakistan development review papers and proceedings: the 32nd conference of the pakistan society of development economists, december 13 15, 2016, pp. 155-170 burki, a. a & khan, a. m. (2012). "agglomeration economies and their effects on technical inefficiency of manufacturing firms: evidence from pakistan", international growth centre working paper chaudhry, a., & haroon, m. (2015). "the economic impact of new firms in punjab". the lahorejournal of economics 20, pp. 143–182 cochran, w. g., & rubin d. b. (1973). “controlling bias in observational studies: a review,” sankhya, the indian journal of statistics, 35(4), pp.417–446. cochrane, w., and s. chambers (1965). “the planning of observational studies of human populations,” journal of the royal statistical society, series a, 128, pp. 234–266. daytyan, k. (2016). "interrelation among economic growth, income inequality, and fiscal performance: evidence from anglo-saxon countries," review of public economics, vol 217(2), pp 37-66. dehejia, r. h., and s. wahba (2002), “propensity score matching methods for nonexperimental causal studies,” review of economics and statistics, 84, pp. 151–161. duranton, g., puga, d. (2004). “micro-foundations of urban agglomeration economies.” in: henderson, j.v., thisse, j.-f. (eds.), handbook of urban and regional economics, vol. 4, north-holland, amsterdam, pp. 2063-117. ellison, g., e. glaeser (1997). "geographic concentration in u.s. manufacturing industries: a dartboard approach", journal of political economy, 105, pp. 889– 927. fambon, s. (2017) "the determinants of inequality and income gap between urban and rural areas in cameroon: evidence from the ecam3 household survey". advances in economics and business, 5(7), pp.394 410. fan, c., and a. scott. (2003). industrial agglomeration and development: a survey of spatial economic issues in east asia and a statistical analysis of chinese regions. economic geography 79 (3), pp. 295–319. flury, b. k., riedwyl, h. (1986) "standard distance in univariate and multivariate analysis". the american statistician. 40 pp. 249–251. gardiner, b., martin, r., & tyler, p. (2010). "does spatial agglomeration increase national growth? some evidence from europe". journal of economicgeography, 11(6), pp. 979-1006 geppert, k., marti, g & axel, w. (2008). "economic growth of agglomerations and geographic concentration of industries: evidence for west germany", regional studies, 42(3), pp. 413-421. glaeser, e., h. kallal, j. sheinkman, and a. schleifer (1992), “growth in cities”, journal of political economy, 100, pp. 1126-1152. https://ideas.repec.org/a/hpe/journl/y2016v217i2p37-66.html https://ideas.repec.org/a/hpe/journl/y2016v217i2p37-66.html https://ideas.repec.org/s/hpe/journl.html https://ideas.repec.org/s/hpe/journl.html journal of applied economics and business studies, volume. 4, issue 4 (2020) 187-214 https://doi.org/10.34260/jaebs.449 211 heckman, j., ichimura, h. & todd, p. (1997), "matching as an econometric evaluation estimator: evidence from evaluating a job training program", review of economic studies, 64(4), pp. 605-654. heckman, j., ichimura, h., smith, j. & todd, p. (1998), "characterization of selection bias using experimental data", economelrica, 66(5), pp. 1017–98. heckman, j., robert, l & smith, j. (1999)."the economics and econometrics of active labor market programs". in handbook of labor economics, 3(1), pp. 1865–209. henderson, v., a. kuncoro, and m. turner (1995), “industrial development in cities”, journal of political economy, 103, pp.1067-1152. imbens, g. (2004): “nonparametric estimation of average treatment effects under exogeneity: a review,” the review of economics and statistics, 86(1), pp. 4–29. jamal, h (2006). "does inequality matter for poverty reduction? evidence from pakistan's poverty trends" the pakistan development review 45(3), pp. 439-459 khan, s., and e. tamer (2010), “irregular identification, support conditions and inverse weight estimation,” econometrica, 78 (6), pp. 2021-2042. kim, s. (2008) "spatial inequality and economic development: theories, facts, and policies". commission on growth and development, working paper no.16 krugman, p. (1991) "geography and trade" cambridge: m.i.t. press, marshall, a. (1890) “principles of economics” london. maurel, f., b. sedillot (1999) "a measure of the geographic concentration in french manufacturing industries", regional science and urban economics, 29, pp. 575– 604. midelfart, k. h. (2004) "does agglomeration explain regional income inequalities?" norwegian school of economics and business administration and cepr naseer, s. & ahmed, a. m. (2016) "determinants of income inequality among the earners in pakistan" s3h working paper series, national university of sciences and technology (nust). normand s.l. t., landrum m.b., guadagnoli e., ayanian j.z., ryan t.j., cleary p.d., mcneil b.j.(2001) "validating recommendations for coronary angiography following an acute myocardial infarction in the elderly: a matched analysis using propensity scores". journal of clinical epidemiology. pp. 387–398. rosenbaum, p. r., &. rubin, d. b. (1983a) "the central role of the propensity scores in observational studies for causal effects." biometrika., 70, pp. 41–55 rosenbaum, p. r., &. rubin, d. b. (1983b). "assessing sensitivity to an unobserved binary covariate in an observational study with binary outcome." journal of the royal statistical society, series b. 45, pp.212–218. rosenbaum, p. r., &. rubin, d. b. (1985) "constructing a control group using multivariate matched sampling methods that incorporate the propensity score." the american statistician 39(1), pp. 33-38 sial, m.s., zheng, c., khuong, n.v., khan, t., usman, m. (2018), "does firm performance influence corporate social responsibility reporting of chinese listed companies?" sustainability, 10(7), pp. 1-12. tabassum, u. (2016)"regional specialization trade and economic size of cities" thesis " university of karachi. uzma tabassum, shaista alam and ambreen fatima 212 weber a, (1909) über den standort der industrien (mohr, tübingen). translated by c. j. friedrich, 1929, as alfred weber’s theory of the location of industries (university of chicago press, chicago) journal of applied economics and business studies, volume. 4, issue 4 (2020) 187-214 https://doi.org/10.34260/jaebs.449 213 annexure table a1: estimates of probit model (if agglomeration=1) standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 source: author's estimations based on lfs 2017-18 variables coefficient average schooling 0.075*** (0.00517) age of highest earner 0.009*** (0.00092) education of highest earner -0.012*** (0.0044) age of head of household 0.005*** (0.00093) dependency ratio 0.0184** (0.00802) migrant household 0.394*** (0.0440) technical/vocational training 0.266*** (0.0410) working age members in household 0.0224*** (0.00589) gender of hoh 0.298*** (0.0445) work hours 0.0099*** (0.00118) female participation rate -0.282*** (0.0351) constant -1.617*** (0.246) occupational dummies most of the dummies appears statistically significant provincial dummies observations 18,722 lr chi2(46) 3143.96*** pseudo r2 0.215 uzma tabassum, shaista alam and ambreen fatima 214 table a-2: distribution of estimated propensity score-livelihood percentiles smallest 1% 0.007168 0.002032 5% 0.014106 0.0023592 10% 0.021964 0.0025063 obs 18722 25% 0.106504 0.0025063 sum of wgt. 18722 50% 0.251298 mean 0.25002 largest std. dev. 0.167869 75% 0.36499 0.8400924 90% 0.466755 0.8439572 variance 0.02818 95% 0.536866 0.8474239 skewness 0.349559 99% 0.678517 0.8526713 kurtosis 2.591341 95% 0.536866 0.8474239 skewness 0.349559 99% 0.678517 0.8526713 kurtosis 2.591341 note: the region of common support is [.002032, .8526713] source: author's estimations based on lfs 2017-18 journal of applied economics and business studies, volume. 4, issue 2 (2020) 141-162 https://doi.org/10.34260/jaebs.427 141 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x an empirical testing of comparative efficiency of static and dynamic factor models towards stock returns’ predictability in capital market of pakistan laila taskeen qazi 1*, atta ur rahman 2, shahid ali 3 & sohail alam4 1 assistant professor, institute of management sciences peshawar. 2 associate professor, institute of management sciences peshawar 3 associate professor, institute of management sciences peshawar 4 lecturer, department of economics, abdul wali khan university mardan abstract efficient market hypothesis has its supporters and critics as it has invited significant attention of research scholarship in recent years. the taxonomy and existence of this hypothesis is widely debated in terms of making economic decisions in the capital markets. stock returns predictability has galvanized researchers to use forecasting models. literature shows that forecasting is possible yet it debates problems associated with the techniques used for forecasting from the time series data. the study relies on stock returns for 67 randomly selected companies listed on the pakistan stock exchange. the static and the dynamic factor models are compared in terms of forecast efficiency. the study also uses eight macroeconomic variables to forecast stock returns by including gold prices, crude oil prices, market capitalization, psx100 index, psx100 index turnover, kibor 1-month rates, kibor 3 years rates and rupee to dollar rates. the results of the hit rates and out-of-sample forecasting technique suggest that dynamic factor model is the best multivariate time series forecasting model in the pakistani context. keywords factor models, arx models jel classification c32, c38, c52, c58 1. introduction evolution seeds counterrevolution and the same is found out to be true about the efficient market hypothesis in finance. the scholarly supremacy of efficient market revolution has been significantly challenged by the economists who claim that the stock market returns to a certain stretch are predictable. this predictability argument of researchers runs against eugene fama’s (1970) market efficiency proposition. in this study, the strength of the predictability argument has been verified which is found to be aligned with the weak form of market efficiency proposed by fama (1970). testing the weak form of market efficiency primarily strive to explore how well the past returns can predict the future returns. fama (1991), in his study of * attaurrahman@imsciences.edu.pk mailto:attaurrahman@imsciences.edu.pk laila taskeen qazi, atta ur rahman, shahid ali & sohail alam 142 efficient capital markets ii has put the weak form of market efficiency tests in more general words as ‘tests of returns predictability’. provision of such tests towards returns predictability formed the basis for this study. such findings may not be novel, but researching it in the local context may bring exciting findings towards efficient market hypothesis. past studies indicate that the use of large models remains limited due to two main visible reasons. firstly, the use of large models lead to computational burden and secondly, not every time series is found to be informative. consequently, a large proportion of research in the field of predictability is more focused on the use of relevant information retrieved from large number of variables. previous research claims that the data declares a factor structure and depicts a common idiosyncratic decomposition, which forms the primary motivation behind the use of factor models in this study for forecasting (boivin & ng, 2006; pesaran & timmermann, 1995). the dimension reduction concept forms the methodological basis of this study. the use of factor models is found to overcome the issue of extensive parameterization prevalent in case of forecasting through vector autoregressive with exogenous variables (varx) models. the idea of dimension reduction achieved through factor modeling facilitates efficient multivariate time series forecasting with the minimal loss of degree of freedom and strives to suggest an alternative multivariate time series forecasting model. the concept of multivariate time series forecasting is applied to data from pakistani stock market, which is exposed to several systematic and unsystematic risk factors inducing major inefficiencies in the market. therefore, in this study both the systematic as well as unsystematic risk factors have been incorporated for multivariate forecasting of stock returns. in order to account for the systematic risk factors, eight macroeconomic variables have been used including gold prices, crude oil prices, market capitalization, psx-100 index, psx-100 index turnover, kibor 1-month rates, kibor 3 year’s rates and rupee to dollar rates. any changes in these macroeconomic factors create an impact on the whole market and serve as the systematic risk factors. for the unsystematic factors the company’s historical data is used to identify a common factor structure, which entails maximum variation in the historical stock returns. through the use of factor models this study strives to fill the gap in literature in terms of multivariate time series forecasting. both static and dynamic factor models are employed to create a comparison between the two approaches. in the light of mixed previous findings, this study works toward fresh evidence towards the forecasting the stock returns. the forecasting results of our suggested models will enable this research effort to conclude whether beating the market through forecasts of the stock returns is possible. 2. literature review the time series forecasting is a challenging topic with an abundance of literature supporting or negating the notion of forecasting in the light of efficient market hypothesis (emh) journal of applied economics and business studies, volume. 4, issue 2 (2020) 141-162 https://doi.org/10.34260/jaebs.427 143 proposed by fama (1965). in order to strengthen the argument of the study, this section provides a comprehensive and critical analysis of the past researches. fama (1965) for the first time gave the definition of an efficient market, in his landmark research work based on empirical analysis of stock market prices. fama (1965) concludes that the stock prices follow a random walk. the argument of random walk is further supported by samuelson (1965), who provides the economic idea of martingale to support the efficiency of markets. this fundamental ideology of efficient markets forms the grounds for emh and on the basis of this hypothesis, markets are further segregated as weakly or strongly efficient (robert, 1967). as a proof of market efficiency, fama et al., (1969) conduct a study and find considerable support to the emh and conclude that the markets are efficient. (fama, 1970) conclude that all supporting arguments lead to the development of a definitive understanding regarding the concept of emh and the categorization of efficient markets emerge as a taxonomy. the literature associated the concept of emh with that of random walk according to which flow of information is uninterrupted and any new information is immediately reflected in the stock prices therefore future prices must be unpredictable and random (charlesa & darnéb, 2009; gupta & basu, 2007; cootner, 1964). kemp and reid (1971) rejects the emh concluding that share price movements are nonrandom proposing that revelation of unequal information prevents risk sharing and allows one party in a trading process to either earn abnormal profits or avoid losses. similarly, several forms of firm announcements also serve as a threat to emh (schole, 1972;). in terms of these economic profits jensen (1978) attempts to define an efficient market as a market where on the basis of an information set, it is impossible to generate economic profits through trading. however, it is impossible for a market to be informationally efficient as generating information is costly and investors tend to spend resources on gathering and processing the information (charlesa & darnéb, 2009; gupta & basu, 2007; grossman & stiglitz, 1980). it is not only the information set that makes a market inefficient rather the response of share prices towards any particular information also creates room for arbitrage and economic profits. literature argues that excessive volatility in stock markets rejects emh (leroy & porter, 1981; stiglitz, 1981; de bondt and thaler, 1985; roll, 1986). in addition to the stock market volatility and price overreaction, another factor that adds to the market inefficiencies is the duration of overreaction or market volatility in response to any new information (fama, 1988; porterbaand & summers, 1988). other similar findings tend to reject the random walk hypothesis for the movements in stock returns and establishes the notion of stochastic trends found in the stock returns depicting mean reverting behaviors (lo & mackinlay, 1988; conrad & kaul, 1988; lo and mackinlay, 2011; bekaert and hodrick, 1992). lo, mamaysky and wang (2000) in their study focus on the use of statistical tools and techniques in pursuit of identifying signals in the stock prices that can be used in the technical laila taskeen qazi, atta ur rahman, shahid ali & sohail alam 144 analysis to determine the future stock prices. their findings support the predictability argument. the predictability argument is further supported by lim & brooks, (2010), charlesa & darnéb, (2009) and gupta & basu, (2007). this argument is further strengthened in the literature by the well-known and heavily cited work of lo & mckinlay, (1988) and fama (1991). lo and makinlay (1988) clearly assert that stock prices do not follow a random walk and claimed that the stock prices are partially predictable. lehmann (1990) finds that the predictability in stock markets is attributable to the predictable changes in equity returns, market inefficiencies, and/or to stock price overreactions. short time intervals such as a week, induces predictability in the stock returns (lim & brooks, 2010). ball (2009) points out the international financial crisis is the best evidence for market inefficiencies. moreover, incorporating the markets of developing countries, lee et al., (2010) investigate the stationarity of the stock prices for 32 developed and 26 developing countries and conclude that the stock markets are inefficient as the return predictability is supported by their findings. also, there is abundance of literature that finds the notion of predictability through various macroeconomic variables including stock market returns (schumacher et al., 2006; chin and lin, 2011; ince & trafali, 2007; stock and watson, 2006; banerjee, et al., 2006; giannone, et al., 2008). the literature that rejects emh creates optimism among the researchers and investors for forecasting of stock returns as well as of several other macroeconomic variables such as gross domestic product (gdp), inflation and industrial production etc (ince & trafali, 2007). in terms of macroeconomic variables forecasting becomes important due to the fact that the current state of the economy cannot be fairly assessed as the important economic indicators are released with a time lag (breitung & eickmeier, 2006). similarly, investors in stock markets strive to acquire firsthand information as quickly as possible in order to take the arbitrage advantage. consequently, forecasting allows for the availability of timely information (schumacher et al., 2006). chin and lin (2011) believe that forecasting in time series through univariate autoregressive model or vector autoregressive models (var) are some common options. amongst these methods, var is preferred in terms of forecasting as it allows for the use of larger information set. the inclusion of larger information sets improve the predictability power of the models (ince & trafali, 2007). a univariate model allows for the use of small subset of the whole information set. therefore, a natural innovation to the univariate models is the var models. in pursuit of using larger information sets for predictability, the model does not make use of all the available variables in the given model (boivin & ng, 2006). using such models is considered unwise due to the fact that they decrease the degrees of freedom, making the estimation ineffective (chin & lin, 2011). apart from the size of information needed for forecasting, another significant issue is the fact that the financial times series are usually found to be non-stationary, complex and deterministically chaotic (boivin & ng, 2006). to incorporate the use of a larger information journal of applied economics and business studies, volume. 4, issue 2 (2020) 141-162 https://doi.org/10.34260/jaebs.427 145 set, the literature stimulates the use of dimension reduction techniques of forecasting (boivin & ng, 2006). according to dimension reduction techniques if movements in all macroeconomic variables are driven by a few common factors then those few common factors must be used to forecast the macroeconomic variables (chin & lin, 2011). in recent studies, factor models are quite successful in forecasting especially in case of panel data (stock and watson, 2006). the large number of variables used in the literature for forecasting the stock market returns indicate that failing to incorporate any significant variable might encounter the problem of omitted variables (marcellino, et al., 2003; den reijer, 2005; schumacher, 2007). in this regard, factor analysis allows the use of large data sets enabling the researchers to look at everything necessary for forecasting (breitung & eickmeier, 2006). the use of factor analysis for forecasting has been significantly confirmed in literature (stock and watson, 2006; banerjee, et al., 2006; giannone, et al., 2007). the estimated factors from the factor models are used for estimation and inference (breitung and eickmeier, 2006; reichlin, 2003). literature also provides evidence on the segregation of factor models into two main forms including static and dynamic factor models (bovin and ng, 2005; stock and watson, 2003). bovin and ng (2005) provide a comparison of the static and dynamic factor models for the purpose of forecasting. factor models are used in literature for forecasting conditional means of market data (stock and watson, 2002; cristadoro, et al., 2001; artis, et al., 2005; marcellino, et al., 2003; den reijer, 2005; schumacher, 2007, alonso et al., 2018). moreover, these models are also used for forecasting conditional volatility. the literature shows the use of factor models in policy analysis as well (bernanke and bovin, 2003; giannone, et al., 2005; favero, et al., 2005; stock and watson, 2005; forni, et al., 2009). ludvigson and ng, (2007) discusses the use of factor analysis for conditioning information and for the term structure analysis. forecasting models based on static factors approach are employed by brisson et al., (2003) to forecast stock returns using the canadian data. these models are also employed for forecasting of british time series data and the spanish macroeconomic data (artis et al., 2005; camacho and sancho, 2003). regardless of the initial success of the factor models in the field of forecasting, in the literature authors tend to doubt the forecasting results of static factor models employing the use of static component factors. giacomini and white (2006) assert in their work that static factor models do not depict superior prediction performance in competition to other models under the moving window simulation studies. similarly banerjee et al., (2005) compares forecasts made by static factor model and single indicator regression models for the european time series data. the results of comparison do not depict any supremacy of forecasts made by static factor models over the single indicator models. schumacher and dreger (2004) use the german stocks laila taskeen qazi, atta ur rahman, shahid ali & sohail alam 146 data to analyze the forecasting capabilities of the static factor models and concluded that the use of factor models do not give any better forecasting results than the usual statistical tests. the apprehensions pertaining to the forecasting ability of the static factor models paves way for the emergence of the use of dynamic factor models. this notion is presented by forni et al., (2001, 2003a, b) through the estimation of factor models in frequency domain. forni et al., (2003b) highlight the theoretical significance of dynamic factor models over the static factor models. the proposition that runs behind the dynamic factor models is the fact that dynamic models connect the variables at different points in time, however, only the contemporaneous variables enter the static models. there is abundance of literature that strives to empirically test the forecasting ability of these models. for instance, den reiger (2005) confirms the forecasting supremacy of these models by using them to successfully forecast the dutch macroeconomic variables. similarly, kapetanios (2004) confirms the immense forecasting capability of these models by using them to forecast the core index for uk inflation. the popularity of their use increased manifold due to their computational advancements and availability of large panels of time series. altissimo et al., (2001) use the dynamic factor models for constructing the coincidence indicator for the euro business area cycle and found out three common factors. similarly kabundi (2004) employ the dynamic factor models for forecasting the french business cycle growth rates. moreover, eickmeier and ziegler (2008) use dynamic factor models for forecasting the output and inflation and also conducted the comparison of forecasts made by the dynamic factor model with other forecasting models. results of their study significantly support the immense forecasting ability of the dynamic factor models. in spite of theoretical and empirical support for these models, some researchers still argue their use on grounds of robustness and misspecification errors. for instance, bovin and ng (2005) assert the significance of static factor models on static principal components claiming that these models are more robust to misspecification errors. for their use, fewer auxiliary parameters are needed to be estimated as compared to the dynamic models. due to the mixed results pertaining to the forecasting ability of both the dynamic and static factor models, this study strives to create a comparison between the more sophisticated dynamic factor models proposed by forni et al., (2001, 2003a, b), and the static factor models proposed by stock and watson (2002). this study intends to assess which of the two econometric models does a better job of forecasting stock returns by using data from pakistani stock markets. in addition, the study works to identify common factors through factor analysis that does forecast the stock returns better. an attempt is made to incorporate certain significant exogenous variables in order to make more accurate forecasts. several studies in the literature are found to study the relationship between stock prices and fundamental macroeconomic as well as firm specific variables. in terms of firm specific variables these studies use dividend yield, earnings to price ratio and size and claim the predictive power of these variables (bassu, 1977; fama & french, 1992). within the context of several international stock markets, journal of applied economics and business studies, volume. 4, issue 2 (2020) 141-162 https://doi.org/10.34260/jaebs.427 147 variables like size, book to market ratio, cash flow yield and earnings to price ratio are found to be significant variables in terms of predictability (jaffe & westerfield, 1985; kato, zeimba & schwartz, 1990). factor models are considered preferable over the traditional forecasting regressions due to a number of its advantages (breitung & eickmeier, 2006, hallin, & lippi, 2013). firstly, factor models can cope with many variables without sacrificing the degree of freedom which is a frequent problem with regressions including more than usual number of variables. secondly, since the factor models allow for the use of a large number of variables, they enable researchers to exploit more information for making more precise forecasts (breitung & eickmeier, 2006). thirdly, idiosyncratic movements, which possibly include measurement error and the local shocks, can be eliminated (breitung & eickmeier, 2006, qazi et al.,2015). fourth, forecasting through factor modeling allows the investors to have a more in-depth knowledge over the stock market and relevant strategies can therefore be developed. in addition to the company specific variables, literature categorizes several macroeconomic variables as the determinants of stock market returns (humpe & macmillan, 2009; paye, 2012; chen, 2009; patelis, 2012; balvers, cosimano & mcdonald, 2012). the most prominent of these variables include oil prices, gold prices, cash reserve ratio, food price inflation, call money rates, dollar rates, foreign direct investments (fdi), foreign portfolio investments and foreign exchange reserves, stock market trade volume, market capitalization and long term and short term interbank offer rates (ghosh, bandyopadhyay & choudhuri, 2011). in the same context, our study checks the predictive power of macroeconomic variables including historical gold rates, dollar rates, 1 month kibor rates, 3 years kibor rates, psx-100 index, psx market capitalization, psx index turnover and crude oil prices. the use of factor models to make forecasts is supported by the literature, however, there is little evidence for implementing these models for the stock returns in pakistan and making less erroneous forecasts. our study attempts to fulfill this gap, it also creates a comparison to evaluate which of the two types of factor models is more effective in forecasting. 3. research methods this study uses exploratory factor analysis for the given time series to identify the most significant factors and to make predictions. the work is relying on meticulously derived econometric models by using historical stock returns data from pakistan stock exchange. 3.1 sample design finding data about stock prices is not a novel thing anymore, as there is abundance of data from stock exchanges around the world. our econometric model setting demands that our unique approach of forecast efficiency is measured using local data which can be taken to broader data sets in later studies. pakistan stock exchange publishes data on listed pakistani laila taskeen qazi, atta ur rahman, shahid ali & sohail alam 148 companies in routine and has historical datasets that can be used by researchers. at the time of our data acquisition, pakistan stock exchange has 651 listed companies. to assure the selection of a random sample, a pilot sample of 50 stock prices is taken and standard deviation in their stock prices is measured. the standard deviation from stock prices of this pilot sample is recorded as 51.478 which is used along a 95% confidence coefficient and allowable error of 1% to estimate the size of the sample. the following formula is used to estimate the sample size; 𝑛 = ( 𝑧𝛼 2⁄ 𝜎 𝑒 ) 2 our resultant objective sample is of 106 companies, which is initially selected randomly using software from 651 listed companies. but during stock data acquisition it was realized that 39 companies are having significant missing values, which are ultimately dropped from the sample. the final obtained sample is found to be 10% of the population, which is deemed acceptable for further analysis. next, there is an important consideration of choosing the timeline for the dataset which is selected from december 15, 2016 to october 2018. this timeline was used to minimize the chances of any time related extraneous effects that may affect our modeling and can affect our forecasts as the effort was to see the efficiency of the different factor models in terms of forecasts. our final sample consists of 489 observations for each company’s stock. this enabled our research to rely on 32763 stock return observations for our sample of 67 randomly selected firms. relative returns 𝑟𝑡have been calculated at time t and t-1 in this study using the following formula, 𝑟𝑡 = 𝑝𝑡 − 𝑝𝑡−1 𝑝𝑡−1 the study employs the use of relative returns due to its empirical nature. the relative differences give interest or percentage yield obtained within period t-1 to t using daily compounding. therefore the average returns here are computed as geometric mean of the 𝑟𝑡with daily compounding. next, exogenous variables are gathered for the selected timelines using publications of different resources like pakistan stock exchange, daily dawn, ministry of finance, and state bank of pakistan. these include eight independent variables consisting of psx 100 index, psx 100 index turnover, psx 100 index market capitalization, rupee to dollar rates, gold rates, 1 month kibor rates, 3 years kibor rates and crude oil prices. the final sample dataset is organized in spreadsheets for further statistical analysis. journal of applied economics and business studies, volume. 4, issue 2 (2020) 141-162 https://doi.org/10.34260/jaebs.427 149 3.2 static factor models there is abundance of literature in the area of static factor models (lawley & maxwell 1971; stock & watson, 1998; croux, et al. 2004; deistler & zinner, 2007). a static factor model may take the following form; 𝑦𝑖𝑡 = 𝜆𝑖𝐹𝑡 + 𝑒𝑖𝑡 (1) in equation. (1) for static factor analysis 𝑒𝑖𝑡 refers to the idiosyncratic error and λi means factor loadings. the model can be further simplified and presented as follows; yt = λft + 𝑒𝑡(2) where yt is the n -dimensional of vector of returns, 𝐹𝑡 are the important principal components (pc), 𝑒𝑡 the n-dimensional vector of noise, and λ ∈ r nxr is the loading matrix. the ft and 𝑒t are uncorrelated, e(ft𝑒t) = 0 as an autoregressive process with input (arx) the one step ahead forecast of factor process ft can be made as under. ft+1 = a(z)ft + d(z)xt + ut+1 (3) a(z)and d(z)are polynomial matrices in the backward shift operator z of order p and q, respectively and the stability conditiondet(i − za(z)) ≠ 0for all |z| ≤ 1 holds. we assume that ut is an r -dimensional white noise, xt is an m -dimensional linearly regular, stationary and ergodic process with mean zero and e (xtϵ́s) = 0 for all t, s ∈ z. one step-ahead forecasts of yt+1 are then obtained as ŷt+1 t⁄ = λ̂f̂t+1 t⁄ (4) where f̂t+1 t⁄ is the one step ahead forecast of ft based on equation (3). 3.3 dynamic factor model this study investigates the viability of dynamic factor models by comparing two approaches of dynamic factor models including the method of stock and watson (2002) and the method of forni, hallin, lippi, and reichlin (fhlr; 2005). according to the dynamic factor model proposed by stock and watson (2002), 𝑋𝑖𝑡 is the observed data for the financial time series understudy at time t, for i=1,2,3…n and t=1,2,3…t. here 𝑦𝑡 is the time series variable to be forecasted. considering a dynamic factor model, 𝑋𝑡 = 𝑏(𝑧)𝑓𝑡 + 𝑒𝑡 (5) laila taskeen qazi, atta ur rahman, shahid ali & sohail alam 150 𝑦𝑡+ℎ = 𝛽𝑓(𝑧) ′𝑓𝑡 + 𝛽𝑤 ′ 𝑤𝑡 + 𝜖𝑡+ℎ (6) in equation (5) 𝑒𝑡 is the idiosyncratic disturbance, h is the forecast horizon, 𝑤𝑡 is an m x 1 vector of observed variables that together with the q-dimensional dynamic factors 𝑓𝑡 are useful for forecasting the 𝑦𝑡+ℎ and 𝜖𝑡+ℎ is the resulting forecast error. according to the method of forni, hallin, lippi, and reichlin (fhlr; 2005) in the dynamic factor model the observations 𝑦𝑖𝑡 are written as the sum of common components 𝑋𝑖𝑡 and the idiosyncratic component𝜖𝑖𝑡. the common component is driven by a q-dimensional vector of common dynamic factors 𝑓𝑡 = (𝑓1𝑡𝑓2𝑡 … . . 𝑓𝑞𝑡) which are loaded with different coefficients and lags. therefore, 𝑦𝑡 = 𝐶𝐹𝑡 + 𝜖𝑡 (7) where 𝐹𝑡 = (𝑓𝑡, 𝑓𝑡−1, … 𝑓𝑡−𝑠) and 𝐶 = (𝐵0,𝐵1,…𝐵𝑆), and in the above equation (7) is a static factor model with r = q(s+1) dimensional static common factors 𝐹𝑡 and q-dimensional dynamic factors 𝑓𝑡. for the purpose of forecasting let us assume 𝑦𝑡 as a stochastic process for which the 𝑋𝑡 is a vector of common components and can be assumed with the following equation, xt̂= [𝛤0 �̂�𝑍 ̂(𝑍 ̂𝛤0̂𝑍 ̂) −1 ] 𝑍 ̂𝑦𝑡 (8) the equation (8) provides consistent in-sample estimators for 𝑋𝑡 . the dynamic factor model employs the use of spectral density matrix for in-sample estimators. moreover out of sample predictors for 𝑋𝑇+ℎ are given by the following equation. �̂�𝑇+ℎ/𝑇= [𝛤ℎ �̂�𝑍 ̂(𝑍 ̂𝛤0̂𝑍 ̂) −1 ] 𝑍 ̂𝑦𝑇 (9) here 𝑍 ̂ = (𝑍 ̂1 … … 𝑍 ̂𝑟) are the r generalized principal components of 𝑦𝑡. therefore under the dynamic factor model one step ahead forecast of 𝑦𝑡+1 is given by �̂�𝑇+ℎ. providing a brief description of both forms of factor models the study utilized both the models to identify the most relevant common factors creating major impact on the stock prices and on the basis of these factors forecasting for the future stock market were performed. under the step 1, model is specified through two simultaneous activities being the estimation of the common components or factors r and the inclusion of the explanatory variables. here the estimation of the number of common factors is done through two variants of factor analysis being static and dynamic factor models. within the two models the number of common components r is determined through using the kaiser criteria. according to the kaiser criteria r is determined on the basis of the number of eigen values in the correlation matrix having vales greater than 1. the number of r determined in the first step through both static and dynamic factor models are then transferred to step 2. journal of applied economics and business studies, volume. 4, issue 2 (2020) 141-162 https://doi.org/10.34260/jaebs.427 151 for the given number of r in step 2, loading and factors are estimated for achieving the one step ahead forecasts for 𝑦𝑡. in this step first pca is calculated as a static factor model which gives the static factors and spectral density matrix gives the dynamic factors. these factors are then estimated through the arx model. the explanatory variables inclusion as input to the arx model and specification of other dynamics are the most important issues in the arx models. in the present study the selection of explanatory variables is made on the basis of aicbic type information criteria. those explanatory variables for which the model’s aic-bic values were found to be the smallest were considered for the final selection for the purpose of forecasting. after the selection of a final arx model forecasting has been done in step 3 on the basis of equation (10) given as under, ft+1 = a(z)ft + d(z)xt + ut+1 (10) this one step ahead forecasting has been done using both dynamic as well as static factors. in step 4 the comparison of the forecasting results from different models using the out of sample forecasting technique. according to the out-of-sample forecasting method for yt+1, the sample is divided into two non-overlapping parts i.e. 1, … , s and s + 1, … , t. the data falling in the first part is termed as in sample and is used for the estimation and the specification of the model. the second part of the data is termed as out-of-sample and has been used for the evaluation of the models and their forecast quality. the one-step ahead predictor of yi,t+1 is ŷi,t+1/t and the equivalent prediction error is ϵ̂i,t+1/t = yi,t+1 − ŷi,t+1/t . following are given different techniques for assessing the forecast quality. 3.4 hit rate a simple concept using the following hit rule is applied to the sampled stock return data of 67 companies. since the observed stock returns data of 489 observations is at hand, which is halved as in-sample and out-of-the sample to monitor the forecast efficiency of our models. the in-sample data is used for estimating the out of sample stock returns at one step only. this suggests that the following conceptual table can be thought of; table 1: hit rate forecasts actual observations profit loss profit a b loss c d the accuracy of the forecast is determined by the forecast hit rate which is measured as (a+d)/(a+b+c+d) and expressed as a percentage in the findings section; the following formula is used to measure the same; laila taskeen qazi, atta ur rahman, shahid ali & sohail alam 152 hi = 1 t − s ∑ 𝕝sign(yi,t+1)=sign(ŷi,t+1/t) t t=s+1 where ŷi,t+1/t is the one-step ahead forecast of yi,t+1. 3.5 coefficient of determination this relative statistical measure is a basic computation to check for model fit. a model is better fitted if the coefficient of determination is higher having a value from 0 to 100%. the following rule is used to compute its value in our models; ri 2 = cor (yiŷi) 2 * sign (cor (yiŷi)) 4. empirical results 4.1 comparison of forecasting results the primary objective of this study is to investigate stock return predictability. factor models are used for predictability. furthermore, the study also strives to establish that which factor modeling technique does a better job in forecasting stock returns. since methodologically the study focuses on testing the forecasting power of factor modeling therefore the study employs the use of two main variants of factor models including static factor models and dynamic factor models. here the forecasting results of these models have been compared using the technique of out-of-sample forecasting. in the out-of-sample forecasting procedure of yt+1, the sample time period has been divided into two non over lapping time periods. the first time period from december 15, 2016 to april 5, 2018 is the in sample which is 70% of the total sample and is used for the estimation of common factors and for model specification. the remaining 30% of the sample time period, is out-of-sample used for evaluation of the forecasting results. based on a moving window of 5 days, out-of-sample 1step-ahead forecasts from april 6, 2018 to october 28, 2018 have been generated. in the static forecasting model the number of principal components or factors, the lag of factors and the number of lags of exogenous variables have been selected on the basis of bayesian information criterion (bic). on the basis of bic criteria the static factor model identifies three principal components significant for forecasting the stock returns. moreover, the 1-step-ahead out-of-sample forecast generated through static factor model incorporates 3 lags of principal components and exogenous variables basing this selection on the bic criteria. within the dynamic factor models the study tested two approaches, one suggested by stock and watson (2002) and the other suggested by forni et al., (2009). in the dynamic factor models the in-sample estimates of dynamic factors are calculated through the spectral density matrix for each data group and the selection of the common dynamic factors is made on the basis of bic criteria. journal of applied economics and business studies, volume. 4, issue 2 (2020) 141-162 https://doi.org/10.34260/jaebs.427 153 the forecasting quality of the understudy models has been evaluated with respect to outof-sample hit rate and 𝑅2. here the forecasting quality has been evaluated on the basis of positive out-of-sample 𝑅2 and hit rates greater than 0.5 for all the under study forecasting models. following a discussion has been provided pertaining to the forecasting results for all 67 companies returns time series arranged in seven groups employed in the study in order to check the multivariate time series forecasting power of the models under study. the forecasting results of each group are provided below. table 2: forecasting results for group 1 sector company names static factor models dynamic factor models r2 hit rate sw flhr r2 hit rate r2 hit rate financial services arif habib limited 0.000 0.16 0.01 0.36 0.1 0.46 bipl securities ltd. 0.01 0.40 0.01 0.50 0.05 0.52 dawood equities ltd. 0.04 0.41 0.00 0.53 0.03 0.53 banks bank al habib 0.06 0.50 0.01 0.55 0.05 0.56 bank al-falah limited 0.01 0.49 0.02 0.51 0.04 0.51 bank of punjab 0.04 0.59 0.01 0.56 0.07 0.58 askari bank 0.01 0.51 0.03 0.53 0.05 0.55 faysal bank 0.02 0.54 0.03 0.54 0.01 0.57 habib metropolitan bank 0.01 0.53 0.04 0.53 0.01 0.55 average 0.0222 0.459 0.0178 0.512 0.0456 0.537 (a) static factor3.5, mva bic moving window, size = 5 (b) sw: q=1, s=2,bic moving window, size = 5 (c) flhr: q=1, s=2,bic moving window, size = 5. table 2 presents a comparison of forecasting results of both static and dynamic factor models for group 1. according to the criteria defined for analyzing out-of-sample forecasting quality of a model, the results depict the fractions of positive out-of-sample 𝑅2 for all the companies of financial services and banking sector included in group 1. the highest average 𝑅2 of 0.0456 has been reported by the flhr method of dynamic factor models. the average 𝑅2 of static factor (0.0222) model is although better than the stock and watson method of dynamic factor models yet it’s less than the flhr method. however in terms of out-of-sample hit rates forecasting quality of the models understudy, both the methods of dynamic factor models outperform the static factor model (table 1). the average hit rate of 53.7% under the flhr method, for the companies of financial services and banking sectors it indicates the supremacy of dynamic factor models in general and the flhr method in particular over the static factor models. static factor model in group 1 reported average hit rate of 45.9% which is indicative of poor forecasting ability of the model. therefore it can be concluded that in group 1 for financial services and banking sector companies the dynamic factor model proposed by flhr method is the best alternate multivariate time series forecasting model. the forecasting results of group 2 are reported in table 3, it can be inferred that again the highest average 𝑅2 of 0.04 has been reported for flhr method (table 2). the average 𝑅2 for laila taskeen qazi, atta ur rahman, shahid ali & sohail alam 154 the static factor model is 0.023 which is better than the stock and watson method of dynamic factor models for which the average 𝑅2 is 0.012. the highest average hit rate of 53.7% for life and non life insurance companies is reported by the static factor models which is greater than the 52.2% average hit rate of dynamic factor models (flhr method). therefore𝑅2 and hit rate criteria provide contradictory results in group 2. according to 𝑅2criteria the dynamic factor models give better forecasts as compared to static models whereas hit rates illustrate dominance of static factor models (table 3). table 3: forecasting results for group 2 sector company names static factor models dynamic factor models r2 hit rate sw flhr r2 hit rate r2 hit rate non life insurance adamjee insurance 0.01 0.57 -0.01 0.53 0.02 0.53 central insurance 0.04 0.56 0.02 0.51 0.06 0.54 century insurance co. 0.00 0.46 0.02 0.53 0.03 0.53 efu general insurance 0.01 0.57 0.02 0.51 0.06 0.54 atlas insurance 0.02 0.51 0.01 0.43 0.02 0.49 life insurance efu life assurance 0.06 0.55 0.01 0.49 0.05 0.5 average 0.023 0.537 0.012 0.500 0.040 0.522 (a) static factor3.5, mva bic moving window, size = 5 (b) sw: q=1, s=2, dynamic factors = 2, bic moving window, size = 5. (c) flhr: q=1, s=2, dynamic factors = 3, bic moving window, size = 5. the forecasting results from factor models of group 3 are reported in table 4. according to the forecasting results for group 3, worst forecasting quality is shown by static factor models (table 4). the average 𝑅2 for static factor model is 0.007 which is far below the average 𝑅2 of 0.016 reported by the dynamic factor models under the flhr method. in addition in terms of hit rates the static factor model failed to satisfy the minimum criteria of hit rates being greater than 0.5 for quality forecasts (table 3). both the variants of dynamic factor models satisfy the hit rate criteria as for stock and watson approach the average hit rate for group 3 companies is 50.8%, however the flhr method does better than the static model and stock and watson method of dynamic factors, depicting average hit rate of 52.4%. therefore the forecasting results of group 3 categorize dynamic factor models under the flhr approach as the best alternate multivariate time series forecasting model for predicting stock returns in the psx. journal of applied economics and business studies, volume. 4, issue 2 (2020) 141-162 https://doi.org/10.34260/jaebs.427 155 table 4: forecasting results for group 3 sector company names static factor models dynamic factor models r2 hit rate sw flhr r2 hit rate r2 hit rate personal goods gadoon textile 0.00 0.50 0.01 0.50 0.02 0.52 colgate palmolive pak. 0.01 0.50 0.00 0.49 0.01 0.51 bata pakistan 0.01 0.50 0.01 0.50 0.01 0.53 azgard nine 0.00 0.52 0.01 0.51 0.01 0.52 leisure goods grays of cambridge -0.01 0.49 0.01 0.54 0.01 0.54 households goods al abid silk 0.00 0.36 -0.01 0.51 0.01 0.51 food producers habib sugar 0.00 0.51 0.02 0.51 0.01 0.54 habib adm limited 0.01 0.48 0.01 0.50 0.01 0.51 dewan sugar 0.04 0.50 0.01 0.51 0.05 0.54 average 0.007 0.484 0.008 0.508 0.016 0.524 (a) static factor3.5, mva bic moving window, size = 5, (b)sw: q=1, s=2, dynamic factors = 3, bic moving window, size = 5. (c) flhr: q=1, s=2, dynamic factors = 4, bic moving window, size = 5. similar forecasting results have been reported for group 4 (table 5). on the basis of both the 𝑅2 as well as the hit rate criteria, dynamic factor model under the flhr method outperforms the static factor model and the dynamic factor model of stock and watson (2002). the average 𝑅2 for flhr method is 0.017 as compared to the average 𝑅2 of 0.01 for static factor model and 0.008 for dynamic factor model of stock and watson (2002) (table 5). although the 𝑅2 for all the three models understudy are positive yet the highest average 𝑅2 has been reported by flhr method. likewise the highest average hit rate is reported for flhr dynamic model being 53.4%. the hit rate report by flhr dynamic model is evidently better than those reported for static (51.2%) and stock and watson dynamic factor model (50.3%). on the basis of these finding, for the companies in group 4, primarily representing the manufacturing sector, it can be inferred that flhr method of dynamic factor modeling is the best alternate for forecasting the multivariate time series data in the pakistani equity market. table 5: forecasting results for group 4 sector company names static factor models dynamic factor models r2 hit rate sw flhr r2 hit rate r2 hit rate pharma and bio tech abbott labs.(pak.) 0.02 0.56 0.01 0.5 0.03 0.56 ferozsons lab -0.01 0.44 0 0.45 0.01 0.53 glaxosmithkline pak. 0.02 0.58 0.02 0.51 0.02 0.58 chemicals arif habib corporation 0.00 0.52 0.01 0.49 0.01 0.51 clariant pakistan 0.01 0.51 0.01 0.51 0.01 0.54 dawood hrc.chems.corp. 0.01 0.53 0.02 0.53 0.02 0.55 dewan salman fibre 0.01 0.56 0.01 0.51 0.05 0.56 engro 0.00 0.5 0.01 0.5 0.02 0.52 fauji fertilizer 0.02 0.43 -0.01 0.51 0.01 0.51 fauji ftlz.bin qasim 0.03 0.53 0.01 0.51 -0.01 0.52 gatron industries 0.02 0.47 -0.01 0.49 0 0.49 ici pakistan -0.01 0.51 0.01 0.52 0.03 0.54 average 0.010 0.512 0.008 0.503 0.017 0.534 laila taskeen qazi, atta ur rahman, shahid ali & sohail alam 156 (a) static factor3.5, mva bic moving window, size = 5, (b) sw: q=1, s=2, dynamic factors = 3, bic moving window, size = 5. (c) flhr: q=1, s=2, dynamic factors = 5, bic moving window, size = 5. the forecasting results for group 5 exhibit complete dominance of flhr dynamic model over the static as well as stock and watson dynamic factor model (table 6). although the average out-of-sample 𝑅2 for all the three models is positive yet the highest is reported by flhr dynamic model i.e. 0.02 (table 6). however, in terms of hit rate criteria, the average hit rates for static factor model as well as for the stock and watson dynamic model, is less than 50% which categorizes the forecast quality of these models as lower to flhr. the average hit rate for flhr dynamic model is 51.3% (table 6). on the basis of out-of-sample hit rate and out-of-sample 𝑅2, flhr dynamic factor model is sorted out as the best alternate multivariate time series forecasting model for pakistani equity market. table 6: forecasting results for group 5 sector company names static factor models dynamic factor models r2 hit rate sw flhr r2 hit rate r2 hit rate construction and materials gharibwal cement 0.01 0.50 0.00 0.50 0.01 0.54 gammon pakistan -0.01 0.22 -0.01 0.34 -0.01 0.41 fecto cement 0.01 0.38 0.01 0.5 0.01 0.5 fauji cement company 0.01 0.51 0.01 0.5 0.05 0.53 dg khan cement company 0.02 0.54 0.06 0.54 0.09 0.6 dewan cement 0.00 0.5 0.01 0.51 0 0.5 dadex eternit 0.01 0.45 0.01 0.5 0.03 0.53 dandot cement 0.01 0.36 0.02 0.5 0.02 0.5 dadabhoy cement 0.02 0.52 0 0.5 0.03 0.54 chearat cement company 0.01 0.5 0.01 0.54 0.01 0.52 bestway cement 0.00 0.21 0.03 0.47 0.03 0.49 balochistan glass 0.06 0.35 0.01 0.41 0.01 0.45 attock cement pakistan 0.03 0.55 0.01 0.56 0.01 0..56 al-abbas cement 0.04 0.54 0.03 0.5 0.01 0.52 industrial metals and mining crescent steel 0.02 0.55 0.01 0.55 0.02 0.54 huffaz seamless pipe 0.03 0.46 -0.01 0.5 0 0.52 average 0.017 0.446 0.013 0.495 0.020 0.513 (a) static factor3.5, mva bic moving window, size = 5, (b) sw: q=1, s=2, dynamic factors = 3, bic moving window, size = 5. (c) flhr: q=1, s=2, dynamic factors = 6, bic moving window, size = 5. the one-step-ahead forecast quality of the understudy models in group 6 depicts significant win of dynamic factor models over the static factor model (table 7). the average out-of-sample 𝑅2 for both the variants of dynamic factor models i.e. flhr method (0.015) and stock and watson method (0.017) are positive and greater than that of static factor model (0.011). the average hit rate for the static factor model is 42.1% which is less than 50% therefore the forecast quality of static model is found out to be poor for the return time series of companies belonging to industrial engineering and automobiles and parts sectors (table 7). in forecasting the multivariate time series for group 5 the flhr dynamic model outperformed journal of applied economics and business studies, volume. 4, issue 2 (2020) 141-162 https://doi.org/10.34260/jaebs.427 157 the static and stock and watson dynamic factor model as the average hit rate for flhr dynamic factor model is reported to be 53.7% (table 7). table 7: forecasting results for group 6 sector company names static factor models dynamic factor models r2 hit rate sw flhr r2 hit rate r2 hit rate industrial engineering hinopak motors 0.02 0.37 0 0.45 0 0.47 dewan autv.engr. 0.04 0.25 0.02 0.45 0.02 0.49 al-ghazi tractors 0.01 0.52 0.01 0.53 0.01 0.56 al-khair gadoon 0.00 0.14 0.01 0.48 0 0.49 bolan castings 0.01 0.41 0.04 0.52 0.02 0.54 automobiles and parts agriauto industries 0.01 0.47 0.02 0.53 0.01 0.57 atlas honda 0.01 0.50 0.01 0.51 0.02 0.56 dewan farooque motors 0.00 0.50 0.01 0.58 0.01 0.58 general tyre & rubber 0.02 0.53 0.01 0.5 0.01 0.53 honda atlas cars (pak.) -0.01 0.52 0.04 0.56 0.05 0.58 average 0.011 0.421 0.017 0.511 0.015 0.537 (a) static factor3.5, mva bic moving window, size = 5. (b) sw: q=1, s=1, dynamic factors = 2, bic moving window, size = 5. (c) flhr: q=1, s=2, dynamic factors = 4, bic moving window, size = 5. the forecasting results of group 7 are given in table 8 below. the forecasting result for group 7 also categorize flhr dynamic model as the best alternate multivariate time series forecasting model which is attributable to the fact that the average hit rate of flhr model in group 7 is 51% (table 8). the average hit rate reported by the flhr model is better than the average hit rate of 50% reported by the stock and watson dynamic model and 48.2% reported by the static factor model (table 8). moreover these results are further confirmed by the positive fractions of out-of-sample 𝑅2. the average out-of-sample 𝑅2 is also reported highest by the flhr dynamic factor model in group 7. table 8: forecasting results for group 7 sector company names static factor models dynamic factor models r2 hit rate sw flhr r2 hit rate r2 hit rate electricity hub power company 0.04 0.50 0.01 0.53 0.02 0.52 forestry and paper century paper 0.01 0.47 0.01 0.47 0.06 0.5 oil and gas devp. attock petroleum 0.02 0.51 0.00 0.49 0.02 0.5 attock refinery 0.01 0.50 -0.01 0.51 0 0.53 byco petroleum pakistan 0.02 0.43 0.00 0.50 0.01 0.5 average 0.020 0.482 0.002 0.500 0.028 0.510 (a) static factor3.5, mva bic moving window, size = 5. (b) sw: q=1, s=1, dynamic factors = 2, bic moving window, size = 5. (b) flhr: q=1, s=2, dynamic factors = 4, bic moving window, size = 5. overall the forecasting results for all the groups suggest the dynamic factor models in general and the flhr method of dynamic factor modeling as the best alternate model for the multivariate time series forecasting. for all the groups the flhr model reported positive average out-of-sample 𝑅2 and hit rates greater than 0.5 or 50%. therefore on the basis of these laila taskeen qazi, atta ur rahman, shahid ali & sohail alam 158 two criteria the forecast quality of dynamic factor models under the flhr method is found out to dominate the other two models understudy. eight variables are included in this study consisting of historical gold rates, dollar rates, 1 month kibor rates, 3 years kibor rates, psx-100 index, psx market capitalization, psx index turnover and crude oil prices. along with the common factors these variables are incorporated in the forecasting model for the out-of-sample data. the out-of-sample forecasting results depict convergence between the literature and the empirical findings of this study. several variables have been found significant in the forecasting of stock returns. the 3-yrs kibor rates are found to be significant on the 1st lag whereas 1-month kibor rates are found significant on the 3rd lag. the returns of psx-100 index are found significant in forecasting the company’s stock returns on the 4th lag. similarly crude oil prices, psx-100 index turnover and dollar rates become significant on the 4th lag which means the impact of these variables become evident on stock returns over time period. in addition the historical gold rates are found to be significant on the 2nd lag. therefore it can be concluded that out of the 8 exogenous variables incorporated in the study seven are found to be significant in forecasting the stock return. 5. conclusion this study utilizes factor models within the framework of dimension reduction in forecasting. within this context, this study focuses on creating a comparison of various forms of factor models in order to suggest a best fit model for forecasting multivariate time series data. the study tests forecasting power of static factor models and dynamic factor models and within the dynamic factor models the study compares two approaches, one proposed by stock and watson (2002) and the other proposed by forni, et al., (2005). in this study, the forecasting quality of each model is analyzed through the use of out-of-sample forecasting technique. two main criteria were applied for analyzing the forecast quality of all the considered models including positive out-of-sample 𝑅2 and out-of-sample hit rates greater than 50%. the empirical findings of the study categorizes dynamic factor models under the method proposed by forni, et al., (2005) (flhr method). the average 𝑅2 for all the data groups for dynamic models were positive fractions with all the average hit rates greater than 50%. overall this study can prove to be helpful for the stock market investors and analysts as it attempts to provide a robust forecasting model for the purpose of effective forecasting of financial time series data. moreover, the study also includes the macroeconomic variables for the purpose of forecasting which broadens the scope of the study and the applicability of the forecasting results of the study. the empirical findings of the study challenges the efficient market hypothesis and finds it less useful in the pakistani context. the study supports the weak form of emh proposed by fama (1991) which entails predictability and also recommends dynamic factor models as an efficient tool to forecast the stock returns. journal of applied economics and business studies, volume. 4, issue 2 (2020) 141-162 https://doi.org/10.34260/jaebs.427 159 references alonso, a. m., bastos, g., & garcía-martos, c. (2018). bias correction for time series factor models. journal of statistical computation and simulation, 88(8), 1576-1602. artis, m., banerjee, a. & marcellino, m. (2005). ‘factor forecasts for the uk’, journal of forecasting, 24(4), . 279–298. ball, r. (2009). the global financial crisis and the efficient market hypothesis: what have we learned? journal of applied corporate finance, 21(4), 8–16. balvers, r. j., cosimano, t. f. & mcdonald, b. (2012). predicting stock returns in an efficient market, the journal of finance, 45(4), 1109-1128.. banerjee, a., marcellino, m., & masten, i. (2005). leading indicators for euro‐area inflation and gdp growth. oxford bulletin of economics and statistics, 67, 785-813. banerjee, a., masten, i. & massimiliano, m. (2006). forecasting macroeconomic variables using diffusion indexes in short samples with structural change, bekaert, g. & hodrick, r. j. (1992). characterizing predictable components in excess returns on equity and foreign exchange markets, the journal of finance, 47(2), 467–509. bernanke, b. & boivin, j. (2003). monetary policy in a data rich environment, journal of monetary economics, 50(3), 525–546. boivin, j. & ng, s. (2006). are more data always better for factor analysis?. journal of econometrics, 132(1) 169–194. breitung, j., & eickmeier, s. (2006). dynamic factor models. allgemeines statistisches archiv, 90(1), 27-42. brisson, m., campbell, b., & galbraith, j. w. (2003). forecasting some low‐predictability time series using diffusion indices. journal of forecasting, 22(6‐7), 515-531. business recorder, (2012). karachi stocks, business recorder, retrieved on july 15, 2012 from, http://www.brecorder.com/market-data/karachi-stocks/ camacho, m., & sancho, i. (2003). spanish diffusion indexes. spanish economic review, 5(3), 173-203. charles, a., & darné, o. (2009). the random walk hypothesis for chinese stock markets: evidence from variance ratio tests. economic systems, 33(2), 117-126. chen, s. s. (2009). predicting the bear stock market: macroeconomic variables as leading indicators. journal of banking & finance, 33(2), 211-223. chen, c. c., & lin, h. w. (2011). the advantages of dynamic factor models as techniques for forecasting: evidence from taiwanese macroeconomic data. international journal of economics and finance, 3(5), 208-215. cootner, p. h. (ed.) (1964). the random character of stock market prices, the mit press, cambridge, ma. cristadoro, r., forni, m., reichlin, l. & giovanni, v. (2001). a core inflation index for the euro area, croux, c., renault, e., & werker, b. (2004). dynamic factor models. journal of econometrics, 119(2). http://www.brecorder.com/market-data/karachi-stocks/ laila taskeen qazi, atta ur rahman, shahid ali & sohail alam 160 conrad, j., & kaul, g. (1988). time-variation in expected returns. journal of business, 409425. de bondt, w. f., & thaler, r. (1985). does the stock market overreact?. the journal of finance, 40(3), 793-805. deistler, m., & zinner, c. (2007). modelling high-dimensional time series by generalized linear dynamic factor models: an introductory survey. communications in information & systems, 7(2), 153-166. eickmeier, s., & ziegler, c. (2008). how successful are dynamic factor models at forecasting output and inflation? a meta‐analytic approach. journal of forecasting, 27(3), 237265.. fama, e. (1965). the behavior of stock market prices journal of business, vol. 38. 34-105 fama, e. f., fisher, l., jensen, m. c., & roll, r. (1969). the adjustment of stock prices to new information. international economic review, 10(1), 1-21. fama, e. f. (1970). efficient capital markets: a review of theory and empirical work. the journal of finance, 25(2), 383-417. fama, e. f., & french, k. r. (1988). permanent and temporary components of stock prices. journal of political economy, 96(2), 246-273. favero, c. a., marcellino, m., & neglia, f. (2005). principal components at work: the empirical analysis of monetary policy with large data sets. journal of applied econometrics, 20(5), 603-620. forni, m., hallin, m., lippi, m., & reichlin, l. (2001). coincident and leading indicators for the euro area. the economic journal, 111(471), c62-c85. forni, m., hallin, m., lippi, m. & reichlin, l. (2003a). do financial variables help forecasting inflation and real activity in euro area? journal of monetary economics, 50(6), 12431255. forni, m., hallin, m., lippi, m. & reichlin, l. (2003b). the generalized dynamic factor model: one sided estimation and forecasting, mimeo. forni, m., hallin, m., lippi, m. & reichlin, l. (2005). the generalized dynamic factor model: one-sided estimation and forecasting, journal of the american statistical association, 100(471), 830-840. forni, m., giannone, d., lippi, m., & reichlin, l. (2009). opening the black box: identifying shocks and propagation mechanisms in var and factor models. econometric theory, 25(1319), u1347. giannone, d., reichlin, l., & sala, l. (2004). monetary policy in real time. nber macroeconomics annual, 19, 161-200. giannone, d., reichlin, l., & sala, l. (2006). vars, common factors and the empirical validation of equilibrium business cycle models. journal of econometrics, 132(1), 257279. giannone, d., reichlin, l., & small, d. (2008). nowcasting: the real-time informational content of macroeconomic data. journal of monetary economics, 55(4), 665-676. journal of applied economics and business studies, volume. 4, issue 2 (2020) 141-162 https://doi.org/10.34260/jaebs.427 161 giacomini, r., & white, h. (2006). tests of conditional predictive ability. econometrica, 74(6), 1545-1578.grossman, s. j. & stiglitz, j. e. (1980).on the impossibility of informationally efficient markets, the american economic review, 70(3), 393–408. gupta, r., & basu, p. k. (2007). weak form efficiency in indian stock markets. international business & economics research journal (iber), 6(3).. hallin, m., & lippi, m. (2013). factor models in high-dimensional time series—a timedomain approach. stochastic processes and their applications, 123(7), 2678-2695. humpe, a., & macmillan, p. (2009). can macroeconomic variables explain long-term stock market movements? a comparison of the us and japan. applied financial economics, 19(2), 111-119.. ince, h., & trafalis, t. b. (2007). kernel principal component analysis and support vector machines for stock price prediction. iie transactions, 39(6), 629-637. jensen, m. c. (1978). some anomalous evidence regarding market efficiency. journal of financial economics, 6(2/3), 95-101.kabundi, a., (2004). estimation of economic growth using business survey data. international monetary fund, 22working paper no. 04/69. kemp, a. g. & reid, g. c. (1971). the random walk hypothesis and the recent behavior of equity prices in britain, economica, 38(149), 28–51. lawley, d. n. & maxwell, a. e. (1971).factor analysis as a statistical method, 2nd ed., london: butterworth and co. leroy, s. f. & porter, r. d. (1981).the present-value relation: tests based on implied variance bounds, econometrica, 49(3), 555–574. lee, c.-c., lee, j.-d. & lee, c.-c. (2010). stock prices and the efficient market hypothesis: evidence from a panel stationary test with structural breaks, japan and the world economy, 22(1), 49–58. lehmann, b. n. (1990). fads, martingales, and market efficiency, the quarterly journal of economics, 105 (1), 1-28. lo, a. w. & mackinlay, a. c. (1988). stock market prices do not follow random walks: evidence from a simple specification test, the review of financial studies, 1(1), 41– 66. lo, a. w. & mackinlay, a. c., (2011). a non-random walk down wall street, princeton: princeton university press. ludvigson, s. & ng, s. (2007). the empirical risk return relation: a factor analysis approach. journal of financial economics, 83(1), 171–222. marcellino, m., stock, j. h. & watson, m. (2003). macroeconomic forecasting in the euro area: country specific versus euro wide information. european economic review, 47(1), 1–18. paye, b. s. (2012). ‘déjà vol’: predictive regressions for aggregate stock market volatility using macroeconomic variables. journal of financial economics, 106(3), 527-546. laila taskeen qazi, atta ur rahman, shahid ali & sohail alam 162 pesaran, m. h. & timmermann, a. (1995). predictability of stock returns: robustness and economic significance, the journal of finance, 50(4), 1201-1228. qazi, l. t., rahman, a. u., & gul, s. (2015). which pairs of stocks should we trade? selection of pairs for statistical arbitrage and pairs trading in karachi stock exchange. the pakistan development review, 215-244. reichlin, l. (2003). factor models in large cross sections of time series. in: s. t. m. dewatripoint and l. p. hansen (eds.): advances in economics and econometrics: theory and applications, vol. 111, 8thworld congress of the econometric society. cambridge: cambridge university press. samuelson, p. a. (1965). proof that properly anticipated prices fluctuate randomly, industrial management review, 6(2), 41–49. schumacher, c. & dreger, c. (2004). estimating large scale factor models for economic activity in germany; do they outperform simpler models? journal of national economy and statistics, 224(6), 731-750. schumacher, c. & breitung, j. (2006). real-time forecasting of gdp based on a large factor model with monthly and quarterly data, schumacher, c. (2007). forecasting german gdp using alternative factor models based on large datasets. journal of forecasting, 26(4), 271-302. stock, j. h., & w watson, m. (2003). forecasting output and inflation: the role of asset prices. journal of economic literature, 41(3), 788-829. stock, j. h. & watson, m. (1998). diffusion indexes. nber working paper 6702. stiglitz, j. e. (1981). the allocation role of the stock market: pareto optimality and competition, the journal of finance, 36(2), 235–251. stock, j. h., & watson, m. w. (2006). forecasting with many predictors. handbook of economic forecasting, 1, 515-554. stock, j. h., & watson, m. w. (2002). macroeconomic forecasting using diffusion indexes. journal of business & economic statistics, 20(2), 147-162. stock, j. h. & watson, m. w. (2005).implications of dynamic factor models for var analysis. nber working paper, 11467. journal of applied economics and business studies, volume. 4, issue 2 (2020) 181-204 https://doi.org/10.34260/jaebs.429 181 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x a quest for viable economic model for peshawar pakistan using etzkowitz conceptual framework of triple helix model romy khan1, sajjad khan2 1 institute of management sciences peshawar, pakistan 2 institute of management studies, university of peshawar, pakistan abstract the study aims to explore existing university-industry-government (u-i-g) linkages in peshawar, pakistan and study the factors that determine these linkages. furthermore, the research examines the extent to which university-industry-government collaboration can address regional industrial problems and promote economic growth. hence, the quest is to seek a viable economic model that not only enhances innovation in the region but also improves industrial competitiveness. therefore, etzkowitz’s triple helix model, based on academia-industry-government collaboration, provides the intellectual context for this research. the study uses semi-structured interviews and a narrative approach; with descriptive and analytical approaches to investigate the nature of university-industry and government linkages in peshawar. a single case study approach is employed, where entrepreneurship development centre (edc) at institute of management sciences peshawar (imsciences) was selected for research purpose. this is because the centre had already initiated u-i-g linkages to some extent in peshawar. research findings suggest that university-industry-government linkages are too weak. the triple helix model, which is considered a suitable conceptual framework for regional development (etzkowitz and ranga, 2010) needs structural changes to make it work in a developing region like peshawar (dzisah and etzkowitz, 2008). the original spiral model of innovation worked well in the developed world because that industry hosts multi-nationals that can afford industry-academia joint ventures. their governments not only facilitated interactive networks but also designed such policies’ frameworks that supported high growth firms (hgfs) (mason and brown, 2013). whereas peshawar is not only a traditional and developing economy but also a war and crisis-ridden region, due to which it cannot attract multi-nationals headquarters. the findings of the research can be treated as an asset that can easily be reused by other developing region for knowledge transfer and economic development. keywords inclusive education, household-based education index, inequality coefficient, inclusion coefficient, socioeconomic determinants jel classification c00, c21, i24, i29 1. introduction the triple helix model, as envisaged by etzkowitz, (etzkowitz, and zhou, 2018) has emerged as a common thesis globally. it is believed enables numerous nations to enhance and romy khan and sajjad khan 182 accelerate their development form this is due to the fact that countries and regions believe that innovative environment and entrepreneurial culture can only be attained if they opt for this specific model. these regions’ common objective is to generate an academic spin-off with the help of university research groups and establish tri-lateral networks for economic development (afzal et al., 2018). therefore, the underlying model regards the university as a significant partner in knowledge-based economy and entrepreneurship (leydesdorff et al., 2017: etzkowitz and leydesdorff, 2000; jones et al., 2014) because knowledge firm cannot be produced by a single entrepreneur due to the lack of technical expertise. hence, the need for collective entrepreneurship, i.e. collaboration of individuals and related institutions emerges. this institutional collaboration for collective entrepreneurship takes place at three phases (etzkowitz, 2001). in the first phase, university-industry-government collaborates for innovation while retaining their traditional identity. thus, when the three helixes enter into reciprocal relationships to enhance the performance of each other, the first step towards the triple helix is taken. for example, industry-academia-government initiates dialogue on improving the local economy and for this purpose the three helixes take their concern responsibilities, such as that university helps in producing graduates relevant for the industry; government supporting new plant construction as well as financing industry research. industry may seek help from both the helixes in regard of starting a new cluster. in the second phase of triple helix, when the quest for intellectual capital becomes important, modification in the role and performance of university and other knowledge producing institutions gets priority (etzkowitz and dzisah, 2007). the more university r&d capabilities and training facilities are needed by industry, the more the influential role of academia emerges in the local economy superseding the dominance of industry and government laboratories. this is because university knowledge and research has a competitive advantage over other r&d institutions. in this phase, industry and governments get involved in establishing research centers and, in order to speed up academic research production, they provide additional funds and resources to academia (etzkowitz, 2001; 2002). once collaboration on a reciprocal basis is established among the three helixes for economic growth, while retaining their primary role and distinct identities, they take the role of each other to ensure innovation (etzkowitz and zhou, 2007). such role-taking comes in the form of university transmitting education to youth as well as spinning new venture (sarpong et al 2017) and doing industrial research this was previously the task of industry. hence, today’s model of innovation is different from its predecessors where statist models emphasized governments’ dominating and controlling industry and academia or the three spheres working independent and separate of each other as in laissez faire (etzkowitz, 2003). the new model of innovation evolves when interaction and interdependence among the three spheres is increased to the point that hybrid institutions, invention and innovation takes place (dzisah and etzkowitz, 2008). therefore, the new triple helix in current industrial society is perceived as a spiral model (figure 1) that is based on mutual and reciprocal public-private and journal of applied economics and business studies, volume. 4, issue 2 (2020) 181-204 https://doi.org/10.34260/jaebs.429 183 academic partnership for knowledge economy. here, hybrid institutions results in the form of r&d, e.g. u-i-g research consortia; science parks; business incubators; financial support institutions, such as venture capital firms; angel networks seed funds and technology transfer offices (etzkowitz, 2003). this paper presents the detailed analysis of mphil research work’s triple helix model, based on etzkowitz’s framework. the model is commonly perceived as a road towards innovation and economic development in the developed world. the purpose of the study was to search for a viable u-i-g model that could contribute to the economic growth of peshawar, pakistan. initially, economic models, i.e. porter’s diamond model, mode 1 and 2(kashani and zarghami, 2019) and the triple helix of university-industry-government (u-1-g) were explored. as academia’s role was observed in the above approaches, therefore university’s role from an evolutionary perspective was investigated; the purpose was to identify the factors that transformed the academic traditional role, i.e. teaching and scholarship into an entrepreneurial university. hence, study of the university transformation revealed that academia’s third mission, i.e. to generate new knowledge, emerged due to global tendencies towards knowledgebased economies (sarpong et al 2017). therefore, the pursuit of innovative knowledge led the us government to adopt mode 1 since 1945 to 1988 and then mode 2 (etzkowitz and leydesdorff, 1997, 2000) to achieve innovation in different sectors. hence, the us government funded basic academic research since 1945 until 1988 under mode 1 and multidisciplinary or trans-disciplinary research under mode 2 in the early and late twentieth century (etzkowitz, 1997). mode 2 was a pluralistic approach involving a "networked" innovation system (sampat and mowery, 2004). moreover, the academic crucial role in industry is also highlighted in porter’s model. it is because competitive advantage (porter, 1998) is achieved due to new knowledge generation which is possible due to academic research capabilities. however, where the porter model regards all factors involved in value chain equally important (gulbrandson, 1997; quoted in etzkowitz and leydesdorff, 2000), etzkowitz’s model focuses on industry-academia-government linkages for economic growth. as the focus of study was on u-i-g linkages in peshawar region, therefore the triple helix model, which in fact is the further development of the porter model, was considered for research. furthermore, peshawar major economic sectors consist of traditional small and medium enterprises. mncs are nominal in peshawar and exist either as franchises or branches. no official mnc headquarters have been established in peshawar due to its law and order situation. in such scenarios, the chances for the economic growth of peshawar can only be bright if an interactive institutional network approach is adopted. therefore, etzkowitz’s thesis of university-industry-government was selected for systematic study (tranfield et al., 2003) so that a more practical institutional set-up could be adopted for economic growth in peshawar, pakistan. a case study of the entrepreneurship development centre (edc); institute of management sciences (im sciences) peshawar, pakistan, is presented in a quest for a suitable romy khan and sajjad khan 184 academic partner that can collaborate with the local industry and government. to provide an in-depth view of the study, this article briefly discusses the triple helix model, the literature review findings and primary data results. the analysis of literature review discusses the success and issues of the triple helix model that are identified by empirical studies worldwide. the analysis of the primary data collected, presents a picture of how edc im sciences can contribute to economic development in the peshawar industry. since the focus of the study is etzkowitz’s triple helix model and its implications worldwide, an attempt was made to infer suitable measures from the model for peshawar, pakistan. in the end, conclusions are drawn from the theoretical and practical research conducted figure 1: the spiral model of university-industry-government interaction for innovation source: etzkowitz (2007), http://ethiopiantriplehelixassociation.org 2. the three institutions in the spiral model the new model of innovation evolves when interaction and interdependence among the three spheres is increased to the point that hybrid institutions, invention and innovation takes place (dzisah and etzkowitz, 2008). therefore, the new triple helix in current industrial society is perceived as a spiral model (figure1) that is based on mutual and reciprocal publicprivate and academic partnership for knowledge economy. 1. here hybrid institutions result in the form of r&d, e.g. u-i-g research consortia; science parks; business incubators; financial support institutions, such as venture journal of applied economics and business studies, volume. 4, issue 2 (2020) 181-204 https://doi.org/10.34260/jaebs.429 185 capital firms; angel networks; seed funds and technology transfer offices (etzkowitz, 2003). as discussed above, tri-lateral ties among academia-industry-government results in mutual benefits for each spiral (figure 2). for example, massachusetts institute of technology takes entrepreneurship as its academic mission for regional economic development, whereas the government acts as venture capital for academic spin-offs and finances industrial research centers at mit. furthermore, industry takes up the education role of academia when it gets involved in the training and development of its employees (etzkowitz, 2002). also, swedish new universities and regional colleges have made knowledge creation as an important part of their academic program that has led to the creation of science parks; research centers; joint student training projects for firm formation and entrepreneurship training programs. here, the swedish government act as a public venture capitalist that instigates and finances new firms based on new technology (etzkowitz, 2007). figure: 2 each spiral benefiting from u-i-g linkages source; www.mattchwierut.com from the above discussion, the new roles of academia, industry and government in the triple helix model are: romy khan and sajjad khan 186 2.1 government the government not only accepts academic role in industry but also in finance and support research centers at academia. it acts as a venture capitalist to provide the seed funding for technology firms. hence, it provides r&d funds to academia and establishes entrepreneurial universities in the region. furthermore, it enacts mutually agreed policies on patent and research commercialization; cultivates entrepreneurial culture in academia by funding enterprise events and formulating enterprise friendly policies for graduates (etzkowitz, 2007; 2003). hence the government not only facilitates university and industry interaction, such as patenting regulation (leišytė & fochler, 2018) it has a strong role in the development of triple helix model (yoda and kuwashima, 2020). 2.2 academic the academic role comes in the form of establishment of entrepreneurship research centers that produce knowledge based spin-offs and conduct industry specific research. the centers in return provide consultancies to industry; commercialize research and provide training and development facilities to industry. furthermore, it acts as a help desk for industry; introduces entrepreneurship curriculum in education that produces industry specific graduates. hence, these centers cultivate entrepreneurial culture; generates hybrid institutions (etzkowitz and ranga, 2010), such as technology and business incubators, science parks, r&d and mentors entrepreneurs (etzkowitz, 2007). 2.3 industry industry funds industry specific research centers at academia accepts and recognizes academia’s potential for industrial collaboration and designs patent laws agreed by academia. moreover, industry benefits from academic research commercialization; human resource development and academic expertise for industrial solutions. u-i-g linkages are mutually beneficial for all; therefore, industry accommodates academic spin-offs; collaborates with government to establish science parks and incubators at academia. it also mentors graduates on business issues; sharing their real life business stories with them; arranging talk shows, open discussions and seminars on business opportunities as well as giving internships to students at industry for practical exposure other industry contribution required by the model (etzkowitz, 2007). 3. institutional theory and spiral model of innovation as the new spiral model is set up on institutional theory therefore university-industrygovernment relations are based on institutional network (leydesdorff et al 2017). hence reciprocal relationship among the three institutions can only emerge if planned, structured and substantial plan for their cooperation is developed. moreover, if institutional barriers, i.e. resistance to change exist among the three institutions, long term cooperation among the three cannot be form (brundin et al., 2008). thus, hybrid institutions (mentioned above) cannot be journal of applied economics and business studies, volume. 4, issue 2 (2020) 181-204 https://doi.org/10.34260/jaebs.429 187 developed. furthermore, people from academia-industry and government require welldeveloped rules and regulations for routine activities. a proper feedback process should be developed that can put pressure on the three institutions to meet their mutual goals. the institutions involved need to synchronize their activities and identify their mutual objectives. for example, if government is only interested in policy formulation while ignoring implementation, industry doesn’t show interest for mutual cooperation while academia research is not benefiting industry, triple helix is likely to fail. institutional barriers, such as diverse organizational goals and objectives; structure and resistance to change might hamper the process of economic growth and innovation (brundin et al., 2008). furthermore, if the three institutions are subject to mimetic, coercive and normative pressure; triple helix from an institutional perspective cannot foster economic growth and innovation (dimaggio and powell, 1983; brundin et al., 2008). university, industry and government in the triple helix model are forced to resemble each other in their institutional set-up; hence, the model leads to isomorphism (dimaggio and powell, 1983). however, academia and government are subject to institutional isomorphism, while industry is forced to competitive isomorphism (dimaggio and powell, 1983; brundin et al., 2008). in this regard, when taking the role of each other in the new spiral model of innovation, the three institutions are required to adopt new beliefs, goals and objectives as well as accept new environment pressure, i.e. technical and institutional control (meyer and scott, 1991; brundin et al., 2008). for example, academia which is not subject to competitive isomorphism has to compete in the new institutional set-up by taking advantage of its research capabilities. the government can utilize its regulatory pressure on the two institutions to meet their set goals. the industry, controlled by technical pressure, should comply with institutional pressure as well to foster economic growth and innovation (brundin et al., 2008). hence, all the helices in the model have common areas that they can mutually develop and strengthen by collaborating with each other without compromising on their independent domain. therefore, they have to recognize their mutual interests and design policies that can ensure their long-term collaboration for economic growth. even though the three spheres have a common interest and have a valid reason to integrate and support each other, there are still issues involved that create barriers in bringing academia, government and industry to a platform from where they can materialize the true nature of triple helix model. these issues were identified during the systematic literature review (macpherson and holt, 2007: tranfield, 2003) that was conducted to check the practical implications or validity of the triple helix model in peshawar, pakistan. 4. literature review the literature review explored in-depth academia-industry-government collaboration in factor-driven (developing); innovation-driven (highly developed) and efficiency-driven (newly romy khan and sajjad khan 188 developed) countries (jones et al., 2014). a total seventy-eight research articles on the triple helix model, published during 2000-2020, were selected for the literature review on the basis of their research rigour. finally, four themes were derived: triple helix and r&d; triple helix and innovation; triple helix and economic development, triple helix and industrial growth. this review was restricted to publish peer reviews; academic articles held within the following databases: isi web of knowledge; business source premier; science direct; scopus and google scholar. these were chosen from amongst others as providing the largest number of returns using a basic keyword search of the triple helix model and developed* developing* newly industrialised countries* and empirical studies. each database was interrogated by the search strings listed above. research interest was limited from the years 2000 to 2020. titles, keywords and abstracts that were published during year 2000 to 2020 were searched, where more than 150 studies were retrieved and exclusion criteria were included in order to refine the search. for example, studies on medical sciences were not included since the concept of triple helix carries other meaning in medical sciences. therefore, inclusion criteria were limited to social sciences, business studies, and computer science. the total number of potentially relevant studies retrieved using search strings was 150. these were exported to refworks, a referencing database where they were further reviewed against the inclusion and exclusion criteria in using key word, searches, year of publication and title analysis. also, duplicate studies were removed. at this stage, a thorough review of the abstracts alone was conducted and the articles that were relevant to the year of publication, title, search strings were selected for review (macpherson and holt, 2007). 5. analysis of literature review analysis of systematic literature review revealed that due to academic r&d potential, industry and government has benefited in innovation, industrial and economic growth worldwide. however, research found that issues do exist in regions where triple helix is not practiced in its true sense. the main issues confronted worldwide while practicing the triple helix model are: 5.1 external funding financial grants are essential for academic r&d activities (acosta et al., 2009); therefore, the government is required to develop a mechanism where the universities can receive research grants without any vested interest. because research funds received from external sources by academic research councils come with strings attached, this affects research practices in the developing and developed world alike. for example, in the us, etzkowitz (2003) and sweden’s (benner and sandstrom, 2000) external research funds directly affect norms of their research councils and influenced research as a whole. hence, the government is required to allocate industrial-academia research funds through fiscal budgets and on a continuous basis without any pre-determined consequences so that unbiased results can be attained. journal of applied economics and business studies, volume. 4, issue 2 (2020) 181-204 https://doi.org/10.34260/jaebs.429 189 5.2 policy level support countries where political and economic conditions (nwagwu, 2008) are not stable, such as nigeria or policy level support is not provided (langford et al., 2006); the innovation process is either slow or cannot take place. policy-level support includes governments providing r&d funds to academics; establishing research centres (shapiro, 2007; boardman and corley, 2008; boardman, 2009); strengthening inter-institutional linkages between academia and industry; design clear, flexible and consistent policies for r&d activities in their regions. research found that countries, which do not meet these basic requirements of the triple helix model, the process of innovation and industrial growth is slow in their regions. for example, malaysia, due to inflexible, rigid, vague and inconsistent government policies, razak and saad (2007), and nigeria’s (nwagwu, 2008) unstable economic conditions, institutional linkages for economic growth cannot be developed. the countries which want to achieve innovation and economic growth need to follow the precedents of the us; mexico (electronic cluster) (vargas, 2010) and west germany (mueller et al, 2005) where high-tech cluster development is possible because their governments have sincerely supported the integration of academia and industry. whereas, portuguese progress on the creative use of technology (de castro et al., 2000) and lithuanian high-tech development (chlivickas et al., 2009) is slow because government support is not present in these regions. 5.3 foreign research collaboration the countries where there is less trust in local academic research expertise, means there are chances that the local industry and government will turn towards foreign research collaborations. in such circumstances, universities’ expertise is not employed in local industrial research. for example, japan (sun and negishi, 2010); thailand (liefner and schiller, 2008) and malaysian industry (razak and saad, 2007) trust in foreign research or mnc’s expertise for quick industrial solutions has weakened academic role in industrial research. unless regional governments and their industries recognise local academic capabilities, such as research, teaching and technology transfer potential, academia cannot play a strong role in innovation and economic development. 5.4 industry and academia’s willingness for innovation innovation largely depends on industry and academia’s willingness to collaborate i.e. if industry is willing to collaborate with academia (acosta et al., 2009), innovation-based research can be carried out. whereas countries where industry or academia show lack of interest in innovation or are reluctant to collaborate cannot achieve success. for example, due to malaysian (razak and saad, 2007) and australian (gunasekara, 2004, 2006) industries’ lack of willingness for technological innovation, strong and effective university-industry interaction could not be developed in these regions. however, the us’s leading position in the bio-tech sector; information technology and new media is possible because of the active contributions of academia in the knowledge economy. the reason why the us achieves this success in these sectors is: 1) government support for involving academia in the innovation process of the romy khan and sajjad khan 190 country; 2) a strong academic desire for industrial collaboration (mayer, 2006). hence, these factors have led to the establishment of research centres in portland and washington. once industry, academia and government realise the mutual gains by collaborating with each other, there are chances that each helix will happily support each other in r&d, innovation and economic growth. the industry needs to realise that by integrating with academia it can save time and money on in-house r&d and will focus more on business while academia can gain financial help from industry and the state can achieve economic growth and a competitive edge. 5.5 research commercialization and patent rights research commercialization, intellectual property rights and patent policies are not clearly designed in the developed and developing world. due to which, confusion over ownership exists between research partners or innovation groups. these common issues are found in finnish universities and industry (tuunainen, 2002); argentinean public research organizations and firms (arza and lopaz, 2011); european universities (acosta et al., 2009) and malaysian universities and industry (razak and saad, 2007). to overcome such issues, a more conscious approach by the government is required while designing commercialization and patent policies and procedures in these regions. 5.6 other in developing countries, especially malaysia (razak and saad, 2007), universities lack industry specific research expertise and equipment; timely solutions to industry problems by academia, i.e. academic staff’s full-time research commitment for industry problems and the malaysian government’s inflexible, rigid, vague and inconsistent policies creates barriers for successfully implementing the triple helix model. in european regions, triple helix cannot take roots due to the absence of a homogeneous legal system that protects industry and university property rights (acosta et al., 2009). figure 3: thematic findings of the literature review diagram created by author literature review's thematic findings r&d issues lack of policy level support industries' lack of trust on local research potential industryacademia's willingness for innovation patent rights and research commerciali s-ation issues other journal of applied economics and business studies, volume. 4, issue 2 (2020) 181-204 https://doi.org/10.34260/jaebs.429 191 6. validity of the model in peshawar the triple helix model, which is considered a suitable conceptual framework for regional development (etzkowitz and ranga, 2010) needs structural changes to make it work in a developing region like peshawar (dzisah and etzkowitz, 2008). the original spiral model of innovation (figure 1, above) worked well in the developed world because that industry hosts multi-nationals that can afford industry-academia joint ventures. their governments not only facilitated interactive networks but also designed such policies’ frameworks that supported high growth firms (hgfs) (mason and brown, 2013). these hgfs not only speed up productivity growth but also create new jobs; increase innovation and promote business internationalization (mason and brown, 2013). peshawar is not only a traditional and developing economy but also a war and crisis-ridden region, due to which it cannot attract multi-nationals headquarters. therefore, it has to depend on its natural resources and local talent or human capital to develop its economy. once peshawar’s local industry starts developing, it will ultimately lead to innovation and increased productivity (reinert, 2007). as economic and political conditions do not fit triple helix’s requirements, it is suggested that the model should be modified to fit the local sub-dynamics (dzisah and etzkowitz, 2008). the modified model of innovation is provided in figure 4. 7. research methods as recommended above in the policy level measures, research centres can promote innovation and knowledge transfer through r&d activities. therefore, the role of the entrepreneurship development centre’s (edc) im sciences peshawar, pakistan is taken as a case study for industry-academia linkages for economic growth in the local economy. for this purpose primary data was collected through researchers personal account and semi-structured interviews taken from edc’s current coordinator, regional head of small and medium enterprise development authority (smeda) and vice president, women chamber of commerce peshawar, khyber pakhtunkhwa (kpk). 7.1 population and the sample the units of analysis (benbasat et al., 1987) of the study comprise of three groups; since it is a case study, from the academia entrepreneurship development centre (edc) im sciences was selected to investigate how an entrepreneurship centre can contribute to regional economy under the framework of the triple helix model. hence, edc’s first coordinator’s personal account is considered while the second coordinator was selected for in-depth interview; vice president of women chamber of commerce and industry of peshawar was selected from the industry and smeda’s regional head in peshawar was chosen as the third party who is actively involved in the process. the initial contact was made with the respondents through telephone calls and a request for informal interviews was made. the respondents agreed with the research objectives and its importance for the local economy was explained. through a telephone conversation, the participants were informed about timescale; data collection romy khan and sajjad khan 192 technique and the ethical considerations of the study. dates were agreed upon for the semistructured interviews and written confirmation was sent. 7.2 data analysis hence, to explore first-hand accounts of the respondents involved, i employed in-depth semi-structured interviews (kajornboon, 2005). however, the data collected from interviews was a challenging task to interpret. initially, an inductive approach (gray, 2013) was used to familiarize with the data collected (roper and shapir, 1999). as the data collected was based on written words, it needed to be coded in descriptive labels and then into analytical labels. hence, the coding process started with marking the data that could potentially address the research question. at the second level of simplification, the data was minimized to a manageable size. in the third phase, categories were identified and once descriptive labels were grouped into smaller sets, themes were developed and finally a conclusion was drawn (saldana, 2009). the centre, established in 2008, was meant to perform the same functions as reflected in the usa and uk enterprise research centers. unlike the centre for enterprise; the business school of manchester metropolitan university and enterprise centers worldwide (jones et al., 2008), edc im sciences could not pursue its goal of industry-academia integration due to the lack of funds. mmubs centre for enterprise, established in march 2001, secured funds from external sources, such as european regional development funds (erdf); esrc (£364,000) and soft structural european funds (jones et al., 2008), while edc im sciences could not find such avenues to secure funds. hence, in its initial stage it could not perform activities that were performed by enterprise centers in the developed world. again, mmubs enterprise centre bids for funding were a team effort that involved the director and external project expertise (jones et al., 2008) whilst the edc head had no other expertise available that could provide expertise for securing funds. other problems identified by primary data collected through semi-structured interviews from small and medium enterprise development authority (smeda) head’s current edc coordinator; the vice-president of the women chamber of commerce peshawar and from the founder of edc personal account, was the lack of institutional context (jones et al., 2008) and policy level support for the triple helix model in the peshawar region. a personal account from the edc coordinator revealed that it was never taken seriously by industry-academics and government. a policy-level support was absent which made the centre ineffective to perform its activities properly. although the coordinator was successful to revitalize edc, no further support was provided to her in the shape of r&d funds and training and the development of edc staff. unfortunately, edc could not generate income to meet its expenses due to the availability of funding agencies, such as esrc and european regional development funds (erdf) (jones et al., 2007). unlike the centre for enterprise, at manchester metropolitan university where they have generated £9m since 2001, from regional development funds and other sources (jones et al., 2008), edc could not explore such avenues because such funding agencies were not present in peshawar. however, one recent journal of applied economics and business studies, volume. 4, issue 2 (2020) 181-204 https://doi.org/10.34260/jaebs.429 193 development regarding securing funds is that edc can seek funds from oric (office for research innovation and commercialization) which were recently established by hec (higher education commission) at im sciences for giving research funds to its internal research centers. earlier, the only support edc received from its own institute was in the shape of student internees; coordinator allowances; separate offices; some funds for refreshments during awareness sessions and seminars arranged at the institute, and travelling allowances for networking. in such circumstances, materializing the triple helix model in the region was far from reality. it needed a lot of effort to bring the three helixes to one platform; convince them to collaborate as well as retain their independence in their own sphere. interviews with the chamber of commerce and industry and smeda’s head, equally show the need of edc’s role in economic development because the local industry of peshawar is based on traditional businesses that are not technology and innovation driven and the industry is full of uneducated business people who lack proper business skills and expertise. the poor conditions of the local economy demand an institutional approach towards economic growth. firstly, because local businesses are established with no proper market research; secondly, the business community, especially female entrepreneurs lack business know-how and have no knowledge about innovation and knowledge firms, which exist nationally and internationally. they have no avenue to look for proper advice and business support help. most of the business community has no proper direction and lack government and institutional support for starting and developing their businesses. apart from a few entrepreneurs who are educated or are running family businesses, the rest of the industry is ignorant of the international market trends – while, if trained properly, the existing businesses can be taken at international level. 8. enhanced role of smeda in peshawar, pakistan the above discussion of the literature review and the primary data found that local industry in peshawar, pakistan, needs academic help to grow. since small and medium enterprise development authority (smeda) has linkages with industry and is actively involved with local chambers of commerce (industry), it can act as an intermediary party (todeva, 2013) by integrating industry and edc on the one hand and convincing government to support this integration on the other hand. one of smeda’s tasks is to provide entrepreneurial trainings and consultancies to new entrepreneurs; it can outsource this activity to edc; by doing so, industry can get to know the potential of the centre. this will establish a long-lasting collaboration between industry and edc for research and training. as smeda already has strong relationships with financial institutions and donor agencies, it can introduce edc to these institutions for research funds. it can also bridge the gap between the government and edc by convincing government to fix research funds for the centre in the budget. one of edc’s achievements is the establishment of the incubation centre because innovation and academic spin-offs in peshawar are only possible if a technology and business incubation system is established. however, incubation is a new and complex process in this region and requires a proper organizational mechanism. edc can involve smeda in getting seed funds romy khan and sajjad khan 194 for technology spin-offs from government and other financial institutions. for incubators and science parks (etzkowitz et al., 2005; jones at al., 2014) to take roots in peshawar, pakistan, academia has to get involved in industrial research; establish joint ventures with the industry. the government needs to support and initiate such science and technology policies which can help in the development of the incubation industry in the region. the innovation process, as reflected in the research findings, requires policy-level support, i.e. establishment of r&d institutes (li,x, 2009; asheim and coenen, 2005); a full understanding of regional industrial demands and the designing of regional base innovation policies. therefore, government support in this regard is highly essential not only to ensure the triple helix and the entrepreneurial culture in peshawar but also to understand the needs of the sme sector, so that r&d institutes can address industry specific problems. therefore, smeda which already is involved with local sme’s can assist edc in getting the required information. since universities ‘have an important role in the economic development of any region (mayer, 2006; liefner and schiller, 2008), a well-defined framework for academia-industry engagement in economic growth should be outlined. from the empirical study, it is proposed that smeda can involve the policy makers in peshawar to promote the rise of academic capabilities, i.e. entrepreneurial training and education (jones et al., 2014) in accordance with the changing needs of its industry. higher education institutions should provide knowledge input in the form of graduates and publications; direct consultancy services and establish research centers that cater to needs of local sme’s. as discussed above, triple helix cannot be replicated in peshawar in its original form; therefore, a modified form of the model (see figure 4) should consist of smeda as an intermediary body that will act as a bridge among the three helixes for economic growth. academia government industry smeda figure 4: smeda as an intermediary body among u-i-g linkages diagram created by author as responses from the primary data confirmed that for an uneducated person to establish a knowledge base firm is not an easy task; therefore, smeda can utilise edc’s potential for the training and development of the novice entrepreneur. edc, while utilising phd staff’s expertise of im sciences for industrial research and consultancies, can address industrial needs. policy making for u-i-g collaboration economic interest edc: nurture innovation intermediary body journal of applied economics and business studies, volume. 4, issue 2 (2020) 181-204 https://doi.org/10.34260/jaebs.429 195 all the three institutes if integrated can develop a conducive environment for knowledge economy by producing knowledge firms in peshawar. thus, such institutional collaboration will result in the triple helix model that started in the us and in rest of the developed world. due to successful results of the model, it is gradually spreading in the developing world as well. since im sciences peshawar has the potential for developing the local economy, first it has a highly educated faculty specialising in business education; secondly, it has edc on its credit. under the model, governments can provide funds for industrial research to the centre. with funds, the centre can offer business specific trainings to potential and existing entrepreneurs. the centre, with the help of smeda, can do sector mapping of all the businesses in peshawar; identify their problems and needs and then can offer consultancies and business specific solutions to the industry. furthermore, an interview with smeda’s chief revealed that until a date is reached when governments overcome that they are unaware of the academic role in the economic development there is an issue; therefore, no deliberate policies have ever been developed at local and national level that will promote and facilitate industry-academia collaboration. however, the government now want to include academia by establishing entrepreneurship development centers at academia for training and industry research. therefore, the chances are that once the process of u-i-g interaction starts, edc im sciences which already exist (inspired by the uk; the usa and the european model to contribute to the local economy) will be taken as a pilot project. hence, once successful, can be replicated in other parts of the country for economic growth. one positive outcome of interviews with the smeda head and women chamber of commerce vice president was that both industry and government badly need academic expertise for industrial research; training and development of entrepreneurs and industrial specific human resources. hence, if the original model cannot be replicated in peshawar, pakistan, due to the absence of mncs and high-tech firms as the main actors, there are still chances that the model can be modified to suit the local dynamics. figure 2, shows that industry actors are mncs; the chamber of commerce; venture capital and sme spin-offs; academia actors are faculty, students and administration while government means local and federal. industry and government provide research funds and sponsor events like trade fares, business ideas, etc., these two also act as venture capitals for academic spin-offs. in return, academia churns out entrepreneurs; produces industry specific graduates; generates knowledge firms and trains human resource professionals for industry (etzkowitz, 2003; 2007). such ui-g collaboration in peshawar has yet to take roots despite the fact that edc staff have really worked hard to initiate the processes of such linkages. the reasons for poor collaboration and the gaps identified during primary data analysis are mentioned in the following table: romy khan and sajjad khan 196 table 4: gaps identified in u-i-g collaboration in peshawar, pakistan academia industry government • low level university-industry interaction exists in the region. • academia-industry mutual trust is lacking. • no formal framework based on clear goals and objectives for mutual collaboration is designed for edc-industry collaboration. • u-i-g mutual benefits in monetary term need to be identified • no mutual projects have been taken up by both. • no r&d funds have been provided to academia. • no government recognition of edc’s potential for industrial role exists. • edc and industry needs a catalyst/coordinator for bridging the two. • edc hybrid institutions, i.e. incubators need government support. • industry needs academic help for consultancies, skill and capacity building and business ideas. • industry needs a framework to collaborate with academia. • no formal linkages have been developed between industry and academia. • industry needs to identify its benefits in developing linkages with academia. • no joint research ventures between industry and academia have started. • industry need to collaborate with edc economic growth in the region. • the government never took edc’s crucial role in industry seriously. • no r&d funds were ever planned in the fiscal budget for edc. • no industry-academia joint venture ever funded. • not aware of the importance of hybrid institutions. • needs a catalyst/coordinator to bridge u-i-g. 9. critical analysis of the model in peshawar pakistan the study found that initial linear linkages that can be established between edc and industry can be service-oriented (schiller, 2006). that would include consultancies; training of human capital and technical expertise. producing industry specific graduates; providing interns and joint research activities can be another service provided by edc to the industry. however, to cultivate suitable grounds for the modified model, and develop linear modes among the three, smeda needs to bring the three helixes to the negotiating table for possible collaboration in the region. this process is not radical rather evolution-based (see figure 5), e.g. in the first phase smeda will introduce the three helices to each other and edc’s potential for local economy will be acknowledged by industry and government. in the second phase, smeda will initiate and facilitate dialogues among the three spheres for local economic growth. at this stage, common grounds for mutual relationships and collaboration will be established. in the third stage, smeda’s role as facilitator and coordinator will be acknowledged and a formal policy framework for future collaboration among the four partners will be designed. finally, an action plan for reciprocal relationships will be agreed. hence, smeda will act as an initiator, coordinator and enhancer (figure 6) of trio linkages. initially, mutual dialogues will focus on the uplift of the traditional sme sector; then gradually all the journal of applied economics and business studies, volume. 4, issue 2 (2020) 181-204 https://doi.org/10.34260/jaebs.429 197 helixes will design policies for high-tech research and knowledge firms that ultimately leads to innovative smes. 10. positive impacts of the triple helix model in peshawar pakistan once the model is developed and starts functioning, it will have the following implications for all the partners: implication of triple helix model for the government of peshawar the government until now was depending on smeda for industrial development in the region, while smeda lacked expertise that could train and guide the local entrepreneurs towards innovative businesses. furthermore, smeda had not enough time and resources to conduct business specific research and provide solution to the local economy. therefore, edc’s expertise can lessen smeda’s burden to some extent and can play an important role in the growth of local economy. the government by providing research funds to edc will not only identify industrial problems and solve them but also can minimise the business failure rate; therefore, successful businesses can result from academia-government collaboration. industrial research by edc will help the government to grow and flourish the traditional businesses in a more modern and innovative manner with which these businesses can be promoted at international level. edc im sciences with the help of smeda will establish an ‘idea village’, which will generate innovative business ideas that can help the local economy to develop more knowledge base firms. moreover, government should patronise edc for incubation and science parks that results in technology-based local businesses. the government should assist in bridging the gap between academia and industry (jones-evans et al., 1999) will begin a new phase of economic development where more confidence between the two will be developed. hence, the triple helix model taking charge of the local economy will enable governments to focus more on the law and order situation and will be relieved from focusing on the industrial problems. in short, the model will not only contribute to local economic development but will generate new jobs. the government, in this regard, will have less worries for accommodating fresh graduates. 10.1 implications of triple helix model for the industry the model will not only overcome institutional differences in the region but will help industry to address its problems through edc’s platform. edc will train new business entrants in the region by teaching them business specific skills. as new businesses will be research based, this will encourage innovative businesses to sprout up in peshawar. edc consultancies’ services will assist industry to overcome its hurdles in start-up and growth stages. edc training and development and r&d capabilities can help industry to save its r&d budget. hence, the industry will have more time for developing its businesses and will be relieved from financial burdens. romy khan and sajjad khan 198 10.2 implications for academia especially edc im sciences: u-i-g linkages will not only enhance edc’s role in peshawar’s economic development but will also increase its earnings through research funds and entrepreneurial trainings. moreover, edc’s image at government and industry level will be boosted on the basis of its research and staff capabilities. therefore, it can be a pioneer in new knowledge creation in the region as well as from its incubation centre and new firms will spin-off. figure 5: presents the three phases of evolution of u-i-g collaboration in peshawar pakistan figure 6: illustrates phase-wise development of the local economy by triple helix and smeda as the facilitator. diagrams created by author 11. conclusion this research acknowledges the challenges and difficulties in bridging the gap among university-industry-academia (jones-evans et al., 1999) in peshawar, pakistan, for economic growth. the research has shown that there is no awareness among the three helixes about the possible collaboration that can result in the economic uplift of the region. the government role as a venture capitalist (etzkowitz, 2003) and facilitator of the u-i-g linkages has not taken roots. industry is not aware, neither willing to accept academia’s new role in the economy. the edc needs governments and im sciences to provide funds for research and its routine introduce the partners inititiate dialogues for mutual collaborations design policy framwork for future action • need for viable economic model is realised • actors in the model, e.g. academic government and industry are identified • edc, as potential partners, is acknowledged smeda as facilitator to bridge u-i-g collaboration • leads to local sme uplift • focus on existing industrial problems and needs • research focus on existing industrial problems • avenues for future collaboration for economic growth are explored reciprocal relationships developed among the helixes • partnership developed for regional economic development • focus diverted towards high growth firms • birth of innovative and technology based smes • competitive industry is developed • create jobs • business internationalisation policy emerges available u-i-g economic model is developed for innovation journal of applied economics and business studies, volume. 4, issue 2 (2020) 181-204 https://doi.org/10.34260/jaebs.429 199 activities. one of the main issues identified during primary data was that edc’s staff are not well-equipped with the latest trends, information and knowledge required for strong u-i-g collaboration. hence, edc staff requires capacity building to run the centre; therefore, the institute and government should arrange for the training and development of edc staff locally and internationally. this will update edc staff with latest methodologies and knowledge that is practiced worldwide in u-i-g linkages. furthermore, training and development of the staff will enable them to respond to industrial needs more effectively, because local industry needs university services for r&d; technical assistance; training and development and innovative ideas for high growth businesses in the region. hence, the human resource development of edc is strongly recommended by the study moreover the quest for a viable and valid economic model guided the research towards a systematic literature review of the triple helix model; however, the study found that although the model is practical in the west, it cannot achieve the same results in the developing region of peshawar. therefore, it is suggested that the model should be modified to make it valid and viable for the region (tveit, 2011). hence, the model should be extended to include an intermediary body (todeva, 2013), i.e. smeda in the peshawar case (as shown in figure 4 above) which will link edc (academia) with the local industry and government, to pool-in their common resources for economic growth in the region. the primary data informs that edc since inception to-date has always involved smeda to bring government officials and industry to its activities whether it was a business plan competition; an inauguration ceremony; seminars or university-industry discussion forums. this was because edc could not involve industry and government in such activities without smeda’s help. the reason was that previously neither industry nor governments were aware that edc could be their common and equal partner in the economic development of the region. whereas, smeda already had close linkages with government and industry; therefore, it can bring industry-academia and government to the negotiating table where the three can initiate dialogues for local economic growth; design a policy framework for their mutual collaboration. as there is a demand for academic expertise in the local industry, edc can initiate im sciences students’ ventures in idea commercialization and knowledge production. moreover, universityindustry linkage processes have already started through edc and smeda: it only needs a proper mechanistic and network approach to be successful. . furthermore, in future research, the scope should be extended to engineering universities and other management colleges in the region so that these institutes’ policies towards industry-academic linkages should be investigated. this will enable future researchers to be able to adopt a more holistic view of the linkages which exist in peshawar in the light of which the triple helix model can be materialized in the region. romy khan and sajjad khan 200 references acosta, m., coronado, d., león, m.d. & martinez, m.á. 2009, the production of university technological knowledge in european regions: evidence from patent data, vereinigtes königreich; united kingdom. afzal, m.n.i., sulong, r.s., dutta, s. & mansur, k. 2018, "an investigation on triple helix model and national innovation systems: the case of malaysia", asian journal of innovation and entrepreneurship, vol. 3, no. 03, pp. 299-313. arza, v. & lópez, a. 2011, "firms’ linkages with public research organisations in argentina: drivers, perceptions and behaviours", technovation, vol. 31, no. 8, pp. 384-400. asheim, b.t. & coenen, l. 2005, "knowledge bases and regional innovation systems: comparing nordic clusters", research policy, vol. 34, no. 8, pp. 1173-1190. benbasat, i., goldstein, d.k. & mead, m. 1987, "the case research strategy in studies of information systems", mis quarterly, , pp. 369-386. benner, m. & sandström, u. 2000, "institutionalizing the triple helix: research funding and norms in the academic system", research policy, vol. 29, no. 2, pp. 291-301. boardman, p.c. 2009, "government centrality to university–industry interactions: university research centers and the industry involvement of academic researchers", research policy, vol. 38, no. 10, pp. 1505-1516. boardman, p.c. & corley, e.a. 2008, "university research centers and the composition of research collaborations", research policy, vol. 37, no. 5, pp. 900-913. brundin, e., wigren, c., isaacs, e., friedrich, c. & visser, k. 2008, "triple helix networks in a multicultural context: triggers and barriers for fostering growth and sustainability", journal of developmental entrepreneurship, vol. 13, no. 1, pp. 77. chlivickas, e., petrauskaite, n. & ambrusevic, n. 2009, "leading priorities for development of the high technologies market", journal of business economics and management, vol. 10, no. 4, pp. 321-328. de castro, e.a., rodrigues, c., esteves, c. & da rosa pires, a. 2000, "the triple helix model as a motor for the creative use of telematics", research policy, vol. 29, no. 2, pp. 193203. dimaggio, p.j. & powell, w.w. 1983, "the iron cage revisited: institutional isomorphism and collective rationality in organizational fields", american sociological review, , pp. 147-160. dzisah, j. & etzkowitz, h. 2008, "triple helix circulation: the heart of innovation and development", international journal of technology management & sustainable development, vol. 7, no. 2, pp. 101-115. etzkowitz, h. 2007, "university-industry-government: the triple helix model of innovation", eoq congresses proceedings. 51st eoq congress, pp. 22. journal of applied economics and business studies, volume. 4, issue 2 (2020) 181-204 https://doi.org/10.34260/jaebs.429 201 etzkowitz, h. 2003, "innovation in innovation: the triple helix of university-industrygovernment relations", social science information, vol. 42, no. 3, pp. 293-337. etzkowitz, h. 2001, "the second academic revolution and the rise of entrepreneurial science", technology and society magazine, ieee, vol. 20, no. 2, pp. 18-29. etzkowitz, h. 1997, "the entrepreneurial university and the emergence of democratic corporatism", 1997), universities and the global knowledge economy: a triple helix of university–industry–government relations.london: pinter, , pp. 141-152. etzkowitz, h. & dzisah, j. 2007, "the triple helix of innovation: towards a university-led development strategy for africa", atdf journal, vol. 4, no. 2, pp. 3-10. etzkowitz, h. & leydesdorff, l. 2000, "the dynamics of innovation: from national systems and “mode 2” to a triple helix of university–industry–government relations", research policy, vol. 29, no. 2, pp. 109-123. etzkowitz, h. & leydesdorff, l. 1997, "universities in the global knowledge economy: a triple helix of academic-industry-government relations", london: cassell, . etzkowitz, h. & zhou, c. 2007, "regional innovation initiator: the entrepreneurial university in various triple helix models", triple helix 6th conference theme paper, singapore. etzkowitz, h. 2003, "research groups as 'quasi-firms': the invention of the entrepreneurial university", research policy, vol. 32, no. 1, pp. 109-121. etzkowitz, h. 2002, mit and the rise of entrepreneurial science [electronic book] / henry etzkowitz, london ; routledge, 2002. etzkowitz, h. 2002, "networks of innovation: science, technology and development in the triple helix era", international journal of technology management & sustainable development, vol. 1, no. 1, pp. 7. etzkowitz, h., de mello, j.m.c. & almeida, m. 2005, "towards “meta-innovation” in brazil: the evolution of the incubator and the emergence of a triple helix", research policy, vol. 34, no. 4, pp. 411-424. etzkowitz, h. & dzisah, j. 2008, "rethinking development: circulation in the triple helix", technology analysis & strategic management, vol. 20, no. 6, pp. 653-666. gray, d.e. 2013, doing research in the real world, sage. gulbrandsen, 1. (1997), "· universities and industrial competitive advantage", • h. etzkowitz, l. leydesdorff (eds.), universities and the global knowledge economy, pinter, london , pp. 121–131, . gunasekara, c. 2006, "reframing the role of universities in the development of regional innovation systems", journal of technology transfer, vol. 31, no. 1, pp. 101. romy khan and sajjad khan 202 gunasekara, c. 2004, "the third role of australian universities in human capital formation", journal of higher education policy & management, vol. 26, no. 3, pp. 329-343. jones, o. & gatrell, c. 2014, "editorial: the future of writing and reviewing for ijmr", international journal of management reviews, vol. 16, no. 3, pp. 249-264. jones, o., macpherson, a., thorpe, r. & ghecham, a. 2007, "the evolution of business knowledge in smes: conceptualizing strategic space", strategic change, vol. 16, no. 6, pp. 281-294. jones, o., macpherson, a. & woollard, d. 2008, "entrepreneurial ventures in higher education analysing organizational growth", international small business journal, vol. 26, no. 6, pp. 683-708. jones, o., macpherson, a. & jayawarna, d. 2014, resourcing the start-up business : creating dynamic entrepreneurial learning capabilities / oswald jones, allan macpherson and dilani jayawarna, abingdon, oxon : routledge, 2014. kajornboon, a.b. 2005, "using interviews as research instruments", e-journal for research teachers, vol. 2, no. 1. langford, c.h. ( 1,2 ), josty, p.(.1.)., hall, j.(.2.)., matos, s.(.2.). & jacobson, a.(.2.). 2006, "indicators and outcomes of canadian university research: proxies becoming goals?", research policy, vol. 35, no. 10, pp. 1586-1598. leišytė, l. & fochler, m. 2018, topical collection of the triple helix journal: agents of change in university-industry-government-society relationships, . leydesdorff, l., etzkowitz, h., ivanova, i. & meyer, m. 2017, "the measurement of synergy in innovation systems: redundancy generation in a triple helix of university-industrygovernment relations", springer handbook of science and technology indicators.heidelberg: springer, . li, x. 2009, "china's regional innovation capacity in transition: an empirical approach", research policy, vol. 38, no. 2, pp. 338-357. liefner, i.(.1.). & schiller, d.(.2.). 2008, "academic capabilities in developing countries-a conceptual framework with empirical illustrations from thailand", research policy, vol. 37, no. 2, pp. 276-293. macpherson, a. & holt, r. 2007, "knowledge, learning and small firm growth: a systematic review of the evidence", research policy, vol. 36, no. 2, pp. 172-192. macpherson, a. & holt, r. 2007, "knowledge, learning and small firm growth: a systematic review of the evidence", research policy, vol. 36, no. 2, pp. 172-192. mason, c. & brown, r. 2013, "entrepreneurial ecosystems and growth oriented entrepreneurship", background paper for the international workshop on entrepreneurial ecosystems and growth oriented entrepreneurship. journal of applied economics and business studies, volume. 4, issue 2 (2020) 181-204 https://doi.org/10.34260/jaebs.429 203 mayer, h. 2006, "what is the role of universities in high-tech economic development? the case of portland, oregon, and washington, dc", local economy (routledge), vol. 21, no. 3, pp. 292-315. meyer, j.w. & scott, w.r. 1991, "the organization of societal sectors: propositions and early evidence", the new institutionalism in organizational analysis, , pp. 108-140. nwagwu, w.e. 2008, "the nigerian university and the triple helix model of innovation systems: adjusting the wellhead", technology analysis and strategic management, vol. 20, no. -, pp. 683-696. porter, m.e. 1998, clusters and competition: new agendas for companies, governments, and institutions, harvard business school press boston. ranga, m. & etzkowitz, h. 2013, "triple helix systems: an analytical framework for innovation policy and practice in the knowledge society", industry and higher education, vol. 27, no. 4, pp. 237-262. razak, a.a. & saad, m. 2007, "the role of universities in the evolution of the triple helix culture of innovation network: the case of malaysia", international journal of technology management & sustainable development, vol. 6, no. 3, pp. 211-225. reinert, e.s. 2007, how rich countries got rich... and why poor countries stay poor, constable london. roper, j.m. & shapira, j. 1999, ethnography in nursing research, sage publications. saldana, j. 2009, "an introduction to codes and coding", the coding manual for qualitative researchers, , pp. 1-31. sampat, b.n. & mowery, d.c. 2004, "universities in national innovation systems", . sarpong, d., abdrazak, a., alexander, e. & meissner, d. 2017, "organizing practices of university, industry and government that facilitate (or impede) the transition to a hybrid triple helix model of innovation", technological forecasting and social change, vol. 123, pp. 142-152. schiller, d. 2006, "the potential to upgrade the thai innovation system by university‐industry linkages", asian journal of technology innovation, vol. 14, no. 2, pp. 67-91. shapiro, m. 2007, "the triple helix paradigm in korea: a test for new forms of capital", international journal of technology management & sustainable development, vol. 6, no. 3, pp. 171. souzanchi kashani, e. & zarghami, h.r. 2019, "the dynamics of university-industrygovernment relationships in nanoscience: investigating the triple-helix differences between iran and switzerland", technology analysis & strategic management, vol. 31, no. 7, pp. 817-832. steiber, a. & alange, s. 2013, the formation and growth of google: a firm-level triple helix perspective. romy khan and sajjad khan 204 sun, y. & negishi, m. 2010, "measuring the relationships among university, industry and other sectors in japan’s national innovation system: a comparison of new approaches with mutual information indicators", scientometrics, vol. 82, no. 3, pp. 677-685. todeva, e. 2013, "governance of innovation and intermediation in triple helix interactions", industry and higher education, vol. 27, no. 4, pp. 263-278. tranfield, d., denyer, d. & smart, p. 2003, "towards a methodology for developing evidence‐ informed management knowledge by means of systematic review", british journal of management, vol. 14, no. 3, pp. 207-222. tuunainen, j. 2002, "reconsidering the mode 2 and the triple helix: a critical comment based on a case study", science studies, vol. 15, no. 2, pp. 36. tveit, g.m. & webjørnsen, e.k. 2011, "university-industry interaction and its contribution to economic development in uganda", . vargas, m.i. 2010, "government influence and foreign direct investment: organizational learning in an electronics cluster", critical sociology, vol. 36, no. 4, pp. 537-553. yoda, n. & kuwashima, k. 2020, "triple helix of university–industry–government relations in japan: transitions of collaborations and interactions", journal of the knowledge economy, vol. 11, no. 3, pp. 1120-1144. journal of applied economics and business studies, volume. 4, issue 4 (2020) 135-160 https://doi.org/10.34260/jaebs.447 135 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x exploring the approaches to measure the value of household work: an experience from pakistan hamna nasir1*, ambreen fatima2 and shaista alam3 1 phd student at applied economics reseach center (aerc), university of karachi 2 associate professor at applied economics research center (aerc), university of karachi 3 associate professor at applied economics research center (aerc), university of karachi abstract the purpose of this research is to fill the gap in the knowledge by estimating the value of time spent on household work especially by women i.e study aims to assess the economic worth of time. specifically, the objective of this research is to estimate the value of time using two approaches, the one estimating opportunity cost and the other estimating the market price/value of household work. the value of time is estimated for different types of household activities. in addition, the study also assesses the determinants responsible for accelerating the opportunity cost and market value of time during the past three decades in pakistan as well critically study assess why the household is not diverting its human capital towards the paid employment though the value of time has increased substantially. the determinants mainly focus on personal and demographic characteristics as these factors are responsible to keep the individuals engaged in household activity. the determinants are explored using pooled ols. results show that the opportunity cost of household work performed by male and female worth around rs.26,082/week and rs.155,126/week; whereas the market price of male and female are estimated around rs.22,908/week and rs.63,813/week respectively. further, the results of this study confirm that the opportunity cost of all non-market activities is greater than the market value indicating that the work performed by individual at home(necessary to be performed) should be recognized as it has substantial economic worth. hence this research contributes to knowledge by enhancing the understanding of policymakers about the economic contribution made especially by women through household work – that often remain unaccountable. keywords labor economics, economics of gender, household economics, opportunity cost jel classification j01,j16, d13, j17 * hamnainfinite@gmail.com this paper is part of phd thesis under process. mailto:hamnainfinite@gmail.com hamna nasir, ambreen fatima and shaista alam 136 “in past, some of the resistance to further work on this subject came from believe that “invaluable” contribution of homemaker would be demanded by being assigning a monetary value. it is however the failure to assign a price for the services of the homemaker which has tended to convey the impression that they are valueless rather than priceless [marianne (1980) pp -387].” 1. introduction according to the labour force survey (lfs) of pakistan, 2017-18 on average nonmarket work hours of women in the rural region are 40 hours/week and in the urban region, it is around 43 hours/week, whereas the average non-market hours of male work in the rural region is 7 hours/week and in the urban region, it is 6 hours/week only. though becker (1965) emphasized estimating the opportunity cost of time. he noted that the time allocated to non-market work (household work) worth more (if its economic value is estimated) for economic welfare than market time. the argument placed by becker is that the substantial time per week allocated to household work, is hence considered as an investment for the economic and social welfare of the country. economists have made many valiant efforts to estimate the dollar value of nonmarket work (household work). the value of household work is usually not accounted for in the calculation of gdp nor is it taken into account when the income of the household is considered. the value is mainly not accounted for because of the measurement issues. this study highlights the measurement issue regarding the economic contribution of the non-market work made in pakistan. the purpose of this research is twofold to gain an understanding of the nature of non-market work performed by individuals in pakistan and to estimate the economic worth/value of that work. precisely study estimates the economic worth of activities that took place within the household. non-market work here refers to those domestic activities that are performed by women mainly and by men occasionally for which they receive no remuneration. a review of past studies shows that time allocated to non-market work is usually underestimated as the economic value of time is not easy to measure/calculate. for example, research on intrahousehold bargaining power assumed that time allocated to domestic work is a necessity, and the household members with the least power or low opportunity cost of time in the market forces to allocate their time to home production. the opportunity cost of housework for women is high therefore a large segment of the society (females) remained deprived and is not allowed to play its role properly in the labour market. a general assessment shows that domestic work is “undervalued, underpaid, unprotected and poorly regulated” and if done by women it is “unrecognized” in spite of the contributions made. their work is never considered in the national accounting system. it leads to prejudice against individuals who perform domestic chores which indicates that the work performed by them at home (necessary to be performed) should be recognized as it has substantial economic worth. the journal of applied economics and business studies, volume. 4, issue 4 (2020) 135-160 https://doi.org/10.34260/jaebs.447 137 activities for which people receive the compensation provide benefit to the economy, while the unpaid activities have remained economically meaningless. further, most of the international studies have focused on the economic value of household work, whereas studies in pakistan by sultana and nazli (1994) and saqib and arif (2012) have mainly focused on time allocated to paid market work and household work – not fully considering the economic worth of non-market activities performed within the household. according to past studies addressed by khan (2007), bhalotra & attfield (1998), naqvi et al. (2002) and syed et al (2013) household responsibilities, labor market discrimination, customs, and traditions are some of the factors which draw female participation from the paid market work. the only study for pakistan has been conducted by arshad (2008) who has investigated the market price of household activities. considering the arguments raised in earlier studies, the specific objective of this research is to estimate the opportunity cost of non-market work hours for a number of activities. these activities include: within household light construction work, such as mud plaster of roofs and walls, construction and repair of boundary walls, rooms, collection of firewood or cotton sticks for use as firewood for household consumption, bringing water, taking food from house to farm, sewing pieces of cloth or leather, knitting, embroidery, mat, rope making, ginning, spinning and weaving, shopping and marketing, washing, mending or pressing clothes, child care/care of adult, child education, house cleaning, and cooking. further, this study also makes an assessment of factors responsible for accelerating the opportunity cost of time allocated to nonmarket work by individuals in the past few years therefore critically study assess why the household is not diverting its human capital towards the paid employment though the value of time has increased substantially. in order to assess the determinants of opportunity cost and the market price of non-market work, personal and demographic characteristics are explored as these factors are responsible to keep the individuals engaged in household activity. personal characteristics include; age, experience, marital status, education attainment, while demographic characteristics include the number of children and adults, household income, regional and provincial dummies. in order to evaluate the above-stated objectives, this study extracted the data from the labor force survey (lfs) of pakistan for various years. further, to assess the opportunity cost and market price of household work a number of techniques are applied, whereas pooled ols is used for empirical assessment. this research hence fills the gap in the knowledge by exploring the time individuals spent on non-market work and its economic worth. moreover, it assesses the contribution made by an individual in the social development of the country through performing non-market work. further, the role of women in the economy is recognized by society only when they participate in the paid labor force, their non-market work remains hidden and unacknowledged and the double burden on women is often not hamna nasir, ambreen fatima and shaista alam 138 addressed. the importance of non-market work though have some recognition but it is limited to children and elder-care. this study raised some thought-provoking points by assessing the economic value of non-market work in pakistan. overall, the study recommends that the contribution of women's household work if highlighted in gdp and other measures of national wealth it would help in raising their status and promoting gender equality in pakistan – ultimately reducing poverty. further, to achieve the target set in sdg goal 5.4 this study recommends that awareness should be raised about women's contribution to national wealth – i.e. highlight the economic value of their work. this paper is organized as follows: section 2 provides a theoretical foundation; section 3 discusses the methodology; in section 4 data sources and model specification are discussed and in section 5 the opportunity cost and market price of non-market work are detail assessed by using a number of techniques and its determinants are also explored for assessment purpose. the last section contains the conclusion and policy implications. 2. theoretical foundations reid (1934) was the pioneer to estimate the monetary value of home production. since then many scholars have attempted to estimate the monetary value of household work with the aim to evaluate unpaid domestic and care work and integrate it with the system of national accounts the so-called “satellite accounts”. as a result, in the national economic accounts, the economic activities are now categorized as: (i) system of national accounts (sna) production activities. (ii) non-sna production activities (food preparation, childcare, elder care, making and care of textiles, upkeep of dwelling and surroundings, repairs and maintenance of dwelling and of household equipment, household management and shopping, gardening, and pet care) and unpaid work for the community (iii) non-economic activities some times called personal activities (physiological and recreational activities and self-education). reid (1934) wrote in her famous book on “economics of household production” that: “the integral part of our whole economic system is household. if we become aware of labor costs and productive activities which are necessary to maintain the present standards of living. unless this is done, we cannot rightly appraise the economic role of home-keeping women or act intelligently in matters concerning the gainful employment of married women. nor can we understand home problems and formulate satisfactory curricula of education for home and family life.” stoetzel (1948) and girard (1958), using time-based studies conducted in france for the year 1950, provided an improved knowledge of the amount of household work journal of applied economics and business studies, volume. 4, issue 4 (2020) 135-160 https://doi.org/10.34260/jaebs.447 139 which is performed by women engaged in paid labor and as full-time housewives. their detailed analysis provided evidence of the main factors which bear the volume of household work performed. the rational decisions made by individuals without constraints with respect to the number of hours spent on market work and if they have full knowledge of potential earnings than the individuals who are not engaged in market work would value their time at home as much or more than the opportunity cost i.e. alternative forgone. therefore, their potential earnings would provide a lower limit but would not measure the worth of home time, this could be higher and could fluctuate considerably while potential earnings remain the same or change in the opposite direction. the value of housework in terms of the number of hours spent presents certain advantages. for instance, it allows direct comparison between total quantities of work supplied in market and household sectors, the need for fictitious imputation of money values is obviated to the latter, as the unit of measure and hours of work are common in both (fourastit 1965). hawrylyshyn (1976) stated that the idea behind the valuation of household activities was “to replace not the mother – homemaker, but merely a portion of her services” (p.112). the notion stated that “the market cost method was closest to the concept of the dollar value of household services”. he concluded that valuation resulting from the estimation is around one-third of gnp on average. to value the household services various estimates were put forward by hawrylyshyn (1976). there exists wide variability in the results obtained, attributed to each approach. among the methods discussed, the highest value obtained is related to the opportunity cost approach estimated for women involved in paid employment, the lower value is obtained from the method based on the price of a single housekeeper and the least value is obtained from the method which is based on pricing individual services. family size, wife’s market-work status, and age of the youngest child are factors that affect very strongly the value of services performed. according to the results, the value of total household services must include the contribution of spouse and children, although the contribution of a wife is dominant. he also concluded that the valuation of the opportunity cost method has an upward biased relative to the market cost method. he argued that the reason for this upward bias was due to comparable levels of home production in separate households which would produce different values if the individuals in each household had different marketable wages. the report surveyed by hawrylyshyn attempts to value household services which neither improved the accuracy nor quelled the debate regarding appropriate methods that estimate the value of home production. hawrylyshyn (1977) considering the value of non-market time stated that “in equilibrium, the value of time spent at home equals its opportunity cost which clearly hamna nasir, ambreen fatima and shaista alam 140 is, its wage on the market”. the rational household therefore will apply the optimization rule by using the factor “time” to the point where its marginal product equals its price. therefore, households' own estimation is done for a non-market time at the margin which is equal to its market hourly wage. since the work of hawrylyshyn (1976 and 1977) measuring productive household activities in monetary terms become a field of study. however, it was argued by peterson (1978) that these methods overestimate home production whereas bell and taub (1982) argue that current valuation methods understate the true value of household production. further, the authors have emphasized over conceptualizing the household production method and various methodological approaches for its evaluation. their evaluation, based on time use surveys, induces thought on time division between different kinds of activities categories. the classification helps to clarify the concept of housework and also emphasizes the economic and social role of women inside and outside the market sector (szalai 1972 and michel 1978). ferber and birnbaum (1980) conclude that to value housework, the market cost approach is preferable to the opportunity cost approach and general household keeper’s wages are preferable over specialized wages as individuals enjoying specialized wages usually performed a variety of household tasks. quah (1986) explains that household production is a non-market activity and its estimation is difficult, therefore its benefits cannot be measured by survey or observation. to estimate how much production at home contributes to total production taking place in an economy one must compute it indirectly, though these attempts produce a myriad of quantitative results and none of these have shown promise as a direct and sensible approach to value household production. the two fundamental approaches are used to value household production (i) the opportunity cost approach emphasizes the time devoted to household production and a value is set for such production as an individual could earn income in the market, whereas it can be measured from services which can be purchased from the market and (ii) the market price approach household activities are valued as a “market price” of hiring someone who could perform these tasks. it is argued by murphy (1978) that both methods produce significantly different results. time devoted to household production refers to productive activities as it has an economic value, while the time devoted to biological and non-productive leisure activities are not valued. chiswick (1982) and wolf (1986) indicated that the market cost method is preferable to the opportunity cost method since the latter method is over biased. whereas, it was observed by murphy (1978) that as compared to the market cost method opportunity cost method is not significantly upward biased. becker (1985) considered a household as a small factory in which every individual specializes in the task that he or she can fulfill most efficiently. so, the relative importance determines who is going to perform unpaid work and who is going for paid work. further, human capital theory and based on the market income it is journal of applied economics and business studies, volume. 4, issue 4 (2020) 135-160 https://doi.org/10.34260/jaebs.447 141 considered rational for the individuals who have high hourly wage and the high opportunity cost of work at the market should perform market work. further, according to the canadian national account system “information on unpaid work can serve i) to monitor and describe more completely how resources are used; ii) to foster a greater understanding of the economy and of the links between its market and non-market sectors; iii) to provide information on what types of work are undertaken, what goods and services result; what costs are incurred, who provides benefits; and iv) to inform the public debate and help in the formulation of public policy" [extracted from “households’ unpaid work: measurement and valuation” published by statistics canada (1995)studies in national accounting]. the issue of women’s economic contribution by valuing their unpaid labor was raised by many scholars, activists, and others for canada, united states, new zealand, and for many other countries. studies have investigated its scale and estimated the value of women’s unpaid work particularly in the domestic sphere (shamim 1996). recently, folbre (2009) explains that as more women are involved in the labour market the amount of time spent on unpaid work is reduced, whereas men do not make up the slack. living standards are lower than gdp which suggests that more income is spent on purchasing food away from home, housekeeping, child, and elder care services that were provided outside the market. living standards and economic welfare do not depend on family income but depend on the amount of time people devote to unpaid work. heymann and beem (2005) explain that as women are more involved in the labour market they usually find low-paid work and usually their children suffer. only earned income is valued by society, while no assistance is provided by the society on the performance of other duties and difficulties faced by household members for which little/no support has been received by employers or state to balance between work and family responsibilities. for pakistan, arshad (2008) found out that a pakistani woman no matter what region they belong to perform a wide variety of tasks. they work 16 hours/day including major and supplemental tasks. most of them do not have any leisure time and they utilize their spare time to complete household chores. it has become clear from this research that women have taken up most of the household responsibilities without any support. even if they have a servant, rearing and assessing children with their school work and cooking is done by women herself. they recognize that their household activities have value, though they would be surprised at the magnitude of their economic worth. when household activities are outsourced the value calculated by researchers shows that it cost rs 48,000/year or us$8,002 in urban areas, while in a rural area it cost rs. 25,800/year or us$430. from the review of the above literature, it can be concluded that researchers have paid little attention to household activities. though research on household activities is currently at the center of theoretical or policy debates but much emphasis is paid on the hamna nasir, ambreen fatima and shaista alam 142 approach. the literature has overlooked activities taking place within the household (such as food preparation, child care, and other domestic chores) as well as the division of labor by age and sex are the neglected areas of research, especially in the case of pakistan. 3. methodology in this article two approaches; opportunity cost and market pricing approach are used to evaluate the worth of home production. the market price method is preferred in literature if the purpose of a valuation is to incorporate estimates into national income accounts. the opportunity cost method, however, is preferable if the purpose is to compensate for welfare loss. a. opportunity cost method the opportunity cost method is itself evolved in literature through many approaches. the below section elaborates on two different techniques employed in this study to estimate the opportunity cost of non-market work. according to hamdad (2003) “when an individual engages in non-market work, s/he has to sacrifice activities that could have been done instead, along with all associated monetary and non-monetary benefits”. i) james (1996) and pandey (2001) in this method, authors have incorporated the wage rate as the opportunity cost of time which is formalized as: opportunity value of non-market work= (total number of hours/week devoted to non-market work * individual's minimum weekly wage rate1) ii) efroymson (2010) this method was given by efroymson (2010) for evaluating the value of non-market activities. the opportunity cost of non-market work is estimated by hours/week spent on each activity multiplied by the minimum weekly wage. opportunity cost of household worki= (value of non-market work for each activity) * (hours allocated to each activity) here the subscript i represents the non-market activities. b. market price method the second method of valuing non-market work/week is market price. in this method, it is assumed that the time spent on non-market activities by household 1 minimum wage rate is used because wages of household workers and unpaid contributing family workers is not available, therefore minimum wages are used. journal of applied economics and business studies, volume. 4, issue 4 (2020) 135-160 https://doi.org/10.34260/jaebs.447 143 members is valued by the earning level of other people who are engaged in similar activities. it is assumed that household members and their “replacements” are “equally productive and responsible”. likewise, the opportunity cost method; the market price method employed in this study, is also estimated using two different approaches. i) global substitute approach global substitute approach can be defined as “a general housekeeper who is chosen as a substitute for all non-market work that household members actually do.” the global substitute method (gl) is given by chadeau (1992) and is formalized as: 𝑯𝑾 = ∑(𝑻𝒊 ∗ 𝑾) where: hw represents the non-market work ti represents time spent on non-market work by the person i per week w is the wage rate of a person trained in general housekeeping duties/household keeper ii) specialist substitution approach the specialist substitution method is defined as “a variety of trained workers who are chosen as substitutes to perform those tasks at home which correspond to specialization in the labor market”. it includes cooks, domestic cleaners, etc. the specialist substitute method (sp) is also given by chadeau (1992) and it is formalized as: 𝑯𝑾𝒊 = ∑ ∑(𝑻𝒊𝒋 𝒋𝒊 ∗ 𝑾𝒊) where • tij represents time spent on performing non-market activities i by person j per week. • wi is the wage rate of a specialist worker performing task i (or its closest substitute) in the market. 4. model specification and data sources the substantial increase in non-market work hours during 1994-95 to 2017-18 for both gender has raised the opportunity cost and market value of non-market value as well. the raised value if not properly socially and economically acknowledge performing these activities will be looked down and individual moves towards paid activities – the recognized activities. the accelerated value of non-market hours of work has grown 2.84% for females and declined for males by 2.26% (see table 1) needs a detailed assessment of factors not encouraging individuals for having paid work though the opportunity cost and market value has increased substantially. this is mainly because the higher the responsibility of an individual at home lower will be his/her participation hamna nasir, ambreen fatima and shaista alam 144 in the labor market and hence higher will be the opportunity cost of time. in order to estimate the factors determining the accelerating value of non-market work for the past few decades following simple models are estimated. specifically, the model for estimating the determinants of opportunity cost and the market price of non-market work are formulated as: 𝑂𝐶ℎℎ𝑤𝑟𝑘𝑔 = 𝛼0 + 𝛼1𝐷𝐶 + 𝛼2𝑃𝐶 + 𝛼3 𝑅𝑒𝑔𝑖𝑜𝑛 + 𝛼4 𝑃𝑟𝑜𝑣𝑖𝑛𝑐𝑒 + 𝛼5 𝐻𝐻𝑌 + 𝜃0 __(𝐼) 𝑀𝐾ℎℎ𝑤𝑟𝑘𝑔 = 𝛽0 + 𝛽1𝐷𝐶 + 𝛽2𝑃𝐶 + 𝛽3 𝑅𝑒𝑔𝑖𝑜𝑛 + 𝛽4 𝑃𝑟𝑜𝑣𝑖𝑛𝑐𝑒 + 𝛽5 𝐻𝐻𝑌 + ∅0 ______(𝐼𝐼) where 𝑂𝐶 represents the opportunity cost of non-market work, 𝑀𝐾 represents the market price of non-market work and subscript ℎℎ𝑤𝑟𝑘 represent the household work and g represents gender. 𝐷𝐶 represents demographic characteristics such as children less than 6 years, adults above 65 years, family size, etc. 𝑃𝐶 represents personal characteristics which are age, age2, education, marital status, etc. the dummies capturing regional and provincial variation influene non-market hours of work. 𝐻𝐻𝑌 represents dummy of household activities and number of years. 𝑂𝐶 and 𝑀𝑃, dependent variables, are measured in pkr/week. 𝑂𝐶 is estimated by the method proposed by james (1996) and 𝑀𝑃 is estimated by using the global substitute method proposed by chadeau (1992). in this study, the opportunity cost and market price across gender and for different household activities are examined in detail. overall, the average opportunity cost of household work for a female is rs.155,126/week and for males is rs.26,082/week; whereas the market price of household work for the male is rs. 19,655/week and that of the female is rs.74, 714/week. as far as explanatory variables are concerned, starting with the personal characteristics, age may have a positive effect on 𝑂𝐶 and 𝑀𝑃 as with the increase in age the household responsibilities of individuals also increases leading to an increase in household work hours ultimately declining the paid market hours of work which lead to increase in 𝑂𝐶 and 𝑀𝑃. experince proxied here by squaring the age hypothesized to have a negative effect on 𝑂𝐶 and 𝑀𝑃 of household work as after a certain age with the increase in age domestic responsibilities are shifted to the younger generation. this also traces out the possible non-linearity in the variable. years of education may have both a positive or negative effect. it may have a negative effect due to their increased participation in labour market – educated females prefer paid market work. due to the increased participation of females in the labor market, some household work is outsourced which increase mp and declines oc. contrary to it, years of education may have a negative effect as most of the household responsibilities are endured by women family members whether they are highly educated or illiterate since the value of time at home has increased substantially as it is journal of applied economics and business studies, volume. 4, issue 4 (2020) 135-160 https://doi.org/10.34260/jaebs.447 145 mandatory to be performed by them instead of being engaged in paid market work, hence oc increases and mp decreases. whereas years of education may have a negative effect on oc and mp on male work as they diversified opportunities are available for them to make use of their education. whatever the marital status of women is, household work is customary and obligatory for them, therefore their 𝑂𝐶 and 𝑀𝑃 are higher as this work is necessary, therefore it should be recognized. contrary to this, most of the household chores are performed by females that decline the market prices of domestic workers hence lower the 𝑀𝑃. 𝑂𝐶 and 𝑀𝑃 are lower as mainly females work as low paid domestic helpers to support their families. further, 𝑂𝐶 and 𝑀𝑃 of males may have negative relation with their marital status of being married like most of the household tasks are performed by females both at home and in the market and males are the main economic providers therefore their non-market work hours are lower, hence 𝑂𝐶 and 𝑀𝑃 are low. as the non-market work hours are mostly influenced by the family composition hence demographic variables considered here are mainly focusing on dependency ratio. the number of dependents, here, is represented by the children under the age of 6, children between the age of 6-10, children between the age of 11-14, and adults above the age of 65. as the number of dependents increases in a household the opportunity cost and market price of female non-market work increases. whereas the increase in the number of dependents may decline the opportunity cost and market price of male work as they are responsible for supporting their families hence their paid market work hours are usually higher than the non-market hours of work. household income may have a positive effect on 𝑂𝐶 as the socio-economic condition of the household become better, therefore paid working hours of female decline while their non-market hours' increases, whereas the paid work hours of males will increase and their non-market hours will decline. contrary to this, it may have a negative effect on 𝑂𝐶 𝑎𝑛𝑑 𝑀𝑃 as most of the household chores could be performed by maids or home saving appliances could be purchased – the affordability of support workers increases with the increase in income. specifically, household income may have a positive effect on the 𝑀𝑃 as household start outsourcing the domestic task. across the region, the urban region is taken as a reference category. 𝑂𝐶 of household work may be lower, whereas mp of housework may be higher in the urban region due to the availability of domestic helpers/ outsourcing of domestic chores and higher participation of females in the labour market. specifically, the oc of household work may be positive and mp of household work may be negative, if most of the household chores are performed by individuals due to vilest social conditions. across the provinces, kpk is taken as a reference category. 𝑂𝐶 𝑎nd 𝑀𝑃 of household work may be positive or negative depending on the customs and traditions of the province. across the household activities, some are time-intensive and therefore hamna nasir, ambreen fatima and shaista alam 146 it may have a positive effect on oc and mp of household work and some may have a negative effect on oc and mp of household work. the data used in this study is taken from the labor force survey (lfs) of pakistan for the years 1994-95, 1999-2000, 2006-07, 2014-15, and 2017-18, which is conducted by the pakistan bureau of statistics2 (pbs). 5. results the result section is divided into two subsections; the first one provides a descriptive assessment while the second one will make the empirical assessment of the model. a. descriptive statistics for assessing the opportunity cost of non-market work per week different techniques are applied as described above. specifically, the monetary value of non-market work is estimated based on opportunity cost (the household’s average income) and replacement cost (the housekeeper cost method). in this section opportunity cost and market value of non-market work are discussed in detail. a. assessment of opportunity cost of non-market work the opportunity cost of non-market work when analyzed by minimum wage as given by james (1996) across years (see table 1) shows that the value of non-market work of female is higher than male. the opportunity cost of female work has increased from rs.13,928 to rs.155,126 per week around 2.53 annual growth; whereas for male work it has increased from rs. 4,778 to rs. 26,082 per week with annual growth around 1.49. over the years the minimum wage has also increased (2.25 annual growth in minimum wage), therefore the opportunity cost of non-market work has also increased for both genders. table 1: opportunity cost of household work/week no. of years male hours/week male oc of hh work/week female hours/week female oc of hh work/week 1994-95 0 rs0 37 rs13,928 1999-00 8 rs4,778 40 rs25,212 2006-07 7 rs7,978 38 rs44,160 2014-15 6 rs17,477 41 rs123,759 2017-18 7 rs26,082 41 rs155,126 growth rate in hours/week -2.257 2.844 growth in oc 1.486 2.534 2 since the labor force of pakistan is a published data, therefore questionnaire and the detail of labor force survey of pakistan is available on the website of pakistan bureau of statistics and it can be downloaded from www.pbs.gov.pk journal of applied economics and business studies, volume. 4, issue 4 (2020) 135-160 https://doi.org/10.34260/jaebs.447 147 further to this, the opportunity cost of non-market work/week by activities is presented in table 2 by gender. opportunity cost of all non-market activities has increased over the years. specifically, over the years it has been observed that the opportunity cost of child care and cooking activities remains higher as compared to the opportunity cost of other non-market activities. the opportunity cost of child care, child education, and cooking (if performed by females) has increased from rs. 3,960/week to rs. 45,000/week, rs. 3041/week to rs. 26,250/week and rs. 5,095/week to rs. 60,000/week. as these activities are time-intensive, therefore its opportunity cost is also higher. cooking hours have grown by 4.44% and its annual growth is around 2.69. the opportunity cost of sewing, house cleaning, and washing has also increased from rs. 2,587/week to rs. 20,505/week, rs. 3,248/week to rs. 30,000/week and rs. 2,876/week to rs. 26,250/week respectively. moreover, the opportunity cost of shopping and marketing, firewood collection, fetching water, and construction work performed by females has increased from rs. 2,220 to rs.15,000/week, rs.2,425 to rs.17,776/week , rs.2,780 to rs.20,715/week and rs.2,154 to rs.13,978/week. however, these activities if performed by males the opportunity cost remains higher only for shopping and marketing activity as compared to other activities. the opportunity cost of shopping and marketing, firewood collection, fetching water, and construction work have increased from rs.3,438/week to rs.22,500/week, rs.4,524/week to rs.20,291/week, rs.4,212 to rs.18,675/week, and rs.4,183 to rs.12,500/week. hours of firewood collection have declined by 9.524% and its annual growth is around 1.16. the annual growth in opportunity cost for all activities shows an increasing trend due to the increase in the minimum wage over the years, whereas growth rate of household hours for some activities have declined over the years. table 2: opportunity cost of household activities/week no.of years household activities male house hold hours/ week oc of male household work/week (in pkr) female house hold hours/ week oc of female household work/week (in pkr) 1994-95 sewing 0 0 7 2,502 shopping & marketing 0 0 6 2,220 washing& pressing clothes 0 0 8 2,875 child/adult care 0 0 11 3,960 child education 0 0 8 3,037 house cleaning 0 0 9 3,248 cooking 0 0 14 5,095 firewood collection 0 0 6 2,425 fetching water 0 0 7 2,780 constructionwork 0 0 6 2,154 1999-00 sewing 0 0 6 4,035 shopping & marketing 6 3,438 5 3,378 washing& pressing clothes 0 0 8 5,000 child/adult care 0 0 11 6,875 child education 0 0 8 5,000 house cleaning 0 0 9 5,625 hamna nasir, ambreen fatima and shaista alam 148 no.of years household activities male house hold hours/ week oc of male household work/week (in pkr) female house hold hours/ week oc of female household work/week (in pkr) cooking 0 0 15 9,375 firewood collection 7 4,524 6 3,910 fetching water 7 4,212 7 4,508 constructionwork 7 4,183 5 3,083 2006-07 sewing 0 0 6 6,348 shopping & marketing 6 6,942 5 5,750 washing& pressing clothes 0 0 7 8,050 child/adult care 0 0 11 12,650 child education 0 0 7 8,050 house cleaning 0 0 8 9,200 cooking 0 0 15 17,250 firewood collection 6 6,720 5 5,811 fetching water 6 7,312 6 7,318 constructionwork 5 5,750 4 5,041 2014-15 sewing 0 0 5 16,173 shopping & marketing 5 14,857 5 15,000 washing& pressing clothes 0 0 7 21,000 child/adult care 0 0 12 36,000 child education 0 0 7 21,000 house cleaning 0 0 7 21,000 cooking 0 0 16 47,963 firewood collection 6 16,941 5 15,515 fetching water 5 15,474 6 17,426 constructionwork 4 10,500 4 11,154 2017-18 sewing 0 0 5 20,505 shopping & marketing 6 22,500 4 15,000 washing& pressing clothes 0 0 7 26,250 child/adult care 0 0 12 45,000 child education 0 0 7 26,250 house cleaning 0 0 8 30,000 cooking 0 0 16 60,000 firewood collection 5 20,291 5 17,776 fetching water 5 18,675 6 20,715 construction work 3 12,500 4 13,978 growth rate in hours/week sewing 0 3.0704 shopping & marketing 0 -8.105 washing& pressing clothes 0 -2.1767 child/adult care 0 3.4059 child education 0 -3.3918 house cleaning 0 -1.9057 cooking 0 4.4432 firewood collection -9.524 -6.6774 fetching water -9.524 -6.374 construction work -19.048 -8.779 growth in oc sewing 0 1.80 shopping & marketing 1.85 1.44 washing& pressing clothes 0 2.03 child/adult care 0 2.59 child education 0 1.91 house cleaning 0 2.06 cooking 0 2.69 firewood collection 1.16 1.58 journal of applied economics and business studies, volume. 4, issue 4 (2020) 135-160 https://doi.org/10.34260/jaebs.447 149 no.of years household activities male house hold hours/ week oc of male household work/week (in pkr) female house hold hours/ week oc of female household work/week (in pkr) fetching water 1.14 1.61 construction work 0.66 1.37 b. assessing market price of non-market work the market price for non-market work is discussed in this sub-section. total non-market time (housework time) is valued at the housekeeper’s market wage rate in gl method. it is evident from figure 1 that the market price of female work is more than male work. the market value of female housework has increased from rs.9,275/week to rs. 63,813/week; whereas for males, it has increased from rs.5,376/week to rs.22,908/week. most of the non-market activities are performed by the female as domestic helpers; therefore her market price is also higher. market value is estimated by the time individual spend on each type of non-market activity with their respective market wages matched with their specialized activity. from figure 2 once again it is obvious that the market value of activities such as child care and cooking is higher as compared to other non-market activities as said earlier these activities are time-consuming, therefore their market price is also higher. though, the market price of all non-market activities has increased over the years. the market rs0 rs20,000 rs40,000 rs60,000 rs80,000 rs100,000 rs120,000 rs140,000 rs160,000 1994-95 1999-00 2006-07 2014-15 2017-18 m a rk e t v a lu e o f h o u se h o ld w o rk /w e e k ( in p k r ) figure 1 : market price of household work by global substitute method female market value/week male market value/week hamna nasir, ambreen fatima and shaista alam 150 value of sewing has increased from rs.2,500/week to rs. 6,718/week, for washing and house cleaning it, has increased from rs.5,367/week to rs.12,600/week and rs. 2,282/week to rs. 12,627/week respectively, for child care and child education it has increased from rs. 2,407/week to rs. 24,340/week and rs.5,438/week to rs. 13,819/week and for cooking it has increased from rs. 7,081/week to rs. 41,778/week. for males, the market value of shopping and marketing has increased from rs.3,867/week to rs.19,762/week and for females, it has increased from rs.1,349/week to rs.12,394/week, for firewood collection it has increased for males from rs.5,326/week to rs.24,304/week and for females from rs.2,583/week to rs.3,264/week, for fetching water it has increased for males from rs.4,959/week to rs.22,369/week and for females from rs.2,153/week to rs.4,087/week and for construction work, it has increased from rs.4,109/week to rs.10,767/week for males and for females it has increased from rs.1,971/week to rs.10,657/week. as compared to other household activities market price of the male is higher for firewood collection. since, it is necessary for household consumption, especially in rural and underdeveloped areas. it is obvious from the above-mentioned figure and tables that the opportunity cost of all non-market activities is greater than the market value since opportunity cost is rs0 rs5,000 rs10,000 rs15,000 rs20,000 rs25,000 rs30,000 rs35,000 rs40,000 rs45,000 s e w in g w a sh in g c h il d e d u c a ti o n c o o k in g w a te r c o ll e c ti o n s e w in g w a sh in g c h il d e d u c a ti o n c o o k in g w a te r c o ll e c ti o n s e w in g w a sh in g c h il d e d u c a ti o n c o o k in g w a te r c o ll e c ti o n s e w in g w a sh in g c h il d e d u c a ti o n c o o k in g w a te r c o ll e c ti o n s e w in g w a sh in g c h il d e d u c a ti o n c o o k in g w a te r c o ll e c ti o n 1994-95 1999-00 2006-07 2014-15 2017-18 m a rk e t v a lu e o f h o u se h o ld a c ti v it ie s/ w e e k ( in p k r ) number of years and household activities figure 2: market value of household activities by specialist substitution method market value of female hh work/week market value of male hh work/week journal of applied economics and business studies, volume. 4, issue 4 (2020) 135-160 https://doi.org/10.34260/jaebs.447 151 estimated by using minimum wage criteria and market value is estimated by using the actual market price of the activities. b. empirical results the empirical assessment of factors accelerating the opportunity cost and market price of non-market work as discussed in equations (i) and (ii) and are presented in tables 3 and 4. in table 3 determinants of the opportunity cost of household work by both gender is discussed and it shows that with the increase in age opportunity cost has increased. in pakistan, women's work starts early in life and continues till the sixties or late. girls in teens besides giving time to their studies also help their mothers in domestic chores. whereas, males perform the household activities in their teens, and with the increase in their age its opportunity cost declines as in traditional society they are responsible for supporting their family by engaging themselves in paid employment. however, age square explores the non-linearity which shows that as the age doubles, the opportunity cost of female household work decreases as their work is shared by young females of the household. as the years of education increase, the opportunity cost of female housework also increases. women are either highly qualified or illiterate their participation in the labour market remains low as in traditional society household work is mandatory for them, hence opportunity cost of their non-market work is higher. besides household responsibilities the factors which hinder their participation in the labour market are gender discrimination, job mismatch, two-tiers labor market in pakistan, lower remuneration offered, glass ceiling, etc. an additional year of female education will increase the opportunity cost of her household work by rs.98.3/week. female either being married or widow/divorced have a positive and significant effect on the opportunity cost as it is the prime responsibility of female to fulfill the household chores. male either being married or divorced has a negative and significant effect on opportunity cost as they have higher paid market hours and are responsible for fulfilling household expenses which leads to a decline in their household work. the opportunity cost of married and widow/divorced female increases by rs.3,005/week and rs.1,610/week; whereas the opportunity cost of married and widow/divorced male decreases by rs.1,564/week and rs.2,401/week respectively. an increase in the number of dependent in a household increases the opportunity cost of female household work as the dependents require more care and nurturing. further, they have higher responsibilities of childbearing, rearing, and taking care of adults especially if they are sick. according to arshad (2008) women have taken most of the household responsibilities without a helping hand. if they hire a maid, still cooking, up-bringing of children, and assessing them with the school homework is usually done by female households. the increase in the number of children under the age of 6, age 6-10, and age11-14 increases the opportunity cost of female housework by rs. 104.5, rs.91.8, and rs.135/week, whereas an increase in the number of adults increases the opportunity cost of female work by rs.286/week. the opportunity cost of hamna nasir, ambreen fatima and shaista alam 152 non-market work is higher in the urban region because due to the change in the urban lifestyle, increase in the nuclear family setup, and vilest social conditions women prefer to perform most of the household chores by themselves. as compared to rural regions opportunity cost of females residing in urban regions increases by rs.449.7/week. as compared to kpk opportunity cost of females residing in punjab increases by rs.3,424/week, whereas the opportunity cost of females residing in sindh and baluchistan decreases by rs.224/week and rs.239/week. the opportunity cost of males residing in punjab, sindh, and baluchistan decreases by rs.3,809/week, rs.3,488/week, and rs.2,433/week respectively. it depends on the cultural norms and traditions of the province that the opportunity cost of non-market for both gender is higher in some provinces and lower in others. as compared to other provinces most of the household chores are performed by females in punjab due to its tradition which leads to an increase in the opportunity cost of their work. all non-market activities and year dummies have a positive and significant effect on the opportunity cost of household work. finally, the value of r2 for the opportunity cost of non-market work for males and females is 39.2% and 54.4% respectively. table 3: determinants for economic value of household work dependent variable: economic value (in pkr/week) female coefficient male coefficient personal characteristics age 248.0*** 187.1** (7.329) (91.21) age2 -4.079*** -1.625 (0.0936) (1.076) education education in years 98.34*** 111.6 (4.665) (68.99) marital status dummy married 3,005*** -1,564* (55.89) (851.7) widow & divorced 1,610*** -2,401* (99.66) (1,247) socio demographic characteristics number of children under 6 years 104.5*** 296.8 (11.77) (188.7) number of children of age 6-10 years 91.80*** 61.18 (14.41) (234.4) number of children of age11-14 years 135.0*** 218.6 (18.70) (306.3) number of adults above 65 years 286.2*** 327.2 (35.82) (547.9) household income (in pkr/week) 0.00562 -0.0170 (0.00362) (0.0476) region dummy journal of applied economics and business studies, volume. 4, issue 4 (2020) 135-160 https://doi.org/10.34260/jaebs.447 153 dependent variable: economic value (in pkr/week) female coefficient male coefficient urban region 449.7*** -2.102 (33.00) (549.5) province dummy punjab 3,424*** -3,809*** (42.18) (623.6) sindh -224.3*** -3,488*** (45.52) (748.9) baluchistan -238.7*** -2,433** (54.22) (1,047) household activities dummy firewood collection 1,633*** 3,182** (195.7) (1,258) fetching water 3,778*** 2,395* (187.5) (1,279) shopping and marketing -2,762*** 3,128*** (167.5) (1,109) sewing 1,783*** (166.9) washing& pressing clothes 3,576*** (161.1) child care/adult care 13,753*** (163.3) child education 2,791*** (185.5) house cleaning 5,580*** (161.1) cooking 23,127*** (161.4) year dummy 1999 1,982*** (60.01) 2006 6,195*** 1,959** (52.08) (791.3) 2014 23,320*** 11,631*** (52.83) (868.7) 2017 29,714*** 15,383*** (52.32) (834.5) constant -12,087*** (194.6) 521.5 (1,873) f-statistics 38449.31* (0.0000) 41.67* (0.0000) observations 869,408 1,312 r-squared 0.544 0.392 standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 hamna nasir, ambreen fatima and shaista alam 154 table 4 shows the determinants of the market price of household work by genders as outline in equation ii. results show that with the increase in age market price has increased. women work in the market as a domestic helper and their participation usually starts from early life. those girls who do not attend school help their mothers as domestic helpers or fulfill domestic responsibilities and look after their siblings by paying the penalty of not attending school. whereas, males are engaged in only those activities which are necessary for household consumption, and these activities are outsourced only in higher-income families. however, age square exploring the nonlinearity shows that as the age doubles, the market value of female household work decreases as after a certain age their work is shared by a new generation as said earlier. further, as the years of education increases market price of female housework also increases as educated female will/may make use of their education and participate in labour market; therefore some household activities may be outsourced. an additional year of education will increase the market price of female work by rs. 40.6/week. female either being married or widow/divorced may have a positive and significant effect on the market price of her household work, whereas male being married has a negative and significant effect on the market value of his household work. the market value of married and widow/divorced females is higher because in order to support their families they work as domestic helpers. further, the market value of women's work may also have increased as in rural or pari urban areas mostly males have seasonal income or they often do not participate in labour market and hence depend on the wages of their spouse. if the women participate in paid market work they usually hire maids to perform mundane tasks or depends on their daughters to support them. if widow/divorced females are from lower-income family then besides performing market work she also has to perform her own domestic work which leads to an increase in the opportunity cost and market price of her work. after marriage responsibilities of male increases therefore they will search for better employment opportunities to earn a better livelihood instead of participating in lower-paid occupation. the market price of married and widow/divorced females increases by rs.1,375/week and rs.708/week; whereas the market price of married males decreases by rs.705/week. as the number of children under the age of 6 years and the number of adults above the age of 65 years increases in a household then the market price of non-market work also increases as the dependents require proper care and attention which leads to an increase in the market price of female work. an increase in the number of children under the age of 6, children between the age of 6-10, and children between the age of 11-14 increase the market price of female work by rs.50, rs.39, and rs.62.68/week, whereas an increase in the number of adults increases the market price of female work by rs.134.8/week. as compared to rural regions females dwellers of the urban region have a positive and significant effect on market price as prices of these activities are higher in urban areas. the market price of female work in urban regions increase by rs.215/week. further, most of the females are engaged in paid market work and change in urban lifestyle especially of higher-income journal of applied economics and business studies, volume. 4, issue 4 (2020) 135-160 https://doi.org/10.34260/jaebs.447 155 families has increased the hiring of maids. female households who have lower income are double burdened as besides performing the domestic duties they are also engaged in market work as domestic helpers. the market cost of females residing in punjab may increase by rs.1,503/week. whereas the market price of females residing in sindh and baluchistan may decrease by rs.128.5/week and rs.69.9/week. the market value of males residing in punjab, sindh, and baluchistan decreases by rs.1,747/week, rs.1,617/week, and rs.1,127/week. non-market activities and year dummies show that the effect of all non-market activities has a positive and significant effect on the market value of female household work except shopping and marketing as there are fewer female who performs this activity or it is mostly performed by male household members, whereas the effect of all household activities has a positive and significant effect on the market price of male household work. year dummies have both positive/negative effect on the market price. the value of r2 for the market price of nonmarket work for males and females is 26.9% and 50% respectively. table 4: determinants of market price for household work dependent variable: market price(in pkr/week) female coefficient male coefficient personal characteristics age 119.3*** 81.10** (3.128) (40.10) age2 -1.937*** -0.698 (0.0400) (0.473) education education in years 40.60*** 48.30 (1.991) (30.33) marital status dummy married 1,375*** -705.1* (23.86) (374.5) widow & divorced 708.3*** -833.8 (42.53) (548.5) socio demographic characteristics number of children under 6 years 50.00*** 142.1* (5.021) (82.98) number of children of age 6-10 years 39.44*** 40.23 (6.151) (103.1) number of children of age11-14 years 62.68*** 83.47 (7.983) (134.7) number of adults above age 65 years 134.8*** 106.7 (15.29) (240.9) household income (in pkr/week) 0.00244 -0.00939 (0.00154) (0.0209) region dummy urban region 215.2*** 86.99 (14.09) (241.6) province dummy punjab 1,503*** -1,747*** hamna nasir, ambreen fatima and shaista alam 156 dependent variable: market price(in pkr/week) female coefficient male coefficient (18.00) (274.2) sindh -128.5*** -1,617*** (19.43) (329.3) baluchistan -69.91*** -1,127** (23.14) (460.3) household activities dummy firewood collection 902.0*** 1,429*** (83.52) (553.2) fetching water 1,925*** 1,106** (80.02) (562.4) shopping and marketing -975.8*** 1,291*** (71.48) (487.4) sewing 973.2*** (71.26) washing& pressing clothes 1,882*** (68.77) child care/adult care 6,503*** (69.70) child education 1,596*** (79.17) house cleaning 2,837*** (68.76) cooking 10,821*** (68.88) year dummy 1999 3,693*** (25.61) 2006 3,045*** -783.6** (22.23) (347.9) 2014 8,852*** 2,413*** (22.55) (381.9) 2017 11,516*** 3,962*** (22.33) (366.9) constant -5,018*** 2,726*** (83.05) (823.6) f-statistics 32256.48* (0.0000) 23.79* (0.0000) no. of observations 869,408 1,312 r-squared 0.500 0.269 standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 6. conclusion and policy implications in this study opportunity cost and market price of household work by gender and for different type of household activities have been estimated and explored in detail. the opportunity cost of male and female household work was found out to be rs.26,082/week and rs.155,126/week respectively; whereas market price of male and journal of applied economics and business studies, volume. 4, issue 4 (2020) 135-160 https://doi.org/10.34260/jaebs.447 157 female is found out to be rs.22,098/week and rs.63,813/week respectively. the opportunity cost and market price of child care and cooking activity is higher as compared to other household activities if it is performed by females. whereas, the opportunity cost of shopping and marketing and firewood collection is higher as compared to other activities if performed by males. as these activities are necessary for household consumption and are performed outside the house and it mostly increases when paid market hours of male work decreases as these activities are mostly performed by male residents of rural and pari-urban areas. moreover, opportunity cost and market price is higher in those regions where the patriarchal system and veiling factors are still being practiced. married and widow/divorced females have a positive and significant effect on the opportunity cost and market price of their household work, whereas married males have a negative effect on the opportunity cost and market price of their household work. since it is the prime responsibility of a female to fulfill the household chores whatever her marital status is, whereas married males are responsible to support their families which increases their paid market hours leading to decline in their non-market hours. in the presence of dependents, the opportunity cost of non-market work may increase the domestic responsibilities of the female; whereas the market value of non-market work increases therefore maids are hired to assess females to fulfill other domestic duties. female residents of punjab have a positive and significant effect on the opportunity cost and market price of their household work, whereas female residents of sindh and baluchistan have a negative and significant effect on the opportunity cost and market price of their household work. male residents of all provinces have a negative and significant effect on the opportunity cost and market price of their household work. hence the increase and decrease in the opportunity cost and market price depend on the cultural norms and traditions of the province as well as on the socio-economic condition of the households. in this study, the emphasis on monetary value is given to highlight the contribution made by women to society. no society can function with the only economic side of life. the most important activity in society is maintaining a household and nurturing its members. women work with affection for their family members so money cannot buy such affection. society would crumble without the daily tasks of caring, nurturing, and supporting others. an economic value of non-market work is assigned not to undervalue it, but to remind ourselves that even in the male world of economics; women make an essential contribution as their work has substantial economic value, hence it should be recognized. women can be seen as subsidizing salaries throughout the workforce, by contributing to domestic work freely rather than demanding a wage. it is vital that officials and others recognize that expenditures on women, whether in the home or at the national level are not a drain or an expense, but rather an investment in the economic hamna nasir, ambreen fatima and shaista alam 158 and social welfare of the country. it is recommended that women’s housework should be valued at the national level as they contribute by taking care of their homes, save their family money and support their family in multiple ways and in this way target set in sdg goal 5.4 can be achieved. it would help in raising the status of women and promote gender equality in pakistan by ultimately reducing poverty and dependency which is faced by many women. by changing the perceived value of women would change the way policymakers approach programs to assess women. as women’s work is given value in india and vietnam, if their work is given value by the government of pakistan then men would have a better understanding of their value, give importance their work and they would likely treat them with more respect. thus, it would be easier to create happier, mutually respectful, and violence-free families. if policymakers and others wish to see more women entering the formal economy they often forget that women who take on a paid job are required to do double duty as they would remain the sole or main responsible person for domestic duties. a society that respects and values the care work for others would also be far more humane than one that values paid jobs. references alian girard (1958).‘the time budget of married women in urban agglomerations’, population, october-december 1958. ann chadeau (1992). ‘what is households’ non-market production worth?’, oecd economic studies no. 18 spring 1992, pp.1-19. ayesha khan (2007), ‘pathways of women’s empowerment south asia programme, pakistan scoping paper’, collective for social science research. bhalotra, s., & attfield, c. (1998), ‘intrahousehold resource allocation in rural pakistan: a semiparametric analysis’, journal of applied econometrics, 13(5), pp.463-480. becker, g.s. (1965) "a theory of the allocation of time,” economic journal, 75, pp. 493-517. becker, g. s. (1985). ‘human capital, effort, and the sexual division of labor’, journal of labor economics, 3(1), pp. 33–58. bell, edward b and taub, allan j. (1982). ‘the value of household services’, american journal of economics and sociology, 41, pp. 214-218. chiswick, carmel. (1982). ‘the value of a housewife's time’, journal of human resources, 17, pp. 413-425. journal of applied economics and business studies, volume. 4, issue 4 (2020) 135-160 https://doi.org/10.34260/jaebs.447 159 debra efroymson (2010). ‘women, work and money: studying the economic value of women’s unpaid work and using the results of advocacy’, report by health bridge and the institute of social development studies. folbre, nancy (2009). ‘valuing unpaid work matters, especially for the poor’, new york times online, september 21, 2009. fourastii, j., les (1965).‘40.000 hours’,laffont, paris, 1965. hamdad, m. (2003).‘valuing households’ unpaid work in canada, 1992 and 1998: trends and sources of change’, presented at statistics canada economic conference in may 2003. hamid shamim (1996). ‘why women count, essays on women in development in bangladesh’, theuniversity presslimited. harvey james (1996). ‘the valuation of household production: how different are the opportunity cost and market price valuation methods?’, presented at workshop of brigham young university. hawrylshyn (1976). ‘the value of household services: a survey of empirical estimates’, the review of income and wealth, 22 (2), pp.101-103 hawrylshyn (1977). ‘towards a definition of non-market activities’, review of income and wealth, 23 (1), pp.79-96. heymann, j and c beem, ed. (2005).‘unfinished work, building equality and democracy in an era of working families’, the new press. margaret reid (1934). ‘economics of household production’, john wiley & sons inc. marianne a. ferber and bonnie g. birnbaum (1980). ‘housework: priceless or valueless’, review of income and wealth, 26 (4), pp.387-400. michel, andrke (1978).‘women in merchant society’, university press of france 1978. murphy, martin (1978). ‘the value of nonmarket household production: opportunity cost versus market cost estimates’, the review of income and wealth, 24, pp.243-256. najam-us-saqib, & arif, g. m. (2012), “time poverty, work status and gender: the case of pakistan”, pakistan development review, vol.51 (1),pp. 23-46. naqvi, z. f., shahnaz, l., & arif, g. m. (2002). ‘how do women decide to work in pakistan’?, the pakistan development review, 41 (4), pp. 495-513 pandey (2001). ‘women’s contribution to the economy through their unpaid labor’, working papers from national institute of public finance and policy, pp.1-33 peterson, r. d. (1978). ‘problems in estimating the value of household services’, american journal of economics and sociology, 37, pp.145-148. hamna nasir, ambreen fatima and shaista alam 160 quah, euston (1986). ‘persistent problems in measuring household production: definition, quantifying joint activities and valuation issues are solvable’,american journal of economics and sociology,45, pp.235-245. statistics canada (1995). ‘households' unpaid work: measurement and valuation’,studies in national accounting, catalogue 13-603e. no 3. stoetzel jean (1948).‘a study of the budgeting of time by womenin urban communities’, population 3(1), january-march 1948, pp.47-62. sultana, n., nazli, h., malik, s. j., & kazi, s. (1994), “determinants of female time allocation in selected districts of rural pakistan”, pakistan development review, vol.33 (2), pp. 1141-1153 syed, j., pio, e., & ali, f. (2013), ‘a multi‐level perspective on equal employment opportunity for women in pakistan. equality, diversity and inclusion’, an international journal, 32 (3), pp. 289-309. szalai alexander (1972). ‘the use of time: daily activities of urban and sub-urban population in twelve countries’, the hague: mouton& co 1972. wolf, robert (1986). ‘assessing the value of household services: three methods’, trial, 22, pp.81-83. zehra arshad (2008). ‘the economic contribution of pakistani women through their unpaid labor’, report by society for alternative media and research and health bridge 2007. journal of applied economics and business studies, volume. 2, issue 1 (2018) 37-46 https://doi.org/10.34260/jaebs.214 37 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x a cointegration analysis between macroeconomic variables and fiscal deficit in pakistan mian nasir uddin1, muhammad tariq2* & saleem khan31 1 ms scholar, department of economics, abdul wali khan university mardan. 2,3 assistant professor, department of economics, abdul wali khan university mardan abstract this paper estimates the short and long run association between selected macroeconomic variables and fiscal deficit in pakistan for the period of 1985 to 2016. macroeconomic variables such as exports, exchange rate, gdp per capita, inflation, gross capital formation have strong implications for the fiscal deficit. this study checks the data for stationarity using the augmented dickey fuller test. johansen co-integration test and vector error correction method are used to investigate both the short and long run relationships. results indicated the existence of both short run and long run relationship between the macroeconomic variables and fiscal deficit. the findings of the study revealed that exports, exchange rate, gdp per capita, inflation, gross capital formation are important determinants of fiscal deficit in pakistan. the study suggested that the government may focus on these factors to overcome fiscal deficit in pakistan. keywords fiscal deficit, johansen cointegration test, vector error correction method jel classification b22, e6, e62, e63, f62 copyright © 2018 saebr all rights reserved 1. introduction fiscal deficit arises when the government's expenditures are more than the government’s revenues. fiscal deficit has remained a major issue in both developed and developing countries. payments needed for fiscal imbalances create variations in the interest rate and increase government loans. fiscal imbalances in underdeveloped economies are considered as a soaring problem (anwar & ahmad, 2012). one of the main reasons of fiscal deficit is an increase in public spending with constant or decreasing collection of revenues through taxes and other sources. reduction in fiscal deficit can help in accelerating the development (romer, 1986). reduction in fiscal deficit can have negative effect on social development as many underdeveloped economies in order to minimize fiscal deficit, they reduce investments in social sectors such as health, education, and infrastructure. others reduce it by increasing the tax rate. most economists agree that sustained deficits can damage a 1*correspondence author: tariq_noor@awkum.edu.pk open access mian nasir uddin, muhammad tariq & saleem khan 38 country’s economy, and high fiscal deficit creates major challenges for middle-income economies (agenor & montiel, 1999). furthermore, rising government spending may be undertaken to boost economic growth and wellbeing of the individuals (peacock &wiseman, 1961). pakistan has been consistently facing the problem of increasing fiscal deficit. high fiscal deficit in the economy is a consequence of high inflation as well as low growth (chaudhary & abe, 1999). figure 1 exhibits the fiscal deficit pattern of pakistan and indicates a widening gap over time (pakistan economics survey, 2016-17). figure 1: expenditure and revenue gap (%age of gdp) pakistan has high population of about 193.2 million in 2016 (wdi, 2018). also, it has small level of gdp per capita amounting to 1641 us dollars (gop, 2017-18). pakistan's budget deficit in the last two decades has been 5.4% to 8.7% of its gdp. pakistan's budget deficit as percent of gdp has been respectively 6%, 7.6% and 6.4% in 1970’s, 1980’s, and 1990's. despite increasing budget deficit, there has not been many tax reforms to generate more revenue for the government. maintaining a reasonable fiscal deficit level requires a balance between expenditure and revenues. however, due to lack of attention and poor governance in the taxation system such as many exemptions, and poor collection system has been affecting the pakistan's economy. it has been recognized that there has been an association between gdp per capita, inflation, exchange rate, exports and gross capital formation and fiscal deficit. therefore, this study explores the factors affecting the fiscal deficit of pakistan and test if short-run and long-run relationships exist between macroeconomic variables and fiscal deficit. 2. literature review agha & khan (2006) investigated the long run association between inflation and fiscal deficit in pakistan using monthly time series data from 1973-2003. they used johansen cointegration test and vector error correction models (vecm) for testing the relationship. journal of applied economics and business studies, volume. 2, issue 1 (2018) 37-46 https://doi.org/10.34260/jaebs.214 39 they found that inflation was associated with fiscal deficit in the long run. also, the vecm model indicated that inflation is affected by the public’s borrowings from banks when fiscal deficit is financed. epaphra (2017) found that the exchange rate and real gdp had a negative role on budget deficit while budget deficit was positively related to money supply, inflation and interest rate in tanzania. usman and adebisi (2017) also investigated the factors responsible for widening the gap between nigerian's government revenue and expenditures. uppal (2011) stated that pakistan’s political process, institutional condition, and budget processes influences the fiscal budget and that the fiscal deficit can be improved by introducing better discipline, institutional checks & balances and better market discipline. better governance and development of political institutions can play a long-term improvement in the fiscal deficit. furthermore, javid et al. (2011) studied the budget deficit of south asia and asean countries using the dynamic panel method. they found that high inflation and high budget to gdp ratio had a positive effect on budget deficit whereas better political situation, growth and social development negative effect on deficits. furthermore, it is also found that corruption and mismanagement were contributing to high budget deficit. anwar & ahmad (2012) examined the short and long run association between deficit and factors such as democracy, economic growth of pakistan. they found that government size had positive impact on the budget deficit. furthermore, democracy was not found to have any impact on the budget deficit of pakistan. also, brima and pearce (2015) investigated the association of financial deficit with selected macroeconomic variables using johansen’s test of co-integration, vector error correction model and the granger causality in sierra leone for the period from 1980 to 2014. they found that exchange rate, gdp and supply of money had an inverse association with the deficit while interest rate and inflation had a positive relationship with the budget deficit. torayeh (2015) employing the auto-regressive distributed lag (ardl) estimated the factors for deficit in egypt using time series data from 1985 to 2013. they found that the interest payments and people income increase the egypt deficits. further, these payments eat away a big chunk of the tax revenues. they suggested that subsidies and interest payments may be curtailed to control the budget deficit. further, tax reforms are also needed to improve the revenue. safdar and padda (2017) also investigated the impact of institutional quality on budget deficit utilizing time series data from 1984 to 2014. they found that weak institutions cause high budget deficit due to mismanagement and leakages in the economy. further, they discovered that open trade and inflation had a positive effect on the financial deficit. real gross domestic product per capita was not found to have a significant effect on budget deficit. mismanagement, fraud, corruption, weak rules and regulation also cause the deficit. they suggested that policymakers should pay attention to improve institution's quality to improve the budget deficit. mian nasir uddin, muhammad tariq & saleem khan 40 3. research methods this section discusses the data and the empirical model of the study. since this study is an empirical in nature, data pattern shall show us the relationship between the fiscal deficit and macroeconomic variables. therefore, we focus on data and empirical model. the study uses the data from 1985 to 2016. all the data about different variables have been collected from economic surveys of pakistan and world development indicators of world bank. this study uses jagirani, et al. (2016) for examining the relationship between the dependent variable and independent variables of the study given as follows: 𝐹𝐷 = 𝛽0 + 𝛽1𝐸𝑋 + 𝛽2𝐸𝑋𝐶 + 𝛽3𝐺𝐷𝑃𝑃𝐶 + 𝛽4𝐺𝐶𝐹 + 𝛽5𝐼𝑁𝐹 + 𝜀 (1) where fd is the fiscal deficit (% of gdp), and ex for exports ((% of gdp), exc for exchange rate (rs. /us$), gdppc for gross domestic product per capita (rs.), gcf for gross capital formation ((% of gdp) and inf for inflation (%). moreover, 𝛽0 is the intercept and 𝛽1, 𝛽2, 𝛽3, 𝛽4, and 𝛽5 are the parameters to be estimated and 𝜀 is the error term. when variables are non-stationary, then we have the problem of spurious results (juseof, et al. 2011). therefore, the data is tested for stationarity using the augmented dickey fuller (adf) test given below: ∆y = 𝛼0+𝛼1𝑦𝑡−1 ∑ αi 𝑝 𝑖=1 ∆yt−1 + εt (2) johansen cointegration test shall be used to test for the cointegration (whether there is a short and long run association). cointegration approach is applicable when the data is integration of order one (i(1)). cointegrating equations can be tested through trace statistics (λtrace) and the maximum eigen statistics (λmax). furthermore, vector error correction model (vecm) is used when variables are cointegrated in the long run otherwise one can use the vector autoregressive model for the variable. finally, wald test is used to find out the existence of short-run relationships. 4. results and discussion 4.1 stationarity test we use the time series data and it is important to check the data for stationarity. we use the adf test to check for unit root. the results of unit root or adf test are listed in table 1 as follows. table 1: augmented dickey-fuller (adf) test results variables trend and intercept order of integration level of variable 1st difference t-test probability t-test probability fd -2.43 0.36 -6.57 0.00 i(1) journal of applied economics and business studies, volume. 2, issue 1 (2018) 37-46 https://doi.org/10.34260/jaebs.214 41 ex -2.23 0.46 -4.86 0.00 i(1) exc 1.72 0.99 -4.40 0.01 i(1) gcf -2.52 0.32 -5.50 0.00 i(1) gdppc -2.30 0.42 -6.91 0.00 i(1) inf -2.30 0.42 -6.91 0.00 i(1) source: own estimation. the adf test results in table 1 shows that all variables are non-stationary at level and they become stationary after at 1st difference. when all variables are i(1), then we have cointegration test to check for the linear relationship between the variables (brima & pearce, 2015). johansen cointegration test shall be used to investigate the long run relationship; however, it is important to determine the lag length using the akaike information criterion (aic). 4.2 determination of the optimal lag selection in order to run the johansen’s test of co-movement and the vecm, the optimal lag selection criteria w adopted to determined the lag length that would be used in estimation. various lag length criteria such as logl, sequential modified lr test statistic, final prediction error (fpe), schwarz information criterion (sc), akaike information criterion (aic), and hannan quinn information criterion (hq) have been used to determine the optimal lag length indicated by *. it is indicated in table 2 that the optimal lag length based on the majority of criteria is two. table 2: lag length criteria lag logl lr fpe aic sc hq 0 -690.26 na 5.81e+12 46.42 46.69 46.51 1 -520.98 259.55 8.47e+08 37.53 39.49* 38.16 2 -466.26 62.04* 3.33e+08* 36.28* 39.93 37.45* source: own estimation. 4.3 co-integration of variables to check the co-integration between the variables, johansen’s co-integration test of trace statistics (λtrace) and the maximum eigen statistics (λmax) were used. the results of the johansen’s co-integration test are presented in tables 3 and 4. mian nasir uddin, muhammad tariq & saleem khan 42 table 3: unrestricted co-integration rank test (trace) hypothesized equations eigenvalues trace statistic critical value prob no 0.93 186.19 95.75 0.00 at 1 0.81 107.42 69.81 0.00 at 2 0.71 59.15 47.85 0.00 at 3 0.35 22.29 29.79 0.28 at 4 0.26 9.65 15.49 0.31 at 5 0.02 0.70 3.84 0.40 source: own estimation. table 4: unrestricted co integration rank test (maximum eigenvalue) hypothesized equations eigenvalues trace statistic critical value prob none 0.93 78.76 40.07 0.0 at 1 0.81 48.27 33.87 0.00 at 2 0.71 36.85 27.58 0.00 at 3 0.35 12.64 21.13 0.48 at 4 0.26 8.94 14.26 0.29 at 5 0.02 0.70 3.84 0.40 source: own estimation. results of both trace and maximum eigen values indicate three co-integrating equations at 5% significance level. when variables are found cointegrating at i(1), then we use both (λ trace) and (λ max) values in estimating the vector error correction model (vecm). 4.4 vector error correction model (vecm) the results of vector error correction model (vecm) are presented in table 5. it shows that there are long-run relationships between fiscal deficit and selected macroeconomics variables. this test also indicates the speed of adjustment from disequilibrium to the equilibrium. coefficient values of all the coint equations 1, 2 and 3 are negative and statistically significant indicating a trend towards equilibrium adjustment. table-5: vector error correction model (vecm) variables coefficient s. error t-value p-value cointeq1 -2.000001 0.270230 -7.509379 0.0000 cointeq2 -0.610761 0.257857 -2.368602 0.0340 cointeq3 -0.228549 0.062938 -3.631337 0.0030 d(fd(-1)) 0.538951 0.193443 2.786105 0.0154 d(fd(-2)) 0.409828 0.171841 2.384927 0.0330 d(ex(-1)) -0.577595 0.155129 3.723324 0.0026 d(ex(-2)) 0.095028 0.177754 0.534606 0.6019 d(exc(-1)) 0.130318 0.142519 0.914392 0.3772 d(exc(-2)) -0.039340 0.086768 -0.453388 0.6577 d(gcf(-1)) 0.436087 0.287419 1.517255 0.1531 d(gcf(-2)) 0.272770 0.261824 1.041807 0.3165 d(gdppc(-1)) -0.000197 7.89e-05 2.491265 0.0270 d(gdppc(-2)) -1.61e-05 7.31e-05 0.219743 0.8295 d(inf(-1)) 0.487991 0.152115 -3.208034 0.0069 d(inf(-2)) 0.317594 0.074637 -4.255179 0.0009 c -1.181520 1.217501 -0.970447 0.3495 r2 0.905955 adjusted r2 0.7974 source: own estimation. journal of applied economics and business studies, volume. 2, issue 1 (2018) 37-46 https://doi.org/10.34260/jaebs.214 43 similarly, wald test is used to check for the pair wise relations of the fiscal deficit with the selected macroeconomic variables given in table 6. wald test indicates that there is no short-run relationship of gcf and exc with the fiscal deficit while ex, gdppc, and inf have a statistically significant short run relationship with the fiscal deficit. table 6: wald test results variable chi -square d.f p-value ex-fd 13.87492 2 0.00 gcf fd 2.358778 2 0.30 gdppc fd 7.190731 2 0.02 inf fd 18.24464 2 0.00 exc fd 2.919783 2 0.23 source: own estimation. 4.5 diagnostic and other stability tests of the model once the model is estimated, it is important to check the model for autocorrelation and heteroskedasticity problems prior to using it for prediction or policy analysis. the results of the autocorrelation and heteroskedasticity are given respectively in table 7 and 8. table 7 and 8 indicate that based on the breusch-godfrey serial correlation lm test, we conclude that the model does not have a serial correlation heteroskedasticity problems. table 7: breusch-godfrey serial correlation lm test results f-stat 1.89 prob. f(2,11) 0.19 obs*r-squared 7.432 prob. chi-square(2) 0.02 source: own estimation. table 9: heteroskedasticity test: breusch-pagan-godfrey f-statistic 1.089610 prob. f(18,10) 0.46 obs*r-squared 19.20700 prob. chi-square(18) 0.37 scaled explained ss 4.013151 prob. chi-square(18) 0.99 source: own estimation. finally, the data is checked for the normality and for that jarque-bera test is used and the results are provided in figure 3. the residual of this model is distributed normally because the p-value of the jarque-bera is more than 5% and we accept the null hypothesis of normality. mian nasir uddin, muhammad tariq & saleem khan 44 0 1 2 3 4 5 6 7 8 -1.00 -0.75 -0.50 -0.25 0.00 0.25 0.50 0.75 1.00 1.25 series: residuals sample 1988 2016 observations 29 mean 4.14e-16 median 0.019964 maximum 1.016457 minimum -0.890475 std. dev. 0.427998 skewness 0.090431 kurtosis 3.079530 jarque-bera 0.047169 probability 0.976691 figure 3: normality assessment the stability test is another important part of the diagnostics tests which is estimated by cusum test. cusum technique is shown in the figure 4 below. cusum test indicates that the model is stable with the 5% level of significance as the blue line in the above figure is within both critical lines i.e. dash lines. -12 -8 -4 0 4 8 12 04 05 06 07 08 09 10 11 12 13 14 15 16 cusum 5% significance figure 4: cusum stability test results 5. conclusion the study analyzed the short and long-run relationship between selected macroeconomic variables namely exports, gdp per capita, exchange rate, inflation, gross capital formation and fiscal deficit for pakistan over the period 1985 to 2016. augmented dickey-fuller (adf) test was applied to the data for checking the stationarity in data. the adf the results showed that all the variables were stationary at first difference. knowing the data was stationary at i (1), johansen co-integration test and vector error correction method were used for investigation the long run relationship between dependent and independent variables. finally, the existence of short-run relationship was checked by using the wald test. the results showed that exports, exchange rate, gdp per capita, inflation, gross capital formation are the most important determinant of fiscal deficit in pakistan. this study concluded that exports, gdp per capita and inflation has a high impact on fiscal deficit in pakistan. journal of applied economics and business studies, volume. 2, issue 1 (2018) 37-46 https://doi.org/10.34260/jaebs.214 45 references agenor, r, p., & montiel, j. p. (1999), development macro-economics, princeton university press, 2nd edition. agha, a. i., & khan, m. s. (2006). an empirical analysis of fiscal imbalances and inflation in pakistan. sbp research bulletin2(2). anwar, m., & ahmad, m. (2012). political determinants of budget deficit in pakistan: an empirical investigation (no. 135). hwwi research paper. brima, s., & pearce, e. (2015). budget deficit and macroeconomic variables in sierra leone: an econometric approach. journal of economics and sustainable development, 6(4) chaudhary, m. a., & abe, k. (1999), pakistan economy: current situation and future prospects. chiba university economic journal, 14(1),49-85. epaphra, m. (2017). analysis of budget deficits and macroeconomic fundamentals: a varvecm approach. journal of economics & management, 30, 0-57 government of pakistan (2016/2017). pakistan economic survey http://www.finance.gov.pk jagirani, t., abro, k., &tahir, h. (2016). a vector error correction model (vecm): an approach to short-run and long-run causality between solvency and liquidity: a study of oil & gas development company limited (ogdcl), pakistan. international journal of management sciences and business research, 5(4). javid, a., arif, u., & arif, a. (2011). economic, political and institutional determinants of budget deficits volatility in selected asian countries. the pakistan development review 50(4), 649–662. juseoff, k., shamsudin, n., mohamad, z., jusoh. n., baharuddin, n. s., & asari, h. (2011), a vector error correction model (vecm) approach in explaining the relationship between interest rate and inflation towards exchange rate volatility in malaysia, world applied sciences journal, 49-56. peacock, a. t. (1961). and j. wiseman. the growth of public expenditure in the united kingdom. romer, c. (1986). spurious volatility in historical unemployment data. journal of political economy, 94(1), 1-37. safdar, f., & padda,i. (2017). impact of institutions on budget deficit: the case of pakistan. numl international journal of business & management. 12(1),2410-5392. mian nasir uddin, muhammad tariq & saleem khan 46 torayeh, n. m. (2015). fiscal deficits in egypt: is it a macroeconomic or politicoinstitutional problem? the journal of developing areas, 49(1),365-380. uppal, j. (2011). government budget deficits and the development of the bond market in pakistan: issues and challenges. the lahore journal of economics 16.159-198 usman, o. a., & adebisi, d. g. (2017). a structural break analysis of fiscal deficit process in nigeria. the review of black political economy, 44(3-4), 341-352. wiseman, j., peacock, a. t., & veverka, j. (1961). the growth of public expenditure in the united kingdom. princeton up. world bank indicator (2016). https://data.worldbank.org. journal of applied economics and business studies, volume.4, issue 3 (2020) 85-112 https://doi.org/10.34260/jaebs.434 85 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x development of citizens’ trust through social media platforms in a public food standards agency in pakistan: role of transparency and responsiveness saman arshad1 & sobia khurram*2 1research scholar, institute of administrative sciences, university of the punjab, lahore, pakistan 2assistant professor, institute of administrative sciences, university of the punjab, lahore, pakistan abstract the problem of citizens’ mistrust in government institutions has been prevailing in various countries around the world. nowadays, government institutions attempt to curb this issue by narrowing the gap of communication with citizens through online channels such as social media, which allows citizens to access timely and authentic information. the purpose of this paper is to assess citizens’ perceptions about transparency and responsiveness in the context of social media use of a public agency in pakistan and to investigate whether the perception about transparency and responsiveness helps in the development of citizens’ trust in a government agency. an empirical investigation is conducted by employing a quantitative survey design. the survey data was taken randomly from 393 social media followers of a government agency in pakistan (punjab food authority), which was then analyzed by using the structural equation modeling technique (cb-sem) to obtain the findings. the results indicate that the transparency of information and active responsiveness ensured by the pakistani agency on social media contributes to the development of followers/citizens’ trust in that agency. this study contributes to the literature and provides empirical evidence for practical implications. multiple government institutions in pakistan and elsewhere, particularly in developing countries, can practically implement such practices in their online communication with citizens to gain citizen trust. keywords social media, trust, perceived responsiveness, perceived transparency, pakistan, cbsem jel classification o38, h11, h83, l40 * sobia.ias@pu.edu.pk saman arshad & sobia khurram 86 1. introduction in the last few decades, trust in government institutions has seen a downward trend in various countries all around the world (denhardt and denhardt, 2009; gracia and arino, 2015) including developed as well as developing countries (cheema and popovski, 2010; khan, 2016) which makes it of utmost importance for scholars and practitioners to find ways to alleviate this issue. citizens’ mistrust in government institutions makes them dysfunctional in terms of service delivery as for the successful implementation of any program, initiative, reform, or service, citizens’ support and belief is an essential requirement (gracia and arino, 2015). without the presence of trust, public institutions fail to acquire legitimacy of their decisions which has implications not only for those institutions but for the social and economic growth of the entire country (porumbescu, 2016). there is a relation of interdependence where citizens depend on government institutions to resolve social issues, make decisions, formulate policies, and to evaluate their effectiveness, and government institutions are also dependent on citizens to abide by the laws, follow the rules and regulations, and to cooperate with the state for its policies to work effectively (grimmelikhuijsen and knies, 2015), all of which is not possible without the presence of trust. for this reason, it is of crucial importance that the existing cycle of mistrust be shattered. lack of trust widens the gap between citizens and government giving rise to problems such as decreased voter turnout, non-compliance, cynicism, tax evasion, and disinterest in political participation (tolbert and mossberger, 2006; kim, 2010). citizens develop trust in government institutions based on their personal characteristics, experiences, or perceptions (cook and gronke, 2005; song and lee, 2015). scholars have identified multiple sources of citizens’ trust in government institutions (chanley et al., 2000). some citizens put a greater value on economic outcomes and performance (donovan and bowler, 2004), others are more concerned about prevailing socio-cultural issues and political factors (christensen and laegreid, 2005), while some citizens believe that government institutions are worthy of trust if they include citizens in public decisions (kim, 2010). therefore, restoration of citizens’ trust in public institutions is a complex process, however, maximum efforts can be exerted to reach and communicate with as many citizens as possible so as to gain their trust. in this day and age, the use of the internet and communication technologies have proliferated in both personal and professional lives of people especially in the way they communicate. government institutions have not stayed behind in this regard, ‘the use of the internet and other digital media to deliver government information and services journal of applied economics and business studies, volume.4, issue 3 (2020) 85-112 https://doi.org/10.34260/jaebs.434 87 to citizens’ is termed as e-government (song and lee, 2015, p.143). e-government platforms were introduced to bring efficiency in service delivery and to minimize the communication gap between citizens and public institutions to eliminate any misperceptions and misunderstandings regarding government decisions in order to restore citizen trust. many scholars have studied the influence of e-government on citizens’ trust in government institutions (holzer et al., 2004; morgeson et al., 2011; smith, 2011; sharoni, 2012). most of these studies conclude that e-government platforms enhance citizens’ trust in government institutions. social media is one of the fastest-growing communication channels having 2.62 billion monthly users globally (statista, 2018) and it has also proved to be one of the most effectively utilized egovernment platforms. as mentioned by linders (2012), “social media has become ‘a central component of e-government in a very short period of time’ (bertot, jaeger, and hansen, 2012)” (p. 446). further, highlighting the importance and role of social media in e-governance dwivedi et al. (2017) mention that, “social media has played a massive role in e-government as it has remarkably used a network of social interactions and has widely connected with the diversified population resulting in their enhanced participation, higher engagement, swift and effective communication, easier crowdsourcing and superior trust-building that enabled the government to take the popularity of e-government systems use a step further” (p. 102). therefore, governments all around the globe are becoming aware of the benefits of social media, and government entities at all levels are gradually turning towards social media platforms not only for information provision but also for interaction with citizens in an efficient way (criado et al., 2013; bonson et al., 2012; mergel and bretschneider, 2013). one of the hallmarks of government institutions’ social media is the heightened perception of citizens regarding transparency as the information is directly delivered by the institution and directly accessed by citizens (song and lee, 2015). another hallmark is the convenience for the institutions to remain actively present and respond to the public on social media, which enhances the perception of citizens that the institution is concerned about and responsive to the needs, demands, issues and complains of the citizens communicated through social media (al-aufi et al., 2017). therefore, assuming the presence of these two factors in the context of social media of public institutions, the study aims to investigate whether the perception about transparency and responsiveness helps in the development of citizens’ trust in a government institution in pakistan. the number of monthly active social media users in pakistan have reported to be 35 million in the year 2018 (wearesocial, 2018), which makes it one of the fastestsaman arshad & sobia khurram 88 growing channel of communication. the concept and practice of e-government as whole is still novel in developing countries like pakistan (al-aufi et al., 2017; memon et al., 2015), the government of pakistan mostly use social media platforms for one-way information dissemination (mostly news and updates). however, a change in the trend is being detected and the government is steadily moving towards the effective use of social media to stay connected with citizens (memon et al., 2015). therefore, it is significant to examine if the institutions which are employing such mechanisms are reaping favorable outcomes such as the heightened perception of citizens regarding transparency and responsiveness ultimately leading towards the development of citizens' trust in those institutions. as mentioned by diamond (2007), to build trust, “government must govern better: more transparently, responsibly, accountably, and responsively, with more active engagement with the public and in particular more rigorous respect for the law and the public interest”. the present study aims to contribute to the literature and provide empirical evidence for practical implications. the concept and practice of e-government, particularly through the channels of social media, is still novel in developing countries like pakistan. therefore, it was considered significant to examine if the institutions employing such mechanisms in pakistan are reaping favorable outcomes, the most crucial of which is trust. the study takes the case of punjab food authority which is a food safety and hygiene agency in punjab, pakistan. the next section provides a brief literature review regarding trust in government institutions, perceptions about transparency and responsiveness of the institution, and their relationship with trust as explained by previous studies. further, the hypotheses are stated, and the research model is illustrated. the section after that provides a statistical analysis of the data. lastly, the results are discussed and are aligned with the literature, and the conclusions are drawn from the key findings of the study. 2. literature review citizenship is essentially an existence of a relationship between the government and an individual. however, the concept of citizenship includes multiple aspects which can be elucidated from the definition by joppke (2007), “citizenship as status, which denotes formal state membership and the rules of access to it; citizenship as rights, which is about the formal capacities and immunities connected with such a status; and in addition citizenship as identity, which refers to the behavioral aspects of individuals acting and conceiving of themselves as members of a collectivity, classically the nation or the normative conceptions of such behaviour imputed by the state” (p. 38). therefore, citizenship is more than just identity, although identity is an important aspect of it. in journal of applied economics and business studies, volume.4, issue 3 (2020) 85-112 https://doi.org/10.34260/jaebs.434 89 post-colonial societies there are several groups with different identities of their own, therefore having a national identity is not enough, rather there should be a legitimate sense of shared citizenship across groups (lall, 2015). pakistan is a post-colonial society with people having multiple identities and a weak understanding of citizenship. lall (2015) mentions that the situation in pakistan about citizenship could be classified as “parochial citizenship” as termed by almond and verba (1963). “in developing countries, they argue that the everyday needs of life and limited social skills create parochial citizens, unaware and uninvolved in politics” (lall, 2015, p. 73). however, research also indicates that people in developing countries are more politically aware than expected. “in pakistan, despite the high awareness of rights, duties and political processes, political activity and action to assert the rights are low” (lall, 2015, p. 74). the people “are parochial in action but not in their understanding” (lall, 2015, p. 74). the results of the empirical study on pakistani citizens by lall (2015) concluded that the reason that despite having high political awareness, the people of pakistan have low political involvement is the mistrust in the system due to high corruption. “it seems here that this disengagement from politics stems from the perception of the corrupt nature of all political leaders. it seems to indicate that these … people would rather disengage entirely from what they perceived to be a ‘dirty’ system” (lall, 2015, p. 84). therefore, in pakistan, people have a strong sense of national identity but there is a distance/gap between citizens and government who function separately from each other. this happens because of the lack of trust in government officials and institutions, poor governance, and the existence of alienation and resentment among citizens towards the government (lall, 2015). 2.1. trust in government agency scholars of public administration all around the world have been increasingly studying the concept of citizens’ trust in government institutions (song and lee, 2015). levi and stoker (2000) define trust as dependent upon "whether or not political authorities and institutions are performing in accordance with normative expectations held by the public" (p.498). grimmelikhuijsen and knies (2015) links the concept of trust with trustworthiness and states that "trust in a government organization' actually consists of the extent to which it is considered 'worthy of trust' by its citizens" (p.587). citizens and government institutions have a trust-based relationship because of the existence of two basic conditions of trust i.e. risk and interdependence. there always remain uncertainty and a sense of risk among citizens regarding the government institution’s actual fulfillment of duties and responsibilities. also, its position of power and authority adds to the risk as this power can be used positively or it can be abused as saman arshad & sobia khurram 90 well. citizens commonly are not well aware or capable to evaluate the strategies behind governmental decisions. therefore, they rely on these institutions to act in the best interest of the whole society (houston and lauren, 2013). moreover, both governments and citizens are dependent upon each other for the effective and smooth functioning of the system, thus fulfilling the condition of interdependence. while citizens depend on government bodies to resolve social issues, make decisions, formulate policies, and to evaluate their effectiveness, they are also dependent on citizens to abide by the laws, follow the rules and regulations, and to cooperate with the state for its policies to work effectively (grimmelikhuijsen and knies, 2015). moving on to the perception of trustworthiness, there is a general agreement across all disciplines that it is a multidimensional concept. the three most common dimensions studied by a number of organizational trust scholars and which are also relevant to citizens' perception of government trustworthiness are competence, benevolence, and integrity (grimmelikhuijsen and meijer, 2011; grimmelikhuijsen et al., 2013; porumbescu, 2016; porumbescu, 2017). perceived competence encompasses citizens' perception of government institutions’ skills, abilities, expertise, and resources to make decisions, meet citizens' expectations, and fulfill their duties to society. while competence is a rational dimension of trust, the second dimension, perceived benevolence takes on an effective and ethical approach (houston and harding, 2013). it refers to the perception that government institutions sincerely care about the welfare of citizens, and works for their best interest without having any opportunistic motives of their own. thirdly, perceived integrity or honesty is the belief that government institutions are sincere with the public, reveals the truth, and fulfill their promises and commitments. a higher evaluation of these dimensions of trustworthiness by citizens reflects a higher level of citizens' trust in government bodies (grimmelikhuijsen and meijer, 2011; grimmelikhuijsen et al., 2013; porumbescu, 2016; porumbescu, 2017). the existence of trust enables citizens to conform with the rules, laws, and policies formulated by the government institutions which increases the acceptability of the government decisions forming a relationship of trust between citizens and government institutions (im et al., 2012; park et al., 2015; houston and harding, 2013; porumbescu, 2017). however, many scholars and practitioners have been discussing the issue of mistrust in government officials and agencies and the possible causes of it have been identified to be poor economic performance, corruption, scandals, and published information such as wikileaks (yildiz and saylam, 2013; gracia and arino, 2015). lack of trust results in alienation and distance between citizens and the government. government institutions lose their legitimacy as well as compliance with policies and journal of applied economics and business studies, volume.4, issue 3 (2020) 85-112 https://doi.org/10.34260/jaebs.434 91 collaboration by citizens which affects the functioning of an entire economy (porumbescu, 2016). since independence, pakistan has faced challenges in achieving good governance in the country. issues like political instability, inefficient policies, and corruption prevent the practice of good governance (javaid, 2010; sarfaraz, 2007). according to the transparency international (ti) report (2019) of 180 countries, pakistan ranked at 120 with a corruption perception index (cpi) of 32 thus falling in the red zone. therefore, the prevalence of such problems creates a gap/distance between citizens and government which adds to the lack of confidence and mistrust of citizens towards the government. in recent years, internet and communication technologies (icts) have paved ways for governments to directly reach and interact with citizens. egovernance is still at an early stage in pakistan, the country which approved its first it policy and action plan in the year 2000. it is majorly employed for semi-automation of service delivery and publicity and dissemination of updates (sarfaraz, 2007). it is expected that e-government channels such as social media can bridge the communication gap and open ways of bi-directional communication which can facilitate in erasing misunderstandings and confusions that might exist in the minds of citizens regarding their government and leading towards the development of trust in the governmental institutions. transparency and responsiveness are two of the pillars of good governance (yousaf et al., 2016) which can be enhanced by means of using effective communication tools such as social media platforms leading to the development of trust (jameel et al., 2019). "through identifying the factors required to improve the trust, the government can improve the governance to meet the expectations of citizens, thus, can increase the citizens' trust to improve the legitimacy of government” (yousaf et al., 2016, p. 201). so, it is important to know by what means citizens’ trust in government institutions can be reinstated and this study attempts to investigate this matter in the setting of social media platforms of a government agency in pakistan. the studies regarding governance in the context of social media in pakistan are sparse. the previous research that has considered the factors like trust in government have studied them in the context of e-government as a whole (rehman, esichaikul, and kamal, 2012; ahmad, markkula, and oivo, 2013). therefore, this study aims to contribute to the literature by focusing specifically on trust development in the social media context. 2.2. perceived transparency and trust porumbescu (2015) defines transparency as “the extent external stakeholders are afforded regular access to information about the way their public organizations operate” (p. 02). ball (2009) provides a more comprehensive and broad understanding of the saman arshad & sobia khurram 92 concept by stating that government transparency exists “when organizations promote visible decision-making, are open to public input, allow the public the maximum choice of services, and work in cooperation with other organizations for common public purposes” (p. 302). several authors have discussed the relationship between transparency and trust in government bodies (bertot et al., 2010; song and lee, 2015). according to song and lee (2015), citizens’ access to timely and updated information of a government institution on social media enhances their perception of transparency which contributes to the development of citizens’ trust in that institution. thus, the transparency of decision processes diminishes the misunderstandings that citizens might have, which allows them to understand, accept, comply, and trust those decisions (mettler, 2011; cook et al., 2010; porumbescu, 2015; buell et al., 2018). worthy (2013) assessed the effect of transparency on trust in a study conducted in the united kingdom. the study concludes that not only the availability of information (i.e. transparency) but the medium through which information is accessed also influences citizens' trust. tolbert and mossberger (2006) conducted a study in the united states regarding the effects of e-government use on citizens’ trust in government. they found out that citizens’ access to relevant information through e-government websites makes them aware of the government's actions which are taken to address the concerns of the public resulting in enhancement of citizens’ trust in the government. hence, their study supports the view that transparency contributes towards the enhancement of public trust in government institutions in a way that it enables citizens to assess and comprehend the several means through which these institutions attempt to ensure public welfare. similarly, social media and e-government websites have been investigated by porumbescu (2017) with regards to government transparency and their association with perceived government trustworthiness. he found out that citizens' use of government social media is positively associated with perceptions of trustworthiness, however, such association was not found in the case of e-government website use. in their study on the effects of different forms of transparency on citizens' perceptions of trustworthiness of south korean and dutch governments, grimmelikhuijsen et al. (2013) report a somewhat opposing view. they found out that generally, transparency does not have any influence on trust in government. however, if the disclosed information is negative, it can have negative implications for citizens’ trust in government. besides improvements in the level of trust and confidence of citizens in government, transparency can benefit the government in other ways as well (armstrong, 2011; grimmelikhuijsen and meijer, 2012; porumbescu, 2015). coglianese (2009) highlights that transparency facilitates a two-way flow of information between policymakers and journal of applied economics and business studies, volume.4, issue 3 (2020) 85-112 https://doi.org/10.34260/jaebs.434 93 citizens regarding policy decisions. moreover, the publics' eye on government matters decreases the chance of careless mistakes or corruption by the officials due to the threat of being questioned. therefore, transparency helps in the enhancement of efficiency in the accountability and financial management of the government. additionally, transparency fosters greater legitimacy of governmental decisions which makes it easier for citizens to monitor, accept, and adopt those decisions and policies (cucciniello et al., 2017). as discussed above, pakistan has faced the problem of poor governance since it came into being in 1947. the prevalence of corruption and poor governance is also attributed to the absence of transparency and openness regarding public affairs i.e. absence of citizens’ access to public information and records which restricts citizens as well as overall civil society to monitor and bring government institutions to account (ali, 2006). "this lack of transparency leads to arbitrary and non-participatory decisionmaking, inefficient project execution, and rampant financial corruption in public bodies. the lack of transparency and access to information also contributes to the sustaining of excessive bureaucratic controls and non-functioning democratic institutions" (ali, 2006, p.4). with increasing issues comes increasing criticism (ali, 2006). therefore, in response to rising criticism, the government of pakistan is making efforts to adopt transparent practices. according to pakistan vision 2025, the government aims to be more transparent, open, and responsive in order to improve governance (planning commission 2014). it has been established from the literature that e-government channels including social media platforms prove to be effective in terms of openness and easy accessibility of government information. jameel et al. (2019) and yousaf et al. (2016) in their studies reported that the factors of good governance including transparency positively influence the citizens’ trust in the government in pakistan. however, as for the investigation of the relationship between perceived transparency of government institutions and citizens’ trust through online channels in the context of pakistan, a sufficient amount of literature is not available. as already mentioned above, previous research in pakistan has mainly focused on the role of trust in government in the adoption of e-government as a whole (ahmad, markkula, and oivo, 2013, rehman, esichaikul, and kamal, 2012; rehman, kamal, and eschaikul, 2012). the development of citizens’ trust in the context of social media remains an understudied area. therefore, it is important to find out the relationship between the two in the social media context in pakistan and the present study attempts to contribute to the gap in the literature in this regard. saman arshad & sobia khurram 94 h1: perceived transparency of a government agency on social media is positively associated with citizens’ trust in that agency. 2.3. perceived responsiveness and trust hobolt and klemmensen (2005) suggest that governments need to be responsive in order to retain the trust of their voters. once elected, if government entities stop listening to the concerns of citizens, they may lose their trust and thus the chance to get reelected. according to niemi et al. (1991), perceived government responsiveness is interrelated with citizens' trust in the political institutions and the whole political system. similarly, pinkleton et al. (2012) mention in their study the association of external efficacy with trust in government. when citizens perceive their government to be responsive to their needs and demands, they feel cared for and hence develop trust in the government. furthermore, esaiasson et al. (2016) suggest that the responsiveness of government helps in developing and maintaining citizen trust even when policy decisions made by the government are undesirable. this happens because the regular and timely responsiveness of the government entities ensures citizens that their views are taken into consideration while making a decision. esaiasson et al. (2016) also mention in their study that responsiveness is not always about adopting the views of citizens in the policies, rather it also includes listening to the views and providing reasonable explanation if the citizens' opinions and ideas cannot be adopted. according to them, responsiveness encompasses three aspects. first, listen i.e. to become aware of the issues and concerns of citizens. second, explain i.e. to provide justifications of the actions taken and policies formulated. third, adapt i.e. to alter and adjust policies in accordance with the majority of citizens' views. if all of these aspects are taken care of, they can assist in the acceptance of governmental decisions. when the government listens and explains, the citizens get a feeling of being heard and their opinions being considered while formulating policies. moreover, when the government adopts the opinions of the majority of citizens in its decisions, citizens find no further reason to protest and complain. therefore, it can be said that when contacted by the citizens, the feedback provided by the government institutions, whether in terms of listening, explaining, or adapting, is regarded as responsiveness. therefore, citizens find it easier to trust these institutions for making the right decisions that may be unwanted by them but are good for the whole society. the findings of the study conducted by catterberg and moreno (2006) in 80 different societies across the world report that among other factors, external efficacy also has a significant and positive influence on citizens' trust. moreover, porumbescu (2015) highlights the increased attention given by governments all over the world to e-government and e-participation initiatives. he suggests that governments journal of applied economics and business studies, volume.4, issue 3 (2020) 85-112 https://doi.org/10.34260/jaebs.434 95 introduce such services to enhance citizens' perception of government responsiveness which ultimately is anticipated to bolster their level of trust in government institutions. hence, from this discussion, it can be inferred that the perceived responsiveness of government institutions helps in the development of citizens’ trust. however, the relationship between the perceived responsiveness and citizens’ trust is still understudied as a sufficient amount of studies investigating the association between these two concepts, particularly in the context of social media, could not be found. the scholars have mainly studied each of these constructs separately or in relationship with other factors such as transparency, political participation, or e-government (porumbescu, 2015; porumbescu, 2017; demoor, 2015; grimmelikhuijsen et al., 2012). their association with each other is investigated by only a few scholars over the past two decades (catterberg and moreno, 2006; pinkleton et al., 2012; esaiasson et al., 2016). in pakistan, jameel et al. (2019) and yousaf et al. (2016) studied the influence of responsiveness, as a factor of good governance, on the citizens’ trust in government and found a positive relationship between the two. however, as per the knowledge of the researchers, responsiveness in the online context such as social media platforms and its influence on citizen’s trust has not been studied before in pakistan, as previous studies in pakistan have mainly focused on e-government adoption (ahmad, markkula, and oivo, 2013, rehman, esichaikul, and kamal, 2012; rehman, kamal, and eschaikul, 2012). therefore, keeping the insufficient body of literature in view, it is proposed that citizens’ trust in government institutions in pakistan is influenced by the perceived responsiveness of those institutions. h2: perceived responsiveness of a government agency on social media is positively associated with citizens’ trust in that agency. figure 1: the proposed research model trust in public agency perceived transparency perceived responsiveness + + saman arshad & sobia khurram 96 3. methodology the target population for the present study is the social media (facebook and twitter) followers of an agency i.e. punjab food authority which regulates and monitors food and hygiene-related concerns within the punjab province in pakistan. as the matter of food and hygiene is important for all, therefore citizens are more active in their communication, inquiries, and are more concerned about the acquirement of foodrelated knowledge disseminated by the agency on social media. also, the selection of this agency was considered appropriate due to their active usage of social media in terms of regular posts (more than 13-16 posts per week), comments, and responses to followers’ inquiries. in pakistan, very few agencies have a social media presence, and even among those agencies having social media accounts, not many of them actively use these platforms to disseminate updates and information. the social media accounts of punjab food authority were comparatively active than other agencies who had social media presence. a random sampling method was used, and a sample was drawn out of a total of 318,454 combined followers of official facebook and twitter accounts of the agency. during the time period of april to september 2019, an online survey was conducted which used a five-point likert scale (from “strongly disagree” to “strongly agree”) to assess the perception of respondents regarding transparency and overall trust in the agency. for assessing perceptions about responsiveness, the items were negatively worded and therefore reverse coded (from “strongly agree” to “strongly disagree”). the followers were contacted at random by the social media team of pfa via means of direct messages. approximately 500 followers were contacted, out of which 393 responses were collected (78.6% response rate) {a smaller version of this dataset was used to test another comprehensive model in arshad and khurram (2020)}. to measure these constructs, existing tools were adapted; for perceived transparency and responsiveness, the tools developed by al-aufi et al. (2017) were used, and for measuring trust in the agency a tool developed by park et al. (2015) was used. once the data was collected, it was then screened for any missing values or any other problems. after the assurance, the data was then taken up for analysis. table 1 outlines the demographics characteristics of the study participants. covariance based structural equation modeling (cb-sem) technique was applied using amos 20 software (gefen et al., 2011). a two-step approach was used for analysis in which at the first step, confirmatory factor analysis was conducted in order to assess the validity of measurement items, and afterward, the structural model was tested to determine path coefficients (hair et al., 2014). the purpose of using cb-sem journal of applied economics and business studies, volume.4, issue 3 (2020) 85-112 https://doi.org/10.34260/jaebs.434 97 is that the hypotheses in this study are developed through existing literature (hair et al., 2017) i.e. the aim was to test the theory instead of developing a new one. maximum likelihood estimation (mle) technique is used as hair et al. (2014) suggests that for a large sample (more than 300), the problem of multivariate non-normality does not hold. table 1: characteristics of participants characteristic category percent (%) gender male 73.5% female 26.5% age (years) 18-21 10.4% 22-25 29.0% 26-29 22.9% 30-33 17.3% 34-37 7.1% above 37 13.4% education level high school 25.7% graduation 37.9% post-graduation 36.4% n= 393 4. results 4.1. measurement model validation prior to testing a structural model, the validation of a measurement model is essential which is achieved by performing confirmatory factor analysis. this step allows filtering the model of any items with poor factor loadings. this process helps in identifying the best indicators of latent constructs which can be then further taken into the structural model (schreiber et al., 2006). therefore, the three items demonstrating poor factor loadings (<0.6) with their latent constructs were eradicated from further analysis (field, 2005) i.e. 1 item from perceived transparency (pt4 = 0.580), 1 item from perceived responsiveness (pr1 = 0.528), 1 item from trust in agency (ta4 = 0.582) (see table 2.). table 2: descriptive statistics and factor loadings construct item code statements factor loading perceived transparency (m= 3.91, sd = 0.765, α = 0.786) pt1 “i follow government news on social media because of its clarity” 0.724 pt2 “i make sure to follow the government (agency) on social media because i believe it provides accurate information.” 0.746 pt3 “i believe that government (agency/official) provide a satisfactory level of transparency on social media.” 0.748 saman arshad & sobia khurram 98 pt4 “i believe that transparency between citizens and the government is obtainable through social media.” 0.580* trust in public institution (m= 4.11, sd = 0.746, α = 0.846) ta1 “i have confidence in the agency and its service.” 0.800 ta2 “i have confidence in the contents that the agency provides on its social media (facebook/twitter).” 0.857 ta3 “(i feel that) the agency’s social media (facebook/twitter) provides better government services to contacting citizens” 0.759 ta4 "the agency's social media (facebook/twitter) promoted me to have a positive outlook towards government authorities." 0.582* perceived responsiveness (m= 3.23, sd = 0.959, α = 0.879) pr1 “this government agency withdraws (escapes/ fails to respond) when there is an increase in the number of inquiries addressed on social media.” 0.528* pr2 “this government agency only clarifies its position on social media without following-up subsequent posts (or any comments of people).” 0.719 pr3 “this government agency does not respond to inquiries addressed on social media.” 0.844 pr4 “this government agency is not following up on citizen complaints and concerns on social media.” 0.826 pr5 “there is no specialized employee or representative authorized to respond to what is posted by the public on social media.” 0.736 pr6 "the government does not take real actions in practice that reflects what has been addressed by the public on social media." 0.686 note: *item eliminated as factor loading was lower than 0.6. m = mean, sd = standard deviation, and α = cronbach's alpha. mean, standard deviation, and cronbach alpha are calculated after the elimination of items with poor loadings. the study utilized multiple tests to assess the reliability and validity of the measurement model i.e. tests of convergent and discriminant validity. further, the overall fitness of the measurement model was also assessed to inspect if the measurement model fits the data. all the threshold values followed in the study are provided in table 3. it can be seen from the table that the values measuring convergent journal of applied economics and business studies, volume.4, issue 3 (2020) 85-112 https://doi.org/10.34260/jaebs.434 99 validity, discriminant validity and the overall model fit are within the suggested ranges. therefore, the measurement model appears to be reliable and valid. table 3: construct validity, and model fit indices measure threshold values result perceived transparency perceived responsiveness trust in public institution convergent validity factor loadings > 0.61 yes yes yes composite reliability > 0.71,2,3 0.787 0.874 0.851 average variance extracted > 0.53,4 0.553 0.584 0.657 discriminant validity maximum shared variance msv < ave5 0.601 (no) 0.036 (yes) 0.601 (yes) correlation among constructs r < √ave5 no yes yes model fit measurement model structural model normed χ2 (χ2/df) < 33 2.368 2.371 gfi > 0.903 0.957 0.956 cfi > 0.946 0.974 0.973 tli > 0.903 0.964 0.964 srmr < 0.053,6 0.043 0.058 rmsea < 0.083,6 0.059 0.059 note: r = correlation coefficient; gfi = goodness-of-fit index; cfi = comparative fit index; tli = tucker-lewis index; srmr = standardized root mean square residual; rmsea = root mean square error of approximation. 1bagozzi and yi (1988), 2byrne (2001), 3hair et al. (2006), 4fornell and larcker (1981), 5fornell et al. (2009), and 6hooper et al. (2008). however, the discriminant validity indices of perceived transparency do not meet the suggested criteria. it can be seen from table 4 that the correlation of perceived transparency with trust in public institutions is higher than its ave. it should be noted that "discrimination is harder to demonstrate when two or more constructs should be highly correlated, yet distinct, according to the theory" (bagozzi and yi, 2011, p.18), which might be the cause as these concepts are generally intermixed in governance table 4: fornell and larcker criterion variables perceived responsiveness perceived transparency trust in public institution perceived responsiveness 0.764 perceived transparency 0.096 0.743 trust in public institution 0.190 0.775 0.811 note: the bold values in diagonal depict the square root of the ave of each construct. the other values depict correlation among constructs which are statistically significant at p < 0.01, two-tailed. saman arshad & sobia khurram 100 literature. hence, to further ensure that there is a distinction between these two constructs, another method (anderson and gerbing, 1988) was utilized in which the correlation between the two variables was constrained to unity in one model and then it was compared with the model in which the correlation was allowed to be freely estimated. the χ2 difference test between constrained and unconstrained models was significant and there was a substantial improvement in model fit indices which proves the discrimination of the two constructs (see table 5). 4.2. assessment of structural relationships after the validation of the measurement model, the structural model was assessed to test the hypothesized relationships (hair et al., 2014). to test the overall validity of the structural model, model fit was assessed, and to test the strength, direction, and significance of the relationships, the parameter estimates along with p values were considered. the model fit of the structural model was assessed in the same manner as the measurement model and the fit indices of the structural model were evaluated on the same criteria as mentioned in table 3. table 3 outlines the fit indices of the structural model and it can be seen that all of the relevant indices are within the suggested ranges. prior to the testing of hypotheses, it was ensured that the issue of multicollinearity does not exist between independent variables. the analysis showed that the variable inflation factor (vif =1.010) was smaller than the accepted maximum value of 5 and the value tolerance (0.990) was above the accepted minimum value of 0.20 (hair et al., 2010). also, the variables were not highly correlated (> 0.80) which further ensures the absence of multicollinearity (berry and feldman, 1985). therefore, it was safe to proceed and test the hypotheses. next, the parameter estimates were examined to test the hypotheses. the results of path analysis (see figure 2) reveal that the followers’ perception regarding transparency of the agency on social media is positively related to their trust in the agency (β = .769, p = .000, two-tailed), which supports h1 of the study. also, a positive but a relatively weak relationship exists between followers’ perception regarding the agency's table 5: discriminant validity (constrained and unconstrained model comparison – perceived transparency and trust in public institution) model gfi cfi tli χ2 (df) ∆χ2 (df =3) constrained .849 .823 .759 201.009 (11) unconstrained .976 .980 .963 29.352 (8) 171.657* note: *p = 0.00. journal of applied economics and business studies, volume.4, issue 3 (2020) 85-112 https://doi.org/10.34260/jaebs.434 101 responsiveness on social media and their trust in the agency (β = .134, p = .003, twotailed), which provides evidence for h2 of the study. together these two variables explain 61% variance (r2 = 0.61) in the dependent variable i.e. trust. figure 2 illustrates the final structural model along with standardized parameter estimates. 5. discussion and conclusion the present study contributes to the literature as previously the studies in pakistan have mainly focused on the adoption of e-government as a whole (ahmad, markkula, and oivo, 2013, rehman, esichaikul, and kamal, 2012; rehman, kamal, and eschaikul, 2012) and therefore, trust development in the context of social media remains understudied. the findings of the study reveal that the heightened perception of citizens of pakistan related to the transparency of public agency on social media facilitates the development of citizens’ trust in the agency. this finding is consistent with the claims made in the previous studies on developed as well as developing countries (song and lee, 2015; grimmelikhuijsen et al., 2013; halachmi and greiling, 2013; buell et al., 2018). the basic idea behind this relationship is that transparency of information disseminated by government institutions on social media allows citizens to comprehend the context, procedures, and reasons behind the decisions taken by the government and to observe their operations and performance leading to enhanced confidence and trust in those institutions. therefore, it can be stated that the more is the adoption of transparent practices by the government institutions, the more would be the trust of citizens they gain. when government institutions become transparent in their communication with citizens, the citizens feel closer and more connected to these institutions which develop their positive perspective about institutions’ trustworthiness. government institutions in pakistan need to realize the effectiveness of this practice and trust in public agency perceived transparency perceived responsiveness figure 2: the final structural model with standardized path estimates note: ***p < 0.01; **p < 0.05 0.77*** 0.13** saman arshad & sobia khurram 102 move towards its active implementation to restore citizen trust which is important to ensure and maintain efficient governance in the country (yousaf et al., 2016). one important aspect about transparency has been discussed in the literature that the information which is disclosed in order to be transparent will only be effective if that information is intelligible and comprehendible by the general public (porumbescu, 2015), otherwise, the provided information will not result in any positive outcomes rather it will create more confusions. reduction in the communication gap and elimination of misconceptions existing in the minds of citizens not only develops trust but it can also greatly and positively impact the state of the relationship between citizens and government entities (fairbanks et al., 2007). on the other hand, the results also show that the perception of citizens of pakistan about public institution’s responsiveness also has positive implications for the development of trust. the finding of this study is in line with the few studies that have tested the association (niemi et al., 1991; catterberg and moreno, 2006; pinkleton et al., 2012; esaiasson et al., 2016). when citizens observe that the government institution is responsive to the comments and suggestions of the public on social media, they develop heightened perceptions about the responsiveness of the institution to the issues and concerns of citizens. this way citizens feel being taken care of and thus find it easier to trust the institution. however, it can be seen from the results that the effect size of this relationship is quite small, which means that responsiveness is not as important or significant as transparency in predicting citizens’ trust in the public institutions in pakistan. this shows that responsiveness is secondary to transparency when it comes to earning the trust of the citizens of pakistan. while the importance of responsiveness cannot be ignored, government institutions in pakistan first and foremost need to make efforts to adopt transparent practices through social media for it to become an effective tool for the development of trust. unlike developed countries, developing countries like pakistan have many challenges, problems, and constraints, therefore it is not realistic for such countries to address all of them at once to gain citizens’ trust (kim, 2009). while aiming for good governance is an idealistic approach, it may be too ambitious for developing countries (andrews, 2010). research indicates that developing countries should instead set priorities and aim for “good enough governance” or “incremental governance” which includes feasible, realistic, and achievable goal setting (kim, 2010). “its core elements are focus, what needs to be done; sequencing, when it needs to be done; approach, how it needs to be done; selectivity, what is essential and what is not; and pragmatism, what is feasible and what is not (grindle 2004; 2007)” (kim, 2010, p. 275). further kim journal of applied economics and business studies, volume.4, issue 3 (2020) 85-112 https://doi.org/10.34260/jaebs.434 103 (2010) also states that “each developing country needs to find their own solution based on their individual importance of need. a localized solution that draws on localized knowledge and regional experience. each country should build their own best fit and with it take pride and ownership in their achievement, and not solely rely on imported (or imitated) international trends or donor-driven best practices” (p. 292). therefore, to sum up, government institutions in developing countries like pakistan need to set priorities and set achievable incremental goals rather than imitating the practices of developed countries. the use of ict channels like social media is a costeffective and convenient method to reach citizens and to regain their trust by means of simple acts which can lead to heightened perceptions about good governance indicators like transparency and responsiveness. for a developing country like pakistan, such incremental achievements can be much more significant than for a developed country as it shows that the country is moving up the ladder, even if at a slow but steady pace. 5.1. conclusion the study aimed to analyze the relationship of citizens' perceptions about transparency and responsiveness of a pakistani public institution on social media with their trust in that institution. the survey data was collected from 393 social media followers of an agency in pakistan and the structural equation modeling technique was employed to analyze the data. the findings reveal that the perceived transparency of the agency on social media is positively associated with the development of followers' trust in the agency. also, the perceived responsiveness of the agency was found to be a significant predictor of trust but not as strong as transparency. this study raises an important issue that by the use of social media platforms, government agencies in pakistan can effectively and directly communicate with citizens which can facilitate the eradication of misunderstandings and confusions, thus leading to a more conducive environment for restoration of trust. the study contributes to the body of knowledge as in pakistan the usage of social media for governmental communication is at its initial stage. therefore, the study provides empirical evidence of the benefits of effectively utilizing this tool to restore the lost trust of the public in the government institutions in pakistan. citizens’ access to timely, updated, and transparent information regarding the operations and functioning of pakistani government institutions on social media and the institutions' responsiveness to the comments, suggestions, and complaints of the public on social media can aid in the development of their trust. saman arshad & sobia khurram 104 5.2. future directions every research has some shortcomings and that creates a gap and opportunity for future researchers to build on and improve the existing body of knowledge. based on this study, a few recommendations can be made for future researchers. this study was focused on a single agency, in future studies, the data from followers of multiple government institutions can be collected to assess whether the determinants of trust on social media platforms are the same for most institutions or not. future research can be conducted by considering other than online contextual settings and investigating the factors which contribute to the development of citizens' trust in public institutions. trust in the communication channel i.e. social media platforms and its impact on the overall trust in the institution can be tested. this study can also be replicated by employing qualitative methods that allow in-depth analysis of data and the emergence of interesting and useful information. references ovais ahmad, m., markkula, j., and oivo, m. (2013), “factors affecting e-government adoption in pakistan: a citizen's perspective”, transforming government: people, process and policy, vol. 7 no. 2, pp. 225-239. al-aufi, a.s., al-harthi, i., al-hinai, y., al-salti, z. and al-badi, a. (2017), “citizens’ perceptions of government’s participatory use of social media”, transforming government: people, process and policy, vol. 11 no. 2, pp. 174194. ali, m.a. (2006), “lack of transparency and freedom of information in pakistan: an analysis of state practice and realistic policy options for reform”, open society institute international policy fellowship program paper, available at: http://pdc.ceu.hu/archive/00003129/01/mukhtarali_f3.pdf almond, g. and verba, s. (1963), the civic culture: political attitudes and democracy in five nations, princeton university press, princeton. anderson, j.c. and gerbing, d.w. (1988), “structural equation modeling in practice: a review and recommended two-step approach”, psychological bulletin, vol. 103 no. 3, p. 411. andrews, m. (2010), “good government means different things in different countries”, governance, vol. 23 no.1, pp. 7-35. http://pdc.ceu.hu/archive/00003129/01/mukhtarali_f3.pdf journal of applied economics and business studies, volume.4, issue 3 (2020) 85-112 https://doi.org/10.34260/jaebs.434 105 armstrong, c.l. (2011), “providing a clearer view: an examination of transparency on local government websites”, government information quarterly, vol. 28 no. 1, pp. 11-16. arshad, s. and khurram, s. (2020), “can government’s presence on social media stimulate citizens’ online political participation? investigating the influence of transparency, trust, and responsiveness”, government information quarterly, vol. 37 no. 3. bagozzi, r.p. and yi, y. (2012), “specification, evaluation, and interpretation of structural equation models”. journal of the academy of marketing science, vol. 40 no. 1, pp. 8-34. berry, w.d., feldman, s., and stanley feldman, d. (1985), multiple regression in practice, sage publications, thousand oaks, ca. bertot, j.c., jaeger, p.t. and grimes, j.m. (2010), “using icts to create a culture of transparency: e-government and social media as openness and anti-corruption tools for societies”, government information quarterly, vol. 27 no. 3, pp. 264271. bertot, j.c., jaeger, p.t. and hansen, d. (2012), “the impact of policies on government social media usage: issues, challenges, and recommendations”, government information quarterly, vol. 29 no. 1, pp. 30–40. bertot, j.c., jaeger, p.t., munson, s. and glaisyer, t. (2010), “social media technology and government transparency”, computer, vol. 43 no. 11, pp. 53-59. bonsón, e., torres, l., royo, s. and flores, f. (2012), “local e-government 2.0: social media and corporate transparency in municipalities”, government information quarterly, vol. 29 no. 2, pp. 123-132. buell, r.w., porter, e. and norton, m.i. (2018), “surfacing the submerged state: operational transparency increases trust in and engagement with government”, working paper, harvard business school, boston, ma. catterberg, g. and moreno, a. (2006), “the individual bases of political trust: trends in new and established democracies”, international journal of public opinion research, vol. 18 no. 1, pp. 31-48. chanley, v.a., rudolph, t.j., and rahn, w.m. (2000), “the origins and consequences of public trust in government: a time series analysis”. public opinion saman arshad & sobia khurram 106 quarterly, vol. 64 no. 3, pp. 239-256. cheema, g.s. and popovski, v. (2010). building trust in government: innovations in governance reform in asia. united nations university press. christensen, t. and lægreid, p. (2005), “trust in government: the relative importance of service satisfaction, political factors, and demography”, public performance & management review, vol. 28 no. 4, pp. 487-511. cook, f.l., jacobs, l.r. and kim, d. (2010), “trusting what you know: information, knowledge, and confidence in social security”. the journal of politics, vol. 72 no. 2, pp. 397-412. cook, t.e., and gronke, p. (2005), “the skeptical american: revisiting the meanings of trust in government and confidence in institutions”, the journal of politics, vol. 67 no. 3, pp. 784-803. coglianese, c. (2009), “the transparency president? the obama administration and open government”, governance, vol. 22 no. 4, pp. 529-544. criado, j.i., sandoval-almazán, r. and gil-garcía, j.r. (2013), “government innovation through social media”. government information quarterly, vol. 30 no. 4, pp. 319-326. cucciniello, m., porumbescu, g.a. and grimmelikhuijsen, s. (2017), “25 years of transparency research: evidence and future directions”, public administration review, vol. 77 no. 1, pp. 32-44. de moor, j. (2015), “external efficacy and political participation revisited: the role of perceived output structures for state-and non-state-oriented action forms”, parliamentary affairs, vol. 69 no. 3, pp. 642-662. denhardt, r.b. and denhardt, j.v. (2009), public administration: an action orientation, wadsworth, boston, ma. diamond, l. (2007), “building trust in government by improving governance”, in 7th global forum on reinventing government: “building trust in government” sponsored by the united nations session v: elections, parliament, and citizen trust vienna. donovan, t. and bowler, s. (2004), reforming the republic: democratic institutions for the new america, pearson/prentice hall, upper saddle river, nj. dwivedi, y.k., rana, n.p., tajvidi, m., lal, b., sahu, g.p. and gupta, a. (2017), journal of applied economics and business studies, volume.4, issue 3 (2020) 85-112 https://doi.org/10.34260/jaebs.434 107 “exploring the role of social media in e-government: an analysis of emerging literature”, in baguma r., de r., & janowski t. (ed.s), icegov '17: proceedings of the 10th international conference on theory and practice of electronic governance, association for computing machinery, new york, ny, pp. 97-106. esaiasson, p., gilljam, m. and persson, m. (2016), “responsiveness beyond policy satisfaction: does it matter to citizens?”, comparative political studies, vol. 50 no. 6, pp. 739-765. fairbanks, j., plowman, k.d. and rawlins, b.l. (2007), “transparency in government communication”, journal of public affairs: an international journal, vol. 7 no. 1, pp. 23-37. field, a.p. (2009), discovering statistics using spss, sage publications, london, england. fornell, c. and larcker, d.f. (1981), “evaluating structural equation models with unobservable variables and measurement error”, journal of marketing research, vol. 18 no. 1, pp. 39-50. fornell, c., mithas, s. and morgeson iii, f. v. (2009), “commentary—the economic and statistical significance of stock returns on customer satisfaction”, marketing science, vol. 28 no. 5, pp. 820-825. gefen, d., rigdon, e.e. and straub, d. (2011), “editor's comments: an update and extension to sem guidelines for administrative and social science research”, mis quarterly, pp. iii-xiv. gracia, d.b., & arino, l.c. (2015), “rebuilding public trust in government administrations through e-government actions”, revista española de investigación de marketing esic, vol. 19 no. 1, pp. 1-11. grimmelikhuijsen, s. and knies, e. (2017), “validating a scale for citizen trust in government organizations”. international review of administrative sciences, vol. 83 no. 3, pp. 583-601. grimmelikhuijsen, s., porumbescu, g., hong, b. and im, t. (2013), “the effect of transparency on trust in government: a cross‐national comparative experiment”, public administration review, vol. 73 no. 4, pp. 575-586. grimmelikhuijsen, s.g. and meijer, a.j. (2012), “effects of transparency on the saman arshad & sobia khurram 108 perceived trustworthiness of a government organization: evidence from an online experiment”, journal of public administration research and theory, vol. 24 no. 1, pp. 137-157. grindle, m.s. (2004), “good enough governance: poverty reduction and reform in developing countries”, governance, vol. 17 no. 4, pp. 525-548. grindle, m.s. (2007), “good enough governance revisited”, development policy review, vol. 25 no.5, pp. 533-574. hair jr, j.f., babin, b.j. and krey, n. (2017), “covariance-based structural equation modeling in the journal of advertising: review and recommendations”, journal of advertising, vol. 46 no. 1, pp. 163-177. hair, j.f., black, w.c., babin, b.j. and anderson, r. e. (2014), multivariate data analysis. pearson education limited, harlow, uk. hair, j.f., black, w.c., babin, b.j. and anderson, r.e. (2010), multivariate data analysis. prentice hall, upper saddle river, nj. hair, j.f., black, w.c., babin, b., anderson, r. and tatham, r (2006), multivariate data analysis. prentice hall, upper saddle river, nj. halachmi, a. and greiling, d. (2013), “transparency, e-government, and accountability: some issues and considerations”, public performance & management review, vol. 36 no. 4, pp. 562-584. hobolt, s.b. and klemmensen, r. (2006), “government responsiveness in words and actions: policy promises and public spending in comparative perspective”, working paper, nuffield college, oxford university, oxford, uk, available at: nuffield college website: https://www.nuffield.ox.ac.uk/politics/papers/2006/workingpaper_hoboltand klemmensen.pdf holzer, m., melitskin, j., rho, s.y. and schwester, r. (2004), “restoring trust in government: the potential of digital citizen participation”, frontiers of public administration, vol. 6, pp. 6-23. hooper, d, coughlan, j. and mullen, m. (2008), “structural equation modelling: guidelines for determining model fit”, electronic journal of business research methods, vol. 6 no.1, pp. 53-60. houston, d.j. and harding, l.h. (2013), “public trust in government administrators: journal of applied economics and business studies, volume.4, issue 3 (2020) 85-112 https://doi.org/10.34260/jaebs.434 109 explaining citizen perceptions of trustworthiness and competence”, public integrity, vol. 16 no. 1, pp. 53-76. im, t., cho, w., porumbescu, g. and park, j. (2012), “internet, trust in government, and citizen compliance”, journal of public administration research and theory, vol. 24 no. 3, pp. 741-763. jameel, a., asif, m. and hussain, a. (2019) “good governance and public trust: assessing the mediating effect of e-government in pakistan”, lex localis, vol. 17 no. 2, pp.299-320. javaid, u. (2010), “corruption and its deep impact on good governance in pakistan”. pakistan economic and social review, vol. 48 no. 1, pp.123-134. joppke, c. (2007), “transformation of citizenship: status rights, identity”, citizenship studies, vol. 11 no.1, pp. 37–48. khan h.a. (2016), “the linkage between political trust and the quality of government: an analysis”, international journal of public administration, vol. 39 no. 9, pp. 665-675. kim, p.s. (2010), “building trust by improving governance: searching for a feasible way for developing countries”, public administration quarterly, vol. 34 no. 3, pp. 271-299. kim, p.s. (2009), “introduction: the aid—good governance conundrum: searching for more realistic discourse”, international review of administrative sciences, vol. 75 no. 4, pp. 555-563. kim, s. (2010), “public trust in government in japan and south korea: does the rise of critical citizens matter?”, public administration review, vol. 70 no. 5, pp. 801810. lall, m. (2012), “citizenship in pakistan: state, nation and contemporary faultlines”, contemporary politics, vol. 18 no.1, pp. 71-86. levi, m., & stoker, l. (2000), “political trust and trustworthiness”, annual review of political science, vol. 3 no. 1, pp. 475-507. linders, d. (2012), “from e-government to we-government: defining a typology for citizen coproduction in the age of social media”, government information quarterly, vol. 29 no. 4, pp. 446-454. memon, s., mahar, s., dhomeja, l.d. and pirzado, f. (2015), “prospects and challenges saman arshad & sobia khurram 110 for social media in pakistan”, paper presented at international conference on cyber situational awareness, data analytics and assessment (cybersa), london, uk. mergel, i. and bretschneider, s.i. (2013), “a three‐stage adoption process for social media use in government”, public administration review, vol. 73 no. 3, pp. 390-400. mettler, s. (2011), the submerged state: how invisible government policies undermine american democracy, university of chicago press, chicago, usa. morgeson iii, f.v., vanamburg, d. and mithas, s. (2010), “misplaced trust? exploring the structure of the e-government-citizen trust relationship”, journal of public administration research and theory, vol. 21 no. 2, pp. 257-283. niemi, r.g., craig, s.c. and mattei, f. (1991), “measuring internal political efficacy in the 1988 national election study”, american political science review, vol. 85 no. 4, pp. 1407-1413. park, m.j., kang, d., rho, j.j. and lee, d.h. (2016), “policy role of social media in developing public trust: twitter communication with government leaders”, public management review, vol. 18 no. 9, pp. 1265-1288. pinkleton, b.e., austin, e.w., zhou, y., willoughby, j.f. and reiser, m. (2012), “perceptions of news media, external efficacy, and public affairs apathy in political decision making and disaffection”, journalism & mass communication quarterly, vol. 89 no. 1, pp. 23-39. planning commission (2014), pakistan vision 2025: one nation one vision. islamabad, pakistan: ministry of planning, development, and reform government of pakistan. porumbescu, g.a. (2017), “linking transparency to trust in government and voice”, the american review of public administration, vol. 47 no. 5, pp. 520-537. porumbescu, g.a. (2016), “comparing the effects of e-government and social media use on trust in government: evidence from seoul, south korea”, public management review, vol. 18 no. 9, pp. 1308-1334. porumbescu, g.a. (2015), “using transparency to enhance responsiveness and trust in local government: can it work?”, state and local government review, vol. 47 no. 3, pp. 205-213. journal of applied economics and business studies, volume.4, issue 3 (2020) 85-112 https://doi.org/10.34260/jaebs.434 111 porumbescu, g. and im, t. (2015), “using transparency to reinforce responsibility and responsiveness”. hildreth, w.b., miller, g.j. and rabin, j. (eds.), the handbook of public administration, john wiley, san francisco, ca, pp. 120136. rehman, m., esichaikul, v., and kamal, m. (2012), “factors influencing e-government adoption in pakistan”, transforming government: people, process and policy, vol. 6 no. 3, pp. 258-282. rehman m., kamal, m. and esichaikul, v. (2012), “determinants of trust in egovernment adoption: a case study of pakistan”, amcis 2012 proceedings, pp. 12. sarfaraz, h. (2007), “e-governance: a case for good governance in pakistan”, paper presented at conference on communication policy research for improving ict governance in the asia-pacific, asian institute of management, manila. sharoni, s. (2012), “e-citizenship: trust in government, political efficacy, and political participation in the internet era”, electronic media & politics, vol. 1 no. 8, pp. 119-135. schreiber, j.b., nora, a., stage, f.k., barlow, e.a. and king, j. (2006), “reporting structural equation modeling and confirmatory factor analysis results: a review”, the journal of educational research, vol. 99 no. 6, pp. 323-338. smith, m.l. (2011), “limitations to building institutional trustworthiness through egovernment: a comparative study of two e-services in chile”, journal of information technology, vol. 26 no. 1, pp. 78-93. song, c. and lee, j. (2016), “citizens’ use of social media in government, perceived transparency, and trust in government”, public performance & management review, vol. 39 no. 2, pp. 430-453. statista. (2018), “number of social network users worldwide from 2010 to 2021 (in billions)”, available at: https://www.statista.com/statistics/278414/number-ofworldwide-social-network-users/. tolbert, c.j. and mossberger, k. (2006), “the effects of e‐government on trust and confidence in government”, public administration review, vol. 66 no. 3, pp. 354-369. transparency international. (2019), “corruption perception index 2019”, available at: saman arshad & sobia khurram 112 https://www.transparency.org/files/content/pages/2019_cpi_report_en.pdf wearesocial (2018), “digital in 2018 in southern asia”, available at: https://www.slideshare.net/wearesocial/digital-in-2018-in-southern-asia86866282/. worthy, b. (2013), “david cameron's transparency revolution? the impact of open data in the uk”, available at: https://ssrn.com/abstract=2361428 or http://dx.doi.org/10.2139/ssrn.2361428. yıldız, m. and saylam, a. (2013), “e-government discourses: an inductive analysis”, government information quarterly, vol. 30 no. 2, pp. 141-153. yousaf, m., ihsan, f. and ellahi, a. (2016), “exploring the impact of good governance on citizens' trust in pakistan”. government information quarterly, vol. 33 no. 1, pp. 200-209. https://www.transparency.org/files/content/pages/2019_cpi_report_en.pdf journal of applied economics and business studies, volume. 4, issue 4 (2020) 161-186 https://doi.org/10.34260/jaebs.448 161 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x do regulations on smoking limit cigarette demand? an empirical evidence from pakistan assad ullah khan1* and anwar shah2 1 lecturer at university of science & technology, bannu, kpk 2 associate professor at quaid-iazam university (qau), islamabad abstract enacting ordinance no lxxiv 2002, pakistan has developed crucial antitobacco policies in the last two decades. we, therefore, examine in this paper effects of both price (cigarette taxation) and non-price (public regulations on cigarette smoking) anti-tobacco policies on cigarette demand. to accomplish this objective, we examine the short and long run dynamics of cigarette demand in pakistan using auto-regressive distributed lag (ardl) estimator covering the period 1981-2018 (annual observations). the study compares price elasticity estimated with and without regulations on cigarette smoking. the result obtained shows that when price increases by 10%, cigarette consumption decreases by 5% in the short run while it decreases by 6.9% in the long run. this finding confirms that cigarette demand model, in pakistan, is in-elastic. more interestingly, the study finds that non-price regulations on smoking and cigarette demand have negative and statistically significant association. this finding confirms that non-price regulations influences the long-term dynamics of cigarette smoking in pakistan. furthermore, we obtain low price elasticity with non-price regulations and high price elasticity without non-price regulations while estimating cigarette demand equation. this empirical result is an evident of the fact that estimated cigarette price elasticity without incorporating nonprice regulations into the demand model, are upward biased. the study therefore, concludes that smoking regulation policy based on overstated cigarette price elasticity would produce ambiguous outcomes. hence, relying only on cigarette taxation (price policy) to regulate cigarette smoking would not produce desirable outcome. in addition, university education is positively and significantly associated with cigarette consumption. this finding show that our university education do not properly convey anti-smoking message to students. to reduce cigarette smoking, pakistan will have to implement stronger, more comprehensive and better enforced non-price regulations along with taxes on cigarettes. keywords smoking regulations; cigarette demand; cigarette price elasticity; cigarette advertisement, promotion; pakistan tobacco board (ptb); ardl; ecm * asadwazir82@yahoo.com mailto:asadwazir82@yahoo.com assad ullah khan and anwar shah 162 1. introduction cigarette smoking is harmful to health of all individuals (smokers as well as nonsmokers). every year cigarette smoking ( the most commonly used form of tobacco) causes more than 8 million deaths worldwide due to lung cancer, cardiovascular diseases and hypertension (who, 2020). to get rid of the menace of cigarette smoking, the who framework convention for tobacco control (fctc) advocates imposition of high taxes on cigarette production, as well as other non-price regulations on smoking such as restriction on cigarette advertisement and promotion, sales to minors, and prominent warning on packs of cigarettes (fctc, 2015). this report calls for government, policy makers and researchers to design policies that raises cigarette prices through taxation, restrict cigarette smoking in public places so that demand for cigarette may be reduced substantially. in addition, restricting cigarette advertisement and promotion, placing health warning on cigarette packages are also helpful in limiting cigarette demand. since, it is now generally believed that cigarette smoking is the single largest preventable cause of premature death. by 2020, 70% of those killed by tobacco will be from developing economies of the world (jha & chaloupka, 1999). in most develop countries, the prevalence of smoking is declining over time and so are the tobacco related illness and deaths. unfortunately, in developing countries like pakistan, the trend of smoking increases gradually (hana ross & al-sadat, 2007; jha & chaloupka, 1999). world health organization (who) reported that, in pakistan, 31.8 % male, 5.8 % female, and 19.1% adults are consuming tobacco products. and among them, 17.9 % male, 1 % female and 9.6 % adults are smoking cigarettes (who, 2015). given the high prevalence of smoking and the resulting premature death and illness, it becomes imperative for the government of pakistan to intervene in the market for tobacco. literature suggests that taxation is the most useful strategy to reduce the use of tobacco. various studies have indicated that cigarette taxation and demand are negatively associated (see: chaloupka & warner, 2000; graham, 2013; marzioni et al., 2020; randell, 2018; stoklosa et al., 2016; teixeira, 2018). literature also finds that compare to develop countries, cigarette elasticity is high in low and middle income countries. for example, walbeek (2005) shows that studies from developed countries obtained price elasticity of cigarette consumption at around -0.4 while it lies in the range of -0.4 to -0.8 for developing economies. this is supported by john, (2008) showing that price elasticity estimates of cigarette, beedi (tobacco rolled in dry leaves) and leaf tobacco in india lie in the range between -0.4 to -0.9. similarly, evidence from advance countries shows that cigarette consumption reduces by 2.5% to 5% in response to a 10% increase in price(chaloupka & warner, 2000b). in addition, we also find varied estimates of cigarette price elasticities from studies conducted in pakistan. for instance, mushtaq et al.,( 2011) finds price elasticity of smoking for pakistan greater than unity that is, -1.17. however, burki et al., (2013) estimated price elasticity is less than unity, that is, it is equal to -0.58. journal of applied economics and business studies, volume. 4, issue 4 (2020) 161-186 https://doi.org/10.34260/jaebs.448 163 a detail review of literature on cigarette consumption suggests the following research gaps with reference to cigarette price elasticity and non-price regulations on smoking. first, studies on cigarette demand, in pakistan, have estimated price elasticity with reference to taxation or price of cigarette only. for example mushtaq et al.,( 2011) regress cigarette consumption on cigarette price and income only. resultantly, the price elasticity found is greater than unity. similarly, studies from south asian region provide price elasticities to be greater than unity. for example, ( guindon et al., 2011) for india (-1.03); nayab et al., (2018) for pakistan (-1.06); del carmen et al., (2018) for bangladesh (-1.3). these studies have completely ignored the role of non-price regulations in reducing cigarette demand. we therefore, deduce that the estimated cigarette price elasticities owing to the cigarette prices only could be upward biased (over-stated). second, in pakistan the tier system of federal tax on cigarette is highly complex. the complexity arises when government increases tax on high brands of cigarettes, smokers shift to other low brands or low price cigarettes. this results in increasing volume of cigarette smoking as low brands cigarettes are afforded by large number of people. this high tax may also give rise to illicit cigarettes production and increasing demand for unreported cigarettes(nayab et al., 2018; spdc, 2018). as non-price regulations on smoking like ban on cigarette advertisement and promotion and placing health warning on cigarette packets are generally applicable on all brands of cigarette. we therefore, believe that non-price regulations on smoking along with cigarette taxes will prove highly significant in reducing cigarette smoking. in this line, one can see a number of developments in pakistan tobacco control policies since the promulgation of ordinance no lxxiv 2002. the ordinance includes restrictions on smoking in public places, restriction on tobacco companies’ marketing, and placing health warnings on cigarette packages (nayab et al., 2018). pakistan, became a member of who fctc in 2004, since then the government issued a number of ordinances and sros (statutory regulatory orders) to regulate cigarette smoking(burki et al., 2013). however, tobacco control policies in pakistan are hardly monitored and poorly implemented. beside, in pakistan, we find very little evidence on the effectiveness of non-price regulations in cigarette demand models. literature gives a varying range of estimates of the impact of increasing prices through taxation on cigarette consumption. extensive demand models for cigarette have been empirically examined in studies from developed and developing countries as well. beside price effect, these models of cigarette demand have introduced government regulation on smoking like, restrictions on cigarette smoking in work place and or public place, banning cigarette advertising and promotion, health warning on cigarette packets and restriction on selling cigarettes to minors, all of which could play a significant role in reducing demand for cigarette. in pakistan, one can find very little research studies that have examined the effect of regulations on cigarette smoking. this study attempts to empirically examine the effects of price as well as non-price cigarette regulations on the assad ullah khan and anwar shah 164 demand for cigarette in pakistan. more specifically, the study test the hypothesis that do non-price public restrictions on cigarette smoking are helpful (significant) in limiting cigarette demand? to see more detail analysis of cigarette demand, sections of the study are given as under. section 2 provides relevant literature on cigarette demand analysis. in section 3, we have presented a brief overview of cigarette production, taxation and regulations on smoking in pakistan. section 4 gives understanding of econometric model of cigarette demand. this section explain ardl estimator including a dummy variable for regulation on smoking. section 5 presents empirical estimates of the coefficients of variables included in the cigarette demand model. suggestions and policy implications are reported in the final section. 2. literature review empirical findings of the studies that have analyzed cigarette demand, gives inconclusive results. these studies give varying estimates of cigarette price elasticities owing to differences in data and estimation techniques. for instance, warner (1981) obtain price elasticity of -0.37 by conducting aggregate time series data during the period 1947 to 1978. similarly, using ridge regression techniques, fuji (1980) conduct an empirical study to find numerical values of the coefficients of cigarette demand equation. the empirical findings of the study show negative price elasticity of -0.47 while positive income elasticity of 0.22. a recent study conducted by cetin (2017) confirms that regulation on cigarette smoking matters in the estimation of cigarette elasticity. the study apply ols estimation method to monthly and quarterly data and observe that whether taxation and regulations have affected cigarette demand. empirical findings of the study confirms that pre-and post-taxation and regulations elasticities are different. there seems a policy dilemma of the kind that whether additional taxation on cigarette is desirable. on the one hand, cigarette consumption is generally considered an evil which causes serious diseases like lung cancer, heart disease, brain strokes and other skin problems. on the contrary, cigarette production and sale is a source of revenue for the national ex-checquer. hu & mao (2002) address this issue by examining the impact of cigarette taxation on cigarette demand. the study suggests additional duty on cigarette smoking to attain good health and economic gain. more specifically, simulation of the results show that when cigarette elasticity is -0.54, cigarette consumption would fall by 4.74 billion packets to a tax increase of 40 percent. this increase in taxation would add additional amount of 24.74 billion (yuan) to the revenue and would save lives of 1.44 to 2.16 million people. in addition, as a result of taxation, addition to the national ex checquer would be larger than total industrial and farmer’s income loss. similarly, hana ross & al-sadat (2007) confirms that taxation reduces cigarette consumption as well as tobaccorelated deaths. in addition, tobacco taxation is also a source of increasing revenue for the government. applying error correction journal of applied economics and business studies, volume. 4, issue 4 (2020) 161-186 https://doi.org/10.34260/jaebs.448 165 method to the time series data collected during 1990-2004, the study finds tax elasticity of -0.57 and income elasticity as 0.08. simulation of the results indicate that cigarette smoking falls by 3.37% when cigarette tax is increased by 0.40 ringgit (malaysian currency) on a packet of cigarette. this increase in cigarette tax would further add 20.8% to the government revenue. beside excise tax on cigarettes (prices), many studies have analyzed the role of nonprice government regulation, including restrictions on cigarette advertisement and promotion, health warning on cigarette packages, information campaigns, and clean indoor air restrictions. for example, it has been remained a controversial issue that whether or not advertising affects tobacco consumption. government, health-care officials and tobacco control advocates suggest that advertisement and tobacco consumption are positively associated and banning advertisement can reduce tobacco use substantially. on the contrary, tobacco industry argue that advertisement does not encourage cigarette smoking but it only increases the relative importance of a particular brand of cigarette in the market. in this line, literature provides rich empirical evidence to highlight the link between cigarette advertisement and cigarette smoking. for example, saffer & chaloupka (2000), using data on 22 oecd countries during the period 1984-1992, examines the relationship between restrictions on tobacco advertisement and tobacco use. the study concludes that tobacco advertisement and tobacco consumption varies positively. moreover, result of the study shows that a complete set of restriction on tobacco advertisement can reduce tobacco use substantially. on the contrary, a limited or weak ban on tobacco advertisement will have little or no effect at all. the result further indicates that in case all oecd countries had used complete restrictions, it will reduces 5.4% and 7.4% tobacco and cigarette use respectively. similarly, (wasserman et al., (1991) estimates a generalized linear model to empirically estimate demand for cigarette of adult and teenage. the results show that adult and teenage price elasticity are almost equal in magnitude. adult demand indicates an unstable price elasticity ranging from 0.06 in 1970 to -0.23 in 1985. furthermore, the study shows that public restrictions on smoking have statistically significant effect on both demand models for adult and teenage. unlike saffer & chaloupka (2000), nelson (2003) analysis of tobacco demands finds that restrictions on tobacco advertisement have no effect on tobacco use. the study identify that prior studies suffered from two problems. first, these studies ignore the fact that restriction on tobacco advertisement and tobacco use are simultaneously determined. that is, countries had legislated advertising bans only when tobacco consumption had fallen substantially. second, nelson (2003) indicates that in crosscountry studies, almost all countries suffer from auto-correlation. therefore, the study empirically analyze a simultaneous equations and treating restriction on tobacco advertisement as endogenous variable. beside, literature also indicates that cigarette demand is more sensitive to price, income and non-price smoking restrictions (like health warning, media bans, bans on assad ullah khan and anwar shah 166 smoking in public place ) in developing countries relative to developed countries of the world. one possible explanation for this could be larger price elasticity of cigarette demand in developing economy. for example, walbeek (2005) confirms this statement by arguing that almost all researchers are of the view that cigarette elasticity for developed counties equals -0.4 while it lies in the range of -0.4 to -0.8 for developing economies. another reason could be that individuals in developing countries have low education levels which in turn may be the reason for little or no understanding of the negative health consequences of smoking. empirical studies from advance economies indicates that only comprehensive bans on advertisement play a role in reducing cigarette demand. the reason could be that in case of limited bans the industry may be able to shift advertisement away from the banned media towards those which is yet free from such bans. however, this may not be the case for developing economies. for instance, blecher (2008) finds very interesting result for developing economies, that is, both comprehensive as well as weak bans play an important role in limiting cigarette demand. the study examines cross countries demand equations to see whether restriction on advertisement and cigarette smoking are correlated. data set comprising 51 countries (21 developed; 30 developing) during the period 1990 to 2005 was collected. the empirical result for developing countries of the study shows that both comprehensive and weak ban are effective in reducing cigarette consumption. in addition, comparative result of comprehensive and weak ban indicates that the former have a far greater negative influence (23.5%) than the later (13.6%). unfortunately, developing economies lacks effective implementation of tobacco control polices due to economic benefits in the form of tobacco agriculture, manufacture, and tax revenue. however, these economic benefits to the farmer and industry are very low compared to the harms caused by tobacco consumption. keeping in view the rising social and healthcare costs of tobacco consumption, researchers in pakistan have attempted to empirically estimate price elasticity of tobacco demand. for example, (mushtaq et al (2011) empirically estimates price elasticity of cigarette demand in pakistan. using ardl estimation techniques to the annual data from 1981 to 2009, the study finds that increasing cigarette taxation and therefore prices are helpful in limiting cigarette smoking. more specifically, the study result shows that cigarette smoking falls by 4.8% in the short-run and 11.7% in the long-run to a 10% increase in cigarette taxes (cigarette prices). ahsan et al. (2015) conduct cross-sectional study to see whether, in pakistan, health warning on a packet of cigarette are effective in reducing smoking. for this purpose, self-structured questionnaire was distributed in karachi from july to october 2014. the result indicates that out of total 1500, 1330(88.7%) did notice health warnings on cigarette packages. moreover, the study adds that 730 (54.8%) responded positively that pictorial warnings are more effective in reducing cigarette consumption. finally, the study suggest that to reduce cigarette consumption significantly, the government of pakistan should make the graphical warnings more clear and prominent. a similar study journal of applied economics and business studies, volume. 4, issue 4 (2020) 161-186 https://doi.org/10.34260/jaebs.448 167 conducted by rasool et al. (2011) finds that compliance of anti-smoking regulations by cigarette industry for showing health warning is 39% which is very low than desired. cigarette selling shops were observed in abbottabad with the aim to determine total cigarette brand availability and their compliance of anti-smoking regulations. the results show that total 18 cigarette brands are available in abbottabad city. in addition, the study finds that 38.4% varieties of cigarette brands carry health warnings on cigarette packets. mostly cigarette customers are between 20-40 years of age and buy cigarette brands that do not carry health warnings. the study suggest to the ministry of health (moh) for ensuring health warnings on cigarette packages so that smokers be aware of negative health consequences of smoking. literature has identified various social and environmental factors responsible for smoking initiation. these factors include peer pressure, domestic and occupational stress relief, nicotine dependency, parental smoking, media influence, easy availability and affordability of cigarette. in this line, nizami et al. (2011) distributes a questionnaire among 170 participants with the aim to determine factors contributing to smoking initiation and propagation. the study finds occupational stress relief mainly responsible for cigarette smoking followed by peer pressure. further, the study finds that although stress or nicotine dependency appears to be a factor of cigarette smoking in the middle age, however, smoking initiation is mainly due to peer pressure at a young age. another study conducted by alam et al., (2008) investigates the socio-economic and demographic determinants of tobacco use during 2004 to 2005. a cross-sectional survey based on self-structured questionnaire was carried out in the rawalpindi district of pakistan. using multi-stage cluster sampling with rural and urban stratification, a total of 1018 respondents (1038 rural; 980 urban) were face to face interviewed. result of the cross sectional analysis indicates that 16.5% of the respondents (33% male; 4.7% women) regularly uses tobacco. the study also finds cigarette smoking as the most commonly used form of tobacco (68.5%) followed by oral tobacco (13.5%). as regard to the demographic factors, the study finds that rural area, male and low educational status all have positive association with tobacco consumption. literature gives mixed and inconclusive results concerning the effect of taxes (cigarette price) and regulations on cigarette smoking (non-price public restrictions) on cigarette demand. few studies have ignored the effect of non-price regulations on smoking, therefore, their estimated price elasticity is more than unity. others have only focused on individual cigarette regulation like, media ban on cigarette advertisement and or placing warning on a packet of cigarette. in pakistan, we find very little empirical evidence which has analyzed the role of non-price regulations on cigarette smoking. we therefore, address this issue in detail. in this line, we attempt to estimate both the effect of price as well as non-price regulations on cigarette consumption. in the next section, we present a brief history of taxation and regulations on cigarette smoking in pakistan and compare these with the estimates of our model. 3. overview of cigarette production, taxation and regulations in pakistan assad ullah khan and anwar shah 168 one reason for the growing prevalence of smoking in pakistan might be the cheap tobacco prices and increasing affordability of cigarettes. affordability of cigarettes in pakistan, as measured by the ratio of average cigarette pack price to per capita income, was low in the early 1990s as real income was falling more rapidly. this reduction in affordability of cigarettes contributed to low cigarette consumption during this period. however, rapid increase in the real income and thereby increasing affordability might be a reason for the rise in cigarette consumption during the late 1990s (burki et al., 2013). beside, in pakistan, the federal excise duty (fed) on cigarette is highly complex, imposing different amount of taxes on the basis of retail prices. for example more recently in 2017, the re-introduction of three tier tax structure on cigarette gave rise to the increasing sale of low-brand cigarettes in the country (nayab et al., 2018; spdc, 2018). figure 1, provides trends of the domestic production of cigarettes from 1981 to 2018. the graph shows that domestic production increases from 35.8 billion sticks in 1980-81 to as high as 75.6 billion sticks in 2008-09. however, from 2009 onwards cigarette production show a declining trend till 2016-17, reaching to a level as low as 34.3 billion sticks. after the declining trend, the production of cigarette picked up to 59 billion sticks in 2017-18. this surge in the production of cigarette was mainly due to the re-introduction of third tier excise duty for low-priced brands. the fbr officials and the tobacco industry defended this move by arguing that the third tier excise duty will prove fruitful at curbing the illicit cigarette production. however, in reality a loss of 36 billion (pakistan rupees) occur to the revenue department as a result of introducing three tier excise duty on cigarettes. this loss occurred mainly because cigarette industries in pakistan shifted high tax brand cigarettes to low tax or low price brand cigarettes. this brand shifting of the cigarette companies resulted in the reduction of fed on several brands of cigarette from rs.32.98 to rs.16 in the fiscal year 2017-18 (nayab et al., 2018). figure 1 production of cigarette in millions 20,000 30,000 40,000 50,000 60,000 70,000 80,000 1985 1990 1995 2000 2005 2010 2015 cigarette production (million sticks) data source: pakistan tobacco board below, table 3.1 show the actual trend of cigarette consumption, average prices of cigarettes, and the tax revenue collected by the government of pakistan during the period 2015-18. table 3.1 cig consumption, taxes and prices (2015-2018) journal of applied economics and business studies, volume. 4, issue 4 (2020) 161-186 https://doi.org/10.34260/jaebs.448 169 year cig. consumption (million sticks) total taxes (million rs) cig. price/packet 2015 62687.24 102890 45.85 2016 53545.23 114202 57.75 2017 34350.92 83693 65.44 2018 59065.35 87450 50.86 data source: pakistan tobacco board. cigarette prices per packet have been taken from economic survey of pakistan. as evident from the table, average prices of a packet of 10 cigarettes have declined from pkr 65.44 to pkr 50.86 during the period 2017-18. however, volume of cigarette production increased from 34 billion sticks to 59 billion during this period. the government regain its lost revenue nearly equal to rs. 3757 million. however, statistics indicate that contrary to the government’s claim, the three tier structure failed to generate additional tax revenue even though cigarette production has increased considerably. beside taxation, non-price regulations on smoking also play a significant role in decreasing cigarette consumption. these regulations includes printing of health warnings on cigarette packages, prohibition of smoking in enclosed places, ban on cigarette advertising and promotion, restrictions on cigarette sale near educational institutions and sale to minors. although, the legislation of tobacco use in pakistan can be traced with the introduction of the “motor vehicle ordinance” in 1965 and later cigarette (printing of warning) ordinance no. lxxiii, 1979. however, due to high literacy rate in the country, these warning ordinances did not effectively convey the message of negative health consequences of smoking. later on, in 2002, the government of pakistan issued a comprehensive health ordinance “restrict smoking in indoor places and health warning ordinance no. lxxiv”. the ordinance restrict smoking in indoor work places such as schools, hospitals, transport and all other government buildings and offices. it bans advertisement and promotion of tobacco, restrict sales to children and prohibits sales in the premises of academic institutions (burki et al., 2013; nayab et al., 2018). in this line, pakistan signed the who fctc in 2004, which resulted in quite significant improvement in tobacco regulation policies. to fulfill its obligations required by fctc, the country issued various statutory rules and orders (sros) to amend and further regulate production and consumption of tobacco. for example, the general label warning of 1979 ordinance “warning: smoking is injurious to health” was replaced in 2002 with more specific and effective health warning that “tobacco causes cancer and heart diseases”. later, a 2010 sros called for the size (top 40% of the front and back) and rotation of the text warnings. more recently, the warning label was further strengthen by a 2018 sros which requires 50% pictorial warning on both sides of cigarette packages. assad ullah khan and anwar shah 170 similarly, health ordinance no. lxxiv, 2002 prohibits smoking in government places, job places, and transport vehicles. this restriction was relaxed by 2008 sro that allowed for specifying special smoking rooms at all government places and or job places except educational institutions and transport services. however, the controversial 2008 sro was rolled back through another sro issued in 2009, requiring all places of public interest 100% free of cigarette smoking. in the same manner, ordinance lxxiv, 2002 contains some weak restrictions regarding cigarette advertising, promotion and sponsorship. these restrictions were strengthen by a 2007 sro that provides guidelines for tobacco products advertisement. for example, it further restrict the times during which advertisement was allowed in television. in addition, the sro banes bill board advertisement, and limit the size of tobacco advertisement in print media. later, a 2009 sro further amends these tobacco advertisement guidelines by restricting tobacco companies from distributing free samples of cigarette packages, and other promotional discounts. it also restrict tobacco company sponsorship of events like, sports, dramas and movies. empirical evidence from advance economies indicates that only strong and comprehensive tobacco regulation policies can play a significant role in decreasing tobacco consumption while weak policies will have no or very little impact on tobacco consumption. however, evidence from developing economies demonstrates that even weak tobacco control policies play a role in reducing tobacco use (saffer & chaloupka, 2000). the statistical data of this paper shows that since 2009, cigarette consumption in pakistan have declined from 75 billion sticks to as low as 34 billion sticks in 2017. this demonstrates that from the last two decades, by strengthening its tobacco control policies, cigarette consumption in pakistan have declined significantly. however, with the introduction of third tier excise tax in 2017, which lowered cigarette prices of low brand, cigarette consumption picked up again reaching to 59 billion sticks in 2018. therefore, to reduce cigarette consumption, pakistan will have to enact uniform excise taxation as well as implement stronger, comprehensive and better enforced non-price smoking regulations. in this line, this study is an attempt to see whether the adoption of non-price smoking regulations play a significant role in reducing cigarette smoking. in the next section, we develop econometric model which is best capable for modeling taxation and smoking regulations on cigarette consumption. 4. methodological framework 4.1 theoretical model the utility function of our representative smokers is given by ),,( exquu = (1) subject to constraint mxpqp =+ 21 where q indicates number of cigarettes consumed, x stands for baskets of all other market goods, e shows education level, 1 p indicates cigarette prices and 2 p prices of all other goods, m shows nominal money journal of applied economics and business studies, volume. 4, issue 4 (2020) 161-186 https://doi.org/10.34260/jaebs.448 171 income (yuanliang & zongyi, 2005). assuming that individual’s educational level is constant, then the following lagrange function is maximized: )(),,(max 21 xpqpmexqul −−+=  (2) taking first partial derivatives with respect to xq, and  respectively, we obtain 0 1 =− pu q (3) 0 2 =− pu x (4) 0 21 =−− xpqpm (5) solving equations (3), (4) and (5) respectively, we get optimal quantity of endogenous variables q and x given as under )( )( 12 ** 21 ** emppxx emppqq = = (6) substituting optimal quantities * q and * x into the objective function (1), we obtain )(),(()( 21 * 21 * 21 * emppxemppquemppu = (7) equation (7) is the indirect utility function, showing maximum value of the utility that an individual derives from optimal consumption of cigarettes and all other basket of goods. to know how educational level and cigarette consumption are related, consider the dual objective function given as under: )(),(),( * euequeqz −= (8) differentiating equation (8) with respect to ''w and solving we get, * ee uu = (9) since we know that marginal utility of education is again a function of education, therefore in the manner of equation (7), equation (9) can be written as under: )),(()( ** eequeu ee = (10) in view of the relationship between cigarette consumption and educational level, two hypothesis can be formulated. that is, 1) the consumer thinks that education reduces the marginal utility of cigarette consumption and thereby increase his/her total utility 2) with the increase of education, cigarette consumption will increase to derive the former hypothesis, we optimize (maximize) equation (8) as under: 0 0 * * −= =−= eeeeee eee uuz uuz therefore, assad ullah khan and anwar shah 172 * eeee uu  (11) now by differentiating both sides of equation (10) with respect to '' e we get, eeeqee u e q uu +   = * by utilizing equation (11), the above equation can be written as 0   e q u eq (12) we know that with the increase of education, marginal utility of cigarette consumption decreases, hence 0 eq u and therefore for equation (12) to hold 0 *    e q . this clearly indicates that with the increase of educational level cigarette consumption will decrease. to prove the second hypothesis, we minimize the function given by equation (8) as under: 0 0 * * −= =−= eeeeee eee uuz uuz therefore, * eeee uu  eeeqee u e q uu +   = * 0   e q u eq we know that with the increase of education, marginal utility of cigarette consumption decreases, hence 0 eq u and therefore for equation (12) to hold 0 *    e q . this clearly indicates educational level increase cigarette consumption. theoretical model gives inconclusive result about how educational level is associated with the cigarette consumption. in the next section of the study we therefore, empirically analyze the relationship among cigarette prices, regulation on smoking, educational level and cigarette consumption (yuanliang & zongyi, 2005). 4.2 empirical model the study follows the framework in mushtaq et al. (2011) which examines the effects of price and income on cigarette consumption. however, the study seems to have missed important variables which can significantly affect cigarette demand. it might be journal of applied economics and business studies, volume. 4, issue 4 (2020) 161-186 https://doi.org/10.34260/jaebs.448 173 the reason for the high cigarette price elasticity (greater than unity) obtained in the study. we, therefore also include regulations on smoking as a controlled variable to accurately measure price elasticity and examine whether, in pakistan, regulations on smoking play additional role in reducing cigarette smoking. in equation 1, we show that cigarette consumption depends on cigarette price, real income and regulations on cigarette smoking. cigarette ),,.( sregulationinocmepricescigfnconsumptio = (1) data related to cigarette production, prices, gross domestic product (gdp), excise taxes and regulations on smoking were obtained from domestic and international institutions. more specifically, data on cigarette production, exports and imports were obtained from pakistan tobacco board (ptb). cigarette consumption was calculated by subtracting net cigarette exports from cigarette production. cigarette prices data used in the analysis for each year, were the average prices (average prices of 17 centers of one particular brand of cigarette) of a package of cigarette having 10 cigarettes. cigarette prices data were received from economic survey of pakistan 20017-18. real gross domestic product per capita (real gdp per capita) was defined by dividing real gdp on population (15 years or older). world development indicator (wdi) gives data for real gdp per capita. government of pakistan issued an ordinance titled ordinance no lxxiv, 2002. this ordinance provides for restriction of tobacco and cigarette smoking in indoor places, job places and public transport vehicles. the ordinance also prohibits advertising and promotion of cigarette, sale to minor, and distribution near educational institutions. hence, to examine the effect of regulations on smoking, the study introduces a dummy variable into the regression equation. the binary variable is taking 1 value for the period 2002 and onwards till 2018 and zero otherwise. in the next step of estimation, we includes all the relevant sros that restrict cigarette smoking. furthermore, the study includes enrollment in primary school and university education to proxy awareness against anti-smoking behavior. we conduct cigarette demand analysis, using aggregate time series data on macroeconomic variables from 1981 to 2018. the study estimates the effects of cigarette prices, real income and regulations on smoking on cigarette demand. in accordance with the economic theory, cigarette consumption was taken as a function of cigarette prices, real income and regulation on smoking. more specifically, we estimate the following cigarette demand model: tttt gypq  ++++= relnlnln 3210 (2) where t q is the quantity (millions of cigarette sticks) demand and or consumed in time period (year) ,t t p is the average price of 17 centers of a particular brand of cigarette in time period ,t t y is gdp per capita in time period ,t ‘reg’ is a dummy variable, taking 1 value for regulation on smoking for years 2002 to 2018 and 0 (zero) otherwise, and t  is the random disturbance or white noise process. assad ullah khan and anwar shah 174 to produce short-run and long-run cigarette elasticities, the study apply autoregressive distributive lag (ardl) co-integration method. this estimation method is used due to its several advantages. first, ardl method can be applied even though data series have different order of integrations. that is, pesaran & pesaran, (1997) argued that ardl method can be used even if various series are integrated of order zero i(0) or one i(1). however, if data series are integrated of order i(2) or above, the estimator thus obtained would not be valid (ouattara, 2004). in addition, in case of simultaneity, ardl gives valid estimation of the coefficients. this is most relevant as nelson (2003) points out that regulation on smoking and cigarette smoking are simultaneously determined. that is, governments suggest restriction on cigarette smoking only when reduction in cigarette demand has already been observed. moreover, h. pesaran & shin, (1999) noted that in case of small sample like ours (38 annual observations), ardl estimation method produce true as well as consistent parameters compare to johansen and juselius’s co integration techniques. therefore, this study uses ardl estimation method for cigarette demand analysis. the study specify ardl equation as under: tit p i it q i ttt pcgpcc  ++++++= − = − = −−  lnlnrelnlnln 0 2 1 1312110 (3) in equation (2), 21  and show long-run association of variables while 21 , captures immediate or short-run impacts of variables on cigarette consumption. in addition, the coefficient 3  indicates the effect of smoking regulations on cigarette demand. for complete analysis of cigarette demand, using ardl estimation, following steps are required. first, to see whether co-integration exist, the study applies f-test to the above equation (3). our null hypothesis is 0 h : .0 321 ===  this shows that there is no co-integration. on the contrary, our alternative hypothesis is 0: 3211  h , showing existence of co-integration. the study then compares the estimated f-statistic with the critical values suggested by m. h. pesaran et al. (2001). if for example, f-statistic ),0(i we accept the null hypothesis of no co-integration. however, if f-statistic > ),1(i in this case we are bound to accept the alternative hypothesis, that is, long-run co-integration exist among the variables. and if estimated f-statistic lies between the suggested critical values of )0(i and ),1(i our result will be inconclusive. second, estimate of long run co efficient is obtained by using akaike information creterion (aic) and schwarz creterion (sbc). to verify convergence towards steady state equilibrium, we estimate short term elasticities by estimating error correction mechanism (ecm) given as under: ttit p i it p i t ecmpcgc  +++++= −− − − =  1 1 3 1 210 lnlnreln (4) journal of applied economics and business studies, volume. 4, issue 4 (2020) 161-186 https://doi.org/10.34260/jaebs.448 175 in the above equation (4), the coefficient of lag ecm shows adjustment factor towards long-run equilibrium when disequilibrium or shock occur in the short-run. finally, to see over all fitness of the model, the study conduct diagnostic tests. for example, for checking serial correlation, the study uses lm test. similarly, to check heteroscedasticity, we use bresh-pagangodfrey test. and for stability of the model, we use cumulative sum of recursive residuals (cusum) line. if cusum line lies within the boundaries of 5 % significance level, it would indicate that our model is stable. 5. results and discussion table 5.1 descriptive statistics ln(cig con) ln(cig pr) ln(real inc) ln(p edu) ln(uni edu) mean 10.75 2.40 10.68 16.37 12.27 median 10.77 2.21 10.66 16.48 11.55 max 11.23 4.18 11.03 16.95 14.28 min 10.27 1.17 10.31 15.52 10.66 std. dev 0.29 0.86 0.19 0.42 1.33 skew -0.09 0.48 -0.10 -0.58 0.36 kurt 1.59 2.27 2.19 2.13 1.45 j-bera 3.20 2.31 1.10 3.33 4.65 prob 0.20 0.32 0.58 0.19 0.10 obs 38 38 38 38 38 above table 5.1 shows descriptive statistics of the included variables. all variables are log-transformed. since coefficients of log transformed are directly interpreted in percentage, hence, it makes the data more easily interpretable. in addition, the log transformed data produce more stable variances of time series (li et al., 2017). based on the probability value of jarque-bera, and the associated values of skewness and kurtosis, we reject the null hypothesis that all included variables are not normally distributed. since, all the included variables are normally distributed, the study therefore, formally estimate the ardl equation (3) to calculate cigarette price elasticities with and without regulations on smoking. however, before estimating elasticities, we apply augmented dickey-fuller (adf) test, to see whether data series are stationary or non-stationary. below, table (1) shows results of the adf test, indicating that none of the variables are stationary at level with intercept. however, by taking first difference, all series become stationary. interestingly, none of the series have integration order (2) or above, we therefore, are justified in using the ardl estimation method. table 5.2 adf unit root test at level (intercept and trend) at first difference (intercept and trend) variables t-statistics p-value t-statistics p-value cig. consumption -1.2085 0.6601 -7.3369 0.0000 cig. price -0.2641 0.9731 -4.9628 0.0003 real income -0.3422 0.9085 -3.6127 0.0104 primary education -1.9423 0.6134 -4.2038 0.012 assad ullah khan and anwar shah 176 university education -1.5821 0.7808 -4.7928 0.002 we applied ardl long run bound test procedure and obtained numerical values of the long run coefficients and joint f-statistic estimate. table 3, shows f static value and critical values at 1%, 5% and 10% respectively. the f-statistic value (12.97) is well above the suggested critical values of both i(0) and i(1) at all significance level, indicating that long run co-integration exist. table 5.3 bound testing ftest null hypothesis: no long-run relationship test statistic value significance. i(0) i(1) f-statistic 12.97197 10% 3.03 4.06 5% 3.47 4.57 1% 4.4 5.72 below table 4, shows long-run estimates of ardl. as evident from the table, cigarette prices negatively and significantly affect cigarette demand in the long-run. more specifically, coefficient of cigarette price which is –0.69, indicating that cigarette demand decreases by 6.9% to a 10% increase in cigarette prices. this finding shows that cigarette demand model in pakistan is in-elastic. our result is inconsistent with mushtaq et al. (2011), which shows an elastic cigarette demand model (lnp= -1.173). however, it is possible that mushtaq et al. (2011) may have overestimated the price elasticity of cigarette demand. to examine this, we re-estimated cigarette demand model without government regulations. the result confirms upward bias in mushtaq et al. (2011), that is, we obtained price elasticity equal to -1.22. the study concludes that regressing cigarette consumption on cigarette prices only gives upward bias in the estimation of cigarette price elasticity. to accurately measure price elasticity of cigarette, we need to include all relevant determinants of cigarette consumption including regulations on cigarette smoking. table 5.4 long run ardl estimates variables coefficients std. error t-statistic prob ln cig. price **69.0− 0.27 -2.57 0.02 ln real income *88.3− ** 0.93 -4.20 0.00 reg*ln price *50.0− 0.28 -1.82 0.08 reg*lnincome *98.3 ** 1.25 3.18 0.00 ***, ** and * indicate 1%, 5% and 10% significance levels below table 5, shows short-run effects of variables on cigarette demand. signs of both short -run and long-run coefficients are almost same, however, the short run estimates are smaller in magnitude. that is, the short-run cigarette price elasticity (50%) is smaller than long-run (69%). table 4, shows that when cigarette prices are increased by 10%, smoking will decreased by almost 5% in the short run. this indicates that compare to short run we can substantially decrease smoking in the long run. the lagged ecm coefficient )( 1−t ecm is given in the last row of table (5). the ecm coefficient is journal of applied economics and business studies, volume. 4, issue 4 (2020) 161-186 https://doi.org/10.34260/jaebs.448 177 negative and statistically significant at all levels. the negative and significant 1−t ecm verifies that long run relationship exist among variables. the negative coefficient of 1−t ecm (-0.72), indicates that every next year 72% shocks of the last year are adjusted back to the long run equilibrium. in addition, the interactive term “reg*lnprice” show negative and statistically significant effect on cigarette consumption. the interactive term indicates that cigarette consumption decreases by 50% higher when regulation on smoking are imposed along with the increase in cigarette prices. however, interactive term “reg*ln income” is insignificant. this result indicates that the effect of income on cigarette consumption does not depend on smoking regulations. that is, the magnitude of income effect on cigarette consumption will not change even if regulation are imposed. table 5.5 estimates of error correction model variables coefficients std. errors t-statistic p-value c *26.37 ** 8.42 4.42 0.00 ∆ln(cig cons) *36.0− 0.19 -1.88 0.07 ∆ln(cig. price) **50.0− 0.19 -2.64 0.01 ∆ln(r. income) ***81.2− 0.75 -3.73 0.00 ∆(reg*lnprice) *36.0− 0.19 -1.88 0.07 ∆(reg*ln(r.income)) ***89.2 0.82 3.52 0.00 ordinance 2002 ***99.29− 8.39 -3.57 0.00 ecm (-1) ***72.0− 0.084 -8.59 0.00 note: ***, ** and * indicates 1%, 5% and 10% significance levels respectively. furthermore, we see that regulations on smoking have negative and significant association with the cigarette consumption. the result obtained shows that cigarette consumption decreases by 29.99 million sticks when regulations on cigarette smoking are imposed by the government. this clearly indicates that beside cigarette price (taxation), regulations on smoking like health warnings, restriction on advertising and promotion of cigarette etc. significantly reduces cigarette smoking. this finding is consistent with (cetin, 2017; saffer & chaloupka, 2000; wasserman et al., 1991). the coefficient of income is although significant but negatively associated with cigarette consumption. income elasticity indicates that cigarette consumption decrease by 2.8% and 3.9%, in the short-run and long-run respectively, to a response of 1% increase in income. this finding is inconsistent with the empirical evidence from other studies on developing economies (see: cetin, 2017; siman et al., 2020; tansel, 1993). however, evidence from develop countries also show a negative relationship of income with the cigarette demand (see: koffarnus et al., 2015). the negative association of income with cigarette demand could be justified on the ground that as individual real income increases, living standard and therefore, awareness about health hazard assad ullah khan and anwar shah 178 increases. when awareness about the negative health consequence of smoking increases, individuals respond to it by reducing cigarette consumption. we conduct several diagnostic tests and report the results below in table (6). more specifically, we conduct lm test to check whether errors are serially correlated. the study also checked whether errors have normal distribution and constant variances. the statistics obtained, indicated that errors have normal distribution and have no serial correlation. furthermore, the result indicated no evidence of heteroscedasticity, that is, errors have constant variances. table 5.6 model diagnostic tests breusch-godfrey serial correlation lm test: f-statistic 1.655597 prob. f 0.2098 obs*r-squared 4.041878 prob. chi-square 0.1325 heteroskedasticity test: breusch-pagan-godfrey f-statistic 1.445588 prob. f 0.2257 obs*r-squared 9.570956 prob. chi-square 0.2142 in addition to the diagnostic tests, the study checked whether model’s parameters are stable. to observe this, the study used cumulative sum of recursive residuals (cusum) line. the graph shows that the cusum line lies inside 5% significance boundaries. the cusum line indicated that model’s parameters are stable. figure 2 plot of cumulative sum of recursive residuals by estimating equation (4), we are able to test the hypothesis that whether various government sros (anti-smoking policies) are helpful in reducing cigarette smoking. in addition, we also investigate that whether our national educational policies properly convey anti-smoking message to the youth of our country. we therefore, examines long run and short run relationships (scenario i and ii respectively) among variables in different models. in the first scenario, we examines that whether government ordinance journal of applied economics and business studies, volume. 4, issue 4 (2020) 161-186 https://doi.org/10.34260/jaebs.448 179 2002 and various sros have significantly affected cigarette smoking. in the second scenario, we investigate that whether our educational institutions convey anti-smoking message to students. using this strategy, we examines nine (9) different cigarette demand models under two different scenarios. the study uses dummy variables to examine the effects of non-price regulations (ordinance and sros) on cigarette smoking. below, table (5.7) and (5.8) show long run and short run empirical results of included variables in (5) different models under the scenario (i). in the first scenario, the study finds that cigarette prices, across all models (equations) both in the short run and long run, are negatively and significantly associated with cigarette consumption. more specifically, cigarette price elasticities are higher in the long run (varies from -0.79 to -0.96) than the short run (varies from -0.42 to -0.57). the result finds indicate that cigarette prices (tax on cigarettes) is an important determinant of cigarette demand that play a more significant role in reducing cigarette smoking in the long run. in the same line, real income has negative and statistically significant effects on cigarette demand across all equations (except few models) of the first scenario. this shows that wealthy individuals avoids risky behavior like cigarette smoking. results of table (5.8) shows that in the short run government ordinance 2002 and all sros have negative and statistically significant effects on cigarette smoking. the estimated value of the effects of nonprice regulations on cigarette demand varies in magnitude from -21.62 (million sticks) to -40.09 (million sticks). in the long run only ordinance 2002 and sro 2003 (restriction on cigarette advertisement) have negative and statistically significant effects on cigarette smoking (as shown in table 7). this clearly indicates that government non-price anti-smoking policies are more effective in the immediate period. for long run dynamic effect of non-price cigarette regulations, the government of pakistan will have to pursue and monitor its proper implementation. in the second scenario, the study test the hypothesis that whether national educational institutions of pakistan properly convey anti-smoking education (awareness) among students. the study proxy national educational level with total enrollment in primary school and total enrollment of students in pakistani universities. model (2) and (3) of the scenario ii, regress cigarette consumption on primary and university educational level along with price and income variable. however, model (4) and (5) further includes regulation ordinance 2002 beside primary and university educational level. empirical results of educational level in both primary school and universities are shown below in table (5.9) and (5.10), under the scenario ii. throughout the second scenario and across all models, the study finds very interesting and consistent results. the result obtained shows that both primary and university educational levels are significantly associated with cigarette consumption. although primary educational level shows negative effect on cigarette consumption, however, on the contrary university education has positive effect on cigarette smoking. the result obtained indicates that cigarette smoking increases with the increase of enrollment in pakistani assad ullah khan and anwar shah 180 universities. this further highlight the fact that university education, in pakistan, do not properly convey anti-smoking message to students. the reason for the positive outcome could be that students in university are far away from parental affluence (home restrictions on smoking). second, in university usually peer pressure (friends and classmates) influence students towards risky behavior like cigarette smoking and alcohol use. this finding of positive association between university education and cigarette smoking is consistent with (wechsler et al., 1998; yuanliang & zongyi, 2005). table 5.7 long run estimates of ardl scenario-i variable model 1 model 2 model 3 model 4 model 5 lncig. price -0.79*** -0.68** -0.73*** -0.90** -0.96* (0.206) (0.267) (0.242) (0.454) (0.563) lnreal income -1.96* -3.88*** -.3.30*** -2.10* -1.89 (1.032) (0.925) (0.726) (1.198) (1.469) regulations ordinance (2002) -41.38*** (12.902) sro (2003) -27.27* (13.993) sro (2009) -50.92 (38.39) sro (2010) -91.35 (66.162) bound test critical values i(0) 3.12 and i(1) 4.25 at 5% level of significance f-static 10.62 11.10 11.06 7.00 6.79 standard error (s.e) are in parenthesis ( ). ***, **, and * represent coefficient significant at 1%, 5%, and 10% respectively. table 5.8 short run estimates of ardl scenario-i variables model 1 model 2 model 3 model 4 model 5 cons 18.82** 37.26*** 36.03*** 17.46** 14.07* (8.343) (8.423) (8.231) (8.255) (8.521) ∆ ln(cig. price) -0.47*** -0.49** -0.57*** -0.47** -0.42** (0.106) (0.179) (0.178) (0.208) (0.210) ∆ln(real income) -1.17 -2.81*** -2.61*** -1.10 -0.87 (0.723) (0.753) (0.714) (0.719) (0.737) regulations) ordinance (2002) -29.89*** (8.392) sro (2003) -21.62** (10.164) sro (2009) -26.63* (15.63) sro (2010) -40.09** (18.953) journal of applied economics and business studies, volume. 4, issue 4 (2020) 161-186 https://doi.org/10.34260/jaebs.448 181 ecm term(-1) -0.59*** -0.72*** -0.52*** -0.43*** -0.79*** (0.102) (0.080) (0.075) (0.063) (0.089) standard error (s.e) are in parenthesis ( ). ***, **, and * represent coefficient significant at 1%, 5%, and 10% respectively. table 5.9 long run ardl estimates scenario-ii variable model 1 model 2 model 3 model 4 ln cig. price -1.08*** -0.76** -0.49* -0.98** (0.238) (0.135) (0.288) (0.232) ln real income -0.97 -1.70** -.3.32*** -3.36*** (1.02) (0.701) (0.988) (0.827) ln primary education -1.00** -1.03* (0.400) (0.596) ln university education 0.20*** 0.22** (0.061) (0.119) regulations ordinance (2002) -44.78*** -31.24*** (13.632) (11.619) bound test critical values i(0) 2.87 and i(1) 4.00 at 5% level of significance f-static 11.39 11.82 10.97 10.85 standard error (s.e) are in parenthesis ( ). ***, **, and * represent coefficient significant at 1%, 5%, and 10% respectively. table 5.8 short run ardl estimates scenario-ii variables model 1 model 2 model 3 model 4 cons 23.32*** 21.65*** 41.49*** 34.62*** (7.789) (7.599) (8.336) (8.201) ∆ ln(cig. price) -0.67*** -0.62*** -0.33* -0.47** (0.122) (0.109) (0.199) (0.181) ∆ln(real income) -0.61 -1.38** -2.44*** -2.71*** (0.692) (0.697) (0.776) (0.724) ∆ ln(primary edu) -0.62** -0.69* (0.230) (0.358) ∆ ln(university edu) 0.16*** 0.22* (0.059) (0.119) regulations) ordinance (2002) -30.22*** -25.19*** (8.018) (8.446) ecm term (-1) -0.66*** -0.80*** -0.67*** -0.80*** (0.067) (0.112) (0.069) (0.083) standard error (s.e) are in parenthesis ( ). ***, **, and * represent coefficient significant at 1%, 5%, and 10% respectively. conclusion and recommendations: from the estimated model of the study we concludes, that beside cigarette taxation, regulations on smoking play a significant role in reducing cigarette consumption. however, in pakistan, research concerning the effects of smoking-regulation on cigarette demand is missing. therefore, this study is an attempt to examine whether nonassad ullah khan and anwar shah 182 price regulations on smoking, in pakistan, are helpful in reducing cigarette demand over the period 1981-2018. the findings of this study confirm the hypothesis that regulations on smoking are playing highly significant and quite effective role in limiting cigarette demand. furthermore, the estimated price elasticities are 5% in the short-run and 6.9% in the long run. this show that demand for cigarette smoking, in pakistan, is in-elastic.. beside, enrollment in universities has positive effects on cigarette. that is, cigarette smoking increases with the increase of enrollment in university education. this is alarming and show that education at university level do not educate students about the danger of cigarette smoking. based on findings of the study, following major policy suggestions are recommended. first, empirical findings of this study is evident of the fact that regulations on cigarette smoking have negative but highly significant impact on the demand for cigarette consumption. regulations on smoking is more effective policy as it is generally applicable to all brands of cigarette. therefore, to reduce cigarette consumption, pakistan will have to announce and ensure implementation of stronger, more comprehensive and better enforced regulations on cigarette smoking. second, although price and real income both are negatively associated with cigarette consumption. however, cigarette affordability in terms of real income has also increased. therefore, to further discourage smoking, the study recommends to the government to increase tobacco taxes to make cigarettes less affordable. third, as university education is positively associated with cigarette consumption. this means that national educational policies do not properly convey the message that smoking causes negative health consequences. therefore, the study recommends that seminars should be arranged at university level which disseminate information about the negative effects of smoking on health. although, this study analyzes important determinants which play significant role in reducing cigarette demand. however, there are few limitations of this study as well. fist, tobacco companies usually manipulate the actual data. for example tax on cigarette is an important determinant of cigarette demand, however, illicit cigarette production tactics are used by tobacco companies to minimize the role of taxes in limiting cigarette sale. however, the study is limited by the availability of illicit cigarette production data. second, in pakistan, large number of cigarettes sale is unreported, therefore the calculated cigarette price elasticity in this study may be overstated. when price of taxable cigarettes increases, individual might shift to unreported or non-custom (smuggling cigarette) cigarette consumption. therefore, it is possible that the estimated price elasticity may not have reflected the actual decline in cigarette consumption. again, this study is limited by the availability of unreported or illegal cigarette consumption data. however, unlike other studies conducted in pakistan, this study beside cigarette prices have analyzed the role of nonprice regulations on smoking. since, regulations on smoking are generally applicable on all types of cigarettes journal of applied economics and business studies, volume. 4, issue 4 (2020) 161-186 https://doi.org/10.34260/jaebs.448 183 (reported or unreported), we therefore, have made efforts to actually show the decline in cigarette smoking as a result of both cigarette price and regulations on smoking. third, the study is limited by analyzing the role of religion in reducing cigarette consumption. pakistan is a majority muslim populated country and the religion “islam” promotes good health of human being. islam forbids every addicted and bad goods like use of alcohol and cigarette smoking. however, the study is limited by the availability of data to empirically examine the role of religion on cigarette smoking. references ahsan, m., saad, m., jawed, m., akhtar, s., haseeb, a., bilal, m., & ahmed, a. (2015). impact of tobacco health warnings on smokers in pakistan. journal of the pakistan medical association, 66(1), 59–62. alam, a. y., iqbal, a., mohamud, k. b., laporte, r. e., ahmed, a., & nishtar, s. (2008). investigating socio-economic-demographic determinants of tobacco use in rawalpindi, pakistan. bmc public health, 8, 1–9. https://doi.org/10.1186/14712458-8-50 blecher, e. (2008). the impact of tobacco advertising bans on consumption in developing countries. journal of health economics, 27(4), 930–942. https://doi.org/10.1016/j.jhealeco.2008.02.010 burki, s., pasha, a., pasha, h., john, r., jha, p., baloch, a., kamboh, g., cherukupalli, r., & chaloupka, f. (2013). the economics of tobacco and tobacco taxation in pakistan. https://tobacconomics.org/wpcontent/uploads/2014/05/pakistanreport_may2014.pdf cetin, t. (2017). the effect of taxation and regulation on cigarette smoking: fresh evidence from turkey. health policy, 121(12), 1288–1295. https://doi.org/10.1016/j.healthpol.2017.09.015 chaloupka, f. j., & warner, k. e. (2000a). chapter 29 the economics of smoking. in handbook of health economics. https://doi.org/10.1016/s1574-0064(00)80042-6 chaloupka, f. j., & warner, k. e. (2000b). the economics of smoking. handbook of health economics, 1, 1539–1627. del carmen, g., fuchs, a., & genoni, m. e. (2018). the distributional of impacts of cigarette taxation in bangladesh. in the distributional of impacts of cigarette taxation in bangladesh. world bank. https://doi.org/10.1596/30011 emmanuel guindon, arindam nandi, frank chaloupka, p. j. (2011). why india should increase taxes on cigarettes | vox, cepr policy portal. https://voxeu.org/article/why-india-should-increase-taxes-cigarettes fctc, w. (2015). the who framework convention on tobacco control: an overview background. fuji, e. t. (1980). the demand for cigarettes: further empirical evidence and its implication for public policy. applied economics, 12, 479–489. assad ullah khan and anwar shah 184 graham, k. m. (2013). eating on an interconnected planet. environmental research letters. hu, t. w., & mao, z. (2002). chinese economy. china report, 2(6), 8–12. https://doi.org/10.1177/000944556600200602 jha, p., & chaloupka, f. (1999). curbing the epidemic. governments and the economics of tobacco control. washington (dc): the world bank. john, r. m. (2008). price elasticity estimates for tobacco products in india. health policy and planning, 23(3), 200–209. https://doi.org/10.1093/heapol/czn007 koffarnus, m. n., wilson, a. g., & bickel, w. k. (2015). effects of experimental income on demand for potentially real cigarettes. nicotine and tobacco research, 17(3), 292–298. https://doi.org/10.1093/ntr/ntu139 li, x., fan, y., & wu, l. (2017). co2 emissions and expansion of railway, road, airline and in-land waterway networks over the 1985–2013 period in china: a time series analysis. transportation research part d: transport and environment, 57, 130– 140. https://doi.org/10.1016/j.trd.2017.09.008 marzioni, s., pandimiglio, a., & spallone, m. (2020). excise gap and price in the italian cigarette industry. applied economics. https://doi.org/10.1080/00036846.2020.1733480 mushtaq, n., mushtaq, s., beebe, l. a., mushtaq, n., mushtaq, s., & beebe, l. a. (2011). economics of tobacco control in pakistan : estimating elasticities of cigarette demand stable url : http://www.jstor.com/stable/41320190 economics of tobacco control in pakistan : estimating elasticities of cigarette demand. 20(6), 431–435. nayab, d., nasir, m., memon, j. a., khalid, m., & hussian, a. (2018). economics of tobacco taxation and consumption. nelson, j. p. (2003). contributions to economic analysis & policy cigarette demand , structural change , and advertising bans : international evidence , cigarette demand , structural change , and advertising bans : international evidence ,. nizami, s., sobani, z., raza, e., baloch, n.-u.-a., & khan, j. (2011). causes of smoking in pakistan : an analysis of social factors. journal of the pakistan medical association, 61(2), 198–201. ouattara, b. (2004). www.econstor.eu. pesaran, h., & shin, y. (1999). an autoregressive distributed lag modelling approach to cointegration “chapter 11. econometrics and economic theory in the 20th century: the ragnar frisch centennial symposium. pesaran, m. h., & pesaran, b. (1997). microfit 4.0: interactive econometric analysis. oxford university press. pesaran, m. h., shin, y., & smith, r. j. (2001). bounds testing approaches to the journal of applied economics and business studies, volume. 4, issue 4 (2020) 161-186 https://doi.org/10.34260/jaebs.448 185 analysis of level relationships. journal of applied econometrics, 16(3), 289–326. randell, m. (2018). australia can help indonesia kick the habit australia-indonesia centre. australia-indonesia centre. rasool, a., ahmad, s., farooq, u., rasool, m., shah, z., hammad, m., utmanzai, a., wahab, f., sadique, i., & ahmad, j. (2011). compliance of anti-smoking regulations by cigarette industry in pakistan. journal of ayub medical college, abbottabad : jamc, 23(3), 94–96. ross, h., & al-sadat, n. a. m. (2007). demand analysis of tobacco consumption in malaysia. nicotine and tobacco research, 9(11), 1163–1169. https://doi.org/10.1080/14622200701648433 saffer, h., & chaloupka, f. (2000). pii: s0167-6296(00)00054-0 2011070814815705.pdf. health economics 19 2000 1117–1137, 1124. http://storage.globalcitizen.net/data/topic/knowledge/uploads/2011070814815705 .pdf siman, s., tawakal, m. a., risamasu, p. i. m., & kadir, r. (2020). effect of household size, working hours, health and income on consumption expenditure of poor household. enfermeria clinica, 30, 512–515. https://doi.org/10.1016/j.enfcli.2019.10.131 spdc. (2018). macroeconomic impacts of tobacco use in pakistan. december. https://doi.org/10.13140/rg.2.2.12229.81128 stoklosa, m., drope, j., & chaloupka, f. j. (2016). prices and e-cigarette demand: evidence from the european union. nicotine and tobacco research. https://doi.org/10.1093/ntr/ntw109 tansel, a. (1993). cigarette demand, health scares and education in turkey. applied economics, 25(4), 521–529. https://doi.org/10.1080/00036849300000060 teixeira, a. p. (2018). the government willingness to legislate tobacco control and government income from tobacco taxes in brazil. journal of global oncology. https://doi.org/10.1200/jgo.18.94700 walbeek, v. (2005). the economics of tobacco control in south africa. file:///c:/users/4star/downloads/thesis_com_2005_van_walbeek_c.pdf warner, k. e. (1981). the effects of the anti smoking campaign on cigarette consumption. american journal of public health, 67(7), 645–650. https://doi.org/10.2105/ajph.67.7.645 wasserman, j., manning, w. g., newhouse, j. p., & winkler, j. d. (1991). the effects of excise taxes and regulations on cigarette smoking. journal of health economics, 10(1), 43–64. https://doi.org/10.1016/0167-6296(91)90016-g wechsler, h., rigotti, n. a., gledhill-hoyt, j., & lee, h. (1998). increased levels of cigarette use among college students: a cause for national concern. journal of the american medical association, 280(19), 1673–1678. assad ullah khan and anwar shah 186 https://doi.org/10.1001/jama.280.19.1673 who. (2015). who | tobacco control in pakistan. https://www.who.int/tobacco/about/partners/bloomberg/pak/en/ who. (2020). tobacco. https://www.who.int/news-room/fact-sheets/detail/tobacco yuanliang, b., & zongyi, z. (2005). aggregate cigarette demand and regional differences in china. applied economics, 37(21), 2523–2528. https://doi.org/10.1080/00036840500358640 journal of applied economics and business studies, volume.4, issue 3 (2020) 185-198 https://doi.org/10.34260/jaebs.438 185 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x food security assessment in azad jammu and kashmir: empirical analysis of dietary diversity, current and project food demand and supply-demand gap ghulam sadiq afridi1*, muhammad ishaq2** & abdul jabbar3 1 principal scientific officer social sciences division pakistan agricultural research council islamabad-pakistan 2 director agricultural marketing and trade social sciences division, pakistan agricultural research council islamabad-pakistan 3 assistant professor, international institute of islamic economic international islamic university islamabad-pakistan. abstract this study attempts to determine the dietary diversity score for different food groups and estimates the current demand for different food groups in ajk. data for this study was collected from 1250 households covering all 10 administrative districts of ajk by following multi-stage sampling techniques through a comprehensive interview schedule. ten commodity groups (wheat, rice, milk, meat, vegetables, fruits, pulses, oil/ghee, tea and other food were developed for analysis. linear approximate-almost ideal demand system (la-aids) model was used for estimating food demand elasticities through zellner’s seemingly unrelated regression by using stata software. all the 10 food groups were found normal goods, except milk and fruits (income elasticity higher than unity). this implies that in future demand for milk and fruits would increase more proportionately as income would increase. the own price elasticities of all 10food groups were also found congruent to economic theory (wheat-0.28, rice0.65, milk-0.92, meat-0.34, fruits-0.61, vegetables-0.21, pulses-0.29, oil and ghee-0.04, tea-0.35 and other food-0.32). the income and own price elasticities depicted that wheat, vegetables, pulses and oil and ghee are necessities as compared to other food groups. the cross-price elasticity illustrated that rice could be a substitute for wheat and wheat for rice in the state of ajk. the berry’s index (82.43) depicted that food was diverse on average with a minimum index value of 61.31. the study of different factors in relation to berry’s index depicted that education and income significantly and positively affect berry’s index while location of household in urban or rural area do not affect diversification of food. inclusion of all necessities in food stamp program and ramadan package and efforts by agriculture and livestock department regarding improved production of food commodities are recommended. keywords dietary diversity, food demand, supply-demand gap, la-aids, ajk jel classification c3, c31, d01, d11, d12 * sadiqafridi@gmail.com ** ishaqecon@gmail.com ghulam sadiq afridi, muhammad ishaq & abdul jabbar 186 1 introduction according to unicef-mics survey about 28 percent children were underweight and 37 percent children were stunted in azad jammu and kashmir due to nutritional status (unicef 2008). consumption diversification is a widely used indicator especially in developing countries to capture various health indicators like, child underweight, stunting, infant mortality rate or improved birth rate, and also as an indicator in reduction of other child diseases. higher living standard of the household and balanced diet is indicated by diversity in food consumption. similarly, balanced life is signified by consumption expenditures on the non-food items including electricity, clothing, furniture, housing, transport and education; etc. therefore, analysis of consumption patterns provides an insight into the status of human resource of the country. empirical research on food consumption patterns can provide evidence on consumers’ responsiveness to price and expenditure changes that are useful in designing a country’s food policies. estimates of price and income elasticities of different foods play key role and not only help in setting administered prices but also help in designing policies regarding subsidy and tax policies. these policies also have great impact on poverty and food security. to formulate a long-term policy for food security and poverty reduction in a developing country, there is a need to understand how different groups of households respond to changes in the prices of different foods commodities. ali and farooq (2004) have found that “dietary diversity improves nutritional balance of the diet, which enhances productivity through possible improvement in health”. ajk is food deficit state in respect of production of cereals (wheat and rice), and fruit and vegetables. so far, no in-depth study has been carried out regarding production or consumption diversification at the state level. this study has been conducted in the state of ajk with objectives to examine and estimate the dietary/consumption diversification and dietary diversification score for different food groups at the state level; and to estimate current demand for different food groups in the state of ajk. this research study was the ever firs attempt to analyze household food demand as no study prior to this could be found in this neglected area. numerous research studies could be found of similar nature but all have been conducted in pakistan mostly using hies data set. like haq et al., (2011) examined household food demand pattern in punjab pakistan using hies data by employing flexible la-aids model both for rural and urban areas for wheat, rice, fruits, vegetables, milk, cooking oil, meat, and other food products. they found that households in both rural and urban areas with head of journal of applied economics and business studies, volume.4, issue 3 (2020) 185-198 https://doi.org/10.34260/jaebs.438 187 family having agriculture as profession consume less of all foods with the exception of wheat. households in both rural and urban areas with literate head of family consume more of all food products with the exception of vegetables and wheat. they further investigated that demand for all eight-food groups was price inelastic with wheat having the most price inelastic demand. all of the expenditure elasticities were positive suggested that all goods are normal with the largest expenditure elasticities found for milk followed by fruits, other food products, meat, rice, vegetables, wheat and cooking oil. similarly, ahmad et al., (2012) analyzed food consumption pattern of pakistan at various income levels of the consumers at national as well as provincial levels by using household integrated economic survey (hies), 1998-99 data. 2 methods 2.1 universe and sample size this study was aimed to estimate dietary diversity, current food demand for ajk, therefore consideration in heterogeneity in consumption pattern was kept in mind and project area was split in two different strata, i.e., northern ajk and southern ajk, as there is substantial difference in the consumption pattern between the two strata. northern ajk, comprised of neelum, muzaffaabad, hatian, bagh, haveli, sadhnoti and poonch districts, is considered as rice consuming districts and southern ajk, comprised of kotli, bhimber and mirpur is considered as wheat consuming districts. for this purpose, district neelum, muzaffarabad and poonch were selected from northern ajk while district kotli and mirpur were selected from southern ajk. district selection within strata was mainly based on income differences and/or development indicators. through multi-stage sampling techniques, a sample of 1250 households was selected with representative sample from each district as per population of the respective district. within district village and household selection was totally random. this study was conducted under “research for agricultural development research” in the year 201819. 2.2 food groups food items for in the analysis included i) wheat and wheat flour, ii) rice, iii) pulses, iv) fruits, v) vegetables, vi) milk and milk products, vii) meat, viii) tea and beverages, ix) oil and ghee, and x) other food. a single utility function for all the households was assumed to estimate the expenditure elasticities within all income level, and that preference ordering does not change across families. it was also assumed that for every commodity all the households ghulam sadiq afridi, muhammad ishaq & abdul jabbar 188 face the same prices. the last assumption in this regard is that expenditures were used as proxy for income. the household dietary diversity score (hdds) was calculated according to the frequency of all above mentioned food groups. besides hdds, berry’s index (berry, 1970) was estimated as following; 𝑩𝑰𝒋 = 𝟏 − ∑ 𝒘𝒊𝒋 𝟐𝑵 𝒊=𝟏 (1) where wij is the expenditure share of food commodity i consumed by household j. n is the total number of food items. the index was different from household to household and from rural to urban. the factors responsible for these differences were captured through the following model. 𝑩𝑰𝒋 = 𝜶 + 𝜷𝟏𝑿𝟏𝒋 + 𝜷𝟐𝑿𝟐𝒋 . . . +𝜷𝒌𝑿𝒌𝒋 + 𝝐𝒋 (2) where α and β1 to βk are the parameters to be estimated, x1 to xk are the explanatory variables while εj is the error term assumed to be normally distributed with zero mean and constant variance. to estimate the elasticities of different food groups the following linear approximate almost ideal demand system (la-aids) was employed developed by (deaton and muellbauer (1980). 𝑊𝑖 = 𝛼𝑖 + ∑ 𝛾𝑖𝑗𝑗 ln 𝑃𝑗 + 𝛽𝑖 ln (𝑥 𝑃 ∗)⁄ + 𝜙𝓏 (3) where 𝛼𝑖, 𝛽𝑖, 𝛾𝑖𝑗 are the parameters to be estimated, wi is budget share of good i ( i = 1, 2,...,n ), 𝑃𝑗 is price of good j (j = 1, 2,..., n ) and x is total expenditure and p* is stone’s price index: ln p* = σj wj ln pj this is the socioeconomics flexible linear approximate (la) version of aids and it has been used extensively in consumer analysis, and the elasticities estimated using la-aids are generally similar to those estimated using aids. expenditure elasticity: 𝜼𝒊 = 𝜷𝒊 𝝎𝒊 + (4) marshalian own and cross price elasticity for good i with respect to good j: 𝜺𝒊𝒋 = 𝜸𝒊𝒋− 𝜷𝒊𝝎𝒋 𝝎𝒊 − 𝜹𝒊𝒋 (5) hecksian own and cross price elasticity for good i with respect to good j: 𝖊𝒊𝒋 = 𝜸𝒊𝒋 𝝎𝒊 + 𝝎𝒋 − 𝜹𝒊𝒋 (6) journal of applied economics and business studies, volume.4, issue 3 (2020) 185-198 https://doi.org/10.34260/jaebs.438 189 where δij is the kronecker delta and equals 1 for own price and 0 for cross-price elasticities. projected demand for different food groups was estimated using following formulae. 𝑫𝒕 = 𝒅𝟎 × 𝑵𝒕(𝟏 + 𝒚 × 𝒆) 𝒕 (7) where 𝐷𝑡 is household demand (million metric tonnes) of a commodity group in year 𝑡, 𝑑0 is per capita consumption (kg) of the commodity group in base year i.e. the year 2017, 𝑁𝑡 is the projected population (million) in the year 𝑡, 𝑦 is the growth in per capita income (gdp), 𝑒 is expenditure/income elasticity of demand for the commodity group, and 𝑡 represents years 1, 2, 3……..n. for base year 𝑡 = 0. 𝑷𝒐𝒑𝒇𝒖𝒕𝒖𝒓𝒆 = 𝑷𝒐𝒑𝒑𝒓𝒆𝒔𝒆𝒏𝒕 × (𝟏 + 𝒊) 𝒏 (8) where 𝑃𝑜𝑝𝑓𝑢𝑡𝑢𝑟𝑒 is the future population, 𝑃𝑜𝑝𝑝𝑟𝑒𝑠𝑒𝑛𝑡 is the present population 𝑖 is population growth rate and 𝑛 is number of year 3 results and discussion table-1 divulges the descriptive statistics of the sampled respondents. according to the table on average about 85 kgs of wheat and wheat flour, about 16 kgs of rice, 90 liters of milk and milk products, about 13 kgs of meat, 43 kgs of vegetables, 64 kgs of fruits, about 8 kgs of pulses, 11 kgs of oil and ghee, 2 kgs of tea and about 4 kgs of other food was consumed per capita per annum in ajk. the average household size was found to be 6 persons per household with minimum of 2 persons and maximum of 14 persons. comparing these estimated statistics for ajk with pakistan, it is evident that per capita wheat consumption is lower in ajk as compared to pakistan (92 kgs), but consume more rice than pakistan on average. table-1: per capita per annum consumption of different food groups in ajk variables mean minimum maximum wheat and wheat flour 84.63 30.00 170.31 rice 15.53 01.50 75.23 milk and milk products 90.16 04.80 112.15 meat 12.63 01.33 29.07 fruits 43.27 03.43 114.11 vegetables 64.04 07.20 122.38 pulses 08.39 00.75 18.57 ghee and oil 11.03 20.40 28.71 tea (green and black) 02.31 00.25 6.19 other foods 04.00 01.50 10.32 size of household 06.00 02.00 14.17 source: author’s own calculation based on primary data for the year 2017-18. ghulam sadiq afridi, muhammad ishaq & abdul jabbar 190 3.1 results of the berry’s index results of the dietary index (berry’s index) are depicted in table 2. the dietary diversity score on average (82.32) revealed that diet is diversified on average with a little bit but not significant variation across region and districts. the results further divulged that income level and education affect diversification positively except family size. food diversity is adversely affected by family size means that as family size increases food diversity decreases or people with large family size restricted to specialized food groups. this revealed that people having large family size could not afford diversified food. table-2: results of the berry’s index dep. var (bi) = 82.32 coefficient st. error t-value constant 0.797 0.010 81.64 family size 0.003 0.001 4.451* education 0.732 0.371 1.973* region (rural/urban) 0.003 0.001 1.55 income 0.004 0.002 2.003* source: author’s own calculation based on primary data for the year 2018-19. * show estimates are statistically significant at 5 % level of significance. 3.2 income and price elasticities estimates of the la-aids model, income elasticities own price elasticities (marshallian and hicksian), cross price elasticities (marshallian and hicksian) for ajk have been given in the following table-3 to table-6. all the 10 food groups were found normal goods having positive income elasticity (wheat-0.33, rice-0.57, milk-1.37, meat0.78, fruits-1.45, vegetables-0.89, pulses-0.11, oil and ghee-0.66, tea-0.40 and other food-0.87), however, milk and fruits were found to be luxurious as divulged by their income elasticities higher than unity. this means that in future demand for milk and fruits would increase more proportionately as income would increase. the own price elasticities of all 10-food groups were also found congruent to economic theory means were negative [wheat (-0.28), rice (-0.65), milk (-0.92), meat (-0.34), fruits, (-0.61), vegetables (-0.21), pulses (-0.29), oil and ghee (-0.04), tea (-0.35) and other food (0.32)]. the income and own price elasticities depicted that wheat, vegetables, pulses and oil and ghee are highly necessities as compared to other food groups. the crossprice elasticity illustrated that rice could be used as substitute of wheat and wheat of rice in the state of ajk. similarly, other food groups exhibit substitution and or complementary effect for one another. our results are in line with the findings of naveed et.al., (2016), ahsan et.al., (2011), haq et.al., (2011), and farooq et.al., (1999). journal of applied economics and business studies, volume.4, issue 3 (2020) 185-198 https://doi.org/10.34260/jaebs.438 191 naveed et.al., (2016) found that food grains, pulses, ghee, sugar and vegetables are necessities, while milk and meat are luxuries. pulses and vegetables, ghee and meat, milk and sugar are identified as gross complements on the basis of uncompensated cross-price elasticities. similarly, ahsan et.al., (2011) found that that supplyside factors (subsidies and world food prices) have a significant impact on food prices, whereas demand-side factors, such as money supply, are the main cause of the increase in food prices in the short as well as the long run. table-3: expenditure elasticities for various food groups (ajk) food groups expenditure elasticities wheat and wheat flour 0.33 (0.051)* rice 0.57 (0.089)* milk 1.37 (0.048)* meat 0.78 (0.056)* fruits 1.45 (0.066)* vegetables 0.89 (0.049)* pulses 0.11 (0.061) oil/ghee 0.66 (0.052)* tea 0.40 (0.062)* other food 0.87 (0.066)* source: author’s own calculation based on primary data for the year 2018-19. * show estimates are statistically significant at 5 % level of significance. 3.3 current and projected demand estimates of the current and projected demand for various food groups have been presented in table-7 and table-8. the estimates of projected demand for the state of ajk in the year 2017 to 2050 depicted that wheat and wheat flour for the year 2017 was found to be 0.343 million tons while it would be 0.400, 0.485, 0.858 and 1.255 million tons in the year 2020, 2025, 2040 and 2050 respectively. the current and projected demand for meat was found to be 0.051 million tons in 2017 and 0.068, 0.129, 0.244 and 0.462 million tons in 2020, 2030, 2040 and 2050 respectively. the current and projected demand for milk was found to be 365 million liters, 1426 million liters, 3777 million liters and 10006 million liters in the year 2017, 2020, 2030, 2040 and 2050 respectively. while calculating the supply and demand gap for major food groups, it was found that the state of ajk is deficit in wheat and wheat flour by 0.194 million tones currently while this gap would be more widened by the year 2025 at 0.306 million tons and 0.939 million tons by 2050. similarly, the supply and demand gap for rice was found to be 0.056 million tons while this gap would be increased at 0.102 million tons in 2040 and at 0.169 million tons in 2050. the state of ajk was found surplus in milk ghulam sadiq afridi, muhammad ishaq & abdul jabbar 192 and meat production but it would in the year 2030 ajk would be deficit in both these food groups at figures of 214 million liters of milk and 0.029 million tons of meat. the state of ajk was also found deficit in fruits and vegetables currently but after 2030 gap in demand and supply would be turned deficit. 4 conclusions and recommendations nonetheless the state of ajk is deficit in food production but still people of the state have diverse food. based on the findings of the study it was suggested that agriculture department might suggest subsidized prices for necessities like wheat, pulses and vegetables especially in the month of ramadan. all necessities food items might be included in food stamp program for poor from the government side. agriculture and livestock departments might align their plans and policies in accordance with the increased demand for food with growing population in future. if legislation is not possible at least efforts might be focused for crops enterprise selection in northern and southern parts of the state. references ahmad, n., a. raza and a. saleem, (2012). “food consumption analysis in: expenditure elasticities approach using hies data”. interdisciplinary journal of contemporary research in business. vol. 4 (4), 466 – 475. retrieved from https://journals.sagepub.com/doi/abs/10.1177/1391561415621826 ahsan, h, i. zainab and m. k. ali, (2011). the determinants of food prices. pide working papers 2011:76. retrieved from pide.org.pk/pdf/working%20paper/workingpaper-76.pdf ali m. and u. farooq (2004). dietary diversity and rural labor productivity: evidence from pakistan. selected long paper (116971) prepared for presentation at the american agricultural economics association annual meeting, denver, colorado, july 1-4, 2004. https://ageconsearch.umn.edu/record/20310/files/sp04al09.pdf berry, c. h., (1971). corporate growth and diversification. journal of law and economics vol. 14, 371-383. retrieved from https://econpapers.repec.org/repec:ucp:jlawec:v:14:y:1971:i:2:p:371-84 burney, n. and m. akmal, (1991). “food demand in pakistan: an application of extended linear expenditure system” journal of agricultural economics. vol. 42, no 2, pp185-195. burney, n. and a.h. khan, (1991). “household consumption patterns in pakistan: an urban rural comparison using micro data” pakistan development review, vol.30, no.2, pp 145-171. https://journals.sagepub.com/doi/abs/10.1177/1391561415621826 https://journals.sagepub.com/doi/abs/10.1177/1391561415621826 https://www.google.com.pk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=2ahukewijsmro85lhahxb1qqkhygycwqqfjaaegqibxac&url=http%3a%2f%2fpide.org.pk%2fpdf%2fworking%2520paper%2fworkingpaper-76.pdf&usg=aovvaw3nuhmnyxp6kkquzottd6fg https://www.google.com.pk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=2ahukewijsmro85lhahxb1qqkhygycwqqfjaaegqibxac&url=http%3a%2f%2fpide.org.pk%2fpdf%2fworking%2520paper%2fworkingpaper-76.pdf&usg=aovvaw3nuhmnyxp6kkquzottd6fg https://ageconsearch.umn.edu/record/20310/files/sp04al09.pdf https://ageconsearch.umn.edu/record/20310/files/sp04al09.pdf https://econpapers.repec.org/repec:ucp:jlawec:v:14:y:1971:i:2:p:371-84 https://econpapers.repec.org/repec:ucp:jlawec:v:14:y:1971:i:2:p:371-84 journal of applied economics and business studies, volume.4, issue 3 (2020) 185-198 https://doi.org/10.34260/jaebs.438 193 deaton, a. and l. muellbauer, (1980). an almost ideal demand system. amer. econ. rev. 70:312–26 farooq, u., t. young, and m. iqbal. 1999. an investigation into the farm households consumption patterns in punjab, pakistan. the pakistan development review. 38:293–305 haq, u. z,. h. nazli, k. meilke, m. ishaq, a. khattak, a. h. hashmi and f. u .rehman. (2011). food demand patterns in pakistani punjab. sarhad journal of agriculture. vol. 27 (2) 305-311. retrieved from https://www.aup.edu.pk/.../24food%20demand%20patterns%20in%20paki naveed, h., a. hussain and h. yousaf, (2016). food demand in pakistananalysis and projection. south asia economic journal. vol. 17, (1), 94-113. https://doi.org/10.1177/1391561415621826 https://www.aup.edu.pk/.../24-food%20demand%20patterns%20in%20paki https://www.aup.edu.pk/.../24-food%20demand%20patterns%20in%20paki https://doi.org/10.1177%2f1391561415621826 ghulam sadiq afridi, muhammad ishaq & abdul jabbar 194 table-4: parameters estimates of the la-aids model explanatory variable wheat rice milk meat fruits vegetable pulses oil/ghee tea other food log of price of wheat 0.058* (0.008) -0.020* (0.006) -0.029*** (0.015) -0.024* (0.005) -0.026** (0.012) -0.008 (0.006) -0.016*** (0.009) 0.002 (0.004) 0.013* (0.005) 0.009 (0.006) log of price of rice -0.020* (0.006) 0.018* (0.005) 0.023** (0.010) -0.009* (0.004) 0.005 (0.004) -0.015* (0.005) 0.005** (0.002) -0.010* (0.003) 0.002 (0.003) 0.009* (0.004) log of price of milk -0.029*** (0.015) 0.023** (0.010) 0.050*** (0.030) 0.030** (0.014) -0.027** (0.012) -0.030* (0.012) -0.003 (0.006) -0.020* (0.009) -0.003 (0.006) -0.032* (0.012) log of price of meat -0.024* (0.005) -0.009* (0.004) 0.030** (0.014) 0.069* (0.005) 0.012 (0.008) -0.014* (0.004) -0.006* (0.002) -0.011* (0.003) -0.014* (0.005) 0.010*** (0.005) log of price of fruits -0.026** (0.012) 0.005 (0.004) -0.027** (0.012) 0.012 (0.008) 0.073* (0.010) -0.030* (0.005) -0.000 (0.002) -0.02** (0.007) -0.005*** (0.003) -0.011 (0.009) log of price of vegetables -0.008 (0 .006) -0.015* (0.005) -0.030* (0.012) -0.014* (0.004) -0.014* (0.004) 0.085* (0.006) -0.003 (0.007) -0.008** (0.003) -0.004 (0.004) 0.006 (0.005) log of price of pulses -0.02*** (0.009) 0.005** (0.002) -0.003 (0.006) -0.006* (0.002) -0.000 (0.002) -0.003 (0.007) 0.028* (0.003) -0.004* (0.002) 0.002 (0.004) 0.001 (0.001) log of price of oil/ghee 0.002 (0.004) -0.010* (0.003) -0.020* (0.009) --0.011* (0.003) -0.016** (0.007) -0.008** (0.003) -0.004* (0.002) 0.076* (0.003) 0.002 (0.004) -0.001** (0.000) log of price of tea 0.013* (0.005) 0.002 (0.003) -0.003 (0.006) -0.014* (0.005) 0.005*** (0.003) -0.004 (0.004) 0.002 (0.004) 0.002 (0.004) 0.028* (0.002) 0.000 (0.001) log of price of other food 0.009 (0.006) 0.009* (0.004) -0.032* (0.012) 0.010*** (0.005) -0.011 (0.009) 0.006 (0.005) 0.001 (0.001) -0.001** (0.000) 0.000 (0.001) 0.009* (0.001) income (expenditures) -0.032* (0.007) -0.004 (0.005) 0.099* (0.013) -0.024* (0.006) 0.071* (0.010) -0.012** (0.005) -0.042* (0.002) -0.028* (0.004) -0.027* (0.002) -0.001** (0.001) household size -0.002* (0.001) 0.000 (0.000) -0.001 (0.002) 0.001 (0.000) 0.005* (0.002) 0.001 (0.001) -0.000 (0.000) -0.001 (0.000) -0.002* (0.000) -0.000** (0.000) constant 0.630* (0.119) 0.004 (0.078) 0.136* (0.218) -0.116 (0.102) -0.493* (0.176) 0.272* (0.090) 0.328* (0.043) 0.153** (0.070) 0.082*** (0.047) 0.004 (0.015) observations 1250 1250 1250 1250 1250 1250 1250 1250 1250 1250 r-squared 0.145 0.068 0.140 0.261 0.152 0.276 0.390 0.510 0.269 0.035 chi 125.26 54.57 121.16 262.37 133.11 282.92 473.86 769.75 272.23 177.93 source: author’s own estimation with survey data. standard errors are reported in parentheses. *, ** and *** show estimates are statistically significant at 1, 5 and 10% respectively. journal of applied economics and business studies, volume.4, issue 3 (2020) 185-198 https://doi.org/10.34260/jaebs.438 195 table-5: estimated uncompensated own and cross-price elasticities of demand food group wheat rice milk meat fruit vege. pulses oil/ghee tea other food wheat -0.286* (0.071) 0.186* (0.057) -0.153** (0.065) -0.174* (0.048) -0.042 (0.057) -0.025 (0.065) -0.067 (0.079) -0.082** (0.041) 0.117* (0.046) 0.100** (0.051) rice 0.028 (0.106) -0.649* (0.086) 0.130 (0.190) -0.053 (0.068) -0.054 (0.081) -0.14*** (0.075) -0.20*** (0.333) -0.17*** (0.058) 0.147** (0.067) 0.173** (0.075) milk -0.159* (0.060) -0.111** (0.049) -0.915* (0.111) -0.160* (0.043) -0.206* (0.050) 0.371* (0.057) -0.12*** (0.066) -0.180* (0.038) -0.139* (0.042) -0.222* (0.046) meat -0.12*** (0.067) 0.050 (0.055) 0.310* (0.123) -0.342* (0.045) 0.005 (0.055) -0.046 (0.062) 0.11*** (0.066) -0.019 (0.039) -0.088** (0.045) 0.104** (0.050) fruit -0.235* (0.080) -0.127** (0.065) -0.22*** (0.145) 0.002 (0.055) -0.605* (0.066) 0.015 (0.075) 0.109 (0.090) 0.145* (0.048) 0.161* (0.054) 0.119** (0.060) vegetables -0.057 (0.059) 0.088** (0.040) -0.254** (0.112) -0.199* (0.040) -0.264* (0.048) -0.208* (0.055) -0.006 (0.068) -0.066** (0.034) -0.032 (0.040) 0.075*** (0.045) pulses -0.038 (0.072) 0.176* (0.059) -0.032 (0.137) 0.104** (0.048) 0.022 (0.058) 0.029 (0.067) -0.294* (0.083) -0.032 (0.042) 0.060 (0.049) 0.075*** (0.044) oil/ghee 0.053 (0.062) -0.022 (0.050) -0.122 (0.116) -0.112* (0.042) -0.037 (0.050) 0.044 (0.058) -0.058 (0.071) -0.04*** (0.021) 0.034 (0.042) -0.015 (0.046) tea 0.246* (0.073) 0.079*** (0.045) -0.048 (0.139) 0.099** (0.049) -0.076 (0.059) 0.053 (0.068) 0.075 (0.085) -0.017 (0.043) -0.351* (0.049) 0.026 (0.055) other food 0.052 (0.078) -0.070 (0.064) -0.293** (0.148) -0.004 (0.053) 0.019 (0.064) 0.037 (0.073) 0.058 (0.091) -0.093** (0.045) 0.046 (0.053) -0.315* (0.059) source: author’s own calculation based on primary data for the year 2018-19. *, ** and *** show estimates are statistically significant at 1, 5 and 10% respectively. ghulam sadiq afridi, muhammad ishaq & abdul jabbar 196 table-6: estimated compensated own and cross-price elasticities of demand food group wheat rice milk meat fruit veget. pulses oil/ghee tea other food wheat -0.398* (0.069) 0.117** (0.057) 0.032 (0.133) -0.086*** (0.046) 0.047 (0.056) 0.099*** (0.060) -0.093 (0.081) -0.001 (0.040) 0.157* (0.047) 0.094*** (0.053) rice 0.088 (0.105) -0.601* (0.086) 0.721* (0.200) -0.068 (0.071) 0.251* (0.085) -0.180** (0.088) -0.099 (0.122) -0.113** (0.055) 0.122*** (0.071) 0.202** (0.079) milk 0.060*** (0.045) 0.040 (0.047) -0.548* (0.108) 0.038 (0.038) 0.058 (0.046) -0.169* (0.053) 0.034 (0.066) 0.003 (0.033) 0.002 (0.039) -0.106** (0.043) meat -0.025 (0.066) 0.078*** (0.050) 0.553* (0.125) -0.258* (0.044) 0.180* (0.053) 0.003 (0.062) 0.128*** (0.076) 0.049*** (0.031) -0.082*** (0.045) 0.107** (0.050) fruit -0.044 (0.078) -0.004 (0.046) 0.119 (0.149) 0.182* (0.053) -0.376* (0.064) 0.203* (0.073) 0.010 (0.091) -0.017 (0.045) -0.030 (0.053) -0.056 (0.059) vegetables 0.050 (0.058) 0.129* (0.048) 0.002 (0.111) -0.023 (0.039) -0.118* (0.045) -0.111** (0.054) 0.019 (0.068) 0.004 (0.034) 0.007 (0.039) 0.075*** (0.044) pulses 0.040 (0.072) 0.186* (0.059) 0.194*** (0.131) -0.037 (0.048) 0.138* (0.050) 0.097*** (0.062) -0.294* (0.083) -0.017 (0.042) 0.055 (0.050) 0.036 (0.055) oil/ghee 0.146** (0.060) 0.002 (0.050) 0.118 (0.116) -0.031 (0.041) 0.094** (0.043) 0.126** (0.057) -0.044 (0.071) -0.162*** (0.098) -0.018 (0.042) -0.026 (0.046) tea 0.338* (0.073) 0.104*** (0.059) -0.193*** (0.125) 0.181* (0.049) 0.055 (0.059) 0.136** (0.068) 0.089 (0.085) 0.037 (0.042) -0.333* (0.049) 0.012 (0.055) other food 0.064 (0.077) -0.058 (0.064) -0.280** (0.148) 0.008 (0.053) 0.032 (0.063 0.049 (0.073) 0.071 (0.091) -0.080*** (0.045) 0.058 (0.053) -0.303* (0.059) source: author’s own calculation based on primary data for the year 2018-19. *, ** and *** show estimates are statistically significant at 1, 5 and 10% respectively. journal of applied economics and business studies, volume.4, issue 3 (2020) 185-198 https://doi.org/10.34260/jaebs.438 197 table-7: current and projected demand of different food groups in ajk year pop (m) wheat (mt) rice (mt) milk (ml) meat (mt) fruit (mt) veget. (mt) puls (mt) oil/ (mt) tea (t) oth (mt) 2017 4.045 0.343 0.063 365.16 0.051 0.175 0.259 0.034 0.045 0.009 0.016 2020 4.360 0.400 0.087 538.20 0.068 0.262 0.344 0.035 0.056 0.010 0.021 2025 4.781 0.485 0.097 875.99 0.094 0.436 0.489 0.038 0.075 0.012 0.030 2030 5.242 0.586 0.109 1425.79 0.129 0.725 0.695 0.042 0.099 0.015 0.043 2035 5.748 0.709 0.123 2320.65 0.177 1.206 0.989 0.047 0.132 0.018 0.061 2040 6.303 0.858 0.139 3777.17 0.244 2.007 1.406 0.051 0.176 0.022 0.086 2045 6.911 1.037 0.158 6147.84 0.336 3.340 2.000 0.056 0.234 0.028 0.121 2050 7.588 1.255 0.179 10006.41 0.462 5.557 2.844 0.062 0.312 0.034 0.171 source: author’s own calculation based on primary data for the year 2018-19. table-8 (a): demand and supply situation of different food groups in ajk year wheat (million tons) gr. rate = 2.3% rice (million tons) gr. rate = 1.3% fruits (million tons) gr. rate = -0.004% demand supply gap demand supply gap demand supply gap 2017 0.343 0.149 0.194 0.063 0.007 0.056 0.175 0.971 -0.796 2020 0.400 0.160 0.240 0.087 0.007 0.080 0.262 0.936 -0.675 2025 0.485 0.179 0.306 0.097 0.007 0.090 0.436 0.814 -0.378 2030 0.586 0.201 0.385 0.109 0.008 0.102 0.725 0.640 -0.085 2035 0.709 0.225 0.484 0.123 0.008 0.115 1.206 0.455 -0.751 2040 0.858 0.252 0.606 0.139 0.009 0.130 2.007 0.293 -1.714 2045 1.037 0.282 0.755 0.158 0.009 0.148 3.340 0.170 -3.169 2050 1.255 0.316 0.939 0.179 0.010 0.169 5.557 0.090 -5.467 source: author’s own calculation based on primary data for the year 2018-19. ghulam sadiq afridi, muhammad ishaq & abdul jabbar 198 table-8 (b): demand and supply situation of different food groups in ajk year vegetables (million tons) gr. rate =2.89% meat (million tons) gr. rate = 1.09% milk (million liters) gr. rate = 1.32% demand supply gap demand supply* gap demand supply* gap 2017 0.259 0.020 0.240 0.051 0.087 + 0.036 365 1024 + 659 2020 0.344 0.021 0.322 0.068 0.090 + 0.022 538 1065 + 526 2025 0.489 0.025 0.464 0.094 0.095 + 0.001 876 1136 + 2 60 2030 0.695 0.029 0.666 0.129 0.100 0.029 1426 1211 214 2035 0.989 0.033 0.956 0.177 0.106 0.072 2321 1292 1028 2040 1.406 0.038 1.368 0.244 0.112 0.132 3777 1378 2399 2045 2.000 0.044 1.956 0.336 0.118 0.218 6148 1470 4677 2050 2.844 0.051 2.793 0.462 0.125 0.337 10006 1568 8438 source: author’s own calculation based on primary data for the year 2018-19. * figures seem exaggerated but based on documented figures of livestock department (milk=2.77 million liters/day, meat=0.087million tons/year). journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 1 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x role of education in economic growth of pakistan: a sectoral analysis faiza hassan1, hafsa hina2, abdul qayyum (late)3 & anwar hussain*4 1 phd scholar, pakistan institute of development economics, islamabad, lecturer, department of economics, university of malakand. 2 assistant professor, department of economics and econometrics pakistan institute of development economics, islamabad. 3 joint director/ professor, pakistan institute of development economics, islamabad. 4 associate professor, department of economics and development studies university of swat. abstract education and economic growth nexus is one of the abundantly researched topics in economics. the social returns of education in addition to its private returns makes it public good and justifies the use of public funds. while most of the studies conclude the positive relationship between the two, examples also exist of the negative relationship. however, there is a gap in the literature to find and compare the effect of education on the growth of agriculture, industry and services sector separately. this paper is particularly aimed to analyze the impact of different levels of education of employed persons on level and growth of national output, agriculture, industry & services sector output in pakistan. the method of analysis is the autoregressive distributed lag model (ardl). each level of education is found to have a positive effect on the output per employed person both in the short-run and long-run except for agriculture sector. in the agriculture sector, each education level is negatively associated. the deeper analysis showed that the greater negative effect of employment evades the positive effect of education in the agriculture sector. the comparison of different sectors shows that primary education contributes more to the industrial sector. while the contribution of the secondary & tertiary education is highest in the services sector. keywords education; economic growth; economic development; sectoral analysis; human capital, economic impact; pakistan; agriculture; industry; services sector. jel classification i21;i25; i26; o11, q10 * anwar@uswat.edu.pk faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 2 1. introduction education as the main determinant of human capital is considered a fundamental factor of economic growth. although in economics, the recognition of the role of education in improving productivity and economic growth is as old as the subject itself, the recent growth theories acknowledge the role of education in the following three distinct ways. first, the neoclassical growth theories that illustrate education as a source of increasing human capital embedded in the labour force which is important for increasing labour productivity and thus economic growth, for example, mankiw et al., 1992; barro, 1991. second, the endogenous growth theories that demonstrate education as the basis for innovation and new knowledge, technology and improved methods of production that stimulates economic growth, for example, lucas, 1988; romer, 1990; aghion & howitt, 1998. third, theories that acknowledge the role of education in economic growth as its ability to speed up technological catch-up and diffusion (e.g. nelson and phelps, 1966; benhabib and spiegel, 1994). the literature of economics elaborates that in addition to private returns of education to individuals, it also has social returns due to its spillover effect. the benefits of education spread to other workers in a firm or industry, to other community members in a community, to members of the city or region and the economy in general. it is established that in addition to increase productivity, education increases economic growth by having positive externalities. for example, barro and lee (2001) stated that “the level and distribution of educational attainment has a strong impact on social outcomes, such as child mortality, fertility, education of children and income distribution.” furthermore, galor & zeira(1993) described that the main source of inequality in income is the result of inequality in education. similarly, education helps in good parenting, wider and better political participation of individuals in a society, better community participation and curbing crime and negativity in a society (oecd, 1998). this spillover characteristic and social returns makes education a public good and provides the ground for the allocation of government funds to promote education in a country (sianesi & reenen, 2003). the inability of microeconometric analysis to capture the social returns of education as well as testing the emerging growth theories motivated the use of aggregate level analysis in recent past (krueger & lindahl, 2001). the macroeconometric analysis of education and economic growth increased with exponential rate during the past few decades. most of the empirical studies regarding education and economic growth have journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 3 used cross-country regressions using average years of education or enrollments as a measure of education or human capital. holland et al. (2013), sianesi and reeenen(2003), temple(2001), krueger and lindahl(2001), topel(1999) and renelt(1991) provides a comprehensive review of literature in this regard. there are comparatively too few studies based on time series analysis to analyse the relationship. temple(1999) prefers and suggests time series analysis over cross country regressions for studying growth dynamics due to the problem of heterogeneity. moreover, majority of empirical work used average years of schooling as a proxy for measuring education or human capital and there are relatively few studies that are devoted to analyzing the impact of distinct education levels on economic growth. the problem with using average years of schooling as a proxy for education level or human capital is that it does not differentiate between an additional year of higher education and elementary education levels. it treats the additional year of primary education same as an additional year of college or university. it implies that the increase of education in any level of education whether its primary, secondary or tertiary will have an equal effect on economic growth. this underlying assumption cannot be justified. therefore, it is more appropriate to use different education levels instead of aggregating it to a single measure in the form of average years of schooling. few of the studies that are based on time series analysis and education segregated in three different levels are summarized here. nnyanzi & kilimani(2018) examined the effect of enrollments education levels on gdp growth of sub-saharan africa. the study was based on period 1995-2016. it is concluded that all levels of education have significant positive effect on growth and the impact is highest of secondary education. kyophilavong et al.(2018) concluded that the long-run association exists at all three levels of education and economic growth using data for laos over the years 1984 to 2013. kotaskova et al.(2018) confirm the positive impact of primary, secondary & higher education on india’s economic growth by utilizing the data from 1975 to 2016. jenkins(1995) analysed data for the uk for years 1971-1992. it is found that as compared to a worker with no education the highly educated worker produce two times more output. asteriou & agiomirgianakis(2001) used data from 1960 to 1994 to find the long-run association of education and gdp per capita of greece. all educational variables were found cointegrated with gdp per capita. loening(2005) by using data 1951-2002 for guatemala established that education has a positive effect on economic growth and 50% of the growth of output is explained by education. sari & soytas(2006) used time-series data from 1937-1996 of turkey. it is concluded that primary & faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 4 secondary education causes national income, while the causality is bidirectional in case of university-level education. the results of vecm confirms the existence of cointegration between education variables and gdp. lin (2003) examined the link of education and economic growth keeping in view the effect of technological advancement on economic growth in taiwan for years 1965-2000. education is found to have a significantly positive impact on economic growth. self and grabowski(2003) analyzed the impact of different education levels as well as vocational education on the growth of national income in japan in post and pre-war era. it is concluded that primary education has a causal effect on economic growth in both periods. while secondary & higher education is found to have a causal effect in the post-war era. on the other hand, omodero & nwangwa(2020) by utilizing data from 2000 to 2018 concluded that longrun association exists between tertiary education and economic growth in nigeria. however, no evidence of causality is found between tertiary education expenditure variable and economic growth or enrollment ratio and economic growth. tsamadias and pegkas(2012) reported a negative relationship exists between education and economic growth in greece during 1981-2009. however, the negative coefficient was found insignificant except the case of human capital measured in the form of enrollment rates. the cointegration test showed no long-run association of education and output growth. pakistan is a developing country and is facing enormous challenges on economic, social and political fronts. the ability of education to deal with economic, political and social issues simultaneously makes it essential to focus on education and thereby increasing the human capital. there are numbers of studies with reference to pakistan which analyzes the impact of education as a component of human capital on economic growth. hafeez & rahim(2019) found that enrollments in primary, secondary & higher education are important in determining the economic growth of pakistan. the analysis was based on the period 1971 to 2013. iqbal(2018) by using the data for years 1972 to 2014 for pakistan and implying the ardl technique concluded that the long-run association between total literacy rate and economic growth is significant but negative. afridi(2016) attempted to find the link of human capital and the economic growth in pakistan. enrolments rate at primary level education, infant mortality rate and birth rate are used as proxies for human capital while gdp per capita is used as a proxy for economic growth while physical capital is used as a control variable. the study concludes that human capital has an important role in accelerating economic growth. jangraiz et al(2015) used data from 1971-2012 to analyze the link between human capital and the economic growth of pakistan. they used expenditure on research & journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 5 development, enrolment rates in primary, secondary & higher education, gross enrollment rate, and health as a proxy to human capital. by use of the granger causality test, they concluded that the causal association between human capital variables and economic growth exists. they found that the direction of causality is from research and development to economic growth while in the case of education (in the form of gross enrolment), it is from growth to education. jalil & idrees(2013) attempted to analyze the impact of education on economic growth by not only taking average years of schooling as a measure but also analyzing the impact through enrolments in different education levels. the analysis was based on data from 19602010. nonlinear 2sls was used to estimate the error correction model. it was found that average years of schooling, as well as enrolment rates in primary, secondary & tertiary education, positively affected the economic growth of pakistan. secondary education was reported to be more contributing than primary & tertiary education. kiani(2013) used data from 1980 to 2007 and ols technique for analysis. the study used primary, middle, high & other school enrollments to capture the effect of education. it is concluded that each education level contributes significantly to economic growth. amir et al(2012) used primary, secondary, college, university, vocational enrollment rates for capturing the effect of education on the economic growth of pakistan. by using the johansen cointegration test and error correction mechanism it is established that a long-run association between human capital and economic growth exists. abbas & peck(2008) attempted to find the association by using the secondary education per worker and the ratio of government expenditure on health to gdp as proxies for human capital. they used data from 19602005 and johansen cointegration test. it is found that human capital has a positive contribution to gdp from the 1960s to 1990 while from 1990 to 2000 it showed the negative impact and they elaborated that this result is due to the negligence of human capital in economic policies. khan(2005) analyzed the association between human capital and economic growth by utilizing panel data for 72 developing countries including pakistan for the years 1980-2002. average years of schooling, gross secondary school enrollment, adult literacy rate and life expectancy at birth were used as proxies of human capital. they concluded that along with capital formation, quality of institutions the education and health variables have a significant effect on economic growth. from the review of studies with reference to pakistan, it is evident that while attempts have been made to analyze the impact of different education levels on economic growth, still there is a gap in the literature to find its effect on the growth of faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 6 agriculture, industry & services sector separately. moreover, our study will be unique in considering the impact of education of those who are employed and actively participating in the production of output instead of considering overall enrollment rates in the country or a particular sector of the economy. the current study is particularly aimed to analyze the impact of different levels of education (i.e. primary, secondary & tertiary) of employed persons on level and growth of national output in pakistan. it is also designed to find the possible differences in the impact of education in different sectors of the economy i.e. agriculture, industry & services sector. a distinct feature of this study is the use of output per employed person as a dependent variable, as most of the growth models when solved have dependent variable as y/l denoted by y. although per capita output is the most widely used proxy for y/l, it is more appropriate to measure it as output per worker or employee, if there are no data issues. benhabib and spiegel(1994), pritchett(1997), bils and klenow(1997), hall and jones(1999), klenow-rodriquez(1997) and barro and lee(1994) used output per worker as a dependent variable. this study is also distinct in considering the education of employed persons instead of overall enrollment rates. the number of employed persons with a particular education level i.e. primary, secondary & tertiary education in each year for the overall economy, agriculture, industrial & services sector is derived from labour force survey of pakistan. if the dependent variable is output per worker then it is more appropriate to analyze the educational level of those who are employed and participating in the production of gdp rather than to consider the education of all the people living in the economy. although, like all the previous studies this paper aims to find the impact of education on overall economic growth, it is distinct in using different variables for the analysis. moreover, it is a first attempt in finding the impact of education for different major sectors of the economy. the core aim is to analyze the impact of different education levels on the economic growth of pakistan during the period 1985-2018. more specifically the objectives are to analyze and compare the effect education segregated in different levels of education on economic growth in pakistan. also to analyze and compare the impact of all three levels of education on sectoral growth in pakistan (i.e. agriculture, industry & services sector) separately. journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 7 2 theoretical model it is well established in the economic literature that physical capital accumulation alone is not capable of explaining all the growth that occurs in per capita gdp (solow, 1957; kendrick, 1956). the new growth theories well acknowledge the role of education and provide a variety of theoretical frameworks in which education enhances economic growth. one of the basic implications of the endogenous growth theories is that technological changes bring increasing returns to scale and therefore economies tend to grow without converging to steady-state. in contrast, the mankiw, romer & weil(1992) model augments human capital as a separate input in the basic solow model and does not imply unbounded growth. fedderke(2002) referring to the work of mankiw et. al(1992) reports that “introduction of human capital into the solow model successfully enhances its explanatory power to such a degree as to preclude the necessity of resorting to endogenous growth models of either the romer (1986) or (1990) variants” following the footsteps of mankiw, romer & weil(1992) our theoretical model includes human capital as a third input in production function along with labour and capital. the basic production function is given by human capital augmented solow model well illustrated by mankiw, romer & weil(1992) and given below 𝑌(𝑡) = 𝐾(𝑡)𝛼 𝐻(𝑡)𝛽 (𝐴(𝑡)𝐿(𝑡)) 1−𝛼−𝛽 (1) where y is representing output, k is physical capital, ‘h’ is human capital, l is labour and a is representing the level of technology. α + β is assumed to be less than 1 which indicates decreasing returns. labour and technology grow at exogenous rates and represented by textbook symbols n and g respectively. 𝐿(𝑡) = 𝐿(0)𝑒𝑛𝑡 (2) 𝐴(𝑡) = 𝐴(0)𝑒𝑔𝑡 (3) output can either be consumed, invested in physical capital or can be used to increase human capital. it is assumed that a constant fraction ‘sk’ is invested in physical capital while fraction sh is devoted to human capital. the capital accumulation process can be written as 𝐾 =̇ 𝑠𝑘 y(t) δk(t) (4) where δ is denoting depreciation. faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 8 the human capital accumulation suggested by mankiw, romer & weil ( 1992) assumes the same pattern as it is assumed for physical capital accumulation and is illustrated by the following equation. 𝐻 =̇ 𝑠ℎ y(t) δh(t) (5) the dynamics of the economy are represented by the following equations. where y = y/ al , k = k/ al and h = h/al. 𝑘(𝑡) = ̇ 𝑠𝑘 𝑦(𝑡) − (𝑛 + 𝑔 + 𝛿)𝑘(𝑡) (6) ℎ(𝑡) = ̇ 𝑠ℎ 𝑦(𝑡) − (𝑛 + 𝑔 + 𝛿)ℎ(𝑡) ( (7) steady state values are obtained by solving equation (6) and (7) and are listed below k* = ( 𝑠𝑘 1−𝛽 𝑠ℎ 𝛽 𝑛+𝑔+ 𝛿 ) 1 1−𝛼−𝛽 (8) h* = ( 𝑠𝑘 𝛼 𝑠ℎ 1−𝛼 𝑛+𝑔+ 𝛿 ) 1 1−𝛼−𝛽 (9) by substituting steady-state values of k and h (i.e. equation 8 and 9) in production function (given by equation 1) and by taking logs the following equation for per capita income is derived. ln [ 𝑌 (𝑡) 𝐿(𝑡) ] = ln a(0) + gt 𝛼+𝛽 1− 𝛼−𝛽 ln (n + g + δ) + 𝛼 1− 𝛼−𝛽 ln(sk ) + 𝛽 1− 𝛼−𝛽 ln(sh) (10) equation (10) illustrates that per capita income depends not only on population growth and physical capital accumulation but also on human capital accumulation. 2.1 econometric methodology to empirically test the above illustrated theoretical model, we will have to make certain assumptions about the proxy of human capital. it is difficult to get exact data about share of output which is devoted to human capital i.e. to know the investment in human capital which is not only made by governments but by family and individuals themselves and also cost in form of foregone earnings which is low in case of a worker having low education and high in case of a worker having higher education (mankiw, romer and weil, 1992). keeping in view this difficulty most of the studies have used average years of schooling or percentage of the working-age population with secondary enrollments as a proxy of human capital (barro, 1996, 2001; mankiw, romer and weil, 1992). the current study used the number of employed persons with different education levels as a proxy for human capital. journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 9 ardl approach suggested by pesaran and shin(1999) and pesaran et. al(2001) is deemed fit for data analysis. the advantages of the ardl method are that it is capable of identifying cointegration irrespective of the order of variables being one or zero. it appropriately deals with the problem of serial correlation. it is capable of dealing with the problem of endogeneity of variables under consideration (pesaran and shin, 1999). it is consistent in the case of small samples (pesaran and shin, 1999; narayan, 2005). it is not affected by the potential bias of the cointegration test or unit root test. therefore, the following ardl model is used to analyse the data δyt = α + ∑ 𝛽1𝑖 𝛥𝑦𝑡−𝑖 𝑚1 𝑖=1 + ∑ 𝛽2𝑖 𝛥𝑘𝑡−𝑖 𝑚2 𝑖=0 + ∑ 𝛽3𝑖 𝛥𝐸𝑑𝑢_𝑝𝑡−𝑖 𝑚3 𝑖=0 +∑ 𝛽𝑘𝑖 𝛥𝑍𝑗,𝑡−𝑖 𝑚𝑗 𝑖=0 + 𝜆1yt-1 + 𝜆2kt-1 + 𝜆3edut-1 +𝜆𝑗 zj,t-1 + μt (11) δyt = α + ∑ 𝛽1𝑖 𝛥𝑦𝑡−𝑖 𝑚1 𝑖=1 + ∑ 𝛽2𝑖 𝛥𝑘𝑡−𝑖 𝑚2 𝑖=0 + ∑ 𝛽3𝑖 𝛥𝐸𝑑𝑢_𝑠𝑡−𝑖 𝑚3 𝑖=0 + ∑ 𝛽𝑘𝑖 𝛥𝑍𝑗,𝑡−𝑖 𝑚𝑗 𝑖=0 + 𝜆1yt-1 + 𝜆2kt-1 + 𝜆3edut-1 + 𝜆𝑗 zj,t-1 + μt (12) δyt = α + ∑ 𝛽1𝑖 𝛥𝑦𝑡−𝑖 𝑚1 𝑖=1 + ∑ 𝛽2𝑖 𝛥𝑘𝑡−𝑖 𝑚2 𝑖=0 + ∑ 𝛽3𝑖 𝛥𝐸𝑑𝑢_ℎ𝑡−𝑖 𝑚3 𝑖=0 +∑ 𝛽𝑘𝑖 𝛥𝑍𝑗,𝑡−𝑖 𝑚𝑗 𝑖=0 + 𝜆1yt-1 + 𝜆2kt-1 + 𝜆3edut-1 + 𝜆𝑗 zj,t-1 + μt (13) where: y = real gdp per employed person k = capital stock per employed person edu_p = number of employed persons having primary education edu_s = number of employed persons who have secondary education edu_h = number of employed persons with higher education zj = set of j control variables used in the model δ represents the first difference, m is the maximum number of lags included in the model and μt is the error term. whereas, λi symbolizes parameters of the long-run association. the bounds testing approach proposed by pesaran et al (1999) is used for testing the presence of the long-run association between non-stationary variables. to show the short-run dynamics, error correction model (ecm) is estimated. note: the same models will be re-estimated for agriculture, industry & services sector. however, capital, output and education variable will represent the capital, output and education of the respective sector, and control variables will also be different. faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 10 2.2 variables and data to execute the analysis time series data for the years, 1985-2018 is gathered. the variables used along with the definitions and sources are listed in appendix a. the dependent variable is the real gdp per employed person. the main variable of this study the education level is taken as the number of employed persons who have a specific education level i.e. primary, secondary or tertiary education. the data is gathered from various issues of the labour force survey of pakistan (lfs). it is important to mention here that all the data is available in data files of lfs so the data which was not printed is acquired from data files e.g. education levels of employed persons in agriculture, industry & services sector. the details of control variables used in the study are listed in appendix a. the data for capital per employee is generated by utilizing the series of gross fixed capital formation through the method explained below. the capital series is generated for the overall economy as well as for agriculture, industry & services sectors separately. 2.3 construction of capital stock series the method proposed by berlemann and wesselhöft(2014) is followed to construct the capital stock series. the proposed method is based on the perpetual inventory method. however, it unifies three existing approaches nehru and dhareshwar (1993), de la fuente and domenech(2000) and kamps(2006) to have merits of all three. the method is briefly explained as under. step 1: following nehru and dhareshwar(1993) the initial value of the investment is derived by following the regression equation reported below ln it = 𝛼 + βt + 𝑡 (14) complete investment series from period two (t2) to t is used in estimation. once the estimates of α and β are calculated. the value of an investment in time-period 1 is calculated by substituting t=1 in the equation. as the estimation utilized the logarithmic values of investment so the exponential value of α + β will give the investment in time period 1. step 2: the second step is to find the growth rate of investment. the method utilizes the value of β as a measure of the growth of investment. step 3: following kamps(2006), depreciation is considered as time-varying and not fixed. journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 11 after making these three modifications the method estimates the initial capital stock by the following formula 𝐾0 ≈ 𝐼1 𝑔𝐼 + 𝛿 (15) and then follows the perpetual inventory method 𝐾𝑡 = (1 − 𝛿) 𝑡−1𝐾0 + ∑(1 − 𝛿) 𝑖 𝐼𝑡−(𝑖+1) (16) 𝑡−1 𝑖=0 note: we in our study followed all the steps mentioned above except step 3 that suggests the use of time-varying depreciation rates, but due to unavailability of depreciation rates for each year we used a fixed depreciation rate of 5%. 3. results and discussion section 3 reports and discusses the results comprehensively. in section 3.1, the results of unit root tests are registered. the second subsection illustrates the impact of different education levels on growth and output of the agriculture sector. the third and fourth sections are allocated for analysis of the impact of different education levels on the industry & services sector respectively. the fifth subsection is reserved for analysis of the overall economy. while in the last subsection, the results of different sectors are compared. 3.1 unit root test in order to check the stationarity of variables, the augmented dicky fuller test is employed. the results of adf τ-statistics along with their p-values are stated in appendix a1. it is found that all the variables used in the models for the overall economy, agriculture, industry & services are integrated of order 1 i.e. they are nonstationary at levels but stationary at first difference. 3.2 agriculture sector 3.2.1 agriculture sector’s output per employed person as the dependent variable the impact of different educational levels on output per employed person of the agriculture sector is analyzed by estimating the model for each education level separately. in addition to the education variable, capital per employed person in the agriculture sector, land, fertilizer, improved seed distribution, water and exports of agriculture sector’s products are included in each model. faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 12 the bounds test results (reported in appendix b1) does not validate the presence of a long-run association. hence, the results of long-run in case of models with output per employed person as the dependent variable in the agriculture sector are not reported. the results of the short-run effects of education levels on the output per employed person in the agriculture sector derived by fitting the ardl model is given below in table 1 along with the summary of diagnostic tests to check the accuracy of the model. the stability of the model is checked through cusum and cusum of squares and are reported in appendix c. the results show that capital per employed person has a significantly positive impact on output per employed person in the agriculture sector in the models estimated for primary, secondary as well as tertiary education. the coefficients associated with the education levels are found negative in all the three models. however, in the model where primary education is considered the coefficient is found insignificant, while in the case of secondary & tertiary education the coefficients are highly significant. the coefficients of land, fertilizer, seed, water and exports of agriculture product all appeared with a positive sign in all the three models for modelling the agriculture sector output per employed person with different education levels. table 1: impact of education on growth of agriculture sector in pakistan short run variable primary secondary tertiary education -0.149597 (0.1406) -0.268610 (0.0226) -0.145109 (0.0670) k 0.782196 (0.0011) 0.391912 (0.1022) 0.249492 (0.1012) land 0.133457 (0.5575) 0.169806 (0.4661) 0.203712 (0.5529) fertilizer 0.088868 (0.3281) 0.132368 (0.2161) 0.116934 (0.2713) seed 0.045310 (0.1820) 0.139053 (0.0157) 0.017853 (0.6931) water 0.081016 (0.7857) 0.665697 (0.0572) 0.524766 (0.1897) exports_agri 0.040790 (0.3513) 0.000550 (0.9899) 0.007061 (0.9116) c 5.215370 (0.0060) 3.305193 (0.0704) 3.387593 (0.0356) diagnostic test results 𝑹𝟐 0.97 0.97 0.91 serial correlation 0.374958 (0.5403) 2.829344 (0.1089) 0.272629 (0.6016) heteroskedasticity 0.963802 (0.5065) 1.550322 (0.1899) 0.659441 (0.7209) ramsey reset test 1.176395 (0.2917) 1.687570 (0.2095) 2.474942 (0.1299) *p-values are given in parenthesis. *each variable is taken in natural log form. journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 13 the results regarding the impact of different education levels on per employed person output of agriculture sector are in contrast with what is generally expected but few examples exist in literature exhibiting the negative impact of education on agriculture sectors output. for example, oduro-ofori et al.(2014) in a micro-level study of farmers found that average years of schooling have a negative but insignificant coefficient in the regression of agriculture output as the dependent variable. besides, it was observed in the study that primary & secondary education have positive but tertiary education have a negative effect on output. similarly, lee(2012) tested the effect of higher education on agriculture output of japan and found that all major groups in higher education that is science, social sciences, humanities and education have a significant negative effect on the growth of agricultural output. although few examples exist of finding the negative coefficient of education, the issue is further investigated and discussed in the subsequent section. 3.2.2 agriculture sector’s output as dependent variable to further investigate the effect of different education levels in the agriculture sector, the analysis is repeated by taking the output of agriculture sector as dependent variable instead of output per employed person and adding the number of persons employed as an explanatory variable in the model. it helps to investigate if the negative association between education and output per employed person found in the preceding section is due to increased education or because of the surplus labour in the agriculture sector. f-bounds test (reported in appendix b2) confirms the presence of a long-run association in all the three models with different education levels. the speed of adjustment coefficients reported in appendix b2 is highest for the model with tertiary education taken into account, while the speeds of adjustment in case of primary & secondary education are not very different from each other. the results of long-run, as well as the short-run association between different education levels and agriculture’s sector output, are presented in table 2 along with the results of the diagnostic tests. while cusum and cusum of squares test showing the stability of the models are reported in appendix c. other variables included in the model are the capital series generated for the agriculture sector, number of persons employed in the agriculture sector, land, fertilizer, seed, water and agriculture sector exports. the results show that primary, secondary & tertiary education all have positive coefficients in the models of the output of the agriculture sector in the long-run. faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 14 table 2: impact of education on output of agriculture sector in pakistan long-run variable primary secondary tertiary education 0.015845 (0.9345) 0.031673 (0.8937) 0.097506 (0.0854) capital 0.269164 (0.0444) 0.274227 (0.0555) 0.108617 (0.2305) employment -0.131292 (0.7461) -0.153764 (0.7324) -0.033281 (0.8194) land 0.462884 (0.2416) 0.473059 (0.2474) 0.439196 (0.0303) fertilizer 0.168229 (0.2484) 0.161490 (0.3111) 0.135417 (0.0939) seed 0.114471 (0.0470) 0.107215 (0.1836) 0.148739 (0.0003) water 0.499750 (0.2428) 0.474214 (0.3359) 0.895037 (0.0003) exports_agri 0.009942 (0.8920) 0.013398 (0.8650) -0.003521 (0.9214) c 17.94870 (0.0003) 18.16064 (0.0007) 17.37172 (0.0001) short run education 0.008603 (0.9341) 0.017092 (0.8923) 0.093824 (0.0993) capital 0.146138 (0.0709) 0.147985 (0.0079) 0.104516 (0.2493) employment -0.071283 (0.7393) -0.082978 (0.7233) -0.032024 (0.8194) land 0.251315 (0.2671) 0.255284 (0.2650) 0.422612 (0.0515) fertilizer 0.091337 (0.2832) 0.087147 (0.3515) 0.130303 (0.1034) seed 0.062150 (0.1100) 0.057858 (0.2675) 0.074639 (0.0276) water 0.271330 (0.3095) 0.255907 (0.3980) 0.313988 (0.1975) exports_agri 0.005398 (0.8911) 0.007230 (0.8630) -0.003388 (0.9217) c 9.744930 (0.0090) 9.800301 (0.0093) 16.71576 (0.0003) diagnostic tests 𝑹𝟐 0.99 0.99 0.99 serial correlation 2.762678 (0.1113) 2.769776 (0.1109) 0.420900 (0.5165) heteroskedasticity 2.151098 (0.1425) 2.554386 (0.1100) 0.940793 (0.3321) ramsey reset test 1.158977 (0.2939) 1.134694 (0.2989) 0.375970 (0.5470) *estimation technique is ardl, p-values are given in parenthesis. however, the coefficient of education is significant only in the model with a tertiary education. the impact of capital is found positive in all the three models where it is found significant in the case of primary and secondary education but found insignificant in the model with a tertiary education. the employment variable has a negative sign in journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 15 all three models, pointing towards surplus labour in the agriculture sector. all the control variables appear having positive signs in the model with primary and secondary education, however, the p-values show that most of the coefficients are insignificant. in the model with the tertiary education level, all other variables i.e land, fertilizer, seed and water appear as significant and positive determinants of agriculture sector output except agriculture sectors export. the analysis of the impact of different education levels in the short-run reported in table 2 also implies the same conclusions. summing up, when the effect of different education levels is analyzed on output per employed person of the agriculture sector of pakistan, education appears to have a negative impact in all the three cases. however, the issue is investigated further by taking agriculture’s sector output as the dependent variable and including the number of employed persons as a determinant in the model. this modification turns the education coefficient positive for all three education levels and employment coefficients appear as negative values. hence pointing towards the well-known ‘surplus labour’ phenomena of developing countries in the agriculture sector. 3.3 industrial sector the impact of different education levels on output per employed person of the industrial sector is analyzed through the ardl method. the bounds test reported in appendix b1 verifies the existence of a long-run association for all the three models. speed of adjustment coefficients reported in appendix b2 shows that 53% of deviations from long-run equilibrium get corrected during a year in model with primary education, it is 64%in case of secondary education and 45% for tertiary education. table 3 reports the impact of education levels on output per employed person in the industrial sector in the short-run and long-run along with the diagnostic test results of the models. cusum and cusum square tests reported in appendix c confirm the stability of the models. industrial sector’s capital per employed person, foreign direct investment, credit to the private sector, the difference between the weighted average of returns on advance and deposits, external debt and exports of the manufacturing sector are included as explanatory variables in the models. in the long-run, the coefficients of each level of education are positive in models of industrial sector output per employed person. the education coefficient is greater in the model of primary education than the coefficient of secondary education as a proxy. the coefficient is found significant in both models. however, in the model where tertiary education is used as a proxy for human capital the coefficient is found insignificant. the faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 16 table 3: impact of education on output per employee in the industrial sector long run variable primary secondary tertiary education 0.411362 (0.0008) 0.236865 (0.0020) 0.034676 (0.7134) k 0.454421 (0.0749) 1.049690 (0.0001) 1.479203 (0.0001) fdi -0.001219 (0.9611) 0.030322 (0.1802) 0.032754 (0.3692) cps 0.014428 (0.1161) -0.000235 (0.9634) -0.005951 (0.4177) rs -0.018721 (0.2093) -0.040559 (0.0002) -0.052690 (0.0022) inf 0.003494 (0.1769) 0.005145 (0.0694) 0.003319 (0.4525) ed -0.006455 (0.0230) -0.005456 (0.0426) -0.011196 (0.0062) export (manufacturing) 0.011988 (0.3221) 0.000797 (0.9403) -0.012751 (0.4334) short run education 0.220785 (0.0003) 0.153366 (0.0169) 0.015775 (0.7274) k 0.871617 (0.0001) 0.679657 (0.0001) 0.672916 (0.0001) fdi -0.000654 (0.9607) 0.019633 (0.1980) 0.014900 (0.3956) cps 0.007744 (0.0234) -0.000152 (0.9635) -0.002707 (0.4542) rs -0.010048 (0.1871) -0.026261 (0.0001) -0.023970 (0.0009) inf 0.001875 (0.2639) 0.003331 (0.1015) 0.001510 (0.4758) ed -0.003465 (0.0131) -0.003533 (0.0271) -0.005093 (0.0032) export (manufacturing) 0.006434 (0.2548) 0.000516 (0.9403) -0.005801 (0.4374) diagnostic tests 𝑹𝟐 0.98 0.96 0.95 serial correlation 0.002395 (0.9610) 0.095054 (0.7578) 0.645253 (0.4218) heteroskedasticity 1.465420 (0.2261) 0.423245 (0.5153) 0.390103 (0.5322) ramsey reset test 1.061815 (0.3151) 0.291053 (0.5950) 0.556117 (0.4637) * p-values are given in parenthesis, k is capital per employed person in the industrial sector capital per employed person is found highly significant and positive in all the models. the deeper look at the results shows that combining a higher education level with capital makes the capital more productive. the coefficient of per employed person capital in a model with secondary education is higher than its coefficient in the model of primary education. similarly, the coefficient is higher in model with tertiary education as compared to primary & secondary education. in conclusion, it is evident from the results that increased level of education makes capital more productive and the overall impact journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 17 on output per employed person will be higher in case of increase of the higher education level as compared to the lower levels of education. the results of the impact of different education levels on the growth of industrial output per employed person in the short-run are also reported in table 3. it is found that, in the short-run, primary education is contributing more to the output per employed person of the sector. in the model with tertiary education as a proxy for human capital the coefficient of tertiary education is found insignificant, implying that the data does not support the active role of tertiary education in increasing the per employed person output. liu and armer(1993) reported a significant and positive impact of primary and junior high school education on the output of the overall economy in taiwan, while senior high school and tertiary education were found insignificant in explaining output growth. the model results for tertiary education as a proxy for human capital is pointing towards the developing country phenomenon where the lower levels of education are more effectively contributing as compared to the tertiary education level because of its underutilization. 3.4 services sector the results of the analysis of services sector output per employed person are reported in this section. the f-bound test reported in appendix b1 approves the presence of a long-run association. speed of adjustment coefficients reported in appendix b2 shows that highest speed of adjustment is in case of secondary education where 71% of deviations from long-run are corrected in one year, followed by 58% for the tertiary education and 48% for primary education. the results of the impact of different education levels on output per employed person in the services sector in the long-run & short-run are illustrated in table 4. cusum and cusum squares tests reported in appendix c authenticate the stability of all the models. the education variable in all three cases; whether it is proxied by the primary, secondary or tertiary education is found positive and significant in the long-run. it implies that an increase in education in the services sector increases the output per employed person. the coefficients of capital per employed person are also positive and significant in all three models. it could be seen that the coefficient of capital increases as it is mixed with a higher education level pointing towards increased productivity of capital. it implies that changes in output per employed person will be higher if a higher education is combined with capital. the results of the control variables in the model are also reported in table 4. similarly, in the short-run, the coefficient of education variable shows that all three education levels entering into separate models as a proxy for human capital have a positive and statistically significant impact. similarly, the coefficient of per employed person capital is statistically significant and positive. faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 18 it is found that in the short-run the highest impact on output per employed person of the services sector is of secondary education as compared to primary & tertiary education. the coefficient of capital per employed person is positive and significant in all the three models of output per employed person in the short run. table 4: impact of education on output per employee of services sector long run variable primary secondary tertiary education 0.322997 (0.0001) 0.300007 (0.0001) 0.166509 (0.0001) k 0.326675 (0.0048) 0.958221 (0.0001) 1.242560 (0.0001) fdi 0.017185 (0.0284) 0.008182 (0.4853) 0.008000 (0.5839) rs -0.011215 (0.0010) -0.012319 (0.0132) -0.007629 (0.1688) ed -0.003066 (0.0166) -0.004349 (0.0001) -0.006880 (0.0001) inf 0.001300 (0.2701) -0.002292 (0.1912) -0.001956 (0.3647) trade -0.009127 (0.0009) -0.001452 (0.5408) -0.003824 (0.1959) cps -0.004821 (0.0054) 0.000918 (0.7430) 0.001505 (0.6705) short run education 0.112493 (0.0304) 0.215527 (0.0001) 0.097352 (0.0001) k 0.832566 (0.0001) 0.688392 (0.0001) 0.726478 (0.0001) fdi 0.008315 (0.0641) 0.005878 (0.4841) 0.004677 (0.5838) rs -0.005426 (0.0036) -0.008850 (0.0213) -0.004460 (0.1728) ed -0.001484 (0.0107) -0.003124 (0.0003) -0.004022 (0.0001) inf 0.000629 (0.2914) -0.001647 (0.1854) -0.001144 (0.3564) trade -0.000247 (0.7883) -0.001043 (0.5385) -0.002236 (0.1976) cps -0.002332 (0.0201) 0.000660 (0.7407) 0.000880 (0.6654) c 2.669612 (0.0040) 3.149972 (0.1190) ------- diagnostic tests 𝑹𝟐 0.99 0.99 0.98 serial correlation 1.479604 (0.2238) 2.817507 (0.1088) 2.139528 (0.1591) heteroskedasticity 0.348786 (0.5548) 0.016674 (0.8973) 0.5324 (0.4656) ramsey reset test 0.261890 (0.6150) 2.521951 (0.1280) 0.670621 (0.4225) * p-values are given in parenthesis, k is capital per employed person in the services sector journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 19 3.5 overall economy the inquiry of the impact of different education levels on the overall economic growth of pakistan is reported and discussed below. in addition to capital per employee and education variable foreign direct investment (fdi), the difference between weighted average rate of return on advances and deposits (rs), inflation (inf), external debt (ed), exports and credit to private sector(cps) are also included in the model. the results of f-bound test reported in appendix b1 validates the presence of a long-run association in the models for primary, secondary & higher education. the speeds of adjustment reported in appendix b2 shows that highest speed of adjustment is in case of primary education where 46% of deviation is corrected during a year. the speeds of adjustment are 39% and 33% for tertiary education and secondary education respectively. the results of the effect of different education levels on the economic growth of pakistan in the long-run are registered in table 5. the effect of each education level is found positive and highly significant. the coefficients of capital per employed person are also positive and decidedly significant in all three models. the highest coefficient of education variable is in the case of primary education, followed by secondary education and then for tertiary education. however, if we also take the coefficients of capital per employed person in consideration, it could be seen that the coefficient of capital per employee is highest in the model with secondary education. so the sum of education and the capital variable is highest in the case of secondary education, followed by primary education and then by tertiary education. our results are in line with the conclusions of petrakis and stamatakis(2002), gemmel(1996), papageorgiou(2003), loening(2005) and tsai et al.(2010). petrakis & stamatakis(2002) concluded that primary education contributes more to the development of less developed countries while higher education plays an active role in the development of advanced countries. similarly, gemmel(1996) found that primary education has significant effects on the economic growth of developing countries while tertiary education plays a vital role in the development of oecd countries. both studies got ambiguous results in the case of secondary education and in some cases, it has a negative sign. moreover, papageorgiou(2003) established based on cross-country regressions that primary education plays important role in growth of final output while post primary education promotes adoption and innovation. loening(2005) using time series data for guatemala ( a developing country) concluded that primary education contributes more to the economic growth than secondary and higher education level. faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 20 tsai et al.(2010) reported that for developing countries the impact of secondary education is positive and significant while it is insignificant in case of developed countries. table 5: impact of different levels of education on economic growth of pakistan long run variable primary secondary tertiary education 0.447095 (0.0001) 0.277853 (0.0126) 0.233058 (0.0014) k 2.009052 (0.0011) 2.234842 (0.0353) 1.857573 (0.0220) fdi 0.053494 (0.0044) 0.057543 (0.0442) 0.059541 (0.0116) rs -0.033272 (0.0134) -0.039448 (0.0752) -0.029917 (0.0731) inf -0.000968 (0.6616) -0.002478 (0.4776) -0.002072 (0.4651) ed -0.001842 (0.4605) -0.005821 (0.1972) -0.005278 (0.1138) exports 0.002852 (0.6383) 0.006079 (0.5293) 0.003098 (0.6878) cps -0.007413 (0.0651) -0.008804 (0.2006) -0.007070 (0.1987) short run education 0.208598 (0.0066 ) 0.093551 (0.1165) 0.092971 (0.0500) k 0.937347 (0.0001) 0.752456 (0.0007) 0.741017 (0.0005) fdi 0.024958 (0.0156) 0.019374 (0.0773) 0.023752 (0.0375) rs -0.015524 (0.0009) -0.013282 (0.0080) -0.011935 (0.0073) inf -0.000452 (0.6571) -0.000834 (0.4663) -0.000827 (0.4556) ed -0.000859 (0.4182) -0.001960 (0.0758) -0.002106 (0.0364) exports 0.001331 (0.6378) 0.002047 (0.5387) 0.001236 (0.6880) cps -0.003459 (0.0762) -0.002964 (0.1776) -0.002820 (0.1846) c -9.593674 (0.0005) -6.966131 (0.0095) -5.924245 (0.0229) diagnostic tests 𝑹𝟐 0.982 0.977 0.979 serial correlation 0.0000532 (0.9942) 0.103105 (0.7481) 0.713092 (0.3984) heteroskedasticity 1.412948 (0.2346) 0.679400 (0.4098) 0.419914 (0.5170) ramsey reset test 0.859234 (0.3645) 0.717925 (0.4064) 1.028599 (0.3220) * p-values are given in parenthesis the results of the control variables in the model are listed in table 5. the impact of foreign direct investment is positive and significant in all three models. the effect of journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 21 the variable ‘rs’ which is showing the difference of weighted average rate of return on advances and deposits is found negative and significant. it means that if the spread between returns on advances and deposits is low, that will mean the loans are available at lower rates and savings are encouraged by higher returns, then the output per employed person will be higher and vice versa. similarly, the coefficient of inflation variable is found to be negative and insignificant in all three models. the impact of external debt is also negative in every model but the parameters are found insignificant. exports coefficients are positive while the credit to private sector appears as having a negative effect on the output per employed person in the long run. the results reported in table 5 show that all three education levels have a positive and significant effect on the economic growth of the overall economy in the short-run. the highest coefficient in the short-run is of primary education, then for secondary education and then tertiary education. capital per employed person also has a positive and highly significant impact in the short run. the coefficient of capital is higher in the model of primary education and almost the same in the model with secondary & higher education. the diagnostic test results for the models are reported in table 5 while cusum and cusum square tests are given in appendix c showing the stability of the models. 3.6 comparison of major sectors and the overall economy the comparison of coefficients of different education levels across sectors and with overall economy gives an important insight about the role of different education levels in the economy of pakistan. the results of the long-run association are summarized in table 6. the comparison of the coefficients of primary education as a proxy for human capital across different sectors in the economy shows that primary education is best utilized in the industrial sector. whereas, secondary education is best deployed in the services sector. similarly, the comparison of the coefficients of tertiary education across sectors shows that in the long-run the best utilization of tertiary education is in the services sector. in the services sector, the coefficient of the tertiary education is highest and is most significant as compared to other sectors. in the industrial sector, although the coefficient has a positive value, it is low and is also highly insignificant. pointing towards the inefficient utilization of tertiary education in the industrial sector. in the case of the agriculture sector, when output per employed person is taken as the dependent variable, the results negate the presence of a long-run association. however, in the model for the output of the agriculture sector, the effect of tertiary education is faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 22 found to be significant and positive. yet, this positive effect is lower than the effect of tertiary education in the services sector. table 6: comparison of different sectors and overall economy (long-run) variable primary secondary tertiary overall economy 0.447095 (0.0001) 0.277853 (0.0126) 0.233058 (0.0014) agriculture sector bounds test results show that long-run relationship doesn’t exist industry 0.411362 (0.0008) 0.236865 (0.0020) 0.034676 (0.7134) services sector 0.322997 (0.0001) 0.300007 (0.0001) 0.166509 (0.0001) agriculture sector (output) 0.015845 (0.9345) 0.031673 (0.8937) 0.097506 (0.0854) *p-values are given in parenthesis. the results of the impact of different education levels on output per employed person in the short run are presented in table 7. if we compare the effect of primary education in each sector of the economy, its greatest utilization is in the industrial sector, followed by the services sector. the coefficient of primary education is negative in the agriculture sector when output per employed person is taken as a dependent variable. however, when the model for agriculture sector is modified by taking employment of agriculture sector into account and agriculture’s sector output as dependent variable the education coefficients turn to positive values but the coefficient in the agriculture sector remains below the coefficient of the industrial and services sector. table 7: comparison of different sectors and overall economy (short-run) variable primary secondary tertiary overall economy 0.208598 (0.0066 ) 0.093551 (0.1165) 0.092971 (0.0500) agriculture sector -0.149597 (0.1406) -0.268610 (0.0226) -0.145109 (0.0670) industrial sector 0.220785 (0.0003) 0.153366 (0.0169) 0.015775 (0.7274) services sector 0.112493 (0.0304) 0.215527 (0.0001) 0.097352 (0.0001) agriculture sector (output) 0.008603 (0.9341) 0.017092 (0.8923) 0.093824 (0.0993) *estimation technique is ardl, p-values are given in parenthesis. similarly, in case of secondary education the greater impact is in case of the services sector, then for the industrial sector and negative in agriculture sector when output per employed person is considered and positive in case of agriculture’s output. in the case of the overall economy, the effect is positive and representative of all sectors. for journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 23 tertiary education again the largest positive impact is in the services sector. the coefficient of tertiary education in the industrial sector is found positive but insignificant. the coefficient of tertiary education is negative and significant in the model of per employed person output of agriculture sector and significant and positive in the model of agriculture’s sector output. for the overall economy, the effect is significant and positive and representative of all sectors. 4 conclusion the analysis of the impact of different education levels of employed persons shows that each education level whether its primary, secondary or tertiary have a positive effect on the output per employed person both in the short-run and long-run. it is true for the overall economy, industrial & services sector but the agriculture sector gives somewhat different results. in the agriculture sector, it is found that each education level is negatively associated with the output per employed person. the deeper analysis showed us that the greater negative effect of employment evades the positive effect of education in the agriculture sector. however, when the effect of education is considered on output, controlling for the effect of employment in the model, the coefficients turn to be positive. the comparison of different sectors for each education level shows that primary education contributes more to the industrial sector. while the contribution of the secondary & tertiary education is highest in the services sector. it implies that primary education is best utilized in the industrial sector as compared to other sectors, while secondary education & tertiary education are efficiently used in the services sector as compared to the industrial & agriculture sector. the analysis of the impact of different education levels on the agriculture sector shows the negative impact of each education level on output per employed person in the short-run. the data doesn’t support the presence of a long-run association in all three models of output per employed person of the agriculture sector. the impact of education on agriculture sector is further analyzed by estimating the model for output of the sector instead of output per employed person and by including the effect of employment along with other control variables in the model. in the long-run and short-run, the coefficients of education are found positive. however, the coefficient of education variable is only significant in the case of tertiary education. for the industrial sector, the impact of primary, secondary & tertiary education on output per employed person is found positive both in the short-run and long-run. faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 24 however, the coefficient of tertiary education remained insignificant. it is also deduced that in the long-run secondary education makes capital more productive as compared to primary education. the coefficient of capital per employed person is found higher in the model of secondary education as compared to primary education. also, the sum of coefficients of capital and education is higher in the model of secondary education as compared to primary education in the long-run. it is found that the speed of adjustment is greater in the case of secondary education as compared to primary & tertiary education. for the services sector, the education variable in all three cases; whether it is proxied by the primary, secondary or higher education is found positive and significant. the coefficients of capital per employed person are also positive and significant in all three models. moreover, the coefficient of capital increases as it is mixed with a higher education levels, it means that a higher education level escalates the productivity of capital. it implies that changes in output per employed person will be higher if a higher education levels is combined with capital. speed of adjustment was found highest in case of secondary education followed by tertiary education and then for the primary education. 5 policy implications the analysis of agriculture sector shows that surplus employment in the agriculture sector is evading positive returns of education. it implies that there is a need to lower employment in the agriculture sector or to make the agriculture sector as broader as to absorb all the surplus labour efficiently. it will require to define policies that provide incentives to one’s that are associated to the agriculture sector to transfer to other sectors or to increase the productivity and market size of the agriculture sector to absorb the surplus labour. the results show that secondary & tertiary education is best utilized in the services sector. it is necessary to analyze and device policies to know the reasons why secondary & tertiary education is contributing less to other sectors. it is also required to define policies to make education at least as useful as it is in the services sector. references abbas, q., & foreman-peck, j. s. (2008). human capital and economic growth: pakistan 1960-2003. lahore journal of economics, 13(1), 1-27. afridi, a. h. (2016). human capital and economic growth of pakistan. business & economic review, 8(1),77-86. journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 25 aghion, p. and howitt, p. (1998). endogenous growth theory. cambridge, ma: mit press. aghion, p., & howitt, p. (1992). a model of growth through creative destruction. econometrica, 60, 323-351. aghion, p., boustan, l., hoxby, c., & vandenbussche, j. (2009). the causal impact of education on economic growth: evidence from us. brookings papers on economic activity, 1, 1-73. amir, m., mehmood, b., & shahid, m. (2012). impact of human capital on economic growth with emphasis on intermediary role of technology: time series evidence from pakistan. african journal of business management, 6(1), 280. asteriou, d., & agiomirgianakis, g. m. (2001). human capital and economic growth: time series evidence from greece. journal of policy modeling, 23(5), 481-489. barro, r. j. (1996). determinants of economic growth: a cross-country empirical study (no. w5698). national bureau of economic research. barro, r. j., & lee, j. w. (2010). a new data set of educational attainment in the world, 1950-2010. nber working paper no. 15902. national bureau of economic research. barro, r.j. (1991). economic growth in a cross section of countries.the quarterly journal of economics, 106(2): 407-43. barro, r.j. & j.w.lee. (1993). international comparisons of educational attainment. journal of monetary economics, 32, 363-94. benhabib, j., & spiegel, m. m.(1994). the role of human capital in economic development evidence from aggregate cross-country data. journal of monetary economics, 34(2), 143-173. bils, m., & klenow, p. j. (2000). does schooling cause growth?. american economic review, 90(5), 1160-1183. fedderke, j. (2002). technology, human capital, growth and institutional development. theoria, 49(100), 1-26. galor, o., & zeira, j. (1993). income distribution and macroeconomics. the review of economic studies, 60(1), 35-52. galor, oded and david n. weil. 1996. ‘‘the gender gap, fertility, and growth.’’ american economic review 86(3): 374–87. faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 26 gemmel, n. (1996), “evaluating the impacts of human capital stocks and accumulation on economic growth: some new evidence”, oxford bulletin of economics and statistics, vol. 58, pp. 9-28. hafeez & rahim (2019) found that enrollments in primary, secondary and higher education are important in determining economic growth of pakistan. the analysis was based on the period 1971 to 2013. hall, r. e., & jones, c. i. (1999). why do some countries produce so much more output per worker than others?. the quarterly journal of economics, 114(1), 83-116. hanushek e a (2015). why standard measures of human capital are misleading. kdi journal of economic policy 2015, 37(2): 22–39 hanushek e a and wöessmann l (2010), education and economic growth. in: penelope peterson, eva baker, barry mcgaw, (editors), international encyclopedia of education. volume 2, pp. 245-252. oxford https://doi.org/10.4236/me.2018.912132 iqbal (2018). human development and economic growth in pakistan. socio economic challenges, volume 2, issue 3), 66-75. jalil, a., & idrees, m. (2013). modeling the impact of education on the economic growth: evidence from aggregated and disaggregated time series data of pakistan. economic modelling, 31, 383-388. jenkins, h. (1995). education and production in the united kingdom. oxford: nuffield college kakar, zaheer khan, bashir ahmad khilji, and muhammad jawad. (2011). relationship between education and economic growth in pakistan: a time series analysis. journal of international academic research, 11(1). kendrick, j. (1956). productivity trends: capital and labor. in productivity trends: capital and labor (pp. 3-23). nber. khan, jangraiz, khattak naeem ur rehman, khan amir. (2015). human capital economic growth nexus: a causality analysis for pakistan. city university research journal, volume 5(2): 279-290 kiani, a. (2013). education is essential for economic growth in pakistan. african journal of business management, 7(26), 2548. journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 27 kobzev kotásková, s., procházka, p., smutka, l., maitah, m., kuzmenko, e., kopecká, m., & hönig, v. (2018). the impact of education on economic growth: the case of india. acta universitatis agriculturae et silviculturae mendelianae brunensis, 66(1), 253-262. krueger, a. b., & lindahl, m. (2001). education for growth: why and for whom?. journal of economic literature, 39(4), 1101-1136. kyophilavong, p., ogawa, k., kim, b., & nouansavanh, k. (2018). does education promote economic growth in lao pdr?: evidence from cointegration and granger causality approaches. the journal of developing areas, 52(2), 1-11. lee, e. k. (2012). higher education expansion and economic growth in japan and south korea (doctoral dissertation, university of pittsburgh). lin, t. c. (2003). education, technical progress, and economic growth: the case of taiwan. economics of education review, 22(2), 213-220. loening, j. l. (2005). effects of primary, secondary, and tertiary education on economic growth: evidence from guatemala, volume 1 of 2. the world bank. lucas, r. e. (1988). on the mechanics of economic development. journal of monetary economics, 22(1), 3-42. mankiw, n. g., romer, d., & weil, d. n. (1992). a contribution to the empirics of economic growth. the quarterly journal of economics, 107(2), 407-437. mincer r, j. (1974), “schooling, experience and earnings”, national bureau of economic research, cambridge, ma. mincer, j., & polachek, s. (1974). family investments in human capital: earnings of women. in marriage, family, human capital, and fertility (pp. 76-110). journal of political economy 82 (2), part ii. narayan, p. k. (2005). the saving and investment nexus for china: evidence from cointegration tests. applied economics, 37(17), 1979-1990. nehru, v., & dhareshwar, a. (1993). sources, methodology and results. revista de análisis económico, 8(1), 37-59. nelson, r. r., & phelps, e. s. (1966). investment in humans, technological diffusion, and economic growth. the american economic review, 56(1/2), 69-75. nnyanzi, j.b. and kilimani, n. (2018) estimation of disaggregated impacts of education expansion on economic growth in sub-saharan africa. modern economy, 9, 2119-2149. faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 28 oduro-ofori, e., aboagye, a. p., & acquaye, n. a. e. (2014). effects of education on the agricultural productivity of farmers in the offinso municipality. international journal of development research vol. 4, issue, 9, pp. 1951-1960. oecd (1998), human capital investment – an international comparison, organization for economic co-operation and development. paris. omodero, c. o., & nwangwa, k. c. (2020). higher education and economic growth of nigeria: evidence from co-integration and granger causality examination. international journal of higher education, 9(3), 173-182. papageorgiou, c. (2003). distinguishing between the effects of primary and post‐ primary education on economic growth. review of development economics, 7(4), 622-635. pesaran, m. h., & shin, y. (1998). an autoregressive distributed-lag modelling approach to cointegration analysis. econometric society monographs, 31, 371413. pesaran, m. h., shin, y., & smith, r. j. (2001). bounds testing approaches to the analysis of level relationships. journal of applied econometrics, 16(3), 289-326. petrakis, p. e. and d. stamatakis (2002), “growth and educational levels: a comparative analysis”, economics of education review, vol. 21, pp. 513-521. pritchett, l. (2001). where has all the education gone? world bank economic review 15(3), 367–391. renelt, d. (1991). economic growth: a review of the theoretical and empirical literature (vol. 678). world bank publications. romer, p. (1990) “endogenous technological change.” journal of political economy 99 (5): s71-s102. romer, p. m. (1986). increasing returns and long run growth. journal of political economy, 94, 1002–1037. sari, r., & soytas, u. (2006). income and education in turkey: a multivariate analysis. education economics, 14(2), 181-196. self, s., & grabowski, r. (2003). education and long-run development in japan. journal of asian economics, 14(4), 565-580. sianesi, b., & reenen, j. v. (2003). the returns to education: macroeconomics. journal of economic surveys, 17(2), 157-200. solow, r. m. (1957) "technical change and the aggregate production function." review of economics and statistics 39:3 12-320. journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 29 solow, r. m.. (1956). a contribution to the theory of economic growth. the quarterly journal of economics, 70(1), 65–94 temple, j. (1999). the new growth evidence. journal of economic literature, 37(1), 112-156. topel, r. (1999). labor markets and economic growth. handbook of labor economics, 3, 2943-2984. tsai, c. l., hung, m. c., & harriott, k. (2010). human capital composition and economic growth. social indicators research, 99(1), 41-59. tsamadias, c., & pegkas, p. (2012). the effect of education on economic growth in greece over the 1981–2009 period. does the proxy of human capital affect the estimation?. international journal of education economics and development, 3(3), 237-251. appendix a: variables definitions and sources economic survey of pakistan (esp) labourforce survey of pakistan (lfs) world developmetn indicators (wdi) handbook of statistics, state bank of pakistan (hbs) variables definition source overall economy y gdp local currency (constant in 2005 prices) / number of employed persons esp capital generated on the basis of gfcf (local currency and constant value) esp k capital/ number of employed persons -------- primary education number of employed persons having primary education in pakistan lfs secondary education number of employed persons having secondary education lfs tertiary educaton number of employed persons having tertiary education lfs fdi foreign direct investment, net inflows (% of gdp) wdi rs the difference between the weighted average rate of return on advances and deposits hbs inf inflation, consumer prices wdi ed external debt stocks (% of gni) wdi exports exports of goods and services as % of gdp wdi cps domestic credit to private sector (% of gdp) wdi agriculture sector output agriculture’s sector value added local currency constant in 2005 prices esp y value added local currency,constant 2005 prices/number of employed persons in agriculture sector esp k capital series generated on the basis of gfcf of agriculture sector and then it is divided by number of employed persons in agriculture sector esp primary education number of employed persons in agriculture sector having primary level of education lfs secondary education number of employed persons in agriculture sector having secondary level of education lfs tertiary education number of employed persons in agriculture sector having tertiary level of education lfs land cropped area (million hactares) esp fertilizer fertilizer offtake( thousand nutrient tonnes) esp faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 30 water water availability (million acre feet) esp seed improved seed distribution (000 tonnes) esp export_agri food exports as percentage of gdp food comprises food, live animals, beverages, tobacco, animal and vegetable oils and fats, oil seeds, oil nuts, and oil kernels.some calculations are performed to take food exports as percentage of gdp wdi industry y industrial sector’s value added(local currency constant in 2005 prices) /number of employed persons industrial sector esp k capital series generated on the basis of gfcf of industrial sector and then it is divided by number of employed persons in industrial sector esp primary education number of employed persons in industrial sector having primary education lfs secondary education number of employed persons in industrial sector having secondary education lfs tertiary educaton number of employed persons in industrial sector having tertiary education lfs export(manufacturing) manufactures exports (% gdp) wdi services sector y services sector’s value added(local currency constant in 2005 prices) /number of employed persons services sector esp k capital series generated on the basis of gfcf of services sector and then it is divided by number of employed persons in services sector esp primary education number of employed persons in services sector having primary education lfs secondary education number of employed persons in services sector having secondary education lfs tertiary education number of employed persons in services sector having tertiary education lfs trade trade is the sum of exports and imports of goods and services measured as a share of gross domestic product. wdi appendix a1: augmented dickey-fuller test results variables levels first difference adf p-value adf p-value overall economy y -1.5189 0.5113 -5.5019 0.0001 k -0.6910 0.8350 -6.5593 0.0000 primary education -1.8601 0.6514 -6.8798 0.0000 secondary education -2.4019 0.3717 -6.2015 0.0000 tertiary educaton -3.4225 0.0661 -6.7074 0.0000 fdi -2.8995 0.0565 -3.7584 0.0077 rs -1.3562 0.5914 -6.5675 0.0000 inf -2.4796 0.1294 -7.0061 0.0000 ed -0.6842 0.8371 -4.9215 0.0004 exports -0.6825 0.8376 -4.3999 0.0015 cps -1.3158 0.6104 -4.4556 0.0013 agriculture sector y -2.3298 0.1692 -6.0502 0.0000 output -0.0761 0.9438 -6.2247 0.0000 k -1.2897 0.8724 -5.4238 0.0001 primary education -1.8041 0.6791 -6.7716 0.0000 secondary education -2.7437 0.2271 -5.6277 0.0001 tertiary educaton -1.2368 0.8853 -4.7779 0.0006 land -2.2917 0.1806 -8.3224 0.0000 fertilizer -0.1209 0.9387 -6.0584 0.0000 water -1.8803 0.6412 -8.3598 0.0000 seed -2.6586 0.2591 -14.4927 0.0000 export_agri -0.4499 0.5122 -6.647544 0.0000 industry y -1.9147 0.3216 -6.5728 0.0000 k -1.0117 0.7370 -5.1166 0.0002 primary education -0.9703 0.9339 -6.1951 0.0000 journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 31 secondary education -1.1989 0.8938 -6.1193 0.0000 tertiary educaton -0.9972 0.7422 -6.5501 0.0000 export(manufacturing) -1.3700 0.5848 -5.0407 0.0003 levels first difference adf p-value adf p-value services sector y -1.7456 0.7070 -8.7591 0.0000 k -2.4617 0.1339 -9.3439 0.0000 primary education -0.2133 0.9269 -6.9893 0.0000 secondary education -1.9917 0.5835 -5.3849 0.0001 tertiary educaton -1.9382 0.6115 -6.1767 0.0000 trade -1.6478 0.4476 -6.8871 0.0000 * y and k represents output and capital per employed person respectively appendix b1: bounds test results primary secondary tertiary agriculture sector (output per worker) 2.841494 3.668981 2.235252 agriculture sector (output) 5.093671 5.097568 7.789563 industrial sector 6.567579 7.453945 5.255022 services sector 8.197842 8.751988 8.418333 overall economy 5.26 3.68 4.09 * f-bound values are listed in the table *critical value for f-bound test are 2.73 and 4.16 at 5% level of significance * critical values for overall economy are 2.22 and 3.39 at 5% level of significance appendix b2: speed of adjustment primary secondary tertiary agriculture sector (output) -0.542932 -0.539645 -0.962240 industrial sector -0.536718 -0.647483 -0.454918 services sector -0.483860 -0.718406 -0.584663 overall economy -0.4666 -0.3367 -0.3989 faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 32 appendix c agricultural sector (output per capita) cusum test primary education secondary education tertiary education -15 -10 -5 0 5 10 15 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 cusum 5% significance -15 -10 -5 0 5 10 15 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 cusum 5% significance -15 -10 -5 0 5 10 15 96 98 00 02 04 06 08 10 12 14 16 cusum 5% significance cusum of squares test primary education secondary education tertiary education -0.4 0.0 0.4 0.8 1.2 1.6 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 cusum of squares 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 cusum of squares 5% significance -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 96 98 00 02 04 06 08 10 12 14 16 cusum of squares 5% significance agriculture sector (output) cusum test primary education secondary education tertiary education -15 -10 -5 0 5 10 15 96 98 00 02 04 06 08 10 12 14 16 cusum 5% significance -15 -10 -5 0 5 10 15 96 98 00 02 04 06 08 10 12 14 16 cusum 5% significance -16 -12 -8 -4 0 4 8 12 16 92 94 96 98 00 02 04 06 08 10 12 14 16 cusum 5% significance cusum of squares test primary education secondary education tertiary education -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 96 98 00 02 04 06 08 10 12 14 16 cusum of squares 5% significance -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 96 98 00 02 04 06 08 10 12 14 16 cusum of squares 5% significance -15 -10 -5 0 5 10 15 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 cusum 5% significance journal of applied economics and business studies, volume. 4, issue 3 (2020) 1-34 https://doi.org/10.34260/jaebs.431 33 industrial sector cusum test primary education secondary education tertiary education -15 -10 -5 0 5 10 15 98 00 02 04 06 08 10 12 14 16 cusum 5% significance -15 -10 -5 0 5 10 15 96 98 00 02 04 06 08 10 12 14 16 cusum 5% significance -15 -10 -5 0 5 10 15 96 98 00 02 04 06 08 10 12 14 16 cusum 5% significance cusum squares test primary education secondary education tertiary education -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 98 00 02 04 06 08 10 12 14 16 cusum of squares 5% significance -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 96 98 00 02 04 06 08 10 12 14 16 cusum of squares 5% significance -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 96 98 00 02 04 06 08 10 12 14 16 cusum of squares 5% significance services sector cusum test primary education secondary education tertiary education -15 -10 -5 0 5 10 15 2000 2002 2004 2006 2008 2010 2012 2014 2016 cusum 5% significance -15 -10 -5 0 5 10 15 98 00 02 04 06 08 10 12 14 16 cusum 5% significance -15 -10 -5 0 5 10 15 98 00 02 04 06 08 10 12 14 16 cusum 5% significance cusum of squares test primary education secondary education tertiary education -0.4 0.0 0.4 0.8 1.2 1.6 2000 2002 2004 2006 2008 2010 2012 2014 2016 cusum of squares 5% significance -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 98 00 02 04 06 08 10 12 14 16 cusum of squares 5% significance -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 98 00 02 04 06 08 10 12 14 16 cusum of squares 5% significance faiza hassan, hafsa hina, abdul qayyum (late) & anwar hussain 34 overall economy cusum test primary education secondary education tertiary education -15 -10 -5 0 5 10 15 96 98 00 02 04 06 08 10 12 14 16 cusum 5% significance -15 -10 -5 0 5 10 15 96 98 00 02 04 06 08 10 12 14 16 cusum 5% significance -15 -10 -5 0 5 10 15 96 98 00 02 04 06 08 10 12 14 16 cusum 5% significance cusum of squares test primary education secondary education tertiary education -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 96 98 00 02 04 06 08 10 12 14 16 cusum of squares 5% significance -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 96 98 00 02 04 06 08 10 12 14 16 cusum of squares 5% significance -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 96 98 00 02 04 06 08 10 12 14 16 cusum of squares 5% significance journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 143 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x energy consumption: the importance of institutional quality in pakistan rubina ilyas1* 1. pakistan institute of development economics, islamabad, pakistan abstract the energy sector of pakistan has undergone several changes over the previous several years, due to the ever-increasing energy consumption. the objective of this study is to determine the extent to which the consumption of energy sources can be attributed to pakistan’s institutional quality, based on the perception of positive impact of institutional quality on facilitation of public goods. in this context, the demand functions of energy sources (electricity, natural gas and petroleum products) for four main economic sectors(i.e., residential, industry, commercial and agricultural sectors), have been examined by employing bound testing to cointegration (ardl approach), over a period of 35 years (19842019). the analysis reveals that institutional quality has significant positive impact on energy consumption. the results of the study provide implications for energy sector reforms by illustrating the importance of institution-building policy efforts. keywords energy demand; energy prices; demand elasticities jel classification c02, c31, c63 1 introduction pakistan has experienced a rapid growth in consumption of primary energy over the last few decades. according to pakistan country statistics (2018), the primary energy consumption has grown by 150% in the last twenty years. however, the supply has not increased in accordance to the rise in demand due to several constraints such as circular debt, increased reliance on oil/gas, decline in gas reserves, inadequate exploitation of coal reserves, underutilization of power generating units, hydel etc. the energy crisis has become an inevitable challenge due to inefficient resource management, lack of * rubina_ilyas@hotmail.com rubina ilyas 144 investment in the infrastructure and absence of a well-planned policy (alahdad, z., 2012). successive energy policies have focused on the expansion and diversification of energy system through installation of new energy projects, yet the production capacity fails to fulfill the energy needs. the country still faces serious challenges to meet its energy needs (siddiqui, 2004; hye and raiz, 2008). in order to ensure efficient use of energy the pricing policy has also undergone a number of changes. ogra which conduct publics hearings and set prices for natural gas and petroleum products have capped the petroleum products prices in different periods (afia malik, 2008). on the other hand, several reforms have been made in the power sector for improving bill collection and reducing the shortfall in the generation capacity. in the past decade, pakistan spent more than 2 percent of its gdp on electricity subsidies, which not only increased national debt but also weakened the country’s external position. in 2015-16 the electricity subsidies were 0.8% of the gdp (world bank, 2017), which was the same as total expenditure on public health. in order to cover the cost of electricity production at least rs.3 is paid by the government on every kwh consumed by the domestic user (awan, et. al, 2019). however, the government has increased electricity tariffs for the residential consumers and eliminated subsidies for commercial and industrial consumers in october 2013 (world bank, 2016). despite the pricing policy reforms the demand and supply gap has further widened due to the inefficient use and wastage of energy resources (annual report of the state bank, 2018). in view of the current challenges, two important issues have been addressed in this study. firstly, it analyzes why the energy pricing policy reforms have not been successful in increasing the efficient use of energy? in order to address this question, this study extends the existing literature by analyzing the demand functions of energy sources1 of the four main economic groups (i.e., residential, industry, commercial and agricultural sectors). in this context, this study incorporates other potential variables (degree of industrialization, urbanization and institutional quality) in addition to the traditional economic variable (income and price). pakistan was ranked among the most urbanized countries of south asia with a rise of 47% in urban population growth in 2010 (adb). this rise in urbanization has increased the transportation usage by 300% (economic survey of pakistan, 2012), which leads to more energy consumption. in addition to urbanization, institutional quality has also been analyzed as a key determinant of energy specifically electricity consumption. previous studies suggest that institutional quality increase the consumption of energy, as they increase the provision of public goods (pierre and rothstein, 2011; acemoglu and robinson, 2006; deacon et al., 2003 and boix et al., 2003). provision of public goods such as access to electricity and natural gas requires large scale transmission and distribution infrastructure and long term investment which is highly expensive (abbott, 2001). 1 electricity, petroleum products and natural gas journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 145 in most countries electrification programs specifically rural electrification have been conducted through government funding and special national programmes (zomers, 2003). therefore, provision of energy goods (i.e. electricity and natural gas) to the population is completely politically driven and depends on the political and administrative system of a country. the government of pakistan has also conducted several rural electrification and gas provision programs since 2007, under the power sector development programs to raise the socioeconomic standards of rural population. the village electrification and gas provision programs have always been an integral part of the government agenda in pakistan (economic survey of pakistan, 2020). hence, including institutional quality as a determinant of energy consumption will provide useful insights to analyze the extent to which pakistan’s energy consumption is affected by its institutional quality. it will also assess whether the extensions of national grid system ensure the availability of electricity to the targeted population. however, previous studies on pakistan’s energy consumption have ignored this variable. in order to overcome this gap, the institutional quality index has been analyzed as a determinant of energy consumption which has been constructed by using data from international country risk guide (icrg). the index has been constructed by taking an average of pakistan’s government stability, investment profile, law and order, corruption, democratic accountability and bureaucratic quality. however, in case of pakistan very little analytical work has been done in this regard. iqbal m. (1983) analyzed the residential price and income elasticities for electricity and gas for the period 1960 to 1981 using ols. the income elasticity was positive for both the energy goods; however, the own price elasticity was found to be negative in case of gas. later siddiqui and haq (1999) examined the demand function for gas, electricity and petroleum products for aggregate and disaggregate levels using ols estimation technique. they concluded that in general energy consumption is income and price elastic. there are other studies which focus on only one energy item i.e. electricity which use causality test and co integration in order to analyze the relation of electricity consumption and growth of the economy. for example, according to aqeel and butt (2001), siddique (2004) and lee (2005), there is uni-directional causality from energy demand to gdp while, jamil and ahmed (2010) predicted the causality to be unidirectional from economic activity to electricity use. khan, m. a., & qayyum, a. (2009), analyzed the trend of electricity demand for the period 1970-2006. the income elasticity was found to be positive for all the groups while the price elasticity was negative. similarly, shahbaz, et al. (2012) analyzed how gdp and energy demand are related, using the period 1972-2011 and confirmed it to be positive. nawaz, s., iqbal, n., & anwar, s. (2013) checked for the linear as well as nonlinear electricity demand function for the period of 1971-2012 by applying a model of logistic smooth transition regression. they proved that a long run relationship exists between electricity consumption, its prices and gdp per capita. they further concluded that the electricity consumption is basically influenced by development. the nonlinear estimates showed rubina ilyas 146 that the demand for electricity is insensitive to any price change beyond the threshold level. later in 2014, javid, m., & qayyum, a. analyzed the relationship among electricity demand, its prices and real economic activity for aggregate as well as sectoral levels over the period of 1972-2010. the found that the nature of relationship among these variables was stochastic instead of linear and deterministic. recent study by zaman and khalid (2017), have analyzed the causal relationship of energy demand and agricultural technology. their results shows that energy consumption and agricultural technology both have causal effect on each other. however, they have also analyzed the use of electricity ignoring other components of energy. the other energy sources like gas and petroleum products also contributes a considerable portion of the total energy demand so they should not be ignored. moreover, in view of the depleting gas reserves, it has become necessary to analyze how the policy should be formulated. to examine the responsiveness of energy consumption to its determinants (energy prices, gdp, number of users etc.), this study has employed bound testing approach to cointegration (ardl approach) developed by pesaran et al. (2001) considering the non-stationarity of most of the variables. the ardl cointegration approach is a better approach as compared to other cointegration methods such as engle and granger (1987), johansen (1988) and ordinary least square procedures as used in the previous studies of energy demand in case of pakistan. first the ardl procedure has the advantage of being independent of classification of variables into i(0) and i(1) and there is no need of pre testing of variable for unit root. a single reduced form equation is employed to estimate long run relationship instead of a system of equations as used in the conventional cointegration procedures (ozturk, i., & acaravci, a., 2010). secondly, ardl modelling does not need large data samples for validity as required by johansen cointegration techniques. it can test for cointegration with consistent parameters even when the data sample is small (mah, 2000). furthermore, ardl allows for different optimal lags for variables which is not possible while using conventional cointegration procedures (narayan, p. k., & smyth, r., 2005). the rest of the study is organized as follows: section 2 analyzes the consumption trends of the economic sectors; section 3 discusses the methodology and data used and section 4 delves into the results and section 5 gives the conclusion and policy implication of the study. 2. energy consumption in pakistan (sector wise) over the last decades, the pakistan has been facing severe energy crisis due to insufficient energy production to fulfill the escalating demand. the energy supply has increased by more than 40 times during the past 25 years, yet the demand outweighs the supply. moreover, the inter-sectoral patterns of consumption have changed significantly over time (figure 1). the industrial sector contributes to the highest share followed by the transport and household sector. after experiencing a high growth, the industrial journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 147 consumption started to decline in the wake of global oil crises during 2008, with a rise in the transport share correspondingly. the total energy consumption showed a decline during 2008-09 which later on increased at a fast pace after 2014. on the other hand, the household sector has been witnessing a constant rise after 2008 with a peak of 25 percent in 2012-13. figure 1. energy consumption by sectors (% share) source: pakistan energy yearbook (2019) the fall in share of industry and the corresponding rise in share of transport can be attributed to several factors. firstly, the slowing down of industrial growth due to declining growth momentum of the economy reduced demand for energy. secondly, the rising oil prices post 2008 led to substitution of petrol/ diesel for cng in the transportation sector, increasing energy demand in this sector. the energy needs are expected to increase three folds by 2050 while the supply situation is not very inspiring (asif. m, 2009). the primary energy sources and their issues are discussed in details in the following subsections. 2.1 electricity power shortage has become a challenge for pakistan. power shortage has become a challenge for pakistan. for almost a decade, the power crisis is prevailing in the country resulting in routine power outages, called “load shedding” in both urban and rural areas. since 2007, the electricity supply shortfalls (5000-5500mw) have led to load shedding of long hours which in some areas has even averaged up to 17 hours per day (imf, 2013). according to siddiqui, r., et al., (2008), the unserved energy due to power outages has caused an industrial output loss that is estimated to vary between 12% to 0 20 40 60 0 50 100 150 1 9 9 7 -9 8 1 9 9 8 -9 9 1 9 9 9 -2 0 0 0 2 0 0 0 -0 1 2 0 0 1 -0 2 2 0 0 2 -0 3 2 0 0 3 -0 4 2 0 0 4 -0 5 2 0 0 5 -0 6 2 0 0 6 -0 7 2 0 0 7 -0 8 2 0 0 8 -0 9 2 0 0 9 -1 0 2 0 1 0 -1 1 2 0 1 1 -1 2 2 0 1 2 -1 3 2 0 1 3 -1 4 2 0 1 4 -1 5 2 0 1 5 -1 6 2 0 1 6 -1 7 2 0 1 7 -1 8 m t o e p e rc e n ta g e axis title energy consumption by sectors (% share) domestic commercial industry agriculure transport other govt. total energy consumtion (mtoe) rubina ilyas 148 37% for silakot, faisalabad, gujrat and gujranwala. these shortfalls have not only raised the production cost through the arrangement of alternative energy for many firms, but also have caused delay in meeting supply commitments. the supply shortage was estimated to exceed more than 7,000 mw in 2015 (national transmission and dispatch company, ntdc). along with the power supply shortfalls, the shortage of natural gas is increasing, which is a primary energy source for the poor who do not have electricity access. in the hours of power cuts most households use gas fueled generators which adds to the inefficient use. electricity is a second source of energy yet it is indispensable not only for household but also for industry, transport, etc. there was an increase in the generation of electricity by thermal and hydel but the import of electricity remained at 335 gwh in 2016. the installed capacity has increased by 30% in the phase of 2012-18 (from 22812mw to 29573 mw). although the generation has increased yet the inefficiencies in the transmission and distribution (t&d) make the sustained supply of electricity impossible (as shown in figure 2). the t&d loses in pakistan are inevitably high due to a number of reasons such as theft through illegal connections, low voltage of distribution line, outdated generation plants, weak infrastructure etc. the distribution loses range from 9.5%-34.3% which is quite high as compare with the neighboring countries like china whose distribution loses amount to 5.8% (annual report, sbp 2015). figure 2. annual growth rates of electricity generation and consumption source: author’s calculation (pakistan energy yearbook, various issues) on the other hand, the source wise consumption patterns of sectors shows that the domestic sector contributes the largest share in electricity consumption over many years (figure 3). during 2018-19, the domestic share accounted for 49%, while industrial share contributed 26%, commercial share 8% and agricultural share accounted for 9%. there are severe consequences of such a large domestic share as the residential sector -10 -5 0 5 10 15 annual growth rates of electricity generation and consumption annual growth rate of production annual growth rate of consumption journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 149 contribute more to the t&d losses than other sectors (state bank, 2014). moreover, this sector is also highly subsidized. figure 3. sectoral electricity consumption source: pakistan energy yearbook (various issues) about fifty one million (27%) people of the population do not have access to power (iea, 2017). those that are connected have to experience load shedding of some hours on daily basis. in absence of reliable distribution network the rural households use lpg as a source of fuel which is more expensive as compared to the subsidized natural gas used by urban households. 2.2 natural gas analyzing the consumption pattern of natural gas reveals that the industrial sector constitutes the largest share in gas consumption. however, it is seen that from 2005 onwards the consumption share of natural gas of the transport sector has increased immensely. in order to reduce import bill on oil, the use of compressed natural gas (cng) as fuel for automobiles was encouraged through approval of marketing licenses of more than 3416 cng stations by the state. however in view of the depleting gas reserves a ban has been placed on establishment of new gas reserves. the total consumption of natural gas showed a decline during 2010-2015 with decreasing shares of the industrial and transport sector (figure 4). this decline in consumption of natural gas is accompanied by the increase in oil fuel consumption in response to the declining prices of petrol. during 2018-19 the share of domestic sector was 33%, the share of industrial sector was 38 % and that of transport sector was recorded as 8% of total gas consumption. 0 1000 2000 3000 4000 5000 1990 1995 2000 2005 2010 2015 2017 k to e years electricity consumption by sectors industry transport residential commercial and public services agriculture / forestry rubina ilyas 150 figure 4. natural gas consumption by sectors source: pakistan energy yearbook (various issues) despite being sufficient in gas supply pakistan is also facing gas crisis in addition to electricity crisis. according to some estimates by o. rauf et al., (2015), the country has entered the deficiency state after 2006. before 2006, the local resources have fulfilled the demand of the domestic as well as the commercial and industrial consumers with an ample supply. the growth rate of as consumption increased rapidly during 201516 while the growth rate of production was 1.2% (figure 5). figure 5: annual growth rates of gas production and consumption source: author’s calculation (pakistan energy yearbook, various issues) the shortfalls of gas supply have leaded to gas load shedding and supply cuts especially in the winter seasons. the severity of this shortfall can be assessed by the 0 100000 200000 300000 400000 500000 600000 700000 800000 1990 1995 2000 2005 2010 2015 2017 t j natural gas consumption by sectors industry transport residential commercial and public services -10 0 10 20 30 annual growth rates of gas production and consumption annual growth rate of production annual growth rate of consumption journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 151 decision of the government implementing load shedding of two days per week on cng stations and industries since november 2009 to march 2010 (khan, et. al., 2012). the shortage has been substituted by alternative fuels like diesel and kerosene during the phase of 2005-10 increasing the share of oil from 29%to 31%. until 2007, the power sector had the largest share in gas consumption. however, the use of gas for generating power is declining gradually in pakistan. the industrial sector is also facing similar situation of gas supply cut. it has been estimated that 80% of the textile industry relies upon supply of gas from ssgcl and sngpl (all-pakistan textile millers association (aptma)). due to massive gas shortage the annual opportunity loss was more than 5 billion dollars during the last four years (economic survey of pakistan 2018). 2.3 petroleum products in case of petroleum products consumption, the transport sector has the largest share with an increasing trend especially after the fall in world oil prices. during 2018-19 transport sector had the largest share in petroleum products consumption (80%), followed by industry (9%), while commercial sector contributed a 4% share, domestic sector had 3% and agriculture sector had the smallest share of 0.1% (figure 6). figure 6. petroleum products consumption by sectors source: pakistan energy yearbook (various issues) the energy requirements in pakistan are mostly fulfilled by gas and oil. however the indigenous oil resources fall short of the growing demand which has led to the importation of oil and petroleum products in huge amounts from the gulf countries. during fluctuations in the international oil prices, pakistan has focused on reducing the reliance on importation of oil. the production of crude oil in 2016 was 24.02 million barrels while 4.98 mmt (million metric tons) was imported (pakistan economic survey 2015-16). on the other hand the demand for petroleum products is also greater than the oil refining capacity which is why half of the imports constitute of refined products 0 5000 10000 15000 20000 25000 1990 1995 2000 2005 2010 2015 2017 k to e petroleum products consumption by sectors industry transport residential commercial and public services agriculture / forestry rubina ilyas 152 (figure 7). the import of petroleum products contributed to the import bill by 17% in 2017. as international oil prices remained high in 2018, the cost of imported energy escalated by 25% ($13.3 billion) raising the contribution of imported energy to 37% in the import bill (state bank annual report 2018-19). figure 7. annual growth rates of petroleum products source: author’s calculation (pakistan energy yearbook, various issues) in the coming years the demand of petroleum products is expected to grow at a faster pace as compared to the production making is impossible for pakistan to gain selfsufficiency, given the volume of oil resources. during 2018-19, 20.03 million tones, of petroleum products were consumed, out of which, 2.8 million tons were domestically produced. in order to overcome the high cost of imported oil and petroleum products, there is a need to improve the necessary infrastructure of the oil refineries which requires considerable amount of investment. although the government has introduced several incentives to attract investment by the private sector in the petroleum policy 1997, still it has failed to achieve satisfactory results. 3. empirical specification, methodology and data 3.1. data this study uses annual data for the period 1984-2019 collected from different sources to estimate the demand functions of each energy item. data for gross domestic product (gdp) is used from pakistan economic survey, data for consumption of energy sources and their prices is taken from energy year book (various issues). following ahlborg h., et al., (2015), the institutional quality index has been constructed by taking the average of 6 variables from the icrg data that best fits with the theoretical perceptions about the role of institutes in provision of energy goods. the variables -25 -20 -15 -10 -5 0 5 10 15 20 production and consumption of petroleum products annual growth rate of production annual growth rate of consumption journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 153 included are (1) law and order, (2) corruption, (3) government stability, (4) investment profile, (5) democratic accountability and (6) bureaucratic quality. the data for urbanization has been taken from wdi which measures urbanization as percentage of total population. table 1 summarizes the descriptive statistics for the variables used. table 1. descriptive statistics of variables. variables definitions mean std. dev. min max dependent variable ec (h) electricity consumption (residential sector) 14.46 0.52 13.34 16.04 ec (c) electricity consumption (commercial sector) 12.66 0.37 12.06 13.18 ec (i) electricity consumption (industrial sector) 14.38 0.59 13.81 16.11 ec (a) electricity consumption (agriculture sector) 13.34 0.29 12.81 14.10 gc (h) gas consumption (residential sector) 14.60 0.43 13.60 14.70 gc (c) gas consumption (commercial sector) 12.97 0.53 12.08 13.76 gc (i) gas consumption (industrial sector) 14.96 0.30 14.19 15.83 sk consumption of super kerosene oil 12.13 0.29 11.62 12.59 fo consumption of furnace oil 13.74 0.57 12.19 14.56 ms consumption of motor spirit 13.91 0.44 13.13 15.06 hsd consumption of high speed diesel 15.38 0.49 14.34 15.88 ldo consumption of low diesel oil 6.71 0.53 5.87 7.57 independent variables y gross domestic product 15.45 0.47 14.58 16.18 iq institutional quality index 3.94 0.12 2.84 4.08 ur percentage of urban population to total population 3.47 0.04 3.58 3.34 neh no. of electricity users in the residential sector 15.97 0.55 14.93 16.73 nei no. of electricity users in the industrial sector 12.13 0.29 11.62 12.59 nec no. of electricity users in the commercial sector 14.37 0.20 14.19 14.74 nea no. of electricity users in the agriculture sector 12.07 0.32 11.55 12.60 ngh no. of gas users in the residential sector 15.06 0.39 14.65 15.84 ngi no. of gas users in the industrial sector 8.65 0.34 8.37 9.28 ngc no. of gas users in the commercial sector 10.96 0.18 10.72 11.30 tr no. of cars, taxis, motorcycles 14.06 0.57 12.93 15.00 c no. of cars, taxis, motorcycles 12.44 0.45 11.61 13.09 pea price of electricity for agriculture sector 0.44 0.69 0.60 0.45 pec price of electricity for commercial sector 0.89 0.57 0.25 1.06 peh price of electricity for residential sector 0.49 0.54 0.42 0.93 pei price of electricity for industrial sector 0.53 0.12 0..55 0.70 pfo price of electricity of furnace oil 2.13 0.50 2.02 2.63 pgc price of gas for commercial sector 4.84 0.14 3.93 5.02 pgh price of gas for residential sector 4.40 0.13 3.38 4.51 pgi price of gas for industrial sector 4.80 0.29 3.37 5.40 phsd price of high speed diesel oil 2.62 0.19 2.11 2.72 pldo price of low diesel oil 2.42 0.24 2.01 2.58 rubina ilyas 154 pms price of motor spirit 3.06 0.37 2.98 3.61 psk price of super kerosene oil 2.51 0.26 2.11 2.65 3.2. empirical specification the energy sector is divided in two ways i.e. by different energy source (electricity, petroleum products and natural gas) and by different user groups (household, industry, commercial and others). the energy consumption function for each category is specified as under: 3.2.1. electricity consumption following galindo (2005), electricity consumption function is specified for each sector as: 𝑙𝑛𝐸𝑐𝑖𝑡 = 𝛼0 + 𝛼1𝑙𝑛𝑃𝑖𝑡 𝑒 + 𝛼2𝑙𝑛𝑃𝑗𝑖𝑡 + 𝛼3 𝑙𝑛𝑌𝑡 + 𝛼4𝑙𝑛𝑁𝑖𝑡 + 𝛼5𝑙𝑛𝑈𝑅𝑡 + 𝛼6𝑙𝑛𝐼𝑄𝑡 + 𝜀𝑡 (1) where 𝐸𝑐𝑖𝑡 is the electricty consumption for sector “i” respectively, for period t. 𝑃𝑖𝑡 𝑒 is electricity price faced by each sector “i” and 𝑃𝑗𝑖𝑡 is the price of related substitutable energy sources like gas, furnace oil, kerosene oil etc. 𝑌𝑡 gives the total ouput (gdp). 𝑁𝐼𝑡 is the number of users sector i. due to the unavailability of data on price of appliances in time series form it was not included in the model. 3.2.2. gas consumption the consumption function of gas is specified in a similar manner for each sector 𝑙𝑛𝐺𝑐𝑖𝑡 = 𝛼0 + 𝛼1𝑙𝑛𝑃𝑖𝑡 𝑔 + 𝛼2𝑙𝑛𝑃𝑗𝑖𝑡 + 𝛼3𝑙𝑛𝑌𝑡 + 𝛼4𝑙𝑛𝑁𝑖𝑡 + 𝛼5𝑙𝑛𝑈𝑅𝑡 + 𝛼6𝑙𝑛𝐼𝑄𝑡 + 𝜀𝑡 (2) where 𝐺𝑐𝑖𝑡 is the gas consumption for sector “i” respectively, for period t. 𝑃𝑖𝑡 𝑔 is the gas price faced by each sector “i” and 𝑃𝑗𝑖𝑡 is the price of related substitutable energy sources like electricity in each sector. the gas price also includes the surcharge imposed by government. 𝑌𝑡 gives the total ouput (gdp). 𝐼𝑡 is the number of users sector i. the demand for gas and electricity is also affected by the price of appliances using these energy sources but since the data for this variable is not available it is not added in the demand function. 3.2.3. petroleum products consumption the consumption analysis of each petroleum product (super kerosene, furnace oil, motor spirit, high speed and low diesel oil) can be classified according to the consumption of its users. for instance, the industry consumes more than 60% of furnace oil while the domestic sector consumes about 70% of super kerosene oil is consumed by the residential sector. similarly, the consumption of high speed diesel, low diesel oil and motor spirit is predominantly from the transport sector. the consumption function of the petroleum products is given accordingly 𝑙𝑛𝐾𝑐ℎ𝑡 = 𝛼0 + 𝛼1𝑙𝑛𝑃ℎ𝑡 𝑘 + 𝛼2𝑙𝑛𝑃𝑗ℎ𝑡 + 𝛼3𝑙𝑛𝑌𝑡 + 𝛼4𝑙𝑛𝑝𝑜𝑝𝑡 + 𝛼5𝑙𝑛𝐼𝑄𝑡 + 𝜀𝑡 (3) 𝑙𝑛𝐹𝑂𝑐𝑖𝑡 = 𝛼0 + 𝛼1𝑙𝑛𝑃𝑖𝑡 𝑓𝑜 + 𝛼2𝑙𝑛𝑃𝑗𝑖𝑡 + 𝛼3𝑙𝑛𝑌𝑡 + 𝛼5𝑙𝑛𝑌𝑖𝑡 + 𝛼6𝑙𝑛𝐼𝑄𝑡 + 𝜀𝑡 (4) journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 155 𝑙𝑛𝐻𝑆𝐷𝑐𝑟𝑡 = 𝛼0 + 𝛼1𝑙𝑛𝑃𝑟𝑡 ℎ𝑠𝑑 + 𝛼2𝑙𝑛𝑃𝑗𝑟𝑡 + 𝛼3𝑙𝑛𝑌𝑟𝑡 + 𝛼4𝑙𝑛𝐵𝑇𝑡 + 𝛼5𝑙𝑛𝐼𝑄𝑡 + 𝜀𝑡 (5) 𝑙𝑛𝐿𝐷𝑂𝑐𝑟𝑡 = 𝛼0 + 𝛼1𝑙𝑛𝑃𝑟𝑡 𝑙𝑑𝑜 + 𝛼2𝑙𝑛𝑃𝑗𝑟𝑡 + 𝛼3𝑙𝑛𝑌𝑟𝑡 + 𝛼4𝑙𝑛𝐵𝑇𝑡 + 𝛼5𝑙𝑛𝐼𝑄𝑡 + 𝜀𝑡 (6) 𝑙𝑛𝑀𝑆𝑐𝑟𝑡 = 𝛼0 + 𝛼1𝑙𝑛𝑃𝑟𝑡 𝑚𝑠 + 𝛼2𝑙𝑛𝑃𝑗𝑟𝑡 + 𝛼3𝑙𝑛𝑌𝑟𝑡 + 𝛼4𝑙𝑛𝐶𝑡 + 𝛼5𝑙𝑛𝐼𝑄𝑡 + 𝜀𝑡 (7) where 𝐾𝑐𝑖𝑡is the kerosene consumption by the household sector in period t. 𝑃ℎ𝑡 𝑘 is kerosene price and 𝑃𝑗𝑖𝑡 is the price of related substitutable energy sources like electricity and gas for the residential sector. 𝑌𝑡 gives the total ouput (gdp) and pop stands for the population. 𝐹𝑂𝑐𝑖𝑡 is furnace oil consumption of the industrial sector. 𝑃𝑖𝑡 𝑓𝑜 is the price of furnace oil and 𝑃𝑗𝑖𝑡 is the price of other related energy sources. 𝑌𝑖𝑡is the value added by industry. similarly 𝑀𝑆𝑐𝑟𝑡 , 𝐻𝑆𝐷𝑐𝑟𝑡, 𝐿𝐷𝑂𝑐𝑟𝑡 are the motor spirit, high speed diesel and low diesel oil consumptions respectively by the transport sector. 𝑃𝑟𝑡 𝑖 is the price of the ith fuel and 𝑌𝑟𝑡is the value added by the transport sector. since the consumption of these fuels is also affected by the number of vehicles hence the number of buses and trucks (𝐵𝑇) and the number of cars, taxis, motorcycles (𝐶𝑡 )are also included. 3.3. econometric methodology there are two steps in the bound testing approach. in the first step the long run relationship is established between the variables of the consumption function. the long run and short run coefficients are calculated in the following step provided that cointegration exists (pesaren et al., 2001). the ardl procedure to cointergration is explained as under consider a vector of two variable zt where 𝑧𝑡 = (𝑦𝑡 , 𝑥𝑡 ′ )′, 𝑦𝑡is the dependent variable and xt is a vector of regressors. the data generating process of zt is p-order vector autoregression. ∆𝑦𝑡 = 𝛽0 + 𝛽1𝑡 + 𝜋𝑦𝑦 𝑦𝑡−1 + 𝜋𝑥𝑥 𝑥𝑡−1 + ∑ 𝜗𝑖 ∆𝑦𝑡−𝑖 + ∑ ∅𝑗 ′ ∆𝑥𝑡−𝑗 + 𝜃𝑤𝑡 + 𝑢𝑡 𝑞 𝑗=0 𝑝 𝑖=1 (8) here, 𝜋𝑦𝑦and 𝜋𝑥𝑥 are long run multipliers, 𝛽0is the drift, t is the time trend and 𝑤𝑡is a vector of exogenous components. the bound testing procedure entails the following hypothesis 𝐻0, 𝜋𝑦𝑦 = 0, 𝜋𝑦𝑥.𝑥 = 0 ′ 𝐻0, 𝜋𝑦𝑦 ≠ 0, 𝜋𝑦𝑥.𝑥 ≠ 0 ′or , 𝜋𝑦𝑦 ≠ 0, 𝜋𝑦𝑥.𝑥 = 0 ′ 𝑜𝑟 𝜋𝑦𝑦 = 0, 𝜋𝑦𝑥.𝑥 ≠ 0 ′ these hypothesis are estimated using the f statistic. if the f stat is below the lower bound than the null of no cointegration cannot be rejected and if it is greater than the upper bound the null is rejected. the energy demand literature mostly (implicitly) assumes that explanatory variables used in equation (1) to equation (7) are exogenous. however, in the presence of endogeneity an adequate number of lagged values of these variables have to be included (bentzen and engsted, 1999) rubina ilyas 156 4. empirical results before applying the ardl model, the variables have been checked for presence of unit root using ng and perron (2001). the results are given in the appendix (table a1) which show that most of the variables are non-stationary at level as the p-value is insignificant even at 10% significance level. however, at first difference all of the variables turn out to be stationary. 4.1 electricity consumption the existence of long run relationship was tested in eq (1) for electricity consumption. the maximum number of lags were set equal to 2 in the ardl model for all the sectors. the computed f stat are reported in table 2. table 2. results of the bounds test for cointegration electricity consumption variables included f-statistic household 𝐸𝑐𝐻𝑡 , 𝑃𝐻𝑡 𝑒 , 𝑃𝐻𝑗𝑡 , 𝑌𝑡 , 𝑈𝑅𝑡 , 𝑁𝐻𝑡 , , 𝐼𝑄𝑡 12.08* industry 𝐸𝑐𝐼𝑡 , 𝑃𝐼𝑡 𝑒 , 𝑃𝐼𝑗𝑡 , 𝑌𝑡 , 𝑌𝑖𝑡, 𝑁𝑖𝑡, 𝐼𝑄𝑡 7.80* commercial 𝐸𝑐𝑐𝑡 , 𝑃𝑐𝑡 𝑒 , 𝑃𝑐𝑗𝑡, 𝑌𝑡 , 𝑌𝑐𝑡, 𝑁𝑖𝑡, 𝐼𝑄𝑡 6.77* agriculture 𝐸𝑐𝐴𝑡 , 𝑃𝐴𝑡 𝑒 , 𝑃𝐴𝑗𝑡 , 𝑌𝑡 , 𝑌𝐴𝑡 , 𝑁𝐴𝑡 , 𝐼𝑄𝑡 9.86* note: critical values are given in pesaran et al. (2001). * indicates significance at the 1 percent level table 1 shows that the computed f-statistics is greater than the upper bound of the critical values, giving evidence of cointegration among the variables of the household, industrial, commercial and agriculture sectors. now equation 2.1 was further estimated by using the following ardl (m,n,p,q,r) specification. 𝑙𝑛𝐸𝑐𝑖𝑡 = 𝛼0 + 𝛼1 ∑ 𝑙𝑛𝐸𝑐𝑡−𝑖 𝑚 𝑖=1 + ∑ 𝛼2 𝑙𝑛𝑌𝑡−𝑖 𝑛 𝑖=0 + ∑ 𝛼3𝑙𝑛𝑃𝑡−𝑖 𝑒𝑝 𝑖=0 + ∑ 𝛼4𝑙𝑛𝑃𝑡−𝑖 𝑗𝑞 𝑖=0 + ∑ 𝛼5𝑙𝑛𝑌𝑖𝑡−𝑖 𝑟 𝑖=0 + ∑ 𝛼6𝑙𝑛𝐼𝐶𝑇𝑡−𝑖 + ∑ 𝛼7𝑙𝑛𝐼𝑄𝑡−𝑖 𝑡 𝑖=0 𝑠 𝑖=0 + 𝜀𝑡 (9) for each sector (i) a maximum lag was selected according to the minimum value of the sbc criterion. the empirical results obtained for each of the sector for the long run are reported in table 2.2. panel “a” of the table represents the estimated long run results while panel “b” fives the results for short run. the results show the income elasticity for all the sectors is statistically significant with a positive value, except for the domestic sector in the long run. this implies that the income effect on electricity consumption of households is positive but negligible. the own price effect is statistically significant with a negative sign for all sectors. the coefficient of own price effect (in absolute terms) is bigger for the residential sector as compared to other sectors. the price effect of natural gas is statistically insignificant while in case of kerosene oil the price effect is positive and statistically significant. these results imply that the residential sector uses super kerosene as a substitute of electricity. the impact of number of consumer on electricity demand is positively significant. while, the impact of income is significantly positive in the short run. the journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 157 estimated values of own price effect and cross price effect (kerosene oil) have similar signs as of the long run estimates, however they are smaller in magnitude and are statistically significant. the number of users is positive but has an insignificant impact for short run. this means that in case of domestic sector the key factors determining electricity consumption are electricity price, kerosene oil price and number of users. table 3. long-run estimates and short-run error-correction representation dependent variable independent variables coefficient s.e panel (a): long-run estimates ec(h) y 0.48 1.51 peh -0.98** 0.43 pgh 0.08 0.21 pfo -0.06 0.04 psk 0.64** 0.30 iq 1.56*** 0.22 ur 1.41** 0.56 nh 1.29** 0.49 ec(c) y 0.47*** 0.16 pec -0.38** 0.17 pgc 1.04 0.67 pfo 0.01 0.05 iq 0.96*** 0.16 yi 0.33** 0.15 nc 3.35*** 0.85 ec (i) gdp 0.67*** 0.26 pei -2.36* 1.16 pfo -0.80*** 0.23 pgi -0.16 0.11 iq 0.58** 0.21 ur 1.23*** 0.29 ni 4.79** 1.81 ec(a) gdp 4.81** 2.24 pea -0.12 0.08 iq 1.4*** 0.38 na 3.57* 1.96 panel (b): error-correction representation ∆ec(h) d(eh(-1)) 0.38** 0.18 d(y) 0.22*** 0.08 d(y(-1)) 0.16 0.09 d(peh) -0.17** 0.08 d(pgh) 0.07 0.18 d(pfo) -0.05 0.04 d(psk) 0.53* 0.27 d(iq) 0.88** 0.31 d(iq(-1)) 0.86*** 0.28 d(ur) 0.62** 0.23 d(nh) 0.27 0.53 cointeq(-1) -0.82 0.23 ∆ec(c) d(y) 1.41 1.05 d(pec) -0.20*** 0.05 rubina ilyas 158 d(pgc) -0.04 0.19 d(pgc(-1)) -0.58 0.48 d(pfo) 0.003 0.02 d(iq) 0.10 0.27 d(yc) 0.17*** 0.05 d(yc(-1)) 0.01 0.04 d(nc) 1.78*** 0.35 cointeq(-1) -0.53*** 0.15 ∆ec(i) d(y) 0.58*** 0.23 d(y(-1)) 0.32** 0.11 d(pei) -0.39*** 0.05 d(pei(-1)) -1.85** 0.70 d(pfo) 0.22 0.15 d(pfo(-1)) 0.97** 0.42 d(pgi) -0.18 0.12 d(iq) 1.23*** 0.21 d(iq(-1)) 0.58** 0.43 d(yi) 0.28** 0.12 d(ni) 0.74*** 0.14 cointeq(-1) -0.66*** 0.11 ∆ec(a) d(ea(-1)) 0.27 0.17 d(y) 0.28* 0.13 d(gdp(-1)) 2.54 2.69 d(pea) -0.77* 0.36 d(pea(-1)) -0.36 0.27 d(iq) 0.04 0.18 d(ya) 0.24*** 0.23 d(ya(-1)) 0.26** 0.08 d(na) 3.04* 1.64 d(na(-1)) 0.47 1.40 cointeq(-1) -0.57** 0.21 note: *statistical significance at 10% level. **statistical significance at 5% level ***statistical significance at 1% level in case of the industrial sector, the price effect is negative and statistically significant though its value is quite small. the cross-price effect for natural gas and furnace oil are statistically insignificant. the impact of number of consumers is positive and significant but is smaller in magnitude as compared to the commercial and agriculture sectors. while, in case of commercial sector the income elasticity has a significant positive value. the estimated price elasticity has a significant negative value, however it has a very small magnitude. this means that for the commercial sector electricity demand is relatively inelastic and smaller in magnitude than the rest of the sectors. the number of consumers is positive and statically significant in the long as well as the short run. for the agricultural sector the income effect is positive showing an increase in income would raise the consumption of electricity. however the price effect is negative but insignificant. these results coincide with the findings of siddiqui (2004) and qayyum (2014). the reason for insignificant price effect could be the subsidized prices for the agricultural sector. moreover the consumption of the agriculture sector is far less than journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 159 the other sector. however the price effect is significant in the short run but with a very small magnitude showing relatively inelastic demand. the results for the institutional quality index are in line with our theoretical expectations have statistically significant positive values for all the sectors in the long run. these results indicate consistency with the perception that if the politicians are able to implement large electrification projects, the electricity consumption is more likely to increase. 4.2 gas consumption the existence of long run relationship was tested in eq (2) for gas consumption. the maximum number of lags were set equal to 2 in the ardl model for all the sectors. the computed f stat are given in table 4. table 4. results of the bounds test for cointegration gas consumption variables included f-statistic household 𝐺𝑐𝐻𝑡 , 𝑃𝐻𝑡 𝑔 , 𝑃𝐻𝑡 𝑗 , 𝑌𝑡 , 𝑈𝑅𝑡 , 𝐼𝑄𝑡, 𝑁ℎ𝑡 8.61* industry 𝐺𝑐𝐼𝑡 , 𝑃𝐼𝑡 𝑔 , 𝑃𝐼𝑡 𝑗 , 𝑌𝑡 , 𝑁𝑖𝑡 , 𝑌𝑖𝑡 , 𝐼𝑄𝑡 7.99* commercial 𝐺𝑐𝑐𝑡 , 𝑃𝑐𝑡 𝑔 , 𝑃𝐶𝑡 𝑗 , 𝑌𝑡 , 𝑁𝑐𝑡 , 𝑌𝑐𝑡 , 𝐼𝑄𝑡 6.16* * indicates significant at the 1 percent level in each model the computed 𝐹𝐸𝐶 (.) is higher than the upper bound critical value. thus the null hypothesis of no cointegration cannot be accepted. now equation 2.2 was further estimated by using the following ardl (m,n,p,q,r) specification. 𝑙𝑛𝐺𝑐𝑖𝑡 = 𝛼0 + 𝛼1 ∑ 𝑙𝑛𝐺𝑐𝑡−𝑖 𝑚 𝑖=1 + ∑ 𝛼2 𝑙𝑛𝑌𝑡−𝑖 𝑛 𝑖=0 + ∑ 𝛼3𝑙𝑛𝑃𝑡−𝑖 𝑔𝑝 𝑖=0 + ∑ 𝛼4𝑙𝑛𝑃𝑡−𝑖 𝑗𝑞 𝑖=0 + ∑ 𝛼5𝑙𝑛𝑁𝑡−𝑖 𝑟 𝑖=0 + ∑ 𝛼6𝑙𝑛𝑌𝑖𝑡−𝑖 𝑟 𝑖=0 + ∑ 𝛼7𝑙𝑛𝐼𝑄𝑡−𝑖 𝑡 𝑖=0 + 𝜀𝑡 (10) for each sector (i) a maximum of lag 2 was used based on minimizing sbc criterion. the empirical results obtained for each of the sector for the long run are reported in table 5. panel a of the table represents the estimated long run results while panel b fives the results for short run. table 5. long-run estimates and short-run error-correction representation dependent variable independent variables coefficient s.e panel (a): long-run estimates gc(h) y 1.37*** 0.34 pgh -0.24** 0.11 peh 0.75 0.46 psk 0.39** 0.16 iq 0.24*** 0.03 ur 1.77*** 0.57 nh 2.07** 0.79 gc(c) y 0.79*** 0.08 pgc -0.02* 0.01 pec 0.15*** 0.05 rubina ilyas 160 pfo -0.06*** 0.02 iq 0.78* 0.49 yc 0.59** 0.23 nc 0.21 0.19 gc(i) gdp 5.00*** 0.89 pgi -0.63*** 0.08 pei 0.25*** 0.05 pfo -0.44*** 0.08 iq 0.54 0.53 yi 0.15*** 0.04 ni 3.24*** 0.80 panel (b): error-correction representation ∆gc(h) d(y) 1.37*** 0.33 d(pgh) -0.24* 0.09 d(peh) 0.10 0.47 d(psk) 0.33** 0.13 d(iq) 0.26** 0.08 d(ur) 0.24* 0.11 d(nh) 0.24*** 0.03 cointeq(-1) -0.57** 0.21 ∆gc(c) d(y) 0.41** 0.17 d(pgc) -0.01** 0.00 d(pec) 0.08* 0.04 d(pfo) -0.03*** 0.01 d(iq) 0.10*** 0.03 d(yc) 0.13*** 0.03 d(nc) 0.11 0.06 cointeq(-1) -0.52** 0.23 ∆gc(i) d(y) 1.27*** 0.08 d(gdp(-1)) 1.23*** 0.16 d(pgi) -0.02*** 0.01 d(pei) 0.04 0.19 d(pfo) -0.02*** 0.00 d(iq) 0.07 0.06 d(iq(-1)) 0.21** 0.08 d(yi) 0.82*** 0.25 d(ni) 0.14*** 0.00 cointeq(-1) -0.04** 0.01 note: *statistical significance at 10% level. **statistical significance at 5% level ***statistical significance at 1% level in case of the domestic sector the income elasticity has a positive value while, the price elasticity is significantly negative with a small magnitude, showing that gas consumption is relatively price inelastic. the cross price elasticity for electricity is journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 161 insignificant which confirms out earlier result for electricity. however, the cross price effect for kerosene oil has a significant positive value which shows that it is a substitute of gas in the domestic sector. this finding is similar to that of siddiqui (2004), which shows kerosene oil is used as a substitute for gas for cooking and lighting purposes. the number of consumers exert a positive impact on the gas consumption, however the magnitude is very small in the short run. for the industrial sector the income effect has the expected positive sign. the price elasticity has a significant negative value but the small magnitude of price effect shows that the price effect of gas consumption is relatively inelastic in case of industrial sector. the price effect of electricity has a significant positive value, while in case of petroleum products, such as furnace oil, it is significantly negative. these results suggest that natural gas may substitute electricity while furnace oil complements electricity. however, the cross price effect of electricity is insignificant for short run, while cross price effect of furnace oil has a significant negative value. whereas, the gas consumption of the commercial sector is positively related to income and number of users. the price elasticity has the expected sign but the magnitude is very small in case of short run. the institutional quality index is significant for all the sectors in the long run, which coincides with the earlier findings for electricity showing that with the increase in provision of natural gas by the government projects, the gas consumption increases. the value added also shows significant results for all the sectors. these results are in accordance to the previous studies on energy demand (siddiqui (2004) and iqbal (1983)) which conclude high income effect and low price effect for natural gas. 4.3 petroleum products consumption the existence of long run relationship was tested for petroleum products consumption. the maximum lags were set as 2 in the ardl model for all the sectors i.e. households, industrial and transport sector using minimum values of sbc. the computed f stat are given in table 6 which shows that there is cointegration in each model as the computed 𝐹𝐺𝐶 (.) is greater than the critical value. table 5. results of the bounds test for cointegration petroleum products consumption variables included f-statistic kerosene oil 𝑆𝐾𝑐𝐻𝑡 , 𝑃𝐻𝑡 𝑠𝑘 , 𝑃𝐻𝑡 𝑗 , 𝑌𝑡 , 𝑃𝑜𝑝𝑡 , 𝐼𝑄𝑡 4.01** furnace oil 𝐹𝑂𝑐𝐼𝑡 , 𝑃𝐼𝑡 𝑓𝑜 , 𝑃𝐼𝑡 𝑗 , 𝑌𝑡 , 𝑌𝐼𝑡 , 𝐼𝑄𝑡 3.75*** high speed diesel 𝐻𝑆𝐷𝑐𝑇𝑡 , 𝑃𝑇𝑡 ℎ𝑠𝑑 , 𝑃𝑇𝑡 𝑗 , 𝑌𝑡 , 𝑌𝑇𝑡 , 𝑇𝑅𝑡 , 𝐼𝑄𝑡 6.26* low diesel oil 𝐿𝐷𝑂𝑐𝑇𝑡 , 𝑃𝑇𝑡 𝑙𝑑𝑜 , 𝑃𝑇𝑡 𝑗 , 𝑌𝑡 , 𝑌𝑇𝑡 , 𝑇𝑅, 𝐼𝑄𝑡 7.36* motor spirit 𝑀𝑆𝑐𝑇𝑡 , 𝑃𝑇𝑡 𝑚𝑠 , 𝑃𝑇𝑡 𝑗 , 𝑌𝑡 , 𝑌𝑇𝑡 , 𝐶, 𝐼𝑄𝑡 49.28* now equations 2.3 to 2.7 was further estimated by using the following ardl (m,n,p,q,r) specification. 𝑙𝑛𝑃𝑐𝑖𝑡 = 𝛼0 + 𝛼1 ∑ 𝑙𝑛𝑃𝑐𝑡−𝑖 𝑚 𝑖=1 + ∑ 𝛼2 𝑙𝑛𝑌𝑡−𝑖 𝑛 𝑖=0 + ∑ 𝛼3𝑙𝑛𝑃𝑡−𝑖 𝑝𝑝 𝑖=0 + ∑ 𝛼4𝑙𝑛𝑃𝑡−𝑖 𝑗𝑞 𝑖=0 + ∑ 𝛼5𝑙𝑛𝑌𝑖𝑡−𝑖 𝑟 𝑖=0 + ∑ 𝛼6𝑙𝑛𝐼𝑄𝑡−𝑖 𝑡 𝑖=0 + 𝜀𝑡 (11) rubina ilyas 162 for each sector (i) a maximum of lag 2 was used based on minimizing sbc criterion. the empirical results obtained for each of the sector for the long run in table 7. panel “a” of the table represents the estimated long run results while panel “b” fives the results for short run. results show that the income effect is positive for all of the petroleum products and statistically significant indicating that with the increase in income their consumption will also increase. the own price effect is significantly negative in case of all the petroleum products. the number of vehicle has a positive impact on the consumption of petroleum products. in case of cross price elasticities the results match our previous findings of kerosene being a substitute for electricity and natural gas. table.7 long-run estimates and short-run error-correction representation dependent variable independent variables coefficient s.e panel (a): long-run estimates sk y 0.22** 0.08 psk -0.83*** 0.20 pgh 0.95* 0.52 psk 0.29** 0.04 iq 0.27* 0.13 pop 0.81*** 0.16 fo y 1.09** 0.40 pfo -0.49*** 0.14 phsd 1.08747 1.01 iq 0.42** 0.12 yi 0.32** 0.15 ind 6.31** 2.72 hsd y 1.72*** 0.160 phsd -0.40*** 0.094 pms 0.02 0.07 iq 0.73*** 0.11 yt 1.71*** 0.50 ntc 0.59** 0.21 ms y 4.50*** 0.88 pms -0.33*** 0.08 phsd 0.15 0.10 iq 0.42** 0.12 yt 0.53*** 0.16 nc 0.36** 0.16 ldo gdp 1.34** 0.56 pldo -0.27* 0.13 iq 0.02 0.07 yt 0.39* 0.20 ntc 0.23*** 0.07 journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 163 panel (b): error-correction representation sk d(y) 0.10* 0.13 d(y(-1)) 0.17* 0.04 d(psk) -0.02 0.13 d(psk(-1)) 0.28* 0.15 d(pgh) 0.29* 0.15 d(pgh(-1)) 0.33** 0.13 d(peh) 0.42** 0.13 d(iq) 0.27* 0.13 d(pop) 0.64 0.45 cointeq(-1) -0.57*** 0.13 fo d(gdp) 0.74** 0.16 d(pfp) -0.02** 0.01 d(phsd) 0.18 0.12 d(iq) 0.34* 0.19 d(yi) 0.29*** 0.09 d(ind) 1.06** 0.43 cointeq(-1) -0.29** 0.04 hsd d(hsd(-1)) 0.26* 0.13 d(gdp) 1.37*** 0.24 d(phsd) -0.241* 01 d(phsd(-1)) 0.20** 0.06 d(pms) 0.01 0.08 d(iq) 0.34* 0.19 d(ntc) 0.39** 0.16 d(yt) 0.66** 0.21 cointeq(-1) -0.79*** 0.16 ms d(ms(-1)) -0.26979 0.20 d(y) 0.77** 0.22 d(y(-1)) 2.48*** 0.29 d(pms) -0.17** 0.04 d(phsd) 0.03 0.01 d(iq) 0.04 0.10 d(nc) 1.07*** 0.13 d(nc(-1)) 0.97*** 0.09 d(yt) 0.39*** 0.07 d(yt(-1)) 0.54** 0.07 cointeq(-1) -0.66** 0.11 ldo d(ldo(-1)) 1.09** 0.40 d(y) 0.97* 0.23 d(y(-1)) 0.33** 0.13 d(pldo) -0.24*** 0.03 d(iq) 0.25 0.53 d(ntc) 1.29** 0.49 rubina ilyas 164 d(ntc(-1)) 0.73*** 0.11 d(yt) 0.47*** 0.16 d(yt (-1)) 0.38** 0.17 cointeq(-1) -0.53*** 0.15 results show that the income effect is positive for all of the petroleum products and statistically significant indicating that increased income raises the consumption. the own price effect is negative and significant in case of all the petroleum products. the number of vehicle has a positive impact on the consumption of petroleum products. in case of cross price elasticities the results match our previous findings of kerosene being a substitute for electricity and natural gas. the results of institutional quality shows significant results in case of all petroleum products except for ldo in the long run showing a positive impact of institutional quality on petroleum consumption. while the value added by transport sector shows significant impact on petroleum consumption. in case of the remaining fuels the cross price elasticities are not significant. these results are in accordance to the previous finding of siddiqui (2004) on the demand for petroleum products. table 7. results of diagnostic tests model serial correlation functional form normality heteroscedasticity ec(h) 1.04(0.08) 1.17(0.27) 3.36(0.18) 2.38(0.04) ec(c) 1.58(0.45) 1.20(0.29) 0.74(0.69) 3.58(0.81) ec(i) 0.34(0.84) 1.39(0.23) 1.42(0.48) 1.75(0.12) ec(a) 1.46(0.48) 0.03(0.85) 0.25(0.88) 0.39(0.53) gc(h) 2.28(0.31) 0.75(0.29) 1.37(0.02) 6.09(0.41) gc(c) 0.46(0.79) 2.77(0.09) 1.24(0.53) 1.74(0.94) gc(i) 4.41(0.11) 1.53(0.22) 2.07(0.06) 1.26(0.23) sk 2.48(0.28) 0.07(0.94) 1.28(0.52) 0.69(0.75) fo 1.59(0.45) 2.17(0.15) 2.55(0.27) 4.47(0.61) hsd 2.25(0.18) 0.75(0.10) 1.12(0.57) 7.16(0.78) ms 0.03(0.85) 0.22(0.66) 0.25(0.88) 1.93(0.75) ldo 1.12(0.60) 2.15(0.20) 0.02(0.98) 2.60(0.18) table. 7 shows the results of the diagnostic tests, which shows that the models are well fixed. moreover the cumulative tests and cumulative sum of squares test were also used to check for the stability of each model. the results for cusum and susumsq (refer to appendix table a2) show that the parameters remain stable for the period of 1984-2019. 5. conclusion this study analyzes the factors affecting energy demand of the economic sectors (household, commercial, industrial and agricultural) of pakistan. the energy sector of pakistan has undergone several changes over the previous years, due to the increasing energy consumption. the pricing policy has also undergone a number of changes. the main focus of the energy policy has been to increase the supply capacity of energy at the national level. in order to fulfil this objective several rural electrification programs journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 165 and natural gas provision programs have been conducted. this study aims to determine the extent to which the energy consumption is affected by the institutional quality. in this regard it analyzez the demand functions of energy sources (electricity, petroleum products and natural gas) by four main economic groups (i.e., residential, industry, commercial and agricultural sectors) over a period of 35years (1984-2019). in order to examine the relationship between energy consumption and its determinants (energy prices, gdp, number of users etc.) this study has employed bound testing approach (ardl approach) developed by pesaran et al. (2001) as most of the variables analyze are integrated of order 1. the results suggests that the number of users and income have a positive effect on energy (gas, electricity and petroleum products) consumption in all the sectors (household, industrial, commercial and agricultural) while, the own price exerts negative impact on the energy demand with expected signs. the short run elasticities are much lower than the long run estimates which suggests that the demand management policies will have a stronger effect over time. however, the magnitude of the price elasticities is quite small. moreover the income elasticities are found to be insignificant in most of the cases. these results are similar to the findings of alfaris (2002) and narayan and smyth (2005). the results for institutional quality index, measured by using the average of government stability, government accountability, investment profile, law and order and bureaucratic quality, show significant impact on energy consumption. the policy implication that stems out of this study is that targeting energy prices in order to reduce the fiscal burden may be a suitable policy by the government. such a decision should be carefully designed as it will adversely affect the poor adversely. future studies should focus on how the households of different income groups will be affected by increasing energy prices. however, increasing the prices of energy alone may not be affective for energy conservation purposes. the government should create awareness in the public for the use energy saving appliances along with their provision. secondly, the growth rate of population should also be given serious attention. moreover, the results also indicate that the energy sector reforms should also pay attention towards institutional constraints in addition to capacity building efforts. although, village electrification and natural gas provision programs which require large scale infrastructure have been carried out by the state, there is now a need to focus on decentralized development plans. such small scale decentralized generation and distribution efforts will require a lower cost as compared to the expansion of the grid at national level. these results are beneficial not only for policymakers but also private investors as they provide important insights regarding the market for energy consumption. these results are beneficial not only for policymakers but also private investors as they provide important insights regarding the market for energy consumption. rubina ilyas 166 references acemoglu, d., johnson, s., & robinson, j. (2006). understanding prosperity and poverty: geography, institutions and the reversal of fortune. understanding poverty, 19-36. adams, f. g., & shachmurove, y. (2000). the effects of the east asian crisis on the region's energy consumption. university of pennsylvania, center for analytic research in economics and the social sciences. adams, f.g. & shachmurove, y. 2000, the effects of the east asian crisis on the region’s energy consumption, penn caress working papers edn, ucla department of economics ahmad, e., idrees, m., & kazmi, a. a. (1999). the time profile of the cost structure in pakistan's manufacturing sector [with comments]. the pakistan development review, 11011116. alahdad, z. (2012). pakistan's energy sector: from crisis to crisis: breaking the chain. pakistan institute of development economics. al-faris, a. r. f. (2002). the demand for electricity in the gcc countries. energy policy, 30(2), 117-124. aqeel, a. and m. s. butt (2001), the relationship between energy consumption and economic growth. asia pacific development journal 8, 101–110. asif, m. (2009). sustainable energy options for pakistan. renewable and sustainable energy reviews, 13(4), 903-909. beenstock, m. & willcocks, p. 1981, "energy consumption and economic activity in industrialized countries: the dynamic aggregate time series relationship", energy economics, vol. 3, no. 4, pp. 225-32. bentzen, j. & engsted, t. 1993, "shortand long-run elasticities in energy consumption: a cointegration approach", energy economics, vol. 15, no. 1, pp. 916. bentzen, j. & engsted, t. 1993, "shortand long-run elasticities in energy consumption: a cointegration approach", energy economics, vol. 15, no. 1, pp. 916. berndt, e. r., & christensen, l. r. (1973). the translog function and the substitution of equipment, structures, and labor in us manufacturing 1929-68. journal of econometrics, 1(1), 81-113. berndt, e.r., kolstad, c. & lee, j. 1993, "measuring the energy efficiency and productivity impacts of embodied technical change", energy journal, vol. 14, pp. 33-55 berndt, r. & wood, d.o. 1975, "technology, prices, and the derived demand for energy", the review of economics and statistics, vol. 57, no. 3, pp. 259-268. bose, r. k., & shukla, m. (1999). elasticities of electricity demand in india. energy policy, 27(3), 137-146. boix, c., & svolik, m. w. (2013). the foundations of limited authoritarian government: institutions, commitment, and power-sharing in dictatorships. the journal of politics, 75(2), 300-316. brenton, p. (1997). estimates of the demand for energy using cross-country consumption data. applied economics, 29(7), 851-859. burney, n. a., & akhtar, n. (1990). fuel demand elasticities in pakistan: an analysis of households' expenditure on fuels using micro data. the pakistan development review, 155-174. journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 167 caloghirou, y.d., mourelatos, a.g. & thompson, h. 1997, "industrial energy substitution during the 1980s in the greek economy", energy economics, vol. 19, no. 4, pp. 476-491. chan, h.l. & lee, s.k. 1996, "forecasting the demand for energy in china", energy journal, vol. 17, no. 1, pp. 19. chaudhary, m. a., ahmad, e., burki, a. a., khan, m. a., & jehangir, w. a. (1999). industrial sector input demand responsiveness and policy interventions. the pakistan development review, 1083-1100. chishti, s., & mahmood, f. (1991). the energy demand in the industrial sector of pakistan. the pakistan development review, 83-88. cho, w. g., nam, k., & pagan, j. a. (2004). economic growth and interfactor/interfuel substitution in korea. energy economics, 26(1), 31-50. choudhary, i. s., safdar, n., farooq, f., 2012. energy consumption and economic growth, empirical evidence from pakistan. pakistan journal of social science 32(2), 371-382 christopoulos, d. k. (2000). the demand for energy in greek manufacturing. energy economics, 22(5), 569-586. christopoulos, d.k. & tsionas, e.g. 2002, "allocative inefficiency and the capitalenergy controversy", energy economics, vol. 24, no. 4, pp. 305-318. cooper, j. c. (2003). price elasticity of demand for crude oil: estimates for 23 countries. opec review, 27(1), 1-8. deacon, b., ollila, e., koivusalo, m., & stubbs, p. (2003). global social governance. themes and prospects. de vita, g., endresen, k., & hunt, l. c. (2006). an empirical analysis of energy demand in namibia. energy policy, 34(18), 3447-3463. dilaver, z., & hunt, l. c. (2011). modelling and forecasting turkish residential electricity demand. energy policy, 39(6), 3117-3127. engle, r. f., & granger, c. w. (1987). co-integration and error correction: representation, estimation, and testing. econometrica: journal of the econometric society, 251-276. fiebig, d. g., seale, j., & theil, h. (1987). the demand for energy: evidence from a crosscountry demand system. energy economics, 9(3), 149-153. fouquet, r., pearson, p., hawdon, d., robinson, c., & stevens, p. (1997). the future of uk final user energy demand. energy policy, 25(2), 231-240. fuss, m.a. 1977, "the demand for energy in canadian manufacturing: an example of the estimation of production structures with many inputs", journal of econometrics, vol. 5, no. 1, pp. 89-116. galindo, l. m. (2005). short-and long-run demand for energy in mexico: a cointegration approach. energy policy, 33(9), 1179-1185. galindo, l. m. (2005). short-and long-run demand for energy in mexico: a cointegration approach. energy policy, 33(9), 1179-1185. gately, d., & huntington, h. g. (2002). the asymmetric effects of changes in price and income on energy and oil demand. the energy journal, 23(1). hall, v.b. 1986, "major oecd country industrial sector interfuel substitution estimates, 19601979", energy economics, vol. 8, no. 2, pp. 74-89. rubina ilyas 168 holtedahl, p., & joutz, f. l. (2004). residential electricity demand in taiwan. energy economics, 26(2), 201-224. hunt, l. c., & ninomiya, y. (2005). primary energy demand in japan: an empirical analysis of long-term trends and future co2 emissions. energy policy, 33(11), 1409-1424. hye, q.m.a and riaz, s (2008). “causality between energy consumption and economic growth: the case of pakistan”. the lahore journal of economics, vol.13, no. 2, pp. 45-58. iqbal, m. (1983), “residential demand for electricity and natural gas in pakistan”, the pakistan development review, vol. 22, no. 1, pp. 23-36. iqbal, m. (1983). residential demand for electricity and natural gas in pakistan. the pakistan development review, 23-36. iqbal, m. (1986). substitution of labour, capital and energy in the manufacturing sector of pakistan. empirical economics, 11(2), 81-95. jamil, f., & ahmad, e. (2010). the relationship between electricity consumption, electricity prices and gdp in pakistan. energy policy, 38(10), 6016-6025. javid, m. and a. qayyum (2013) electricity consumption-gdp nexus: a structural time series analysis. energy 30, 1–7. javid, m., & qayyum, a. (2014). electricity consumption-gdp nexus in pakistan: a structural time series analysis. energy, 64, 811-817. javid, m., & qayyum, a. (2014). electricity consumption-gdp nexus in pakistan: a structural time series analysis. energy, 64, 811-817. johansen, s. (1988). statistical analysis of cointegration vectors. journal of economic dynamics and control, 12(2-3), 231-254.n jones, c. t. (1996). a pooled dynamic analysis of interfuel substitution in industrial energy demand by the g-7 countries. applied economics, 28(7), 815-821. khan, a. h. (1989). the two-level ces production function for the manufacturing sector of pakistan. the pakistan development review, 1-12. khan, a. h., & rafiq, m. (1993). substitution among labour, capital, imported raw materials, and bank credit in pakistan's manufacturing. the pakistan development review, 32(4), 12591266. khan, m. a., & ahmad, u. (2008). energy demand in pakistan: a disaggregate analysis. the pakistan development review, 437-455. khan, m. a., & qayyum, a. (2009). the demand for electricity in pakistan. opec energy review, 33(1), 70-96. khan, m. u. h., & burki, a. a. (1999). technological change and substitution possibilities in pakistan's large-scale manufacturing: some new evidence. pakistan economic and social review, 123-138. khan, m., & burki, a. a. (2000). estimating allocative efficiency and elasticities of substitution in large-scale manufacturing sector of pakistan. pak j appl econ, 16(1), 2. kouris, g. 1983, "energy consumption and economic activity in industrialized economies-a note", energy economics, vol. 5, no. 3, pp. 207-12. lee. c. (2005) energy consumption and gdp in developing countries: a co-integrated panel analysis. energy economics 27, 415–427. journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 169 lin, b., & ouyang, x. (2014). electricity demand and conservation potential in the chinese nonmetallic mineral products industry. energy policy, 68, 243-253. liu, g. (2004). estimating energy demand elasticities for oecd countries. a dynamic panel data approach. looney, r. (2007). sustaining economic expansion in pakistan in an era of energy shortfalls: growth options to 20351. opec review, 31(2), 125-157. mah, j. s. (2000). an empirical examination of the disaggregated import demand of korea— the case of information technology products. journal of asian economics, 11(2), 237244. mahmood, z. (1992). factor price shocks, factor substitution and their implications for policy. mahmud, f., & chishti, s. (1990). the demand for energy in the large-scale manufacturing sector of pakistan. energy economics, 12(4), 251-254. malik, a. (2008). crude oil price, monetary policy and output: the case of pakistan. the pakistan development review, 425-436. mandal, s. k., & madheswaran, s. (2011). energy use efficiency of indian cement companies: a data envelopment analysis. energy efficiency, 4(1), 57-73. masih, a. m. and r. masih (1996) energy consumption, real income and temporal causality: results from a multi-country study based on cointegration and error correction modeling techniques. energy economics 18, 165–183. medina, j., & vega-cervera, j. a. (2001). energy and the non-energy inputs substitution: evidence for italy, portugal and spain. applied energy, 68(2), 203-214. medlock iii, k. b., & soligo, r. (2001). economic development and end-use energy demand. the energy journal, 22(2). mills, e. (2012). pakistan's energy crisis. us institute of peace. mukherjee, k. (2008). energy use efficiency in the indian manufacturing sector: an interstate analysis. energy policy, 36(2), 662-672. narayan, p. k., & smyth, r. (2005). the residential demand for electricity in australia: an application of the bounds testing approach to cointegration. energy policy, 33(4), 467474. nawaz, s., iqbal, n., & anwar, s. (2013). electricity demand in pakistan: a nonlinear estimation. the pakistan development review, 479-491. ozturk, i., & acaravci, a. (2010). the causal relationship between energy consumption and gdp in albania, bulgaria, hungary and romania: evidence from ardl bound testing approach. applied energy, 87(6), 1938-1943. pesaran, m. h., shin, y., & smith, r. j. (2001). bounds testing approaches to the analysis of level relationships. journal of applied econometrics, 16(3), 289-326. pindyck, r.s. 1979, "interfuel substitution and the industrial demand for energy: an international comparison", review of economics and statistics, vol. 61, no. 2, pp. 169179. pierre, j., & rothstein, b. (2011). reinventing weber: the role of institutions in creating social trust. the ashgate research companion to new public management, 405-416. rubina ilyas 170 popp, d. c. (2001). the effect of new technology on energy consumption. resource and energy economics, 23(3), 215-239. prywes, m. (1986). a nested ces approach to capital-energy substitution. energy economics, 8(1), 22-28. qasim, m., & kotani, k. (2014). an empirical analysis of energy shortage in pakistan. asiapacific development journal, 21(1), 137-166. rapanos, v. t., & polemis, m. l. (2006). the structure of residential energy demand in greece. energy policy, 34(17), 3137-3143. riaz, t., & stern, n. h. (1984). pakistan: energy consumption and economic growth [with comments]. the pakistan development review, 23(2/3), 431-456. rothstein, b. o., & teorell, j. a. (2008). what is quality of government? a theory of impartial government institutions. governance, 21(2), 165-190. roy, j., sanstad, a. h., sathaye, j. a., & khaddaria, r. (2006). substitution and price elasticity estimates using inter-country pooled data in a translog cost model. energy economics, 28(5-6), 706-719. samouilidis, j. e., & mitropoulos, c. s. (1984). energy and economic growth in industrializing countries: the case of greece. energy economics, 6(3), 191-201. schafer, a. 2005, "structural change in energy use", energy policy, vol. 33, no. 4, pp. 429-437. shahbaz, m., zeshan, m., & afza, t. (2012). is energy consumption effective to spur economic growth in pakistan? new evidence from bounds test to level relationships and granger causality tests. economic modelling, 29(6), 2310-2319. siddiqui, r. (1999). demand for energy and the revenue impact of changes in energy prices. working papers & research reports, rr-no. siddiqui, r. (2004) energy and economic growth in pakistan. the pakistan development review 43:2, 175–200. siddiqui, r. and r. haq (1999), “demand for energy and the revenue impact of changes in energy prices”, research report no. 174, pakistan institute of development economics, islamabad. taheri, a. a. (1994). oil shocks and the dynamics of substitution adjustments of industrial fuels in the us. applied economics, 26(8), 751-756. tang, c. f., & tan, e. c. (2013). exploring the nexus of electricity consumption, economic growth, energy prices and technology innovation in malaysia. applied energy, 104, 297-305. the institute of strategic studies (2007), “inaugural address: mr. shaukat aziz”, prime minister islamic republic of pakistan, printed in “energy: sources of regional cooperation and competition”, conference proceedings organized by the institute of strategic studies, islamabad. thompson, p., & taylor, t. g. (1995). the capital-energy substitutability debate: a new look. the review of economics and statistics, 565-569. truong, t. p. (1985). inter‐fuel and inter‐factor substitution in nsw manufacturing industry. economic record, 61(3), 644-653. vlachou, a. s., & samouilidis, e. j. (1986). interfuel substitution: results from several sectors of the greek economy. energy economics, 8(1), 39-45. journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 171 welsch, h., & ochsen, c. (2005). the determinants of aggregate energy use in west germany: factor substitution, technological change, and trade. energy economics, 27(1), 93-111. welsch, h., & ochsen, c. (2005). the determinants of aggregate energy use in west germany: factor substitution, technological change, and trade. energy economics, 27(1), 93-111. wesseh jr, p. k., lin, b., & appiah, m. o. (2013). delving into liberia's energy economy: technical change, inter-factor and inter-fuel substitution. renewable and sustainable energy reviews, 24, 122-130. yoo, s. h., lee, j. s., & kwak, s. j. (2007). estimation of residential electricity demand function in seoul by correction for sample selection bias. energy policy, 35(11), 5702-5707. ziramba, e. (2008). the demand for residential electricity in south africa. energy policy, 36(9), 3460-3466. appendix table a1. results of adf test adf variables level 1st diff gdp -2.71 -4.09*** ec (agri) -2.97** -6.65*** ec (com) -2.97** -9.48*** ec (dom) -1.42 -5.82*** ec (ind) -2.23 -5.64*** gc (dom) -2.21 -9.35*** gc (com) -1.62 -4.41*** gc (ind) -1.30 -3.31** fo -3.13** hobc -1.12 -3.15** hsd -3.01** ldo -2.00 -5.47*** ms -0.18 -7.281*** sk 0.630 -4.80*** en (agr) -0.51 -5.65*** en (com) 0.494 -3.25** en (dom) -1.77 -5.433** en (ind) -0.99 -5.82*** gn (com) -0.48 -3.30** gn (dom) 3.159 -3.76*** gn (ind) -1.93 -3.37** ep (ind) 0.775 -5.34*** ep (com) 1.268 -5.13*** ep (dom) 0.550 -5.47*** ep (ind) 1.243 -4.53*** tc -1.93 -6.18*** gp (com) -0.54 -6.44*** gp (dom) 1.089 -5.24*** gp (ind) -1.48 -5.84*** skp 0.169 -6.43*** hsdp 0.37 -7.13*** fop -3.600** msp -2.34 -4.56*** rubina ilyas 172 table a2 ardl stability tests result model cusum test cusumsq test ec(h) -6 -4 -2 0 2 4 6 2012 2013 2014 2015 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 2012 2013 2014 2015 cusum of squares 5% significance ec(i) -15 -10 -5 0 5 10 15 94 96 98 00 02 04 06 08 10 12 14 cusum 5% significance -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 94 96 98 00 02 04 06 08 10 12 14 cusum of squares 5% significance ec(c) -10.0 -7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 2008 2009 2010 2011 2012 2013 2014 2015 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 2008 2009 2010 2011 2012 2013 2014 2015 cusum of squares 5% significance ec(a) -12 -8 -4 0 4 8 12 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 cusum of squares 5% significance gc(h) -8 -6 -4 -2 0 2 4 6 8 2011 2012 2013 2014 2015 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 2013 2014 2015 cusum of squares 5% significance gc(i) -6 -4 -2 0 2 4 6 2014 2015 cusum 5% significance 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 2014 2015 cusum of squares 5% significance journal of applied economics and business studies, volume. 5, issue 1 (2021) 143-174 https://doi.org/10.34260/jaebs.518 173 gc(c) -6 -4 -2 0 2 4 6 2014 2015 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 cusum of squares 5% significance sk -6 -4 -2 0 2 4 6 2013 2014 2015 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 2013 2014 2015 cusum of squares 5% significance fo -12 -8 -4 0 4 8 12 2000 2002 2004 2006 2008 2010 2012 2014 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 2000 2002 2004 2006 2008 2010 2012 2014 cusum of squares 5% significance hsd -10.0 -7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 1995 1996 1997 1998 1999 2000 2001 2002 cusum 5% significance -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 94 96 98 00 02 04 06 08 10 12 14 cusum of squares 5% significance ms -8 -6 -4 -2 0 2 4 6 8 1997 1998 1999 2000 2001 2002 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 1997 1998 1999 2000 2001 2002 cusum of squares 5% significance rubina ilyas 174 ldo -10.0 -7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 cusum of squares 5% significance journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.511 1 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x afghanistan’s food security: evidence from pakistan and afghanistan wheat price transmission using threshold vector error correction model (tvecm) tahir mahmood1*, afaq ali muluk 2 and seema zubair3 1 department of economics, university of chitral, pakistan olayan school of business, american university of beirut (aub) lebanon 2 department of economics, university of chitral, pakistan 3 the university of agricultural peshawar, pakistan abstract afghanistan's food security mainly depends on pakistan's wheat prospect, circumstances, agriculture policies, and market price dynamics. this study explores the price transmission mechanism of the wheat flour and wheat grain between pakistan and afghanistan using monthly price pairs from january 2003 through october 2017. the paper investigates the existing knowledge of how pakistan’s agricultural policy and wheat market affects the wheat market and food security of afghanistan. the results confirm that the wheat flour price of pakistan is found to be driving the price of wheat flour of afghanistan. this implies that wheat flour price of pakistan evolves independently, and that wheat flour price of afghanistan balances any divergence in the long-run relationship between the two markets prices. the policy implication is to eradicate transaction costs as well as procuring timely wheat grain and flour, in order to maintain price stability between pakistan and afghanistan wheat markets. keywords afghanistan food security; wheat grain & flour; pakistan; market integration; tvecm jel classification q13; f13; q2 1. introduction a timely and adequate supply of food and household food security is a great challenge for many developing countries (chabot & dorosh, 2007). fao defines food security as; physical and economic access to sufficient, safe and nutritious food by the household to fulfill dietary needs and food preference to attain health living standard” (fao, 1996). similarly, jones et al., (2013) classify food security into four dimensions: availability, accessibility, and utilization, and stability of each of these dimensions. * tahirmahmood@uoch.edu.pk tahir mehmood, afaq ali muluk, seema zubair 2 food insecurity has been remained as one of the priorities of national and international policy agendas due to frequent global food crises. consequently, it causes drastic implications for poverty, health, and nutrition (d’souza & jolliffe, 2013). accordingly, the current study explores the wheat shortage of wheat availability and accessibility and its subsequent effect on the food security situation of afghanistan. importantly, the wheat shortage is used as an indicator of food insecurity as wheat is one of the largest source of energy, thus manifesting the importance of the study, from a food security perspective (fao, 2016a). moreover, mostly the impoverished segment of the society’s food security depends on the timely availability of wheat grain and wheat flour. according to the world food program, afghanistan is one of the world’s most vulnerable country in terms of food absorption-suffers from consistent food insecurity and households significant share of income spend on food (d’souza & jolliffe, 2010). moreover, afghanistan is one of the biggest importers of wheat in the world and almost one-third of afghanistan's domestic wheat requirements are fulfilled by imports (see figure 1). on average afghanistan’ wheat consumption is 143 kg/year per head slightly above the regional average and constitutes 66 % of the calories of the domestic diets (fao, 2016a). on average, afghanistan import 2.1 million mt annually mostly from pakistan and kazakhstan (central statistics organization of afghanistan, 2016). similarly, the wheat milling industry is underdeveloped in afghanistan though well developed in pakistan (fao, 2016a). hence, a major portion of wheat imported by afghanistan is wheat flour instead of wheat grain. thus, afghanistan wheat flour imports from pakistan continue in part because of the lack of milling capacity to produce high-quality of wheat flour and in part due to high incomes, hence, mostly urban households demand better quality imported wheat flour (chabot & dorosh, 2007). thus, afghanistan's food security mainly depends on pakistan's wheat prospect, circumstances, agriculture policies, and market prices. figure 1 by countries: afghanistan wheat grain and flour import during 2003-2016 source: central statistics organization of afghanistan (2016), author’s calculation 0% 20% 40% 60% 80% 100% usa turkey china russia iran tajikistan turkmenistan journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.511 3 pakistan is a major contributor of wheat production in the region, however with a population of more than 200 million has also the highest domestic requirements. pakistan often imposes inter-provincial, inter-district and inter-region restrictions on wheat transportation to procure a targeted quantity of wheat for maintaining food price stability. moreover, the pakistan government is constantly regularizing her wheat market by purchasing from farmers, then releasing the wheat to the mills industry at a fixed price. pakistan supply wheat flour to afghanistan, regularly (usda, 2016). moreover, the ban of wheat export of pakistan had drastic consequences on the food security of afghanistan manifested from the escalating wheat and flour prices during the 2009-10 periods (see figure 2). it negatively affected all the stakeholders in the value chain. this could also cause inconsistent supply to afghanistan. similarly, from pakistan, a consistent and abundant wheat supply will keep the wheat price stable as well as keeping in at an affordable rate in afghanistan. moreover, afghan transit trade agreement (atta) is not fully implemented. this could affect trade between the two trading economies and encourage informal and illegal trading. figure 2 pakistan and afghanistan wheat grain &flour prices series trend (2003-2017) note: author’s calculation keeping in view the food security scenario in afghanistan, this study investigate the impact of the regional producer of wheat such as pakistan that could have the greatest expected impact on the wheat markets of afghanistan. the food and agricultural organization (fao) revealed that more than 53.2 percent of households faced food insecurity in 2019, and almost 11million habitants need food aid (samim & zhiquan, 2020) moreover, food insecurity is a widespread and serious issue, due to insufficient, 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00 pak-flour afghan-flour pak-wheat afghan-wheat tahir mehmood, afaq ali muluk, seema zubair 4 limited access and poor utilization of food, in afghanistan (d’souza & jolliffe, 2013). thus, it is important to review the existing knowledge of how pakistan's wheat price fluctuation affects both wheat markets and subsequent food security situation and outlook in afghanistan. in addition, the study fills the existing research gap and will help in improving the food security analysis and the estimation of food demand in afghanistan. moreover, the contribution of the study is the spatial product integration of horizontal price transmission between pakistan and afghanistan wheat grain and wheat flour markets using monthly retail price data from january 2003 to october 2017. the preceding paragraph will discuss the existing literature on spatial price transmission mechanisms. market integration of agriculture goods has acquired colossal attention in empirical field of studies in development economics. similarly, emerging economies have embarked on opening their domestic markets incorporating structural reforms and trade liberalization policies. moreover, price signal transmits smoothly among in an integrated markets. smooth trading of goods among markets helps in transferring of supply of goods from excessive good market to a market of shortage one (barrett, 2008). subsequently, price stability and trade openness results in allocative efficiency which could improve living standard of the people. (srinivasan & jha, 2001). basically, (engle & granger, 1987) advanced the idea of cointegration, highlighting a linear combination of variables. a linear combination of variables exhibits a stable long-run behavior depicted by the vector error correction model. however, economic theory arguments favor that deviation from the long-run equilibrium is not instant, instead, the long-run adjustment happens after a certain threshold level is reached, mainly, because of transaction cost as well as price stickiness (balke & fomby, 1997). in that situation, threshold cointegration comes into play. specifically, the data pairs used in the current study show the properties of nonsymmetry patterns particularly during the 2008/09 period due to the temporary export ban of wheat from pakistan to afghanistan. economic theory put forward another argument in favor of threshold cointegration, which states that the divergence from the equilibrium adjusts asymmetrically instead of symmetric pattern due to market power, menu cost, or some other political reasons (levy et al., 1997). numerous studies are available based on spatial wheat market asymmetric price transmission mechanism (ghoshray, 2007) (ahmed & singla, 2017); (z. bakucs et al., 2015); (l. z. bakucs et al., 2012); (pall et al., 2013) (dawson et al., 2006); (esposti & listorti, 2018); (brosig et al., 2011). in the same way, the law of one price has been investigated for the spatial turkish wheat market by eryigit & karaman, (2011). similarly, goychuk & meyers, (2014) investigated the russian and ukrainian wheat markets and comparing with the us, the eu and canada wheat markets. indian wheat markets also suggest asymmetries in price adjustment (ghoshray, 2007); (ghoshray and ghosh, 2011). moreover, for pakistan and afghanistan, only conventional cointegration methods have been used (chabot & dorosh, 2007) (persaud, 2012) (halimi et al., 2015). journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.511 5 to the best of my knowledge, there is no such studies investigating wheat market integration between pakistan and afghanistan using advanced dynamic models. the rest the paper is organized as follows. section 2 presents the data and empirical method, followed by empirical results and discussions in section 3. finally, section 4 concludes and suggests some policy recommendations. 2. theoretical framework the theoretical framework of the study is based on the law of one price (lop), which considers a frictionless market, meaning that the price of identical commodities will have the same price globally, irrespective of their location (fackler & goodwin, 2001). co-integration between two commodities market implies that price in the two markets may behave differently in the short-run period, however, that converging into a stable equilibrium in the long-run period (rapsomanikis et al., 2006). transaction cost and border restriction are the major cause of price inequality, nevertheless, domestic policies affecting price formation do also affect both vertical and spatial price relations (baffes & ajwad, 2001; cramon-taubadel, 1998). consider prices of a good in two spatially separated markets 𝑷𝟏𝒕 and 𝑷𝟐𝒕 , then the law of one price postulates that allowing for border and domestic policies transaction cost c, suppose transporting the wheat from pakistan wheat market to afghanistan wheat market, the established relationship between the prices is given: 𝑷𝟏𝒕 = 𝑷𝟐𝒕 + 𝒄 (1) a complete price transmission occurs in two markets where changes in one market price such as wheat price in pakistan are transmitted instantaneously to the other price such as afghanistan. thus, indicating spatially separated integrated markets. moreover, this implies that if price changes are not converge quickly, depicting imperfect price transmission in the short run, however, perfect/complete price convergence in the long run. thus, changes in the price at one market that is pakistan, may require considerable time period to be transmitted to other market that is afghanistan due to many reasons, particularly domestic policies, marketing structure, the contractual arrangements, lags caused in transportation and unwarranted price control. similarly, asymmetric response of one market price to another market price implies nonlinear adjustment. thus, asymmetric price responses uses the asymmetric error correction model proposed by granger & lee, (1989) or threshold cointegration models developed by enders & granger, (1998). importantly, the interpretation of the short run adjustment parameters captures the speed of price transmission, whereas, the long run multiplier is interpreted as a measure of the degree of price transmission from one market to another market (prakash, 1999). therefore, the current study will investigate wheat price transmission between pakistan and afghanistan spatially separated wheat market through a linear as well as threshold cointegration models. tahir mehmood, afaq ali muluk, seema zubair 6 3. data description and empirical method 3.1 data and variables the data use in the analysis consists of monthly retail prices of wheat grain and wheat flour from january 2003 through october 2017. the data for pakistan wheat price series are taken from various issues of monthly review of price indices, government of pakistan, and federal bureau of statistics. while the data source of afghanistan comes from global information early warning system (giews) of the food and agriculture organization (fao) of the united nations. finally, the data series of wheat grain and wheat flour for both countries are adjusted with the exchange rate. the subsequent analysis is carried out on the logarithm of prices. 3.2 stationary test and lag selection to start with potentially non-stationary time series data the price series will be initially tested for their order of integration using the (dickey & fuller, 1979)augmented dickey fuller (adf) unit root test (1979, 1981), pp by (phillips & perron, 1988) and kpss by (kwiatkowski et al., 1992). the first two unit root test that is the adf test, and pp test null hypotheses indicate that the price series carry unit root process. in contrast, the kpss null hypothesis test indicates that the price series is level or trend stationary. accordingly if the price series are non-stationary at level, then there is danger of spurious regression and we need to test for the presence of a true cointegration relationship. moreover, the optimal lag length is selected using the aic, bic, epe and the hq. the optimal lag length will be selected based on the aforementioned criterion, having minimum of the test values. 3.3 linear cointegration analysis the wheat price long run cointegration is checked by rank (johansen, 1995) cointegration procedure. the johansen cointegration technique is the most appropriate procedure to employ, for the data series which is integrated of the same order, the. consequently, the (johansen, 1988) johansson (1988, 1995), vector error correction model is specified as: ∆𝑌𝑡 = ∑ 𝜃𝑖 𝑘−1 𝑖=1 𝑌𝑡−𝑖 + 𝜗 𝑌𝑡−1 + 𝜇 + 𝜀𝑡 (2) where ∆ represents difference operator, 𝜇 stand for drift parameter, 𝜗 is the p*p matrix 𝜗 =𝛼𝛽ˊ, where adjustment coefficient is represented by 𝛼 and 𝛽 contains cointegrating vector. the tests are based on calculated by the likelihood test calculate the trace and maximum eigen-values of the johansen test. however, the drawback of this method is that it assumes that the cointegrating vector remains constant during period of analysis. the long run relationship changes between the variables owing to technological progress, economic crisis, political upheaval, institutional development, policy, or regime change. for this reason, we use the (gregory & hansen, 1996) test as a robust check of cointegration incorporating structural break in the data analysis. journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.511 7 3.4 gregory and hansen test (gregory & hansen, 1996) cointegration test allows for possible structural breaks. the simple greogory and hansen model with regime shift is written as: 𝑌𝑡 = 1 + 2∅𝑡𝑘 + 𝛽1𝑋𝑡 + 𝛽2𝑋𝑡 ∅𝑡𝑘 + 𝜀𝑡 (3) where k is the regime shift or breaks date, and ∅ is a dummy variable such as ∅𝑡𝑘 = { 0 1 𝑖𝑓 𝑖𝑓 𝑡 ≤ 𝑘 𝑡 > 𝑘 this test has three statistic such as, 𝐴𝐷𝐹∗, 𝑍𝛼 ∗ and 𝑍𝑡 ∗. the null hypothesis of no cointegration with structural break tests against the alternative hypothesis of cointegrtaion. the single break date determines endogenously. the null hypothesis rejects if the statistic is smaller than the corresponding critical values. 3.5 threshold cointegration finally, (hansen & seo, 2002) two-regime threshold cointegration model is used to investigate the long run relationship between the price pairs. the two-regime threshold model for pakistan and afghanistan price pairs with cointegrating vector and threshold parameter is given by: regime 1 [ ∆𝑝𝑡 𝑃𝑎𝑘 ∆𝑝𝑡 𝐴𝑓𝑔] = [ 𝛼1 𝛼2 ] + ∑𝑖=1 𝑘 [ 𝛽𝑖 𝑝𝑎𝑘,𝑝𝑎𝑘 𝛽𝑖 𝑎𝑓𝑔,𝑝𝑎𝑘 𝛽𝑖 𝑝𝑎𝑘,𝑎𝑓𝑔 𝛽𝑖 𝑎𝑓𝑔,𝑎𝑓𝑔 ] [ ∆𝑝𝑡−𝑖 𝑃𝑎𝑘 ∆𝑝𝑡−𝑖 𝐴𝑓𝑔] + [ 𝜃1 𝑃𝑎𝑘 𝜃1 𝐴𝑓𝑔] [𝐸𝐶𝑇𝑡−1] + [ 𝜀𝑡 𝜖𝑡 ] if 𝐸𝐶𝑇𝑡−1 ≤ 𝛾 regime 2 [ ∆𝑝𝑡 𝑃𝑎𝑘 ∆𝑝𝑡 𝐴𝑓𝑔] = [ 𝛼1 𝛼2 ] + ∑𝑖=1 𝑘 [ 𝛽𝑖 𝑝𝑎𝑘,𝑝𝑎𝑘 𝛽𝑖 𝑎𝑓𝑔,𝑝𝑎𝑘 𝛽𝑖 𝑝𝑎𝑘,𝑎𝑓𝑔 𝛽𝑖 𝑎𝑓𝑔,𝑎𝑓𝑔 ] [ ∆𝑝𝑡−𝑖 𝑃𝑎𝑘 ∆𝑝𝑡−𝑖 𝐴𝑓𝑔] + [ 𝜃1 𝑃𝑎𝑘 𝜃1 𝐴𝑓𝑔] [𝐸𝐶𝑇𝑡−1] + [ 𝜀𝑡 𝜖𝑡 ] if 𝐸𝐶𝑇𝑡−1 > 𝛾 (4) where 𝛾 is the threshold parameter and “pak” and “afg” represent pakistan and afghanistan wheat grain and wheat flour price pairs, respectively. unlike, other methodologies which assume parameters are known before, the method of (hansen & seo, 2002) hansen and seo, (2002) assumes both 𝛽 and 𝛾 are unknown and estimated from the data. similarly, the error correction coefficient shows the speed of adjustment of the market wheat prices back towards equilibrium. furthermore, hansen & seo, (2002), recommended a heteroskedastic consistent lm test statistics for the null hypothesis of linear cointegration against the alternative of threshold cointegration. tahir mehmood, afaq ali muluk, seema zubair 8 suplm = su𝑝𝛾𝑙≤𝛾≤𝛾𝑢lm (𝛽, 𝛾) (5) where 𝛽, is the 𝛽 estimated. finally, (hansen & seo, (2002) explain the sampling distribution either through fixed regressor bootstrap of hansen, (1996) or a parametric residual bootstrap algorithm. 4. empirical results and discussions 4.1 stationary check and lag length selection criteria a suitable lag length of the wheat grain and wheat flour prices pairs for both pakistan and afghanistan markets are determined by the fpe, aic, sbic, hqic, and lr. the optimal lag length chosen by these tests are four (l = 4) for both pakistan and afghanistan wheat grain markets. while, maximum lag length (l = 2) for pakistan wheat flour markets and (l = 4) for afghanistan wheat flour market. in selecting the lag length of the var of bivariate threshold cointegration we use the above mention selection criterion, all of them leading to maximum lag length (l = 4). all the tests used for stationary check, show that the price pairs of wheat (both grain and flour) for both markets is non-stationary at level (trend & constant) except for afghanistan wheat grain and wheat flour prices based on adf test. whereas, the price pairs become stationary in their first difference and significant at 1 % level. the results of both non-stationary and stationarity price pairs are shown in table 1. table 1 unit root test results level form var. adf pp kpss constant trend constant trend level trend pak-f -1.78 -0.77 -1.50 -0.50 15.9-1.21 2.84-0.23 k = 13 afg-f -2.68* -3.50** -2.18 -2.34 10.9-0.96 1.74-0.18 k = 9 pak-w -1.69 -0.83 -1.32 -2.71 15.5-1.2 2.54-0.23 k = 13 afg-w -2.86** -3.89** -2.03 -2.30 11.9-1.01 1.58-0.16 k = 7 first difference form pak-f -6.39*** -6.58*** -10.37*** -10.48*** 0.51-0.26*** 0.19-0.11*** k = 13 afg-f -4.40*** -4.48*** -9.94*** -9.96*** 0.22-0.12*** 0.05-0.03*** k = 13 pak-w -6.53*** -6.81*** -30.13*** -30.61*** 0.03-0.18*** 0.02-0.11*** k = 13 afg-w -3.99*** -4.06*** -10.26*** -10.22*** 0.20-0.09*** 0.06-0.03*** k = 13 where ∝ = 0.01; ∝ = 0.05; ∝ = 0.1, indicated by ∗∗∗,∗∗,∗ 𝑟𝑒𝑠𝑝𝑒𝑐𝑡𝑖𝑣𝑒𝑙𝑦. till lag 13, the price series are trend stationary in their first differences, because the kpss test statistic values are less than the critical values. pak-f = pakistan flour price, pak-w = pakistan wheat price. afg-f = afghanistan flour price, afg-w = afghanistan wheat price. 4.2 johansen test for cointegration first checking the stationarity and optimal lag length selection, johansen cointegration technique is employed to investigate the long run relationship between journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.511 9 the wheat grain and flour markets of pakistan and afghanistan. the findings of the johansen test are illustrated in table 2. the null of no cointegration is rejected as the trace statistic at r = 0 exceeds it critical value for all price pairs. in contrast, the null hypothesis is not rejected, as one or less cointegrating equations exist, because trace statistic at r = 1 is less than its critical value for all price pairs. the result of price series reveal that both markets combinations are cointegrated of order 1 indicating a long run relationship between the two market price pairs. table 2 rank johansen cointergation results series rank = r trace statistic critical value (5%) results pakf-afgf 0 1 16.93 3.045* 15.41 3.76 r = 1 pakw-afgw 0 1 16.75 2.846* 15.41 3.76 r = 1 note: significance at ∝ = 0.05 is indicated by *. pakf-afgf stands for pakistan and afghanistan flour price series, while pakw-afgw stands for pakistan and afghanistan grain price series. 4.3 gregory-hansen test for cointegration with regime shifts similarly, if there is a break in the price series (see figure 2) indicating that the test of linear cointegration is biased towards not rejecting the null hypothesis of no cointegration.1 to handle this issue a test is proposed by (gregory & hansen, 1996) where the data of structural change is estimated endogenously. the test reject the null hypothesis of no structural break and the break dates turns out december 2008. the time period of the break coincides with the ban of wheat exports by pakistan to afghanistan during 2008 and 2009. this ban of wheat exports has been a significant change in the pakistan domestic agriculture policy. the policy change/regime shift threatens food security of afghanistan manifested from the prices increase of afghanistan wheat grain and wheat flour markets during the period 2008 and 2009. table 3 shows the results of the regime shifts. table 3 gregory and hansen structural break test pakistan-afghanistan wheat flour cointegration models break point test statistic critical value reject ho if no ci 1 % 5% 10 % adf* 2008m12 -5.39** -5.47 -4.95 -4.68 yes zt 2009m02 -5.30** -5.47 -4.95 -4.68 yes za 2009m02 42.82* -57.17 -47.04 -41.85 yes pakistan-afghanistan wheat grain adf* 2008m12 -5.08** -5.47 -4.95 -4.68 yes 1 we have already rejected the null of no cointegration above. so even if this bias is present, it cannot be strong enough to change the results of the cointegration tests. however, we still include gregory and hansen test because it helps us in finding the exact regime shift time period, because we are also interested to find out the regime shift due to the export ban of pakistan to afhanistan. tahir mehmood, afaq ali muluk, seema zubair 10 zt 2009m01 -9.05*** -5.47 -4.95 -4.68 yes za 2009m01 -114.42 *** -57.17 -47.04 -41.85 yes note: the critical values are from gregory – hansen (1996a) 4.4 test of null hypothesis of linear cointegration versus threshold finally, hansen & seo, (2002) test confirms the threshold cointegration for both markets: pakistan-afghanistan wheat flour and pakistan-afghanistan wheat grain market prices pairs. test statistic and fixed regressor bootstrap both reject the null of linear cointegration at 5% and 10% significance level. therefore, the threshold model is preferred over other conventional models. table 4 test of linear versus threshold cointegration a. test of linear versus threshold cointegration by hansen and seo (2002)a test statistic: 35.853** (threshold value maximize: -2.098) p-value: 0.035 (fixed regressor bootstrap) critical values: 0.90% 0.95% 0.99% 33.018 35.004 39.594 number of bootstrap replications: 1000 cointegrating value (estimated under restricted linear model): -1.876 b. test of linear versus threshold cointegration of hansen and seo (2002)b test statistic: 24.987** (maximized for threshold value: -1.122) p-value: 0.031 (fixed regressor bootstrap) critical values: 0.90% 0.95% 0.99% 22.487 24.130 26.983 number of bootstrap replications: 1000 cointegrating value (restricted linear model): -1.619 a = pakistan afghanistan wheat flour cointegration b = pakistan afghanistan wheat grain cointegration 4.5 long run and short run relationship between the wheat (flour & grain) market price pairs wheat flour market integration results table 5 presents the long run and short run relationship between the two price pairs utilizing threshold vector error correction model (tvecm). the estimated cointegration relationship is (1, -1.17) and threshold parameter is 𝛾 = -0.27, dividing the data into two regimes. the extreme regime consists of only 15% of the observations (regime 1), while the general regime and consists of 85% of the observations (regime 2). journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.511 11 for the wheat flour markets: both pakistan and afghanistan wheat flour markets depict significant error correction adjustment in regime 1 at 1% and 10% significance level, respectively. specifically, for pakistan the coefficient 𝐸𝐶𝑇𝑡−1 (0.19) indicating the speed of adjustment back to the long run equilibrium, that is nearly 19% of disequilibrium of the previous month shock adjust back to the long run equilibrium in the current month. whereas, adjustment coefficients are not significant for both flour markets in regime 2. thus, confirming the threshold model, as transaction cost or temporarily ban on exports slow down the long run price adjustment between the two flour markets. these findings are in line with the previous studies such as, griffith & piggott, (1994) revealed asymmetries for australian animal markets. the impact of pakistan lagged period prices on current afghanistan flour price are significantly higher than that of the impact of lagged period price of afghanistan prices on current pakistan price in the short run though limited to regime 1. specifically, the lagged one and lagged two period price impacts of pakistan on the current flour price of afghanistan is significant. this mean that pakistan wheat flour price is found to be the driving the price of the afghanistan wheat flour price. this implies that pakistan wheat flour price follows independent pattern, and that the afghanistan wheat flour price adjusts to correct any disequilibrium in the long run between the two markets prices. in addition, the impact of own lagged period price of afghanistan is significant. moreover, in regime 2 only afghanistan current price is affected by its lagged one period price. this shows that afghanistan flour prices are very sensitive to any previous shocks. table 5 wheat flour market integration results dependent variable ∆𝑝𝑡 𝑃𝑎𝑘 ∆𝑝𝑡 𝐴𝑓𝑔 regime-1 regime-2 regime-1 regime-2 𝐸𝐶𝑇𝑡−1 −0.193(0.000) ∗∗ ∗ −0.017(0.208) 0.155(0.020) ∗ 0.003(0.839) intercept −0.048(0.014) ∗ 0.007(0.068) 0.047(0.568) 0.003(0.463) ∆𝑝𝑡−1 𝑃𝑎𝑘 0.220 (0.130) 0.184(0.055) -0.053 (0.626) 0.069(0.458) ∆𝑝𝑡−2 𝑃𝑎𝑘 −0.253 (0.084) 0.007(0.937) -0.310 (0.008)** −0.043(0.668) ∆𝑝𝑡−3 𝑃𝑎𝑘 −0.146 (0.369) 0.041(0.671) 0.249(0.005) ∗∗ 0.103(0.299) ∆𝑝𝑡−4 𝑃𝑎𝑘 −0.091(0.606) −0.079(0.404) −0.192(0.065) −0.074(0.411) ∆𝑝𝑡−1 𝐴𝑓𝑔 0.461(0.009)** −0.003(0.984) 0.407 (0.002) 0.400(0.000) ∗ ∗∗ ∆𝑝𝑡−2 𝐴𝑓𝑔 -0.770 (000)*** 0.022(0.851) 0.151 (0.286) 0.049(0.686) ∆𝑝𝑡−3 𝐴𝑓𝑔 0.108(0.956) −0.030(0.801) 0.590 (0.000) ∗ ∗ 0.053(0.661) ∆𝑝𝑡−4 𝐴𝑓𝑔 0.403 (0.059) −0.074(0.519) 0.096 (0.442) −0.072(0.510) note: parenthesis incorporates eicker-white se. *** p<0.01; ** p<0.05; * p<0.1 regime1: 𝐸𝐶𝑇𝑡−1 ≤ -0.27; regime2: 𝐸𝐶𝑇𝑡−1 > -0.27 tahir mehmood, afaq ali muluk, seema zubair 12 wheat grain market integration results similarly, table 6 presents the results of wheat grain prices of both countries, estimated by maximum likelihood at the var lag length four (l = 4). the estimated cointegration relationship is (1, -1.21) and the threshold parameter is 𝛾 = -0.31, dividing the data into two regime. regime 1 consists of only 15% of the observations, while regime 2 consists of 85% of the observations. the adjustment coefficient for wheat grain market for pakistan is insignificant in both regimes however, significant for afghanistan only in regime 1. the reason could be due to afghanistan wheat grain market integration with other regional markets particularly kazakhstan and uzbekistan. therefore, any deviation from short run shock is quickly adjusted in the long run. for regime 1, the own lagged period prices of both countries have a significant impact on the current prices of respective country. moreover, the lagged four period price influence of pakistan on afghanistan current price is significant in regime 1. this shows that any change in current period wheat price of pakistan affect afghanistan wheat grain price after four months. for regime 2, up to lagged 3 periods, own prices impact current price of both countries. interestingly, lagged one period price of afghanistan affects pakistan current wheat grain price. table 6 wheat grain market integration results dep. va. ∆𝑝𝑡 𝑃𝑎𝑘 ∆𝑝𝑡 𝐴𝑓𝑔 regime-1 regime-2 regime-1 regime-2 𝐸𝐶𝑇𝑡−1 −0.195(0.102) 0.008(0.803) 0.149(0.009) ∗∗ 0.016(0.315) intercept 0.045(0.492) 0.0001(0.991) 0.082(0.011) ∗ −0.0004(0.941) ∆𝑝𝑡−1 𝑃𝑎𝑘 −0.492(0.015) ∗ −0.806(000) ∗∗∗ −0.211(0.375) 0.845(0.0002) ∗∗∗ ∆𝑝𝑡−2 𝑃𝑎𝑘 −0.525(0.004) ∗∗ −0.397(0.0006) ∗∗∗ −0.3263(0.215) −0.0224(0.9233) ∆𝑝𝑡−3 𝑃𝑎𝑘 −0.224(0.332) 0.256(0.021) ∗ −0.042(0.847) −0.286(0.221) ∆𝑝𝑡−4 𝑃𝑎𝑘 −0.241(0.182) 0.126(0.130) −0.245(0.352) −0.246(0.276) ∆𝑝𝑡−1 𝐴𝑓𝑔 0.067(0.486) . 045(0.277) 0.400(0.0006) ∗∗∗ 0.236(0.028) ∗ ∆𝑝𝑡−2 𝐴𝑓𝑔 −0.0540(0.542) 0.0184(0.7351) −0.2390(0.059) 0.1030(0.3573) ∆𝑝𝑡−3 𝐴𝑓𝑔 −0.181(0.103) −0.009(0.855) 0.680(0.000) ∗∗∗ 0.229(0.042) ∗ ∆𝑝𝑡−4 𝐴𝑓𝑔 −0.295(0.0008) ∗∗∗ −0.023(0.553) 0.230(0.070) 0.016(0.879) note: parenthesis incorporates eicker-white se: *** p<0.0; ** p<0.05; * p<0.1 regime1: 𝐸𝐶𝑇𝑡−1 ≤ -0.31; regime2: 𝐸𝐶𝑇𝑡−1 > -0.31 5. conclusion and policy implications afghanistan food security mainly depends on pakistan wheat situation, outlook, policies and prices. the high prevalence of consumption makes wheat grain and flour markets the most significant agricultural market to evaluate in relation to food security. this study investigates the price transmission mechanism of wheat flour and grain markets between pakistan and afghanistan. the finding confirms that there exist threshold cointegration between the two markets. the striking finding which we infer from this study is that wheat flour price of pakistan is found to be the driving the price of wheat flour price of afghanistan. this implies that pakistan wheat flour price follows journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.511 13 independent pattern, and that the afghanistan wheat flour price adjusts to correct any disequilibrium in the long run between the two markets prices. in addition, the recent spike in the wheat price of pakistan is mainly due to production fluctuation rather than market manipulation (inadequate irrigation water supply, post-harvest losses, high cost agriculture inputs, and high cost of doing business hurting competitiveness of pakistan). as the food insecurity is linked to poverty, therefore, dietary quality significantly declines in times of high inflation. thus, pakistan wheat policies have repercussions for both national and households’ food security situation in afghanistan. from policy perspective, the more integrated the markets, the better it convey price signals to the policy and decision makers, marketing of the commodities and efficient allocation of commodities within and among the regions. afghanistan will remain subject to wheat supply disruption and price spike as long as its domestic production is underdeveloped. therefore, reducing transaction cost and implementing an efficient wheat procurement policy will help in the long run equilibrium convergence and price stability. importantly, it is suggested to revisit the procurement pricing policy, facilitating trade openness and smooth running of the wheat transport between pakistan and afghanistan. references ahmed, m., & singla, n. (2017). market integration and price transmission in major onion markets of india. economic affairs, 62(3), 405–417. baffes, j., & ajwad, m. i. (2001). identifying price linkages: a review of the literature and an application to the world market of cotton. applied economics, 33(15), 1927–1941. bakucs, l. z., brümmer, b., von cramon-taubadel, s., & fertő, i. (2012). wheat market integration between hungary and germany. applied economics letters, 19(8), 785–788. bakucs, z., bojnec, š., & fertő, i. (2015). spatial product market integration between two small, open neighbouring economies. agribusiness, 31(2), 171–187. balke, n. s., & fomby, t. b. (1997). threshold cointegration. international economic review, 627–645. barrett, c. b. (2008). smallholder market participation: concepts and evidence from eastern and southern africa. food policy, 33(4), 299–317. brosig, s., glauben, t., götz, l., weitzel, e., & bayaner, a. (2011). the turkish wheat market: spatial price transmission and the impact of transaction costs. agribusiness, 27(2), 147–161. chabot, p., & dorosh, p. a. (2007). wheat markets, food aid and food security in afghanistan. food policy, 32(3), 334–353. tahir mehmood, afaq ali muluk, seema zubair 14 cramon-taubadel, s. von. (1998). estimating asymmetric price transmission with the error correction representation: an application to the german pork market. european review of agricultural economics, 25(1), 1–18. d’souza, a., & jolliffe, d. (2010). rising food prices and coping strategies: household-level evidence from afghanistan. the world bank. d’souza, a., & jolliffe, d. (2013). conflict, food price shocks, and food insecurity: the experience of afghan households. food policy, 42, 32–47. dawson, p. j., sanjuán, a. i., & white, b. (2006). structural breaks and the relationship between barley and wheat futures prices on the london international financial futures exchange. review of agricultural economics, 28(4), 585–594. dickey, d. a., & fuller, w. a. (1979). distribution of the estimators for autoregressive time series with a unit root. journal of the american statistical association, 74(366a), 427–431. enders, w., & granger, c. w. j. (1998). unit-root tests and asymmetric adjustment with an example using the term structure of interest rates. journal of business & economic statistics, 16(3), 304–311. engle, r. f., & granger, c. w. j. (1987). co-integration and error correction: representation, estimation, and testing. econometrica: journal of the econometric society, 251–276. eryigit, k. y., & karaman, s. (2011). testing for spatial market integration and law of one price in turkish wheat markets. quality & quantity, 45(6), 1519–1530. esposti, r., & listorti, g. (2018). price transmission in the swiss wheat market: does sophisticated border protection make the difference? the international trade journal, 32(2), 209–238. fackler, p. l., & goodwin, b. k. (2001). spatial price analysis. handbook of agricultural economics, 1, 971–1024. ghoshray, a. (2007). an examination of the relationship between us and canadian durum wheat prices. canadian journal of agricultural economics/revue canadienne d’agroeconomie, 55(1), 49–62. goychuk, k., & meyers, w. h. (2014). black sea and world wheat market price integration analysis. canadian journal of agricultural economics/revue canadienne d’agroeconomie, 62(2), 245–261. granger, c. w. j., & lee, t.-h. (1989). investigation of production, sales and inventory relationships using multicointegration and non‐symmetric error correction models. journal of applied econometrics, 4(s1), s145–s159. journal of applied economics and business studies, volume. 5, issue 2 (2021) 1-16 https://doi.org/10.34260/jaebs.511 15 gregory, a. w., & hansen, b. e. (1996). residual-based tests for cointegration in models with regime shifts. journal of econometrics, 70(1), 99–126. griffith, g. r., & piggott, n. e. (1994). asymmetry in beef, lamb and pork farm-retail price transmission in australia. agricultural economics, 10(3), 307–316. halimi, g. h., abbott, p. c., & mcnamara, k. t. (2015). price transmission in afghanistan’s wheat markets. ipia report, 1062015. hansen, b. e., & seo, b. (2002). testing for two-regime threshold cointegration in vector error-correction models. journal of econometrics, 110(2), 293–318. johansen, s. (1988). statistical analysis of cointegration vectors. journal of economic dynamics and control, 12(2–3), 231–254. johansen, s. (1995). identifying restrictions of linear equations with applications to simultaneous equations and cointegration. journal of econometrics, 69(1), 111– 132. jones, a. d., ngure, f. m., pelto, g., & young, s. l. (2013). what are we assessing when we measure food security? a compendium and review of current metrics. advances in nutrition, 4(5), 481–505. kwiatkowski, d., phillips, p. c. b., schmidt, p., & shin, y. (1992). testing the null hypothesis of stationarity against the alternative of a unit root: how sure are we that economic time series have a unit root? journal of econometrics, 54(1–3), 159– 178. levy, d., bergen, m., dutta, s., & venable, r. (1997). the magnitude of menu costs: direct evidence from large us supermarket chains. the quarterly journal of economics, 112(3), 791–824. pall, z., perekhozhuk, o., teuber, r., & glauben, t. (2013). are russian wheat exporters able to price discriminate? empirical evidence from the last decade. journal of agricultural economics, 64(1), 177–196. persaud, s. (2012). long-term growth prospects for wheat production in afghanistan. phillips, p. c. b., & perron, p. (1988). testing for a unit root in time series regression. biometrika, 75(2), 335–346. prakash, a. b. (1999). the transmission of signals in a decentralised commodity marketing system: the case of the uk pork market. imperial college london (university of london). rapsomanikis, g., hallam, d., & conforti, p. (2006). market integration and price transmission in selected food and cash crop markets of developing countries: tahir mehmood, afaq ali muluk, seema zubair 16 review and applications. agricultural commodity markets and trade, 187–217. samim, s. a., & zhiquan, h. (2020). assessment of food security situation in afghanistan. svu-international journal of agricultural sciences, 2(2), 356–377. srinivasan, p. v, & jha, s. (2001). liberalized trade and domestic price stability. the case of rice and wheat in india. journal of development economics, 65(2), 417– 441. journal of applied economics and business studies, volume. 5, issue 1 (2021) 131-142 https://doi.org/10.34260/jaebs.517 131 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x evaluation of ridge, elastic net and lasso regression methods in precedence of multicollinearity problem: a simulation study shady i. altelbany1 1. faculty of economic and administrative sciences, al azhar university– gaza, palestine abstract this study aims at performance evaluation of ridge, elastic net and lasso regression methods in handling different degrees of multicollinearity in a multiple regression analysis of independent variables using simulation data. the researcher simulated a collection of data with sample size n=200, 1000, 10000, 50000 and 100000, independent variables p=10. the researcher compared the performances of the three methods using mean square errors (mse). the study found that elastic net method outperforms ridge and lasso methods to estimate the regression coefficients when a degree of multicollinearity is low, moderate and high for any sample size. while, lasso method is the most accurate regression coefficients estimator when data containing severe multicollinearity at sample size less than 10000 observations. keywords ridge, lasso, elastic net, multicollinearity, regression. jel classification c02, c31, c63 1 introduction multiple linear regression is frequently employed is appropriate in particular context to evaluate a model to predict the expected responses, or to explore the link between the dependent variable and the independent variables. the first goal, which is the design's prediction accuracy, is critical; however, the second goal, which is the model's complexity, is more important. common linear regression procedures are popular for generally not carrying out well according to both prediction performance and model involvement (doreswamy and vastrad, 2013). there is a high number of hypotheses about the model in the regression analysis, specially, the most important one is (multicollinearity), in addition to (non-homogeneity of variance, autocorrelation and linearity) . if one or more assumptions are broken, the model becomes unreliable, shady i. altelbany 132 and it is no longer suitable for estimating population parameters (herawati et al., 2018). when there is a close association or interaction between two or more independent variables, multicollinearity occurs in the study of multiple regression. multicollinearity has the potential to produce inaccurate regression coefficient choices, increase regression coefficient standard errors, deflate partial t-tests for regression coefficients, produce wrong, non-significant p-values, and minimize model predictability. (draper and smith, 1998; gujarati 1995). the key issue with multicollinearity is that as the degree of collinearity rises, the coefficient estimates in the regression model become unsteady, and the standard errors for the coefficients become wildly maximized. multicollinearity has two types; the first type is (full/perfect/exact multicollinearity and the second one is partially/less than perfect multicollinearity). the existence of the first type is when independent variables interrupted in a complete way. this means that no particular least squares solution to a multiple regression analysis can be computed under this condition. (slinker, and glantz1985). since we know that multicollinearity is a serious problem when trying to make inferences or find predictive models, it's crucial to figure out the best way to deal with it. (judge 1988). multicollinearity can be detected using a variety of techniques and methods. using pair-wise scatter plots of the independent variables, searching for near-perfect relationships, analyzing the correlation matrix for high interactions and the variance inflation factors (vif), using eigenvalues of the correlation matrix of the independent variables, and testing the signs of the regression coefficients are just a few of the common approaches. (montgomery and peck, 1992; kutner et al., 2005). the reduce of variance at the cost of introducing a group of bias. the scholars call this method as “regularization or shrinkage methods” and is roughly beneficial for the predictive appearance of the model. in the study of current data, regularization is crucial. to overcome the shortcomings of ordinary least squares regression in terms of prediction precision, we introduced regularized regression methods for linear regression. methods of regularization aid in the formalization of a unique solution to this well-posed problem. any coefficients are reduced to zero using these techniques. this does not help with descriptor collection on its own, but it does reduce the gap at the expense of a small increase in bias. this form, on the other hand, improves the estimate's generalization. (doreswamy and vastrad, 2013). the least absolute shrinkage and selection operators ridge, lasso, and elastic net are among the methods. using simulated data, this study examines three different regression methods to see which one works better for coping with multicollinearity obstacles. journal of applied economics and business studies, volume. 5, issue 1 (2021) 131-142 https://doi.org/10.34260/jaebs.517 133 2 materials and methods: at first, we need to consider the basics of regression and what parameters of the equation they changed when using a distinct model. the relationship between a dependent variable and independent variables can be estimated using a multiple linear regression model and the ordinary least square (ols) method. if data comprises in observations,   1 , n i i i y x = each observation has a scalar response yi and a vector of p independents xij for j=1,...,p, we can write a multiple linear regression model as: y x  = + (1) where 1ny  is the dependent vector variable, n px  symbolizes the independent variables, 1p   is the set of regression coefficients that needs to be estimated, and 1p   symbolizes the residuals. ( ) 1ˆ ols x x x y −  = (2) the regression coefficients are calculated using ordinary least square by reducing the squared distances between the predicted dependent variable and the observed (montgomery and peck, 1992). when building a regression model, the model becomes more complicated as the number of data and variables grows, and major optimization issues arise (zou and hastie, 2005). furthermore, classical regression analysis fails when assumptions such as constant variance, multi-collinearity, and normality are not met (ogutu et al., 2012). as a result, high coefficients in the model must be corrected, or penalized. regularized regression is a form of regression in which the coefficient estimates are constrained to zero. it penalizes the magnitude of the error term as well as the magnitude of the coefficients. complex models are discouraged, mainly to prevent over-fitting. a typical least squares model has some flaws, such as the fact that it does not generalize well to data sets other than its training data. regularization greatly reduces the model's variance while having little effect on its bias. the effect on bias and variance is dominated by the tuning parameter λ used in the regularization systems mentioned. as the value of λ increases, the rate of coefficients decreases, lowering the variance. to some extent, this increase in λ is advantageous because it only reduces variance (thus preventing over-fitting) while losing no significant properties in the results, but after a certain value, the model loses significant properties, resulting in bias and under fitting. accordingly, the value of λ they should select it carefully ( biswas, 2019). there are three kinds of regularization systems called the ridge, lasso, and elastic net. ridge regression corrections are made with squared values, while lasso regression corrections are made with absolute values. the ridge and lasso biased estimation regression methods are combined in elastic net regression (zou and hastie 2003). https://medium.com/@imsaikatb?source=post_page-----ad9ff80f9ccc-------------------------------shady i. altelbany 134 2.1 ridge regression: it is obvious that ordinary least square (ols) is unsteady and presents estimates having a lot of variance when multicollinearity appears among independent variables, e.g. the columns of x are strikingly correlated. hoerl et al (1975) develop ridge regression and this approach is the adjustment of the least squares method, which allows for biased regression coefficient estimators. (myers, 1986). ridge regression approach depends on adding a ridge parameter to the diagonal of ( )x x matrix resulting a new matrix ( )x x  +  . since the diagonal of systems in the correlation matrix can be interpreted as a ridge, we call it ridge regression (hoerl and kennard, 2000). the ridge formula to find the coefficients is: ( ) 1ˆ , 0 ridge x x x y   −  = +   (3) when λ equal zero, the ridge estimator appears as the ordinary least square (ols). if they all λ 's are like each other, the estimators that resulted are called the ordinary ridge estimators (hoerl, 1962; hoerl et al., 1975). it is usually acceptable to edit ridge regression in lagrangian form:   2 2 22 ˆ arg min ridge y x     = − + (4) where ( ) 2 2 2 1 n t i i i y x y x  = − = − is the l2 -norm (quadratic) loss function (i.e. residual sum of squares), t i x is the row of x, 2 2 2 1 p j j   = =  is the l2 norm penalty on  and 0  is the tuning parameter (penalty, regularization) that controls the power of the penalty (linear shrinkage) by selecting the relative importance of the data-dependent practical error and the penalty term. the vaster the value of, the greater is the amount of shrinkage. since the value of is reliant on the data, we can find it out using data-possessed techniques, includes; cross-validation (doreswamy and vastrad, 2013). by constraining the coefficient estimates, ridge regression can overpower this multicollinearity, as a result, it may decrease the estimator's variance while also introducing bias. (james et al., 2013). 2.2 lasso regression: they broadly used lasso regression approaches in handling with big databases, such as those used in drug discovery, where efficient and quick algorithms are required (hastie and friedman, 2010) the lasso estimator is also recognized as basis pursuit (chen et al., 1998). still, because there are steep correlations between descriptors, lasso will choose one and ignore the others, and when all descriptors are file:///c:/users/ابودلال%202016/downloads/e.g journal of applied economics and business studies, volume. 5, issue 1 (2021) 131-142 https://doi.org/10.34260/jaebs.517 135 the same, it will decrease. the lasso penalty looks for many coefficients that are similar to zero, with only a small subset of them being the best (and not equal zero). to get a sparse solution to the following expansion problem, the lasso estimator uses the l1 penalized least squares basis (tibshirani, 1996).   2 12 ˆ arg min lasso y x     = − + (5) where 1 1 p j   = =  is the l1-norm penalty on  , that causes the solution to become sparse, and 0  is a parameter for fine tuning. penalizing the absolute values of the coefficients introduces shrinkage towards zero, likewise ridge regression. in contrast, unlike ridge regression, it reduces certain coefficients to zero; such solutions have a large number of identically zero values. the penalty acts as a continuous variable selection tool (herawati, 2018). the lasso estimation method handles both the multicollinearity issue and best feature selection together in the high dimension linear regression model. nonetheless, according to hastie and zou (2005) lasso estimation procedure is unstable if the amount of predictors is greater than the amount of observations. further, the prediction performance of re dominates lasso if there is high multicollinearity among predictors. 2.3 elastic net regression: according to (friedman et al., 2010) this is a continuity of the lasso that is robust to the strongest correlations among the predictor variables.. in order to prevent the imbalance of the lasso solution paths when predictor variables are strongly correlated, they projected the elastic net for assessing high-dimensional data. zou and hastie (2005) recommended elastic net estimator by using a mixture of ridge and lasso, and it is:   2 2 1 21 22 ˆ arg min elastic net y x       = − + + (6) the regularization parameter  is the sum of two nonnegative penalties 1 2   = + , now, let 2 1 2     = + then 1 1 2 1     − = + , where 0 1 . further, it can be defined as: ( )  2 2 1 22 ˆ arg min 1 elastic net y x         = − + + −   (7) note that 0 = , and then elastic net estimator in equation (7) is equivalent to ridge. similarly, 1 = , and then elastic net estimator in equation (7) is equivalent to lasso. if 0 = , so using this method, the elastic net method, decreases to ordinary least squares regression. shady i. altelbany 136 hence, we can write the ridge, lasso and elastic net estimator in a common form in the mis specified regression model as below: ( ) ( )  2 2 1 22 ˆ arg min 1y x          = − + + −   (8) where ( ) ˆ , 0 ˆ ˆ , 0 1 ˆ , 1 ridge elastic net lasso          =   =   = the mse, which is the scheduled prediction error of the estimators is given by: ( )( ) ( )( ) ( )( ) 1ˆ ˆ ˆ new new new new mse y x y x n        = − − (9) where ( ),new newy x includes new observations that are not used to obtain the coefficient estimates ( ) ˆ   . in brief, the following are some salient distinctions between lasso, ridge and elastic net (hastie et al., 2001): • lasso has a sparse selection, unlike ridge which does not have. • ridge regression shrinks the two coefficients towards one another if we have extremely correlated variables. furthermore, lasso is neutral and picks one over the other. in terms of context, no one would know which variable was chosen. elastic net is an adjustment between the two which attempts to shrink and do a sparse selection at the same time. • ridge estimators are neutral to multiplicative scaling of the data if constants multiplied both x and y, the coefficients of the fit do not change for λ parameter. however, for lasso, the fit is not separate from the scaling. in fact, the multiplier must scale the λ parameter up to get the same result. it is more complicated for elastic net. • in a comparison with lasso, ridge penalizes the largest β‘s rather than it penalizing the small ones (as it square them in the penalty term). lasso penalizes the small ones more consistently. sometimes, this is of no consequence. when faced with a forecasting issue involving a strong predictor, the ridge shrinks the predictor’s effectiveness as compared to the lasso. 3 results: simulation study using r package, we simulate the linear regression model for number of data n = 200, 1000, 10000, 50000, 100000 observations and 10 independent variables. to explore the implements of different grades of multicollinearity on the estimators, we journal of applied economics and business studies, volume. 5, issue 1 (2021) 131-142 https://doi.org/10.34260/jaebs.517 137 choose ( )0.70, 0.80, 0.90, 0.99 = which represent low, moderate, high multicollinearity and severe multicollinearity. (mcdonald and galarneau, 1975) generate the independent variables: ( ) 1/ 2 2 1 , 1, 2,..., & 1, 2,..., ij ij ij x u u i n j p = − + = = (10) where ij u are independent, standard normal pseudo-random numbers and  is fixed, so that any two independent variables' theoretical correlation is defined by 2  . performance assessment of ridge, lasso and elastic net approaches are compared on the basis of mse value. cross validation is a technique for determining a value for the  value for ridge, lasso and elastic net, and are displayed in tables 1-4. the estimated mse values of the ridge, lasso and elastic net estimators versus regularization parameter when 0.70, 0.80, 0.90 0.99and   = = = = and the optimal value of regularization parameters are summarized in table 1–4. we can observe in table 1-3, that elastic net was outperforms than ridge and lasso at n = 200, 1000, 10000, 50000, 100000 observations if degrees of multicollinearity are low, moderate, high. in table 4 we can show that lasso method was outperforms than ridge and elastic net when 𝜌 = 0.99 (severe multicollinearity) at n = 200, 1000, 10000 observations, while at n = 50000, 100000 observations was showed that elastic net method was the best. elastic net method is the most accurate regression coefficients estimator. shady i. altelbany 138 table 1 mse values and optimal value of regularization parameter when 0.70 = 0.7 = method  n=200 n=1000 n=10000 n=50000 n=100000  mse  mse  mse  mse  mse ridge 0 2.4943870 1.3542480 2.435132 1.77365 2.393143 2.071687 2.397843 1.90527 2.398618 1.925571 elastic net 0.1 0.8556920 0.8115550 0.5246333 1.09823 0.4280494 1.111337 0.3560721 1.026351 0.3561871 1.011122 0.2 0.6812518 0.8255325 0.4584058 1.094004 0.3740143 1.111348 0.3414572 1.034078 0.3112236 1.013724 0.3 0.4984487 0.8274622 0.3353999 1.086273 0.3003344 1.105096 0.2741909 1.030563 0.2499133 1.009879 0.4 0.4502874 0.8562786 0.2760757 1.08463 0.2472124 1.10109 0.247698 1.034514 0.2257661 1.013563 0.5 0.3953519 0.8766654 0.2919647 1.08985 0.2170523 1.101262 0.1981584 1.029075 0.1982224 1.014189 0.6 0.3294599 0.8792247 0.2670257 1.091589 0.1808769 1.09714 0.1812321 1.031515 0.1651853 1.009974 0.7 0.3099273 0.9037126 0.2085462 1.086522 0.1701533 1.100581 0.1704875 1.035552 0.155392 1.013765 0.8 0.2251437 0.8662195 0.2002693 1.088843 0.1488841 1.098445 0.1359241 1.027507 0.135968 1.011472 0.9 0.2410546 0.9089749 0.1953736 1.092268 0.1452446 1.103449 0.1208214 1.025793 0.1208604 1.008507 lasso 1 0.2169491 0.9112325 0.1602154 1.088178 0.1307201 1.102091 0.1193412 1.031095 0.1087744 1.00979 table 2 mse values and optimal value of regularization parameter when 0.80 = 0.8 = method  n=200 n=1000 n=10000 n=50000 n=100000  mse  mse  mse  mse  mse ridge 0 3.072037 2.377661 2.65657 1.925118 2.72915 2.103151 2.720972 2.013501 2.710478 2.047791 elastic net 0.1 0.9602314 1.001544 0.6281431 1.120764 0.4881492 1.111701 0.4040558 1.034427 0.4417403 1.02362 0.2 0.76448 0.9836472 0.5000907 1.101727 0.4265273 1.110792 0.3874714 1.040867 0.3516877 1.016339 0.3 0.5593439 0.9327243 0.401574 1.091794 0.375896 1.112307 0.3111404 1.03481 0.3099403 1.01822 0.4 0.5052988 0.9567745 0.3305453 1.086529 0.281922 1.100099 0.2810773 1.03758 0.2551193 1.014054 0.5 0.4436518 0.9626045 0.3495691 1.09122 0.2716609 1.10776 0.2467856 1.03770 0.2458337 1.022227 0.6 0.3697098 0.9457097 0.3197097 1.091918 0.2062727 1.096673 0.2257057 1.039964 0.2048614 1.016942 0.7 0.3477909 0.9656147 0.2496922 1.085563 0.2129625 1.107571 0.2123244 1.044483 0.1927159 1.020984 0.8 0.2772824 0.9290207 0.2184807 1.08461 0.1863422 1.105263 0.1692793 1.034339 0.1686264 1.018251 0.9 0.2968779 0.9840657 0.2131398 1.086901 0.1817869 1.111765 0.1504705 1.031538 0.1498901 1.01628 lasso 1 0.2434537 0.9478264 0.1918258 1.086458 0.1636082 1.110342 0.1354234 1.032729 0.1349011 1.014802 journal of applied economics and business studies, volume. 5, issue 1 (2021) 131-142 https://doi.org/10.34260/jaebs.517 139 table 3 mse values and optimal value of regularization parameter when 0.90 = 0.9 = method  n=200 n=1000 n=10000 n=50000 n=100000  mse  mse  mse  mse  mse ridge 0 3.3935400 2.7443080 3.10191 2.095804 3.123437 2.490004 3.105966 2.234529 3.098342 2.260106 elastic net 0.1 1.0607240 1.1089750 0.8834351 1.183933 0.6131434 1.159529 0.4612261 1.037333 0.4600939 1.024278 0.2 0.7694645 0.9855784 0.771914 1.164514 0.5357428 1.147965 0.4422951 1.041303 0.4020136 1.020438 0.3 0.6781246 0.9557141 0.6198487 1.136541 0.4721469 1.141938 0.3897921 1.040322 0.3888352 1.028375 0.4 0.5581806 0.9200789 0.6145527 1.144592 0.3886355 1.128757 0.3208472 1.034618 0.3200596 1.021084 0.5 0.5378645 0.9313477 0.5921847 1.154328 0.3744903 1.137142 0.2817036 1.034401 0.3084104 1.029477 0.6 0.4482204 0.9077476 0.4934872 1.135276 0.3120752 1.126321 0.2576411 1.037091 0.2570087 1.022753 0.7 0.4216468 0.9163924 0.46423 1.140038 0.2935733 1.13074 0.2659968 1.050992 0.2417714 1.027734 0.8 0.3361653 0.8828899 0.4062012 1.130843 0.2819218 1.138457 0.2120706 1.039766 0.21155 1.024674 0.9 0.3599220 0.9240131 0.3610678 1.120571 0.2505971 1.13428 0.1885072 1.03851 0.1880445 1.022657 lasso 1 0.2951528 0.8923225 0.324961 1.124729 0.2255374 1.131318 0.1861978 1.045797 0.1857407 1.030667 table 4 mse values and optimal value of regularization parameter when 0.99 = 0.99 = method  n=200 n=1000 n=10000 n=50000 n=100000  mse  mse  mse  mse  mse ridge 0 3.654888 3.0121850 3.546463 3.483199 3.585424 4.344516 3.557993 3.859463 3.544116 3.739378 elastic net 0.1 1.040925 1.0764300 0.8385574 1.334766 0.7038333 1.387571 0.6984484 1.251602 0.6339181 1.216287 0.2 0.9982006 1.0413740 0.8825415 1.313373 0.7407508 1.391776 0.7350835 1.251386 0.6671684 1.224578 0.3 0.8015574 0.9893414 0.853611 1.305417 0.7164684 1.361745 0.7109868 1.25647 0.7082139 1.246001 0.4 0.7241091 0.9845812 0.8463177 1.295802 0.7796048 1.379147 0.6422896 1.222855 0.7021628 1.25335 0.5 0.6357671 0.9766932 0.8155141 1.300482 0.6844922 1.35112 0.6189121 1.223221 0.676606 1.227457 0.6 0.5814613 0.9817787 0.7458548 1.274246 0.6870611 1.359158 0.6212348 1.237149 0.6188119 1.221965 0.7 0.4983954 0.9641743 0.7016354 1.266754 0.6463274 1.353982 0.5844038 1.235244 0.5821245 1.227421 0.8 0.4786148 0.9831136 0.6737885 1.272795 0.6206756 1.358227 0.5612097 1.244024 0.5590209 1.236977 0.9 0.4669148 1.0107910 0.6573174 1.281283 0.5517117 1.322116 0.5474906 1.260935 0.5453553 1.257195 lasso 1 0.3828918 0.9524901 0.5915857 1.242952 0.4965405 1.296512 0.4927415 1.244684 0.4908198 1.24159 shady i. altelbany 140 4 conclusion according to the outcomes of simulation at p = 10 and n = 200, 1000, 10000, 50000, 100000 observations, containing different degrees of multicollinearity within all independent variables, it can be summarized in: • elastic net method outperforms ridge and lasso methods to estimate the regression coefficients when degree of multicollinearity is low ( )0.70 = , moderate ( 0.80 = ) and high ( 0.90 = ) at all number of data (n = 200, 1000, 10000, 50000, 100000 observations). • if data including cruel multicollinearity within all independent variables, the lasso method outperforms ridge and elastic net methods at (n = 200, 1000, 10000 observations). • elastic net methods for the larger number of observations (n = 50000, 100000 observations) was outperforms ridge and lasso methods if data containing severe multicollinearity. • results suggest that performance of these methods are depending greatly on the values of α. • elastic net regression outperformed the other two methods in case of low, moderate and high level of multicollinearity when 0 < α <1. also, it can be incurred that severe multicollinearity requires higher value (α=1). this is consistent with the theoretical framework. • ridge method was unsuitable regression coefficients estimator compared with lasso and elastic net methods. • general, in studying relationships and interconnecting economic and social factors, we recommend to the decision maker that use elastic net method for any sample size. • we also recommend to the decision maker that use lasso method for when using real data, and examine the relationships between the different variables (severe multicollinearity) at sample size less than 10000 observations. journal of applied economics and business studies, volume. 5, issue 1 (2021) 131-142 https://doi.org/10.34260/jaebs.517 141 references: biswas, s. (2019). how regularization helps in data overfitting. https://medium.com/towards-artificial-intelligence/how-regularization-can-helpin-overfitting-the-data-ad9ff80f9ccc. chen, s. s., donoho, d. & saunders, m. (1998). atomic decomposition by basis pursuit. siam journal on scientic computing, 20(1), 33–61. https://doi.org/10.1137/s1064827596304010 doreswamy, a., & vastrad, c. m. (2013). performance analysis of regularized linear regression models for oxazolines and oxazoles derivatives descriptor dataset. international journal of computational science and information technology (ijcsity), 1(4), 111-123. dio: 10.5121/ijcsity.2013.1408. draper, n.r., & smith, h. (1998). applied regression analysis. 3rd ed., new york: wiley. https://doi.org/10.1002/9781118625590 friedman, j., hastie, t., & tibshirani, r. (2010). regularization paths for generalized linear models via coordinate descent, journal of statistical software, 33(1),1-22. doi: 10.1163/ej. 9789004178922.i-328.7 gujarati, d. (1995). basic econometrics. 4th ed., new york: mcgraw−hill. hastie, t., tibshirani, r., & friedman, j. (2001). the elements of statistical learning: data mining, inference, and prediction. 2nd ed., springer-verlag new york inc. https://www.springer.com/gp/book/9780387848570 herawati, n., nisa, k., setiawan, e., & nusyirwan, t. (2018). regularized multiple regression methods to deal with severe multicollinearity. international journal of statistics and applications, 8(4), 167-172. doi: 10.5923/j.statistics.20180804.02 hoerl, a.e., & kennard, r.w. (2000). ridge regression: biased estimation for nonorthogonal problems. technimetrics, 42, 80-86. https://doi.org/10.2307/1271436 hoerl, a.e. (1962). application of ridge analysis to regression problems. chem. eng. prog., 58, 54-59. hoerl, a., kannard, r., & baldwin, k.f. (1975). ridge regression: some simulations. communication of statistics, 4, 105-123. https://doi.org/10.1080/03610927508827232 https://medium.com/@imsaikatb?source=post_page-----ad9ff80f9ccc-------------------------------https://medium.com/towards-artificial-intelligence/how-regularization-can-help-in-overfitting-the-data-ad9ff80f9ccc https://medium.com/towards-artificial-intelligence/how-regularization-can-help-in-overfitting-the-data-ad9ff80f9ccc https://doi.org/10.1137/s1064827596304010 https://arxiv.org/ct?url=https%3a%2f%2fdx.doi.org%2f10.5121%2fijcsity.2013.1408&v=0a57879d https://doi.org/10.1002/9781118625590 https://www.researchgate.net/deref/http%3a%2f%2fdx.doi.org%2f10.1163%2fej.9789004178922.i-328.7 https://www.bookdepository.com/publishers/springer-verlag-new-york-inc https://www.bookdepository.com/publishers/springer-verlag-new-york-inc https://www.springer.com/gp/book/9780387848570 https://www.researchgate.net/deref/http%3a%2f%2fdx.doi.org%2f10.5923%2fj.statistics.20180804.02 https://doi.org/10.2307/1271436 https://doi.org/10.1080/03610927508827232 shady i. altelbany 142 james, g., witten d., hastie t., & tibshirani r. (2013). an introduction to statistical learning: with applications in r. new york: springer publishing company, inc. https://link.springer.com/book/10.1007/978-1-4614-7138-7 judge, g.g. (1988). introduction to theory and practice of econometrics. new york: john willy and sons. https://doi.org/10.1002/jae.3950050311 kutner, m. h., nachtsheim, c., neter, & william, n. (2005). applied linear statistical models. 5th edition. new york: mcgraw-hill. https://www.amazon.com/applied-linear-statistical-modelsmichael/dp/007310874x mcdonald g.c., & galarneau, d.i. (1975). a monte carlo evaluation of some ridge type estimators. j. amer. statist. assoc., 20, 407-416. https://www.tandfonline.com/doi/abs/10.1080/01621459.1975.10479882 montgomery, d.c. & peck, e.a. (1992). introduction to linear regression analysis. new york: john willy and sons. https://doi.org/10.1111/biom.12129 myers, r. h. (1986). classical and modern regression with applications, 2nd ed, usa: pwskent publishing company. https://lib.ugent.be/catalog/rug01:000851135 ogutu, j. o., schulz-streeck, t., & piepho, h. p. (2012). genomic selection using regularized linear regression models: ridge regression, lasso, elastic net and their extensions. bmc proceedings, 6 (2), 10. https://doi.org/10.1186/1753-65616-s2-s10 slinker, b.k., & glantz, s.a. (1985). multiple regression for physiological data analysis: the problem of multicollinearity. american journal of physiology regulatory, integrative and comparative physiology, 249(1), r1–r12. https://doi.org/10.1152/ajpregu.1985.249.1.r1 tibshirani, r. (1996). regression shrinkage and selection via the lasso. journal royal. statist. soc b., 58(1), 267-288. https://doi.org/10.1111/j.25176161.1996.tb02080.x zou, h., & hastie, t. (2005). regularization and variable selection via the elastic net. journal of the royal statistical society: series b (statistical methodology), 67 (2), 301-320. https://doi.org/10.1111/j.1467-9868.2005.00503.x https://link.springer.com/book/10.1007/978-1-4614-7138-7 https://doi.org/10.1002/jae.3950050311 https://www.worldcat.org/search?q=au%3akutner%2c+michael+h.%2c&qt=hot_author https://www.worldcat.org/search?q=au%3anachtsheim%2c+chris%2c&qt=hot_author https://www.worldcat.org/search?q=au%3aneter%2c+john%2c&qt=hot_author https://www.worldcat.org/search?q=au%3ali%2c+william%2c&qt=hot_author https://www.amazon.com/applied-linear-statistical-models-michael/dp/007310874x https://www.amazon.com/applied-linear-statistical-models-michael/dp/007310874x https://www.tandfonline.com/doi/abs/10.1080/01621459.1975.10479882 https://doi.org/10.1111/biom.12129 https://lib.ugent.be/catalog/rug01:000851135 https://doi.org/10.1186/1753-6561-6-s2-s10 https://doi.org/10.1186/1753-6561-6-s2-s10 https://doi.org/10.1152/ajpregu.1985.249.1.r1 https://doi.org/10.1111/j.2517-6161.1996.tb02080.x https://doi.org/10.1111/j.2517-6161.1996.tb02080.x https://doi.org/10.1111/j.1467-9868.2005.00503.x journal of applied economics and business studies, volume. 2, issue 1 (2018) 47-60 https://doi.org/10.34260/jaebs.215 47 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online): 2663-693x socio economic determinants of child labour in selected districts of khyber pakhtunkhwa, pakistan nabeela begum1*, javed iqbal2 & hina31 1bs student, department of economics, abdul wali khan university mardan. 2assistant professor, department of economics, abdul wali khan university mardan. 3bs student, department of economics, abdul wali khan university mardan. abstract this study examines the determinants of child labour in mardan and nowshera districts of khyber pakhtunkhwa. primary data on socioeconomic characteristics of children engaged and did not engage in child labour were obtained from labour education organization mardan. age of the children and family size are positively and education is negatively and significantly associated with the probability of children participation in labour market. the probability of child labour is more with the household income although with a very low coefficient value which is contrary to our expectations and may ne indicative that child labour could be a major source of household income. this study suggests that subsidies may be provided to families for their children education. family size is also positively related to the child labour, therefore steps may be taken towards encouraging small family sizes and thereby reducing the child labour. keywords child labour, socio-economic characteristics, logit model jel classification j10, j12 copyright © 2018 saebr all rights reserved 1. introduction child labour is a serious issue faced by many developing countries including pakistan. children engaged in labour are deprived from their right to education. labour work in such an age is harmful for their mental and physical development. child labour is exploitative and results in poor standard of living, poor health, attention and education. parents' lack of education and lack of awareness about the negative consequences of child labour could have long-term implications not only for their families but overall country's development is jeopardized (kazmi, 2015). nawaz and shaheen (2017) reported that poverty, migration, education and family were the main factors for the child labour. further, they found that children were exploited and harassed on regular basis at the workplace. in the developing 1*corresponding author: kpari2743@gmail.com nabeela begum, javed iqbal & hina 48 countries most families face challenges to have a very basic shelter; food and clothing and therefore almost all family members are forced to work including children. international labour organization (ilo) in its 2013 report highlight that around 168 million children were involved child labour worldwide. furthermore, according to ilo, about 12 million children alone are employed in pakistan in the year 2015. osment (2014) argued that many countries face the child labour problem because of the extreme poverty. furthermore, child labour can have many consequences for entire families and could also causes issues such a slavery and abandoned labour. they also found that many countries have policies in place to eradicate the child labour although there has been a challenge in enforcing such policies in true spirit. this problem is more serious in developing countries where children are made to work to fulfil their basic needs and absence of the social security net further aggravates the problem. gandapur et.al (2014) reported that pakistani parents also face many economic problems and therefore children are engaged in labour at their school going age. generally, in developing countries where policies either do not exist or enforcement is weak, children are underpaid, and their working conditions are hazardous. they are engaged labour in unhygienic and unsafe working environments. children are also generally exposed to hazardous work. most children are employed in industrial related jobs or technical trade related jobs where they use sharp objects and physically exertive jobs that which could result in long term health related issues in future. pakistan's per-capita income has been around $1900 which is quite low to support a larger average pakistani family. there has been ever increasing inflation that also make the life of people miserable. according to asian development bank, about 24.3 percent of pakistan population lived below the national poverty line in 2015. furthermore, social policy development centre (spdc) reported that in 1999, 33% people lived under poverty which increased 38% in the next two years (zaidi et. al., 2013). this study examines the socioeconomic and demographic status of the families whose children were engaged in child labour and otherwise. it is important to study such characteristics to have an understanding of the localised environment under which children are engaged in child labour and what steps can be taken to effectively deal with it. 2. literature review child labour has been studied in many countries with different polices in places and under various socio-economic characteristics of the households. siddiqi & patrino (1995) noted that the child labour is a huge problem especially in developing countries. they found that child labour was more wide spread in rural areas where monitoring and enforcement was difficult. children work for a variety of reasons among which poverty is the most important reason to meet family needs. children contribute significantly to their families’ income even though they are not paid as per their labour work. they found that journal of applied economics and business studies, volume. 2, issue 1 (2018) 47-60 https://doi.org/10.34260/jaebs.215 49 inaccessibility of schools in rural areas, lack of quality education and other socioeconomic problems contributed to the child labour. social and cultural norms especially about the female education further increased the child labour. this study advocated the establishment of partnerships of governments with humanitarian organizations and the international community for remedying the problems of working children. bhalotra and heady (2001) observed that land-rich families employed more children compared to land-poor families as land created more opportunities for family labour as well as there is more need for children to work. the majority of children works in developing countries are engaged in agricultural work, predominantly on farms operated by their families. land is the most important resource of wealth in agrarian countries and it is typically distributed very unequally. they observed that child labour mainly existed in the poorest households. they suggested that this seeming paradox can be explained by failures of the markets for labour and land. credit market improvement may help alleviating the child labour. they argued that after controlling for household consumption and other covariates, the wealth paradox persists for girls and boys in many countries. beegle et.al (2005) explored the relationship between household income shocks and child labour using a panel data technique on the household survey in tanzania. they found significant relationship between child labour and income shocks and borrowings. mahmood et.al (2005) studied the socio-economic determinants of child labour in the automobile workshops, and other engineering workshops in tehsil samundri, faisalabad, pakistan. they interviewed randomly selected 120 respondents under the age of 15 years. those who aged less than 14 years constituted one half the children interviewed, most had attained primary education, majority lived in rural areas, and had both parents alive but with low income. majority of the respondents were not willing to work but family financial problems forced them to work. they suggested that more educational and more adult employment opportunities could help eradicate poverty and thus child labour. edmonds (2007) noted child labour was influenced by the local labour markets, family conditions, the net return to schooling, and finally the poverty. the study emphasized the importance of the effectiveness of child labour policies, and understanding of the determinants of child labour. ahmad and haider (2012) investigated the factors for child labour using cross-sectional data collected from 100 households and using a binary choice model. it was found that the education of the household, and their income negatively affected the child labour significantly. furthermore, child age, and family size were found to be positively related with child labour although insignificantly. in this study they concluded that parental education was necessary for better future of children. the study suggested that government and private sector should provide education facilities to the poor people of khyber pakhtunkhwa province. the government also needs to establish skill development schools nabeela begum, javed iqbal & hina 50 or centres in rural areas as well as in urban areas so that those who cannot afford formal education can acquire technical skills. zaidi et.al (2013) studied the basic demographic characteristics of those who worked as child labour and also examined the working conditions in which they worked. they collected cross sectional data through a survey which was conducted in district rawalpindi from january-september 2008. the data were collected from 700 children aged between 4-18 years and their employers. the major reasons for the child labour were low family income and the existence of poverty. pervez and mukhtar (2014) noted most children were employed in fan manufacturing industries, ceramics industry and automobile workshops in the district gujarat, pakistan. they used the logit and probit models for estimating the factors responsible for child labour. family income had a negative and statistically significant effect on the child labour while average family size was found to positively affecting it. they suggested that people need to be informed of the adverse effects of child labour and education should be made accessible and affordable. 3. research methods we begin this section with the explanation of factors that could be responsible for the existence of child labour. various factors have been identified in the previous literature which are outlines as follows: a majority of studies in developing countries indicated poverty was the major cause of child labour and that poorer families were more likely to send their children to work compared to non-poor families (pervez & mukhtar, 2014). parent’s profession is also found to be important factor determining the child labour. where parents have been employed in white collar jobs, they were less likely to send their children to work in early ages. parents’ education plays a critical role in their children growth. educated parents can make more rational decisions to decide about the future of their children. when the educational level of parents is higher, they are less likely to send their children to work. generally, age, income and family size are correlated with the child labour. lower is the family income, the higher are the chances of a child to be sent to the work at tender age. furthermore, family size also plays a huge role in child labour because as the size of the family increases so are the needs and therefore that lack of income is made up with the child labour. finally, cost of education is another major problem for poor household’s family. schools need to be affordable and accessible for all the households. after identifying various factors responsible for child labour based on the previous literature, we develop our empirical model. the nature of problem is that whether a child is in labour force or otherwise. therefore, the dependant variable is dummy variable and hence a binary choice variable could be used. generally, logit model is used to empirically analyze such a problem. below is presented a short introduction to the model that will be used in journal of applied economics and business studies, volume. 2, issue 1 (2018) 47-60 https://doi.org/10.34260/jaebs.215 51 this research. choices of the households/children can be represented by a binary variable that takes the value 1 if a child is in labour force and takes the value 0 otherwise. an individual chooses a choice between being a part of child labour or not part of child labour and enjoying the childhood and going school. the choice problem of an individual is given as follows: 𝑦 = { 1 𝑐ℎ𝑖𝑙𝑑 𝑙𝑎𝑏𝑜𝑢𝑟 0 𝑜𝑡ℎ𝑟𝑤𝑖𝑠𝑒 } (1) where y is a random variable and the probability function for y is 𝑓(𝑦) = 𝑝𝑦 (1 − 𝑝)1−𝑦 𝑤ℎ𝑒𝑟𝑒 𝑦 = 0,1 (2) p is the probability that y takes the value 1. this discrete random variable has expected value 𝐸(𝑦) = 𝑝. in regression analysis, dependent variable has a fixed component and a random component and given as follows: 𝑦 = 𝐸(𝑦) + 𝑒 = 𝑝 + 𝑒 (3) 𝐸(𝑦) = 𝑝 = 𝛽1 + 𝛽2𝑥 (4) 𝑦 = 𝐸(𝑦) + 𝑒 = 𝛽1 + 𝛽2𝑥 + 𝑒 (5) the cumulative distribution function of the logit model is given as; 𝑃𝑖 = 1 1+𝑒 −(𝛽1+𝛽2𝑥) (6) where x are the explanatory variables such as age (age of children), ce (child education), w (child age), fs (family size), pe (parent's education), fi (family income), hh (head of household), and fp (father profession). after substituting for the x, we get, 𝑃𝑖 = 1 1 + 𝑒−(𝛽0+𝛽11𝑎𝑔𝑒+ 𝛽2𝐶𝐸+ 𝛽3 𝑤+ 𝛽4 𝐹𝑆+ 𝛽5 𝑃𝐸+ 𝛽6𝐹𝐼𝑖+ 𝛽7 𝐻𝐻+ 𝛽8𝐹𝑃 ) (7) where, pi is the probability of child labour being engaged in it. for ease of exposition, we can write the equation (6) as follows; 𝑃𝑖 = 1 1+𝑒 −𝑍𝑖 = 1 1+ 1 𝑒 𝑍𝑖 = 1 𝑒 𝑍𝑖 + 1 𝑒 𝑍𝑖 = 𝑒 𝑍𝑖 1+𝑒 𝑍𝑖 (8) where zi = (𝛽1 + 𝛽2𝑥). the probability of not working as a child labour is given as follows; 1 − 𝑃𝑖 = 1 − 𝑒 𝑍𝑖 1+𝑒 𝑍𝑖 = 1+𝑒 𝑍𝑖 −𝑒 𝑍𝑖 1+𝑒 𝑍𝑖 = 1 1+𝑒 𝑍𝑖 (9) using equation (8) & (9) we can write the odd of someone work compared to not working as follows: 𝑝𝑖 1−𝑝𝑖 = 1+𝑒 𝑍𝑖 1+𝑒 −𝑍𝑖 = 𝑒 𝑍𝑖 (10) 𝑝𝑖 1−𝑝𝑖 = 𝑒𝛽0+𝛽11𝑎𝑔𝑒+ 𝛽2𝐶𝐸+ 𝛽3 𝑤+ 𝛽4 𝐹𝑆+ 𝛽5 𝑃𝐸+ 𝛽6𝐹𝐼𝑖+ 𝛽7 𝐻𝐻+ 𝛽8𝐹𝑃+ 𝑢𝑖 (11) nabeela begum, javed iqbal & hina 52 now 𝑝𝑖 1 − 𝑝𝑖⁄ is simply the odd ratio of the incidence of child labour implying that the ratio of the probability that family send their child for working compared to those families who will not. taking log on the above equation, we get the following model; 𝐿 = 𝑙𝑛[𝑃𝑖/1 − 𝑃𝑖] = ln (𝑒𝛽0+𝛽11𝑎𝑔𝑒+ 𝛽2𝐶𝐸+ 𝛽3 𝑤+ 𝛽4 𝐹𝑆+ 𝛽5 𝑃𝐸+ 𝛽6𝐹𝐼𝑖+ 𝛽7 𝐻𝐻+ 𝛽8𝐹𝑃 ) (12) 𝐿 = ln [ 𝑃𝑖 1 − 𝑃𝑖] = 𝛽0 + 𝛽11𝑎𝑔𝑒 + 𝛽2𝐶𝐸 + 𝛽3 𝑤 + 𝛽4 𝐹𝑆 + 𝛽5 𝑃𝐸 + 𝛽6𝐹𝐼𝑖 + 𝛽7 𝐻𝐻 + 𝛽8𝐹𝑃 (13) equation (13) is the logit model that we will estimate to analyse the effect of various socio-economic factors on the child labour choice of being engaged in child labour or otherwise. the model is not only linear in explanatory variables but also linear in parameters. the data for this study were obtained from the labour education organization (leo) mardan, khyber pakhtunkhwa. leo collected the data through a questionnaire both in mardan and nowshera districts. a total of 300 respondents were selected from this data based on completeness of the data and other inaccuracies. in this sample, about 154 of them were working children in the age group of 5-14 years at different work places of mardan and nowshera city. these children were involved in different economic activities such as tailoring, workshops, and in hotels. 4. results and discussion 4.1. descriptive analysis the descriptive analysis for all the variables included in this research work is presented in table 1. the minimum age of child in the sample who was working was 5 years & the maximum age of child was 14 years with average age of 11.51 years. it can also be seen in the table that that average years of education was 4.28 years. it implies that the children were taken out from schools by the parents at the very early age. there can be multiple factors that could have made the parents to withdraw their children from schools such as household income, affordability and accessibility of schooling, poor performance of the children, strict and less caring teachers and many others. hopelessness of the parents to get higher earnings or employment from the schooling of their children. education of parents is generally believed to be a factor determining the child labour participation as well. in our sample, parental education on average was 5.8 years with maximum of 16 years of education. family size could also play a role as the bigger the sizes of the family, the more are the needs of the families. the average family size was about 6 members with a maximum family size 14. the household income was generally low and the average income was just rs. 11,350. majority of the income was earned from the activities such as van driving, sewing, garbage and waste collection, fruit and vegetable selling, working abroad, meat shop, and mechanics. the child income has been low so that main reason for child labour journal of applied economics and business studies, volume. 2, issue 1 (2018) 47-60 https://doi.org/10.34260/jaebs.215 53 may not be income but to learn business skills, and hands on experience and technical skills so that in future they could earn the livelihood for themselves. table 1: descriptive statistics variables frequency minimum maximum mean age (years) 300 5 14 11.51 education of the child (years) 300 0 9 4.28 parent education (years) 300 0 16 5.80 family size (numbers) 300 3 14 6.12 total household income (rs. /month) 300 2500 50000 11351.67 child income (rs. /month) 300 0 400 58.30 source: survey data table 2 shows that majority (83%) of the children were aged more than 10 years old. also, about half of them had primary education compared to the other half who had secondary education level. similarly, we had the sample such that about half of them were in child labour. the sample was split almost in half too with 54% female and 46% male children. there is a variation in urban and rural areas such that 73% of respondents lived in rural areas while 27% in urban areas. majority of them (58%) had their own house while 42% lived in rented houses. finally, most of the working children were employed on daily wages and 64.4% had their daily income less than 100 rupees. table 2: characteristics of the respondents categories frequency percentage (%) cumulative% % child age < 10 age of child 51 17 17 > 10 age of child 249 83 100 child education < 5 years(primary) 174 58 58 > 5 years(secondary) 126 42 100 working status in child labour 153 51 51 outside child labour 147 49 100 child status 1 (male) 140 46.67 46.67 0 (female) 160 53.33 100 neighborhoods status urban 81 27 27 rural 219 73 100 child income rs. /day 0 to 100 228 75.98 75.98 100 to 200 34 11.32 87.3 200 to 300 26 8.66 95.96 300 to 400 12 4 99.96 family size (#) < 4 10 3.33 3.33 4-7 232 77.33 80.66 8-10 54 18 98.66 > 10 7 1.33 99.99 income of hh rs. /month < 15,000 234 77.98 77.98 nabeela begum, javed iqbal & hina 54 source: survey data table 3 exhibits a relationship between the child labour status and the neighbourhoods that the children lived in (rural versus urban). out of the total 219 children who lived in rural areas, about 117 (53%) of them were engaged in child labour compared to 102 (47%) of them who were not in child labour. similarly, among the urban children, about 28% were found to be engaged in child labour. the reason for few children working in urban areas could be that the parents had better jobs in urban areas, better educated and more opportunities for children to go to school. the table also indicated that majority of the children who were in child employment came from rural areas (84%) compared to the urban areas (16%). the difference between the child labour based on the neighbourhood status was found to be statistically significant. table 3: relationship between the child labour and neighbourhood status neighbourhood total chi square value rural urban outside of child labour count 102 58 160 % within child labour status 63.8% 36.2% 100.0% 14.88*** % within neighbourhood 46.6% 71.6% 53.3% in child labour count 117 23 140 % within child labour status 83.6% 16.4% 100.0% % within neighbourhood 53.4% 28.4% 46.7% total count 219 81 300 % within child labour status 73.0% 27.0% 100.0% % within neighbourhood 100.0% 100.0% 100.0% source: survey data table 4 shows the relationship between the child labour status and child age. out of the total 249 children whose age was more than 10 years, about 55% of them were in child labour compared to the 45% who were not. similarly, children aged less than 10 years, about 4 % were found in child employment compared to the other 96% who were not engaged in child labour. the table shows that the majority of children in child labour were those who aged greater than 10 years. the difference between child labour status based on child labour age was found to be significant. 15,000 to30,000 57 18.99 96.97 > 30,000 9 3 99.97 parental education < 5 (primary) 94 31.34 31.34 5 to 8 (middle) 98 32.67 64.01 8 to 10 (matrix) 54 18 82.01 10 to 12 (intermediate) 32 10.67 92.68 12 to 14 (bachelor) 13 4.33 97.01 16 (master) 9 3 100 assets (property) own house 176 58.67 58.67 rent house 124 41.33 100 journal of applied economics and business studies, volume. 2, issue 1 (2018) 47-60 https://doi.org/10.34260/jaebs.215 55 table 4: relationship between child labour and child age child labour age total chi square value > 10 years < 10 years outside of child labour in child labour count 111 49 160 45.10*** % within child labour status 69.4% 30.6% 100.0% % within child age 44.6% 96.1% 53.3% count 138 2 140 % within child labour status 98.6% 1.4% 100.0% % within child age 55.4% 3.9% 46.7% total count 249 51 300 % within child labour status 83.0% 17.0% 100.0% % within child age 100.0% 100.0% 100.0% source: survey data table 5 exhibits the relationship between child labour status and children education. about 126 children who got more than 5 years of education, about 15% were in child labour compared to the remaining 85% who were not in child labour. similarly, about 70% who had less than 5 years of education were in child labour. table 5: relationship between child labour and child education child labour education total chi square value >5 years <5 years outside of child labour count 107 53 160 87.08*** % within child status 66.9% 33.1% 100.0% % within child education 84.9% 30.5% 53.3% in child labour count 19 121 140 % within child status 13.6% 86.4% 100.0% % within child education 15.1% 69.5% 46.7% total count 126 174 300 % within child status 42.0% 58.0% 100.0% % within child education 100.0% 100.0% 100.0% source: survey data table 6 indicates the relationship between child labour status and household head of the children. out of the total children, about 19, 277 and 4 had their mother, father and other as their household head. comparing the categories based on household head, all 19 (100%) were in child labour when the mother was household head, 44% in child labour compared 56% for the household having father as a household head. this difference could be explained due to the fact that where mothers were household head, they might have severe financial problems due to mothers being widow. nabeela begum, javed iqbal & hina 56 table 6: relationship between child labour and household head relationship with household head total chi square value mother father other outside of child labour count 0 156 4 160 26.20*** % within child status .0% 97.5% 2.5% 100.0% % within relationship with hh .0% 56.3% 100.0% 53.3% in child labour count 19 121 0 140 % within child status 13.6% 86.4% .0% 100.0% % within relationship with hh 100.0% 43.7% .0% 46.7% total count 19 277 4 300 % within child status 6.3% 92.3% 1.3% 100.0% % within relationship with hh 100.0% 100.0% 100.0% 100.0% source: survey data table 7 shows the relationship between child labour status and parent's education. it is generally believed that parents education plays a huge role in determining the child labour status. it can be seen that where the parent's education was high, then the lower percentage was engaged in child labour compared to the other categories. the chi square value is 35.16 which shows that the child labour status is significantly different for the various categories of the parents’ education. table 7: relationship between child labour and parent's education parent's education (years) total chi square value <5 5-10 10 -16 outside of child labour count 30 86 44 160 % within child status 18.8% 53.8% 27.5% 100.0% 35.16*** % within education 31.9% 56.6% 81.5% 53.3% in child labour count 64 66 10 140 % within child status 45.7% 47.1% 7.1% 100.0% % within education 68.1% 43.4% 18.5% 46.7% total count 94 152 54 300 % within child status 31.3% 50.7% 18.0% 100.0% % within education 100.0% 100.0% 100.0% 100.0% source: survey data table 8 indicates a relationship between the family size and child labour status. a large number of family sizes could lead to more child labour. it shows that as the family size grew up, so was the child labour. also, the difference in child labour status based on the size of family was statistically significant as well. journal of applied economics and business studies, volume. 2, issue 1 (2018) 47-60 https://doi.org/10.34260/jaebs.215 57 table 8: relationship between child status & family size family size (numbers) total chi square value up to 5 6-10 > 10 outside of child labour 74 86 0 160 % within child status 46.2% 53.8% .0% 100.0% % within family size 64.3% 47.5% .0% 53.3% 12.64*** in child labour 41 95 4 140 % within child status 29.3% 67.9% 2.9% 100.0% % within family size 35.7% 52.5% 100.0% 46.7% total count 115 181 4 300 % within child status 38.3% 60.3% 1.3% 100.0% % within family status 100.0% 100.0% 100.0% 100.0% source: survey data the table 9 exhibits a relationship between child status and total income of house hold income groups. the table shows that as household income grew up, the child labour fell down. it is indicated that a higher percentage (58%) of children were in labour force for those households whose family income was less than rs. 15,000. it can be concluded that a higher number of children are in child labour due to low level of household income. our results are in conformity with that of kondylis and manacorda (2012) who also found that less educated parents who did not have suitable jobs, and therefore were in poverty, engaged their children in child labour. similar, results were found in the studies of abrar and ghouri (2010), mahmood et.al (2005) and khan and ejaz (2003) and ray (1999). table 9: relationship between child labour and household income household income (rs.) total chi square value <15,000 15,000 30,000 >30,000 outside of child labour count 100 50 10 160 % within child status 62.5% 31.2% 6.2% 100.0% % within total income 42.6% 90.9% 100.0% 53.3% 50.92*** in child labour count 135 5 0 140 % within child status 96.4% 3.6% .0% 100.0% % within total income 57.4% 9.1% .0% 46.7% total count 235 55 10 300 % within child status 78.3% 18.3% 3.3% 100.0% % within total income 100.0% 100.0% 100.0% 100.0% source: survey data nabeela begum, javed iqbal & hina 58 table 10 shows the relationship between child labour status and the residence status of the households. living in rental houses indicates an additional expense and an indicative of low level of income. it is shown that about 89% children lived in rental house or had no assets compared to 17% children who owned a house. although, the difference was not found to be statistically significant. table10: relationship between child labour and residential status assets total chi square value rent house own house outside of child labour count 14 146 160 % within child status 8.8% 91.2% 100.0% % within assets 11.3% 83.0% 53.3% 1.501 in child labour count 110 30 140 % within child status 78.6% 21.4% 100.0% % within assets 88.7% 17.0% 46.7% total count 124 176 300 % within child status 41.3% 58.7% 100.0% % within assets 100.0% 100.0% 100.0% source: survey data the discussions in the preceding paragraphs indicate that there are many socioeconomic factors that are responsible for the child labour. they were discussed in isolation from the others factors through bi-variate statistics. the following section presents the results together in a more formal econometrics model. 4.2. child labour determinants the estimated results from the logit model are presented in table 11. age of the children has been found to be positive and significantly relating with the log of the odds of the participation compared to not participation. as the children were growing up, they tended to be more involved in child labour. this is in conformity with the results found in like kondylis and manacorda (2012) where they also found a positive relation of child labour and child age. as expected, the relationship between child labour and child education is negative. those children who went to school, they had a lower probability of working compared to those who did not attend school. similarly, parent's education was also found to negatively affect the participation in child labour albeit insignificantly. similar relationship was also found in (mahmood et al., 2005; bhalotra & tzannato 2003). the relationship between family size and child labour is found to be positive although insignificant. generally, as the family size increases, so as the expenditures and need for more income. therefore, it is expected that more children from such families are engaged journal of applied economics and business studies, volume. 2, issue 1 (2018) 47-60 https://doi.org/10.34260/jaebs.215 59 in child labour. a positive relationship between family size and child labour was also found in kondylis and manacorda (2012). the probability of child labour was more with the household income although with a very low coefficient value which is contrary to our expectations. table 11: logit model estimation coefficient standard error wald sig. exp(b) age 2.661 .555 22.961 .000 14.306 child education -4.112 .855 23.160 .000 .016 parent education -.202 .686 .087 .768 .817 family size .269 .311 .750 .387 1.309 total income household .000 .000 5.665 .017 1.000 constant -10.569 3.971 7.085 .008 .000 r2 0.95 log-likelihood 43.63 5. conclusions the major objective of this study was to find out the socioeconomic determinants of child labour in mardan and nowshera districts of the khyber pakhtunkhwa. we used both descriptive and econometrics techniques to analyze the effect of various factors on children participation in child labour. age of the children was found to be positive and significantly relating with the probability of participation. the relationship between child labour and child education was found to be negative. furthermore, parent's education was also found to negatively affect the participation in child labour albeit insignificantly. the relationship between family size and child labour was found to be positive although insignificant. the probability of child labour was more with the household income although with a very low coefficient value which is contrary to our expectations. as it was found that the age was positively related to the participation in child labour, it is pertinent that families are monitored and provided financial support especially for education. as the age increases, the cost of education also generally increases. therefore, targeted subsidies may provide to the families to encourage them to send their children to school. education of both the children and their parents were found to be significant determinants of the child labour. it had a negative effect on the child labour participation. therefore, it is important that the education of not only the children but also their parents may be improved. family size was also positively related to the child labour. the growth in families has been a major problem in pakistan as there are usually a few members who contribute to the income while the rest are dependents. steps may be taken towards encouraging small family sizes and thereby reducing the child labour. references abrar, n., & ghouri, a. m. (2010). comprehensive study of socio economic dimensions of child labour in new vegetable and fruit market karachi. interdisciplinary journal of contemporary research in business, 2(7), p182. nabeela begum, javed iqbal & hina 60 ahmed, s. s., & haider, w. (2012). determinants of child labour in khyber pakhtunkhwa: an econometric analysis. research journal social sciences 4(2) ,53–67. bhalotra, s. r., & heady, c. (2001). child farm labour: theory and evidence. development economics discussion papers series-suntory and toyota international centre for economics and related disciplines, (24). bhalotra, s. r., & tzannatos, z. (2003). child labour: what have we learnt? world bank, social protection. retrieved from the world bank, 1818 h street, n.w., washington, d.c. 20433 usa. visit the social protection website at http://www.worldbank.org/sp. beegle, k., dehejia, r. h., & gatti, r. (2005). child labour and agricultural shocks. journal of development economics, 81(1), 80-96. edmonds, e. v. (2007). child labour. handbook of development economics. 4: 3607-3709. gandapur, s. b., shahab, p., & ehsan qureshi, m. (2014). analysis of various forms of child labour in d. i. khan city, kpk, pakistan. asian journal of management sciences & education, 3(3), 132-140. khan, a., & ejaz, r. (2003). the determinants of child labour a case study of pakpattan and faisalabad (pakistan). department of economics phd thesis, bahauddin zakariya university, multan. kondylis, f., & manacorda, m. (2012). school proximity and child labour evidence from rural tanzania. journal of human resources, 47(1), 32-63. kazmi, s. m. a. (2015). how does socio-economic factors force children into child labour?: a case study of sahiwal district, punjab, pakistan. sustainable development policy institute. mahmood, s., maann, a. a., tabasam, n., & niazi, s. k. (2005). socio-economic determinants of child labour in automobile and engineering workshops. journal of agriculture & social sciences, 1, 64-65. nawaz, i., & shaheen, a. (2017). situation analysis of child labour in commercial areas of islamabad city. international journal of academic research in business and social sciences, 7(2), 79-98. osment, l. (2014). child labour; the effect on child, causes and remedies to the revolving menace. unpublished lup student papers, lund university libraries, utvk01 20132, department of human geography. http://lup.lub.lu.se/studentpapers/record/4275652 pervez, s., & mukhtar, z. (2014). determinants of child labour: a case study of automobiles workshops, fan industry and ceramics industry in gujarat. j. asian dev. stud, 3(3), 56-63. ray, r. (1999). how child labour and child schooling interact with adult labour. the world bank. siddiqi, f., & patrinos, h. a. (1995). child labour: issues, causes and interventions. education and social policy department, human resources development and operations policy, the world bank. zaidi, n., javed, n., & khan, m. s. (2013). child labour in pakistan-a tip of an iceberg. journal of postgraduate medical institute (peshawar-pakistan), 27(3). http://lup.lub.lu.se/student-papers/record/4275652 http://lup.lub.lu.se/student-papers/record/4275652 journal of applied economics and business studies, volume. 2, issue 1 (2018) 61-74 https://doi.org/10.34260/jaebs.216 61 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online): 2663-693x impact of corporate social responsibility guidelines 2013 on the extent of the corporate social responsibility activities and disclosures in pakistan maqsood hayat1*, shehzad khan21 1phd scholar, yunnan university of finance and economics, kunming, china 2assistant professor, institute of business studies and leadership, abdul wali khan university mardan. abstract the main objective of this study was to examine the impact of csr guidelines 2013 on the level of corporate social responsibilities (csr) activities and their disclosure in pakistan. this study analyzed the voluntary disclosure guidelines impact on various stakeholders of the top companies of the year listed in pakistan’s stock exchange for the five years (2011-2015). it is found that the introduction of these guidelines 2013 had a positive impact on the overall level of corporate social responsibilities disclosures (csrd). it was also observed that the overall trends in the level of csrd increased gradually within the sample period. there need to be continuous a requirement through regulations and local pressures on the firms for engaging in ethical business practices and to disclose that information to government organizations and general public. corporations can gain both economically and ethically when they take csr as marketing and public relation opportunity. finally, corporations can give itself edge by distinguishing its operations from others and therefore gain competitive advantage. keywords csr, csr guidelines, csr disclosure jel classification m14, m10, m19 copyright © 2018 saebr all rights reserved 1. introduction corporate social responsibility (csr) has always been in practice in various forms. however, the modern structure of csr originated in the early decades of the 20th century. thereafter, it got extensive attention in corporate practices and academic research. international bodies have been striving hard to unify the reporting format for all the businesses across the globe. misani (2010) examined whether firms adopt convergence or divergence in practicing the csr and argued that firms are expected to incorporate csr in their business model due to pressures from various stakeholders such as its customers, 1*corresponding author: drhayat2019@gmail.com maqsood hayat, & shehzad khan 62 governments and non-governmental organizations. firms in developing countries and emerging economies are also forced to follow similar guidelines with respect to csr as their counterparts in developed countries due to globalization and free trade agreements. socially and environmentally sensitive stakeholders expect more from business communities to be both socially and environmentally responsible besides merely profitable. corporate social responsibility disclosure (csrd) is getting mandatory in many countries, and thus corporations implement csr activities which ultimately increase the corporate value of the firms in the context of local priorities of the society (manchiraju & rajgopal, 2015). corporate social responsibility is an amalgam of various activities such as wellbeing of the general public, environmental and biodiversity sustainability, and improvement in the wellbeing of both the owners and other stockholders. therefore, csr focuses on allowing the firms to operate in such a way to not only generate profit for the owners but to also promote overall prosperity. corporate social responsibility enforces the corporation to behave morally inside the organization premises and act like a good citizen outside the organization boundary. large multinationals such as mcdonald and marks & spencer have adopted comprehensive csr framework that takes into account the economic, social, environmental and other benefits both for the business and stockholders. enterprises vary in terms of the scale of the adoption of the csr as some only adopts a limited set of activities compared to others which adopt a comprehensive one. most of the enterprises disclose their csr reports via different channels to target a wide cluster of stakeholders. information in the disclosure reports with respect to csr activities determine the characteristics and nature of the organizations. corporations are pressurized by stakeholders to be responsible and accountable not only for their economic liability but also for social and environmental accountability. the benefits range of csr are widespread which may include transparency in hiring, proper training, healthy and secure working environment for employees. similarly, the consumer related activities such as provision of safe and reliable products, disclosing the details about the product ingredients and manufacturing process. therefore, it is crucial for enterprises to streamline their corporate strategy and specify the corporate activities that are sensitive towards the local norms, social values, religious beliefs, and organizational strength & features. corporate benefits achieved through adopting csr activities have already been examined by numerous authors in their research studies (e.g. kurucz et al, 2008; mullerat, 2010; perry and towers, 2013; porter and kramer, 2006). corporate social responsibility is voluntarily in some parts of the world while legally required and enforced in other parts of the world. however, corporate social responsibility and corporate accountability need to be differentiated to avoid unnecessary demands from stakeholders and let the organizations to operate within their jurisdictions (cash, 2012). the main drivers behind the csr activities are not only external forces in the shape of journal of applied economics and business studies, volume. 2, issue 1 (2018) 61-74 https://doi.org/10.34260/jaebs.216 63 governmental regulations and stakeholder’s activism but also corporate self-regulation. other forces that pressurize the corporations stem from society, non-profitable organizations, media, religious leaders, suppliers, buyers and employees. campbell (2007) observed that corporations are more inclined to behave in a socially responsible way when their activities are strictly observed by many stakeholders. the issue of csr in pakistan was first time highlighted when the use of child labour and sweatshops were found by the international humanitarian organization in local sports and carpet industries of pakistan. this leads to the multilateral consortium called the “pakistan compliance initiative” in 2003 to monitor the business activities and eliminate the unethical and illegal use of workforce in pakistan. the initiative to implement csr activities in corporate operations was taken first time by a general order from the security exchange commission of pakistan (secp) in 2009. unfortunately, due to the unavailability of specific reporting csr format, the corporations took it as a tool of public relation and corporate promotion. in true sense, there are only a few laws and regulations regarding csr in pakistan, issued by secp and state bank of pakistan (sbp). apart from these orders and guidelines, there are other laws provided under the constitution of pakistan which have some relevancy to csr regulations. few of them are corporate laws, environmental laws; labour laws and consumer protection laws. the csr voluntary guidelines 2013 were issued from secp, with the basic purpose to streamline the corporate activities and measure the corporate efficiency regarding all stakeholders. therefore, enterprises operating in pakistan and especially that are listed in any stock exchange of pakistan should seriously consider the importance of these guidelines to avoid any discrepancy for violation of the rules and regulations of the country. the major objectives of this study is to examine the effectiveness of the voluntary csr guidelines 2013 and its impact on the level of csrd practiced by the firms listed in the pakistan stock exchange. this study is an attempt to examine the pre and post effect of the various dimensions of the local csr guidelines. 2. literature review clark (1939) was the pioneer of the modern csr which was then refined by bowen (1953). although the initial empirical research was done in 1976 by bowman and haier (1976), however, this idea was revitalized by carroll (1991) by introducing four dynamics of csr in hierarchal order such as economic, legal, ethical, philanthropic which is called the csr pyramid. later on, elkington (1997) introduced the concept of the triple bottom line (tbl) approach by integrating three dynamics of csr outputs also called the 3ps model. developing countries due to lack of awareness and expertise in the field of csr and unwillingness of corporations to collaborate with the local community do not follow specific csr guidelines (jamali and miurshak, 2007). these corporations are adopting csr activities in their corporate strategies due to stakeholder’s activism. these csr activities maqsood hayat, & shehzad khan 64 are not helpful if not communicated to stakeholders properly. therefore, the corporation should have a strict observation and enough knowledge to classify their stakeholders needs and to prioritize them to achieve corporate benefits. to reap the maximum benefits from the csr investment, the corporations should have a proper strategy to implement csr in business operations. most of the corporations are unwilling to disclose every aspect of their corporate activities to a wide range of stakeholders in a more transparent way. rawlins (2008) observed that corporate transparency is different than just disclosing information to the public. full and neutral information disclosure can lead to corporate credibility and gain the public confidence. martinson (1996) reported that the basic meaning of transparency is to truthfully communicate the reality of a particular subject-incident-event to the stakeholders. corporations need to release most information related to their stakeholders. many organizations work in close collaboration with corporate community and stakeholders to formulate a reporting format which is acceptable for all type of corporations and in line with stakeholders’ demands. cavanagh (2004) reported that united nations declaration of human rights (1948) and the international labour organization's tripartite declaration (1977) are important for the csr activities too. in addition, other reporting formats such as socially responsible investing groups, dow jones sustainability index, global 100, united nations global compact, the global reporting initiatives, kinder lindenberg and domini are not only useful for investors but also for other stakeholders. sometimes it is difficult for small businesses in developing countries to understand the strict csr criteria, which ultimately discourage the domestic firms to incorporate and report such a complicated guideline in their business activities (perera, 2008). therefore, it is crucial to develop a universal legal agenda for all corporations to design a specific csr format, which is acceptable enterprises globally. other scholars are in the view that csr expectations vary from culture to culture due to their formal and informal regulations, i.e. societal, economic and political circumstances (carroll, 2004; cummings & guthrie, 2007; matten & moon, 2008). local governments in developing countries are taking initiatives to introduce the guidelines with a view to encourage businesses to focus more on social and environmental development (campbell, 2007). several studies have been conducted on the importance of csr across the globe, especially in the developed countries mainly in the us and europe. many researchers (clarkson, 1995; samiee & athanassiou, 1998) stated that there is a need for universal guidelines for csr. wood (2010) criticized that most of research relating to csr is undertaken in developed countries. jamali and miurshak (2007) reported that due to lack of awareness, corporations in developing countries generally do not practice csr. therefore, it is important to study the csr in the developing countries context as well. journal of applied economics and business studies, volume. 2, issue 1 (2018) 61-74 https://doi.org/10.34260/jaebs.216 65 pakistan is faced with many social, economic, political and environmental issues. like other developing countries, pakistan lacks resources to fulfil the basic needs of the general public and to build infrastructure. few studies have been conducted on csr in pakistan due to political uncertainty, widespread corruption, immoral corporate culture, human rights abuse, unemployment, and environmental problems (jhatial et al, 2014). the widespread differences in norms and values are also the biggest hurdles in formulating the csr strategy in pakistan. most of the time, enterprises are reluctant to resolve the local issues because of complexity and sensitivity. naeem & welford (2009) investigated that the csr dimensions in pakistan are not aligned with the standards of the global compact principles (gcp). in pakistan, the concept of csr has been promulgated by multinational corporations (mncs) like nestle and unilever. similarly, local enterprises like various banks provide assistance in grooming talent, and other donations in health and education sectors. lund-thomsen & nadvi (2010) studied the impact of international and domestic regulations in the leather industry of pakistan, while hussain-khaliq (2004) investigated the sports industry of pakistan and explored the exploitation of the child labour in the football industry. there is also a growing awareness of the importance of environmental safety and protection in pakistan. many organizations run campaigns regarding the environment and health every year. some advocacy associations such as the national forum of environment and health (nfeh) presents annual excellence awards for the best performance in environment, health, education, and socio-economic development. pasha and liesivuori (2003) found that most of the company workforce are uneducated and not trained in occupational health and safety (ohs) procedures. similarly, jeswani et al (2008) reported that 75 % of the pakistani enterprises face many hurdles to adopt and implement csr in their corporate strategies. these hurdles are lack of awareness, insufficient financial resources, the absence of regulatory framework and policies, unavailability of technology and shortage of expertise to deal with complex environmental issues. it is clear from the above discussion that all the csr guidelines have a wide scope and cover almost all the csr theories. these guidelines relate to shareholders’ theory as profit on investment and firm value is addressed. similarly, the csr guidelines are in conformity with the stakeholder’s theory, carroll csr pyramid, and legitimacy theory too. 3. research methods csr related data were collected from the annual reports published by the “top 25 companies of the year” for five years (2011-2015). individual list was prepared for all enterprises and only those enterprises were selected which had corporate websites and published annual reports for the sample period. a total sample of 51 companies were selected for this study. maqsood hayat, & shehzad khan 66 different techniques have been used to measure the csr. some of the previous studies used corporate internal information for measuring the csr level such as annual reports and social and sustainability reports (hart and ahuja, 1996; muoghalu et al, 1990). these indicators were compared with the international standardized reporting agencies format such as kld, gri, ilo, iso and looked for availability or unavailability of various measures therein. this study calculates the level of csrd by measuring the economic, philanthropic, legal, ethical, investor, employee, customer, community, health, education, sports, environment and government csr indicators with the help of content analysis (ca hereinafter). ca was used by a number of researchers in their social and environmental reporting for classifying the social reporting dimensions of disclosing corporations (e.g. adams et al, 1995; adams and harte, 1999; milne and adler, 1999; beck et al, 2010). berelson (1952) describes that ca is the research method for the objective, systematic and quantitative description of the manifest content of the communications. this ca concept has been further defined by guthrie et al (2004) as a technique for gathering data contained in the annual reports and it involves codifying qualitative and quantitative information into pre-defined categories in order to derive patterns in the presentation and reporting of information. every document was scanned manually line by line and word by word to assure the authenticity and quality of the desired extracted data. the text was scanned against the predefined parameters and was codified with “1” per sentence, as one sentence is equalized to “1”. this type of technique was also adopted by other researchers (e.g., eng and mak, 2003; haniffa and cooke, 2002). descriptive statistics are used to define, summarize and measure the elementary structures of the collected data and sample in the research study. paired sample t-test, also known as the dependent sample t-test, is used to examine the differences in mean values of the two sets. paired sample t-test is conducted in a context of the unequal variance in both samples. in this way, the pre and post effect of the csr guidelines 2013 was examined in a given sample of the study to certify the projected proposition, with an assumption to be rejected or accepted. 4. results and discussion the statistical summary of various stakeholders and the layers of csr’s pyramid is presented in table 1. investor is the first group of internal stakeholders, which has an average value of 11.93 with a stand deviation of 17.67. the employees have mean value 30.89. further, the customer, community, and environmental disclosure have the average values of 11.79, 10.17, 21.85 respectively. the average values of the economic layer, philanthropic layer, legal layer, and ethical layer were 69.92, 38.71, 64.23, and 7.94 respectively. the results show in the table 1 that the economic aspect of the pyramid was the most important of the csr activities. the firms got to be profitable before it can look for other csr journal of applied economics and business studies, volume. 2, issue 1 (2018) 61-74 https://doi.org/10.34260/jaebs.216 67 activities. further, working under the legal framework was also found to be very important activity of the csr. table 1: descriptive statistics of the relevant csr categories mean median std. dev min max csr 180.81 135.90 155.82 0.00 869.00 investor 16.65 11.93 17.67 0.00 91.75 employee 30.89 21.21 31.52 0.00 185.20 customer 23.68 11.79 28.55 0.00 139.50 community 20.20 10.17 33.66 0.00 280.00 environment 21.85 9.33 37.36 0.00 240.50 economic 69.92 53.66 54.36 0.00 256.19 philanthropic 38.71 20.46 47.66 0.00 312.00 legal 64.23 40.34 73.22 0.00 436.81 ethical 7.94 5.50 8.25 0.00 36.50 figure 1 show csr disclosures for the given period of the study (2011-2015). these dimensions have been divided into 13 different categories to calculate the level of csrd of the top 25 companies operating in pakistan. it is indicated that the total 13 dimensions were measured in 100 percent terms. the workforce disclosure (16.37%) seemed to the most important one followed by financial disclosure (16.37%), product disclosure (13.10%), environment (12.09%), community (11.17%), investors (9.21%), ethics (4.39 %), compliance (4.22 %), education/art (4.19 %), health (2.88 %), donation (2.51 %), government (2.13 %) and sports (0.66 %). figure 1: csr dimensions (in %) figure 2 indicates the year-wise total csrd related to different stakeholders (i.e. investors, employees, customers, community and environment) and pyramid’s layers (i.e. economic, philanthropic, legal and ethical). it shows an increasing trend over time. maqsood hayat, & shehzad khan 68 figure 2: proportionate (in %) of csrd indicators (year-wise) figure 3 shows the year wise fluctuation in the level of total csrd related to different stakeholders (i.e. investors, employees, customers, community and environment) and pyramid’s layers (i.e. economic, philanthropic, legal and ethical). figure 3: yearly changes in csrd indicators figure 4 and table 2 show the effect of voluntary csr guidelines 2013 on the level of csrd. both the pre and post csr guidelines are compared, and the change is calculated. 0.00 5.00 10.00 15.00 20.00 25.00 30.00 2011 2012 2013 2014 2015 -15 -10 -5 0 5 10 15 20 25 30 35 2012 2013 2014 2015 journal of applied economics and business studies, volume. 2, issue 1 (2018) 61-74 https://doi.org/10.34260/jaebs.216 69 table 2: impact of “voluntarily csr guidelines 2013” (pre-2013 v/s post-2013) pre-13 post-13 change total csr 34.3 45.2 31.7 financial 35.0 45.4 29.7 customer 36.0 42.4 17.8 investor 31.0 49.1 58.6 donation 28.8 50.5 75.2 community 33.9 45.8 35.3 health 39.1 42.0 7.4 education/art 36.4 43.3 19.0 sports 30.9 53.4 72.8 environment 35.5 44.8 26.2 compliance 35.0 46.0 31.3 govt 33.7 46.0 36.5 employee 33.6 44.4 32.3 ethics 33.3 46.2 38.9 figure 4: impact of “voluntarily csr guidelines 2013” (pre-2013 v/s post-2013) results above shows the pre and post csr 2013 guidelines changes in absolute terms. following in table 3, we present the comparison of the csr results statistically using the pair t-test. the mean values for all csr categories indicate progressively increasing trend after the implementation of csr guidelines in 2013. the mean value for the total csrd increased to 204.42 post 2013 from the pre 2013 of 155.25, and this difference is also statistically significant at 1 per cent level. the observed increase in the level of csrd over the sample period (2011-2015) can be linked with the stakeholder and institutional theories. gao et al (2005) conducted a study in hong kong and examined that the level of csrd increasing between 1993 and 1997. similarly, mahadeo et al (2011) recognized the maqsood hayat, & shehzad khan 70 increasing trends of the csrd over a period of four years (2004-2007) in mauritius. haniffa and cooke (2005) conducted a study in malaysia and examined that the level of csrd increased in the sample period (1996-2002). in another study in the context of malaysia, esa and ghazali (2012) investigated that the degree of the csrd increased between 2005 and 2007 in the annual reports of government-linked corporations. haji (2013) also confirmed the increased level of csrd in malaysian context in the annual reports of malaysian listed firms from 2006 to 2009. the same results were also observed by ratanajongkol et al (2006) in the context of thailand. similarly, the mean values for the entire stakeholder group increased after the announcement of the csr guidelines in 2013. the investor related disclosure also increased to mean value of 20.44 in post 2013 compared to 12.89 in pre-2013 period. employee-related disclosure, and the customer and community related csrd were higher in the post-2013 era compared to that of the pre-2013 and also this difference were significant at 5 per cent level. environment-related csrd is the last indicator of the stakeholder group which was higher in post 2013 era however this difference has not been statistically significant. regarding carroll’s pyramid, the t-test reveals that all the four layers are statistically significant. the mean values in pre-2013 for economic, philanthropic, legal and ethical layers are 60.11, 33.26, 55.27 and 6.60, which increased gradually to 79.11, 44.15, 71.99 and 9.17 in post-2013 respectively. table 3: pre-post effect of csr guidelines 2013 dimension pre/post mean observations d.f t -stat p-value csr pre-2013 155.25 96 95 -4.29 *** post-2013 204.42 96 investor pre-2013 12.89 96 95 -4.18 *** post-2013 20.44 96 employee pre-2013 25.94 96 95 -3.47 *** post-2013 34.31 96 customer pre-2013 21.32 96 95 -2.22 ** post-2013 25.09 96 community pre-2013 17.1 96 95 -2.31 ** post-2013 23.18 96 environment pre-2013 19.41 96 95 -1.72 post-2013 24.49 96 economic pre-2013 60.11 96 95 -4.91 *** post-2013 79.11 96 philanthropic pre-2013 33.26 96 95 -2.75 *** post-2013 44.15 96 legal pre-2013 55.27 96 95 -2.98 *** post-2013 71.99 96 ethical pre-2013 6.6 96 95 -4.41 *** post-2013 9.17 96 ***, **, and * indicates significant level at 1%, 5% and 10% respectively journal of applied economics and business studies, volume. 2, issue 1 (2018) 61-74 https://doi.org/10.34260/jaebs.216 71 5. conclusion the main objective of this study was to investigate the impact of the voluntarily csr guidelines 2013 on the corporate social responsibility disclosures (csrd). corporate social responsibility (csr) related data were collected from the annual reports published by the “top 25 companies of the year” for five years (2011-2015). individual list was prepared for all enterprises and only those enterprises were selected which had corporate websites and published annual reports for the sample period. a total sample of 51 companies were selected for this study. comparison of the csrd was done both in absolute terms and statistically through using the comparison of the means tests. the total level of csrd increased over the period of time in all five years, with an increase of 43% from the year of inception (2011). the results of the study also suggest that the level of overall csrd increased with the introduction of the “csr guidelines 2013”. similarly, the csrd related to community, environment, economic, philanthropic, legal and ethical indicators increased in all five years too. this can be attributed to the gradual increase in the awareness of the stakeholders and their sensitivity towards ethical business and the importance of reporting csr activities to avoid any discrepancy for violation of the rules and regulations in the country. similarly, the impact was also tested for various groups of stakeholders and layers of csr’s pyramid. it was found that the level of csrd for all indicators was recorded positive and statistically significant in the post-2013 era except the environmental dimension. there need to be continuous a requirement through regulations and local pressures on the firms for engaging in ethical business practices and to disclose that information to government organizations and general public. corporations can take csr as marketing and public relation opportunity. therefore, the corporations need to formulate, strategize and implement these guidelines in corporate operation to achieve a distinguished position among the competitors, even if it is not obligatory by the government. the government should encourage the corporations to report beyond the boundary of the obligatory limits to satisfy the demands of the overall stakeholder, especially the active stakeholder groups. references adams, c. a., hill, w. y., & roberts, c. b. (1995). environmental, employee and ethical reporting in europe. london: certified accountants educational trust. adams, c., & harte, g. (1999). towards corporate accountability for equal opportunities performance. acca occasional research paper no. 26. london: acca. beck, a. c., campbell, d., & shrives, p. j. (2010). content analysis in environmental reporting research: enrichment and rehearsal of the method in a british–german context. the british accounting review, 42(3), 207-222. berelson, b. (1952). content analysis in communication research. bowen, h.r. social responsibilities of businessmen (1953). new york: harper & brothers. maqsood hayat, & shehzad khan 72 bowman, e. h., & haire, m. (1976). social impact disclosure and corporate annual reports. accounting, organizations and society, 1(1), 11-21. campbell, j. l. (2007). why would corporations behave in socially responsible ways? an institutional theory of corporate social responsibility. academy of management review, 32(3), 946-967. carroll, a. b. (1991). the pyramid of corporate social responsibility: toward the moral management of organizational stakeholders. business horizons, 34(4), 39-48. carroll, a. b. (2004). managing ethically with global stakeholders: a present and future challenge. academy of management perspectives, 18(2), 114-120. cash, a. c. (2012). corporate social responsibility and petroleum development in subsaharan africa: the case of chad. resources policy, 37(2), 144-151. cavanagh, g. f. (2004). global business ethics: regulation, code, or self-restraint. business ethics quarterly, 14(4), 625-642. chalmers, k., & godfrey, j. m. (2004). reputation costs: the impetus for voluntary derivative financial instrument reporting. accounting, organizations and society, 29(2), 95-125. clark, j.m. social control of business (1939). 2nd ed. new york: mcgraw hill. clarkson, m. e. (1995). a stakeholder framework for analyzing and evaluating corporate social performance. academy of management review, 20(1), 92-117. cummings, l. s., & guthrie, j. (2007). managerial attitudes toward stakeholder salience within selected western pacific-rim economies. journal of asia-pacific business, 8(1), 7-29. elkington, j. (1997). partnerships from cannibals with forks: the triple bottom line of 21st‐ century business. environmental quality management, 8(1), 37-51. eng, l. l., & mak, y. t. (2003). corporate governance and voluntary disclosure. journal of accounting and public policy, 22(4), 325-345. esa, e., & anum mohd ghazali, n. (2012). corporate social responsibility and corporate governance in malaysian government-linked companies. corporate governance: the international journal of business in society, 12(3), 292-305. gao, s. s., heravi, s., & xiao, j. z. (2005, june). determinants of corporate social and environmental reporting in hong kong: a research note. in accounting forum (vol. 29, no. 2, pp. 233-242). taylor & francis. guthrie, j., petty, r., yongvanich, k., & ricceri, f. (2004). using content analysis as a research method to inquire into intellectual capital reporting. journal of intellectual capital, 5(2), 282-293. ahmed haji, a. (2013). corporate social responsibility disclosures over time: evidence from malaysia. managerial auditing journal, 28(7), 647-676. haniffa, r. m., & cooke, t. e. (2005). the impact of culture and governance on corporate social reporting. journal of accounting and public policy, 24(5), 391-430. haniffa, r. m., & cooke, t. e. (2002). culture, corporate governance and disclosure in malaysian corporations. abacus, 38(3), 317-349. journal of applied economics and business studies, volume. 2, issue 1 (2018) 61-74 https://doi.org/10.34260/jaebs.216 73 hart, s. l., & ahuja, g. (1996). does it pay to be green? an empirical examination of the relationship between emission reduction and firm performance. business strategy and the environment, 5(1), 30-37. hussain-khaliq, s. (2004). eliminating child labour from the sialkot soccer ball industry: two industry-led approaches. journal of corporate citizenship, (13), 101-107. jamali, d., & mirshak, r. (2007). corporate social responsibility (csr): theory and practice in a developing country context. journal of business ethics, 72(3), 243-262. jeswani, h. k., wehrmeyer, w., & mulugetta, y. (2008). how warm is the corporate response to climate change? evidence from pakistan and the uk. business strategy and the environment, 17(1), 46-60. jhatial, a. a., cornelius, n., & wallace, j. (2014). rhetorics and realities of management practices in pakistan: colonial, post-colonial and post-9/11 influences. business history, 56(3), 456-484. 456-484. kinderman, d. (2009). the political economy of corporate responsibility across europe and beyond: 1977-2007. department of political science ithaca, ny, cornell university. kolk, a., & perego, p. (2010). determinants of the adoption of sustainability assurance statements: an international investigation. business strategy and the environment, 19(3), 182-198. lund‐thomsen, p., & nadvi, k. (2010). global value chains, local collective action and corporate social responsibility: a review of empirical evidence. business strategy and the environment, 19(1), 1-13. mahadeo, j. d., oogarah-hanuman, v., & soobaroyen, t. (2011). a longitudinal study of corporate social disclosures in a developing economy. journal of business ethics, 104(4), 545-558. manchiraju, h., & rajgopal, s. (2015). does corporate social responsibility (csr) create shareholder value? exogenous shock-based evidence from the indian companies act 2013. matten, d., & moon, j. (2008). “implicit” and “explicit” csr: a conceptual framework for a comparative understanding of corporate social responsibility. academy of management review, 33(2), 404-424. milne, m. j., & adler, r. w. (1999). exploring the reliability of social and environmental disclosures content analysis. accounting, auditing & accountability journal, 12(2), 237-256. misani, n. (2010). the convergence of corporate social responsibility practices. management research review, 33(7), 734-748. mullerat, r. (2010). international corporate social responsibility: the role of corporations in the economic order of the 21st century. kluwer law international bv. muoghalu, m. i., & robison, h. d. (1990). hazardous waste lawsuits, stockholder returns, and deterrence. southern economic journal, 57(2). naeem, m. a., & welford, r. (2009). a comparative study of corporate social responsibility in bangladesh and pakistan. corporate social responsibility and environmental management, 16(2), 108-122. maqsood hayat, & shehzad khan 74 pasha, t. s., & liesivuori, j. (2003). country profile on occupational safety and health in pakistan (report). international labor organization. retrieved from http://www.ilo. org/safework/countries/asia/pakistan/wcms_187800/lang–en/index.htm. perera, o. (2008). how material is iso 26000 social responsibility to small and mediumsized enterprises? international institute for sustainable development: winnipeg, mb, canada. perry, p., & towers, n. (2013). conceptual framework development: csr implementation in fashion supply chains. international journal of physical distribution & logistics management, 43(5/6), 478-501. kramer, m. p. strategy and society: the link between competitive advantage and corporate social responsibility. harvard business review. ratanajongkol, s., davey, h., & low, m. (2006). corporate social reporting in thailand: the news is all good and increasing. qualitative research in accounting & management, 3(1), 67-83. rawlins, b. (2008). give the emperor a mirror: toward developing a stakeholder measurement of organizational transparency. journal of public relations research, 21(1), 71-99. samiee, s., & athanassiou, n. (1998). international strategy research: cross-cultural methodology implications. journal of business research, 43(2), 79-96. taylor, d., & darus, f. (2006). mandatory disclosure’s impact on voluntary disclosure of proprietary information: evidence from introduction of financial instruments disclosure standard. in international conference on accounting and finance in transition (icaft), university of south australia, (2006, 10-11 april). wood, d. j. (2010). measuring corporate social performance: a review. international journal of management reviews, 12(1), 50-84. journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 15 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x suffering silence while exposed to workplace bullying: the role of psychological contract violation, benevolent behavior and positive psychological capital namra jamshaid1* & sadia arshad2 1research scholar at kinnaird college for women university, lahore, pakistan 2 lecturer at kinnaird college for women university, lahore, pakistan abstract becoming impediment to organizational functioning in several ways, the prevalence of workplace bullying costs much to organizations. as in the current transition phase of pakistan, the intentions to leave doesn’t manifest in actual turnover rates. hence the current study is conducted to analyze the passive coping strategies of employees in the face of workplace bullying. it is theorized that the relationship is mediated by psychological contract violation. moreover, it is predicted that the process of mediation is stronger for individuals who report high levels of benevolent behavior and perceives psychological capital to be low. data is collected from 359 young doctors and nurses of three government administered hospitals. results indicate a significant bullying-silence relationship where psychological contract violation plays a role of partial mediator. moreover, benevolence and psycap are powerful moderators to alter the already established relationship. conclusions of the current study are further elaborated in terms of their practical contribution and future directions. workplace bullying is an organizational reality. hence efforts to make an entire bullying-free environment is next to impossible. so, in addition to make an effort in ending up this maltreatment, managers must limit its consequences by understanding its dynamics. reduce the bullying culture and save precious resources i.e. potential employees. keywords workplace bullying, psychological contract violation (pcv), benevolent behavior, employee silence, positive psychological capital (psycap) 1. introduction a consequential concern faced by contemporary working environments is the event of maltreatment in work settings. representing a major hr issue provided its antagonist influence on career outcomes, productivity of targets or even the witness [(salin, 2013) * namrajamshaid95@gmail.com namra jamshaid & sadia arshad 16 validates the impact of workplace bullying beyond the perpetrator-target relationship, whereas salin and notelaers (2020) observes a significant difference in work-related attitudes of bystanders, who witness uncivil behavior and subsequently develop the feelings of betrayal and anger. hence, it supports the notion of negatively impacting witnesses], the evolvement of the concept has reached a degree where it has been identified and recognized as one of the hot topics in the field of management and organizational psychology of 21st century. following the issue, a major challenge remains for the hr professionals in the development of strategies, procedures and processes required in resolving such profoundly perplexing and strenuous events where management of its consequences which perhaps takes the form of bad morale and turnover of the victim stands in question. (fox & cowan, 2015). while taking various forms, a processual and escalating exposure to predominantly psychological mistreatment makes target tough to retaliate in kind normally due to the power disparity amidst the perpetrator and victim. (einarsen et al., 2011). as reported, 97% individuals experience some kind of psychological mistreatment in their entire work experience (fox & stallworth, 2005). as establishing cut off points between occasional adverse encounters and actual bullying is hard, academic literature, surveys and reports all indicate the rise in the rate of occurrence of workplace bullying in recent years (e.g., 27%, workplace bullying institute survey, 2014) and hence its pervasiveness makes it a global phenomenon to warrant researcher and practitioner attention. the rate of global unemployment, according to international labor office ((ilo), 2019) hit a historic high of 170 million, as well its increase is anticipated in the near future. chung and van oorschot (2010) proves that the economic and labor market conditions of countries influence the individual’s perception of uncertainties (turnover intentions) and higher redundancy rates enhance the perceived risk of job insecurity (in terms of losing) and hence diminishing the probability of finding another valuable job. entailing the detrimental effects of unfavorable episodes on the target’ self confidence, selfimage and on job satisfaction, it is understandable that the one exposed might have decreased commitment and engagement and considers leaving his job, (mccormack et al., 2009) whereas studies already established a link between exposure to workplace bullying and intentions to leave (jiang et al., 2012), but such relationships hold partial confirmation within the present time research. majority of those expressing their intentions to make a turnover do not leave immediately for any exogenous or endogenous reason (hom et al., 2012). when intentions are not manifested in terms of actual employee’s turnover, it hence gives an ascent to various models conjecturing journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 17 numerous processes to explain this discrepancy resulting in other form of deviant behaviors. hence the concept of employee silencea deliberate practice to withhold information, ideas, opinions, and concerns that might be crucial for organizational success (rai & agarwal, 2018), is directed to study as a recent and utmost substantial passive coping strategy by employees in response to workplace mistreatment. as it encounters high costs for organizations, understanding its mechanism is imperatively crucial. as the implementation of strategies by organizations like massive layoffs and reduction of labor costs are changing the traditional employment arrangements, this study proposes a detrimental change in employees’ psychological contracts (the feeling of insecurity of an individual over one’s job (in any economic downturn) leads towards the apprehension of a compromised psychological contract (lack of reciprocity) (piccoli & de witte, 2015)). this economic crisis possesses new challenges for companies, and such downturn aggravates the feelings of insecurity leading to the perceptions of lack of reciprocity, compromising the psychological contracts of employees. as pakistan’s economy is presently passing through the transition phase, rapidly increasing populace, high inflation and swelling unemployment, patterns of employment are rapidly shifting from long-term to temporary, casual and piecerate contracts (ghayur, 2009; costa & neves, 2017), simultaneously giving a pertinent context for studying psychological contract violation. such violations have become a recurrent experience of many employees provided the gigantic change in working patterns. hence understanding the reactions in response to these violations is contemporary concern for academics and employers both. particularly in the health care settings, meeting budgets and quality targets upshots in pressure situations instigating leaders to bully their subordinates, and hence benevolent employees are expected to be exceedingly affected by bullying experience and perceived violations plays a significant role in amplifying their withdrawal behavior. being a collectivistic country, altruism and trust is considered implicit in all the social relationships including the relationship of employer-employee. such cultural profile is rule oriented and highly supports the inequalities of power and wealth, and to pursue any issue or to report mistreatment on the employer’s part is considered as a disloyal and unacceptable behavior by authority holders (khan et al., 2013) even in the face of violation of one’s psychological contract. this picture regarding the relationship’s potential nature best sets in public services (islam, 2004) where negative instead of generalized reciprocal norms are expected. therefore in the health care settings, being benevolent in nature is as important as anything, so studying such namra jamshaid & sadia arshad 18 individual’s reactions to mistreatment is a fascinating call to answer. hence this research offers notions of reciprocity in a significant institutional and cultural context. as the study also requires to fix what is already broken, the development of positive strengths within individuals (psycap, luthans et al., 2007) may help in improving tools to cope with this entire negativity, hence characterizing as a potential safeguard and a powerful buffer. combined with a progressively general move in the course of past 10 years towards the resource model of stress, it perhaps is timely to look for the role of resources (hope, resilience, optimism, and self-efficacy) rather than deficits in connection with interpersonal mistreatment. the contribution of this research to the existing literature is threefold. first, it broadens the existing knowledge on the deleterious impact of workplace bullying and one of the few attempts to examine employees’ passive coping strategies in response to this negative social behavior at work floor rather than leaving or intending to leave the organization. second, by incorporating psychological contract violation of employees as an intervening mechanism, this study seeks to determine why victims of bullying engage in silence behaviors. third, this study sheds light on two important boundary conditions that constrains the mediated relationship. it is argued that benevolent employees (more affected by workplace bullying), who are highly subjected to violation of their psychological contracts, if contains positive resources (psychological capital) within themselves, can mitigate their tendency to engage in silent behaviors. therefore, this research not only complements the extant literature but also highlights the role of positive strengths within individuals and its impact on bullying outcome relationship. hence, the amalgamation of proposed variables and their relationships remains an important oversight and a first empirical study to the best of authors’ knowledge. the synergetic effect of workplace bullying on employee silence offers a very fascinating area of inquiry in organizational behavior. to address the problem, overall study is broken down into the study objectives as: a) to investigate the general relationship between workplace bullying and employee silence b) to explore the intervening effects of psychological contract violation in the relationship between workplace bullying and employee silence and c) to explore the cross level conditional indirect effects of benevolent behavior and psycap on the workplace bullyingemployee silence relationship via pcv. specifically, the study aims at investigating the bullying-silence relationship via pcv across levels of benevolent behavior and psycap. journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 19 2. literature review 2.1 workplace bullying bullying at work means offending, harassing, socially excluding someone or negatively affecting someone’s work tasks. to apply bullying to a particular activity, interaction or process, it has to occur repeatedly and regularly (e.g., weekly) and over a period of time (e.g., about six months). (einarsen et al., 2011). events where a person being exposed to aggression and negative behaviors at work predominantly affect the psychological state of the target (leymann, 1996) through humiliation, intimidation and/or punishment by the perpetrator (einersen et al., 2020) is the concept’s defining line among variety of forms of interpersonal mistreatment researched till yet mainly aggression, harassment, emotional abuse, incivility (zapf, 2004; hershcovis, 2011). while an intention to harm and negativity in actions is common to workplace bullying or any other interpersonal maltreatment, persistency and power disparity holds the position for two distinguishing features to bullying. persistence of behaviors inappropriate in nature including repetition (occurrence must be on regular basis (frequency), duration (over a time period) and patterning (involvement of variety of behaviors) are the key features to bullying. many acts under maltreatment, relatively common in workplaces, when occur in isolation may be considered as a sign of uncivil behavior (lim & cortina, 2005) but when persisted towards a particular target for a certain duration, causes severe harm as it becomes an extreme source of social stress (zapf, 1999). although the negativity this treatment entails is essential to the concept, the core characteristic isn’t the nature of the behaviors per se, but the persistency involved in one’s experience (einarsen et al., 2003). hence the duration and frequency is considered as much for emphasis as it is on what and how it is done. additionally, the existence of power disparity between the perpetrator and target whereby the increasing difficulty for the latter to defend himself is core to bullying. the conceptualization of power imbalance is derived from the perpetrator’s informal power, organizational position, the difference between the personality traits of both or the dependency of target on perpetrator (hoel & cooper, 2001; samnani, 2013; aquino & thau, 2009; zabrodska et al., 2016). 2.2 workplace bullying and employee silence circumstances motivating employees to speak up rather on factors that inhibit their willingness to share potentially useful information has remained the prior research’ main focus (walumbwa & schaubroeck, 2009; liang, farh, & farh, 2012), and namra jamshaid & sadia arshad 20 provided its hostile consequences, understanding the reasons of silence and its occurrences in workplaces is crucial (harlos, 2016). dysfunctional relationships with supervisors have been researched reasoning employee silence, in a huge body of work, when responding to workplace mistreatment (morrison, 2014; greenberg & edwards, 2009; duan et al., 2018; jahanzeb & fatima, 2018). a male respondent at information systems sheds light on his experience to becoming silent “i raised a concern about some policies and i was told that i was becoming a troublemaker. i would have pursued [the issue] further but presently i can’t afford to risk my job. this has made me go into a detached mode, making me a ‘yes man’”. such behaviors being target sensitive, mistreatment victims barely report or react against the perpetrator (supervisors in most cases) due to the reason of the committer being in a position well for counter revenge, the dependency of the victim on the offender for resources as opportunities for advancement or continued employment (xu et al., 2015, tepper et al., 2009) or the perception of senior management to not support such reports (roscigno et al., 2009) which eventually leads victim compromising on creativity and commitment and adopting withdrawal behaviors (guo et al., 2018). it is evident from the past literature that bullying at workplace occurs in many organizations throughout the world (einarsen et al., 2020) particularly in the asian firms, whereby its intensity is relatively high and it adversely affects employees’ outcomes (sheehaan, mccabe & garavan, 2020). approachinhibition theory of power by keltner et al. (2003) elucidates that where individuals with higher positions endowed with more power uplifts the positive emotions of others, it also stimulates their approach behaviors. individuals with lower power and positions on the other hand attend more to threats and monitor more preventative and inhibited behaviors (e.g. silence). this concept of power imbalance, as suggested by morrison and rothman (2009), when inherent in the organizational roles, makes silence for employees such a significant common experience as an adopting strategy. as different hierarchical positions are assumed by supervisors and subordinates, the former simultaneously have more control on resources (sturm & antonakis, 2015; anderson & brion, 2014). considered as a general belief of asymmetric powers between employer and employee within an employment relationship (offe & wiesenthal, 1980), consequently, silence is considered as the best approach to adopt when employees face maltreatment such as workplace bullying. conservation of resources theory (cor) proposes the organization of human behavior around acquisition and accumulation of resources. moreover, what valued is strived to be obtained, retained and protected by people (hobfoll, 1989) known as “resources”. such resources can take form of objects (food, tools, shelter), conditions journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 21 (dignity, status, security, social control), personal characteristics (skills, beliefs, efficacy) and energies (knowledge, time, money) where the central tenet lies on the primacy of resource loss, suggesting the intensity of resource loss over resource gain and the psychological harm for individuals associated with it. in case when a person’s resources are threatened, an actual loss of resource, or the expected return on resource investment does not materialize, stress is presumed to occur. encompassing a motivational element as well, the tenet further submits the engagement of depleted individuals in behaviors helping them avoiding further resource loss or minimizing any probable threat to their resources (halbesleben et al., 2014). the mum effect, by rossen and tesser (1970) was used to describe the lack of ability to communicate undesirable information in job settings due to the fear of jeopardizing interpersonal relationships. furthermore, deaf ear syndrome became under consideration for exploring the phenomenon from managerial perspective which precludes employees from speaking up by peirce et al. (1998). m. knoll et al. (2019) claim these approaches where opinions and genuine information is deliberately suppressed to be informative but relatively fragmented. hence different classifications have been raised to understand the concept in depth. among the new proposition of brinsfield’s (2013) silence classification based on supervisor-subordinate relationship in response to workplace bullying, rai and agrawal’s (2018) recent study endorses the strategical use of silence for victims to escape the adverse consequences allied with speaking up (defensive silence), for the protection of their relationships with supervisors (relational silence) and an approach to indicate their consent to organizational circumstances (ineffectual silence). 2.2.1 workplace bullying and defensive silence fearing the consequences allied with speaking up, such silence is a key motivator allowing one to deliberately omit their voices (brinsfield, 2013). employees being hesitant to communicate about issues at work (milliken et al., 2003), victims of mistreatment tend to remain silent particularly due to the fear of destructive ramifications concomitant with speaking up, especially in case where perpetrator holds a status higher than the victim (aquino et al., 2006). considering power disparity as the distinguishing feature to such maltreatment, victims consider themselves incapable in elevating their voice, and hence opt for defensive silence. speaking up per se requires extra time and energy, from the cor perspective, which comes at individual cost and risk and those intending to speak up stimulate their potential loss of resources (xu et al., 2015). hence the interpersonal mistreatment’ extant literature proposes such victims at the risk of being labeled as trouble makers, simultaneously at the menace of losing any namra jamshaid & sadia arshad 22 desirable professional opportunity or personal resources (harvey et al. 2007; xu et al. 2015). dyne et al. (2003) condition the fearfulness of an individual in a working environment where one attempts to shield oneself from potential threats for the occurrence of defensive silence. employees particularly deem supervisors engaging in mistreatment when they display some hostile behaviors as humiliating employees in public or menacing them with the intimidation of job loss (tepper, 2000). while wielding their status power over followers (aryee et al., 2007), power imbalance between the victim and perpetrator reasons the reluctance of taking any aggressive actions by the sufferers fearing the loss of valued resources or receiving admonishments to their insolent behaviors (tepper et al., 2009). kish-gephart et al. (2009) theorize fear of perpetuity in these hostile behaviors, when inflicted by the authoritative in hierarchy, add to the consequence of remaining silent. the emotional state of mind activates employee’s selfprotective mechanism i.e. “to lay low” (xu et al., 2015). confronting perpetrator verbally further provokes them resulting in deployment of high magnitude of mistreatment. 2.2.2 workplace bullying and relational silence avoiding harm within a relationship and tending to maintain interactions for any relational concerns delineate the root of relational silence (brinsfield, 2013), nevertheless lacks cooperative and altruistic motives (dyne et al., 2003). prior examination submits that, in addition to losing professional opportunities and other valuable resources, voice may too, in some way trouble the status quo or pressure current interpersonal relationships. raising of one’s voice increases the likelihood of questioning the position of an authority figure, subsequently deteriorating the victim’s bond with that authoritative figure (li & sun, 2015), that, in effect, costs employee additional available resources hence contributing to further depletion of resources. (ng & feldman, 2012). as to avoid one’s supervisor is rarely an alternative over the long haul, maintaining and enduring the relationship with them remains the only rational choice (whitman et al., 2014). merkin (2006) argues that in a culture where individuals entail low power distance values, engaging in face threatening conflicts for expressing themselves clearly is a stress free task. whereas high power distant individuals, asian cultures in particular, interpret challenges, interventions and confrontations as behaviors inappropriate to themselves. reluctant in verbally expressing negative emotions (fernández et al., 2000), members fear not being in agreement with, approaching and engaging with their journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 23 supervisors and persist tending to remain peaceful, cooperative and obedient. indirect messages are used considering face issues. such communication tactics in response to bullying at workplaces exhibit relational silence. due to the dependency of employees’ resources on supervisors, engagement in such behaviors is a more realistic communication strategy as compared to engagement in retaliatory behaviors having potential to exacerbate a relationship, as per cor theory (tepper et al., 2007). however, research proves a psychical or psychological distance from supervisors and victims of mistreatment, in this situation engaging in surface acting, considering it as a way to cope the adverse relationship (wu & hu, 2013). moreover, victims respond by reconciliation behaviors designed to reestablish the unappealing relationship on power dependence account (aquino et al., 2006), hence verifying mistreatment’s detrimental effects. 2.2.3 workplace bullying and ineffectual silence a common belief that raising voice will not be effective in causing change in the sense of a focal problem, circumstance or concern and will not have a constructive effect on the situation is the dimension’s defining line (brinsfield, 2013). engagement in such counterproductive silence is practiced as an implied tool to showing a sincere recognition of organizational circumstances, a sense of resignation, a relinquishment of hope for change and taking situations for granted (pinder & harlos, 2001). repetition and endurance in these deleterious acts lead employees to assume their unavoidability in their organization or the reluctance of managers and supervisor in resolving such issues. as a consequence, perceptions regarding possible improvements are deleteriously impacted of employees (d’cruz et al., 2014). anchoring the cor’s perspective, as per halbesleben et al (2014) sufferers may embrace ineffectual silence as a mean to shielding their resources, which might take any form of time, effort and energy from supplementary depletion resulting in the engagement of voice behaviors probably not likely to procure benefits. evidences support that disgruntled employees prefer not to voice their concerns and disappointments as they perceive no positive effect on the situation will transpire (wang & hsieh, 2013). hence timing and johnstone’s (2015) postulation is also in line with the above evidences by which anti-democratic personality characterizes individuals choosing to remain silent as they value obedience for authority and an interaction with dominant figure requires their submissiveness and hence generating conditions conducive to ineffectual silence. hence articulation of all the facts validate the suggestion that; h1: workplace bullying is significantly related to employee silence (defensive, relational and ineffectual). namra jamshaid & sadia arshad 24 2.3 the mediating role of pcv in the bullying-silence relationship workplace bullying has an association with employee silence. a perspective of social exchange theory is introduced to develop an understanding by which mechanism these two phenomena are linked together. in understanding workplace behavior, social exchange theory (set) (blau, 1964) acts as one of the most significant paradigm (cropanzano & mitchell, 2005). social exchange, in essence, is a set and series of interaction generating obligations. one party’s trust that even in the non-presence of a formal contract, the benefits received will be reciprocated is placed on the other. abiding by certain rules of exchange, parties involved to the relationship develop mutual commitment, trusting and loyalty over time, which stands at the core of this theory. the rule of reciprocity; central to researchers’ focus elucidates that actions of one party are contingent upon the actions of the other. it connotes that in any work settings, employees repay a favorable working conditions and environment through constructive job related behaviors and attitudes. whereas downward adjustments are likely to be the consequence of those behaviors and attitudes for the unfavorable treatment (parzefall & salin, 2010). psychological contracts have been started approaching from another perspective of cor theory from a developing body of research. it proposes that unfavorable events at workplaces route to a loss esteemed intrinsic resources. the psychological stress leads them to perceive this as a violation consequently triggering strong negative emotional reactions (lapointe et al., 2013; kiazad et al., 2014; priesemuth & taylor, 2016). as within the cor framework (hobfoll, 2001), dignity, fairness and respect are categorized as resources, the persistency in these uncivil acts such as persistence criticism, being belittled and yelled at and at a perpetual position of receiving insulting remarks have the nerve to drain employees’ intrinsic resources resulting in helplessness, frustration and restlessness (brotheridge & lee, 2010) -referred to a state of psychological resource depletion. such experiences weaken reserved resources rendering them vulnerable against further loss of resource consequential to bullying. such deep distressful situations deplete employees’ energy and other resources either consciously or unconsciously preoccupying them with actions aimed at conserving remaining resources (agarwal & bhargava, 2014). hence alleviating this psychological discomfort, remaining silent is employee’s highest preference in protecting the residual resources (ng & feldman, 2012). current research has surveyed the crucial mediating role of pcv in analyzing the employees’ responses to workplace mistreatment (salin & notelaers, 2017; rai & agarwal, 2018). research advocates the maltreatment from the supervisor’s end is journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 25 contemplated as a conspicuous workplace event fracturing the employer-employee contract (kernan et al., 2016). to survive pcv, rai and agarwal (2018) report silence as the most potential employee response. recent work also proves the experience of mistreatment evokes fear and anger as the negative emotions and in turn, behaviors associated to avoidance and selfprotection in subordinates like silence (nifadkar et al., 2012). hence ample indications gives confirmation to workplace mistreatment when channeled through pcv, fosters employee silence. hence, it can be suggested that; h2: pcv mediates the relationship between workplace bullying and employee silence (defensive, relational and ineffectual). 2.4 the moderating role of benevolent behavior in bullying-pcv relationship a perspective from the equity theory (adams, 1965) situations that individuals seek an equitable balance typically between their personal contributions and what they obtain from the organization. unfolding in another way, based on the reciprocity expectations in social relationships, parties involved have likely an expectation to be repaid in kind. a more in return is expected by the individuals who significantly invest in any relationship within a work context and are therefore, more sensitive towards any form of neglect or bad treatment by the employer (cropanzano & mitchell, 2005). previous research on breach and violation of psychological contracts reveals the individual differences as the foundation to their reaction towards this emotional damage and hence personal disposition embraces great importance (raja et al., 2004). a proposition by huseman et al. (1987) inculcates a personality variable such as equity sensitivity elucidates the individual differences while reacting to inequity. individuals high on equity sensitivity want more than others for prescribed level of input. in a nutshell, these can be labeled as outcome oriented individuals (sauley & bedeian, 2000). on the other end of the spectrum, some individuals pay more attention to their inputs, and in the exchange relationships. such individuals are less sensitive to the equity issues and are the ones low on equity sensitivity. in light of the above, it is interpreted that entitled have preference over output/input ratios exceeding those in comparison with others and hence concentrates on “getting more than giving”. contrasting this, benevolents focus remains on “giving more than getting”. maintenance of relationships with employers and finding satisfactions in contributing expertise and diverse talents to their organizations is the highlighting virtue of such individuals. research ascertains the fact that benevolents have a different reactions to the perceived violations (kickul & lester, 2001). their satisfaction with their jobs is reported with a greater decrease and more negative effects compared to the namra jamshaid & sadia arshad 26 other individuals relating to outcomes as autonomy and control (intrinsic outcomes) while a less strong reaction to the violations involving extrinsic outcomes in the form of rewards and benefits is observed. engagement in excessive monitoring, ridiculing, humiliating, withdrawal of valued tasks or insultingas are nested in a concept of workplace bullying, hence exposure to such perilous acts may, thus, have an ability to change the meaningfulness of work itself in numerous ways. consequently, a decrease in personal worth, autonomy and sense of accomplishment is detected of the targets (macintosh et al., 2010). in line with the previous research, when reporting exposure to workplace bulling, their basic needs are thwarted (trépanier, fernet, & austin, 2013) hence exhibiting a great effect to their intrinsic outcomes, something to which benevolents stands highly sensitive (kickul & lester, 2001). provided the particular focus on the intrinsic outcomes particularly for the benevolent employees and the norms of reciprocity, it is argued in this study due to the norms of reciprocity, those individuals supporting and helping others can reasonably have an expectation of more in return from the organization. the “more” is directed towards relationship quality and intrinsic outcomes (sense of accomplishment and feelings of personal worth) rather than fringe benefits. the key line “giving more than getting” represents a high contribution of such employees in terms of skills and efforts and expects less monetary rewards but equal or high psychological rewards from supervisors/leaders. being a victim to such mistreatment in exchange to their input might cause a severe blow to their worldview hence letting the perceptions of psychological contracts into question. hence a compelling rationale is provided to the proposed relationship. therefore, h3: the relationship between workplace bullying and perceived psychological contract violation is moderated by benevolent behavior. 2.5 the moderating role of psycap in the pcv-employee silence relationship in order to mitigate the deleterious negative impacts of contract violation (through bullying) and employee silence, the positive internal resources of employees are proposed to be a coping strategy towards the adverse situations. positive psychological capital (psycap) a positive psychological state of an individual that is open to development (luthans et al., 2008), a second-order factor of four positive resources comprising a shared variance and hence acts as an integration in the mechanism of which these resources share in common (avey et al., 2011) is theoretically and empirically validated, and entails a constructive outcome bigger than the individual impact of each component (norman et al., 2010) journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 27 hope delineates its basis on an intuitively created feeling of accomplishment of goal directed energy (agency) and planning to meeting goals (pathways) followed by an individual’s positive motivational state (snyder & lopez, 2002). agency, providing a deep insight, inculcates the motivation and energy towards goal direction for remaining successful at a given task provided in a specific work context. whereas the latter component, “pathway”, is considered as a mean and a way in the accomplishment of tasks and goals and hence collectively forming the will and way in such accomplishment. suggestions by snyder et al. (1996) theory and research claim that one component standing alone is not adequate. together the will to succeed in any particular task along with a viable mean to accomplish task defines and operationalizes the possession of hope. in workplace research, it has been found to positively influence performance (youssef & luthans, 2007), satisfaction (peterson & luthans, 2003) and negatively to deviance (costa & neves, 2017). resilience, a positive adaption and coping strategy in the face of risk and significant adversity (masten & reed, 2002). adaption to the workplace, a psychological capacity to bounceback or rebound from uncertainty, failure, adversity or even positive change, increased responsibility and progress expounds the nature of the component (luthans, 2002). resilience can therefore be described by coping responses not only to hostile events but also to the extremely positive events. concluded from the discussion of positive psychologist (seligman, 1988), his theoretical foundation for optimism states a stable, internal and global attribution regarding the positive events as achievement of goals, and an external, relatively unstable and specific cause for the negative events as failed attempt in goal achievement. avoiding the criticism of false optimism, positive organizational behavior (pob) forces on realistic optimism (luthans, 2002; luthans et al., 2007; schneider, 2001). signifying it other way, it is not based on an uncontrolled process having an unrealistic assessment. this state of realistic optimism (opposing to dispositional trait) includes an objective assessment of what an individual can accomplish, provided the available resources at that specific time in a specific situation and hence can vary (peterson, 2000). selfefficacy and its positive construction is based on bandura’s (1997) exhaustive theory and extensive research and hence a current prominence in linking the construct to positive psychology (bandura, 2007). its application to workplaces states as the individuals confidence or conviction about one’s abilities in the mobilization of cognitive resources, motivation and courses of actions required in successful execution of a specific task within a provided context (stajkovic & luthans, 1998). particular relevance to the pob criteria, the enhancement of selfefficacy can be through task namra jamshaid & sadia arshad 28 mastery (a success experience by employee), by modeling (vicarious learning by observing others, by attainment of positive feedback) and by psychological and physiological wellness (bandura, 1997). articulation of facts show that employees who perceive themselves high on psychological capital are more progressively confident in their abilities to accomplish goals, and therefore, any infringement in their psychological contract might be seen as a minor hindrance in the quest for their prime goal. being anticipative and hopeful, generation of various paths and solutions so as to manage workplace bullying and violation of psychological contracts. surveying circumstances as a transitory issue that will soon be reduced, demonstrating that they are optimistic and hopeful. resilience in their attitude helps those facing adversities and should therefore positively cope any mistreatment particularly workplace bullying and pcv. empirical evidences regarding psycap demonstrate consistency with performance (avey et al., 2011) due to being more energized and pouring extra efforts in tasks is reflected in outcome. undesirable behaviors are the ones in which employees high on psycap do not engage because such individuals are more capable of rebounding from the demand of additional output by recognizing (hope) and considering themselves to survive the pressure (selfefficacy) as it is just transient and will before long be reduced (optimism) (norman et al., 2010), therefore it is tested in this study whether psycap prevent employees involvement in negative behaviors due to the internal resources within themselves while dealing with situations. based on its mechanisms and empirical evidence, it is hypothesized: h4: the relationship between pcv and employee silence is moderated by psycap an explicit formulation is made based on the assumption that benevolent behavior and psycap conditionally influence the strength of the indirect relationship (workplace bullying-employee silence via pcv), thereby establishing a pattern of a dual stage moderated mediation among the study variables, hence hypothesizing: h5: benevolence and psycap moderates the positive relationship between workplace bullying and employee silence via pcv, such that the indirect relationship between bullying and silence through violation is stronger when benevolence is high and psycap is low journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 29 h2 h3 h4 h5 h1 figure 2.1 theoretical framework 3. methodology 3.1 research design elements as all variables of interest are under investigation for last many decades, so at this time the researcher is not required to conduct an initial (exploratory) research or to generate any qualitative data. the objective nature (quantitative analysis) of data further enhances the statistical comparisons and summary generation of the findings. the purpose of the study is correlational in nature hence allowing the statistical test of the proposed relationship and to know the extent to which these variables might be related. the type of investigation for the variables under study is quantitative as no open ended questions are used and no structured and unstructured interviews are conducted. “individuals” are considered as unit of analysis as variables are linked to the personality of individuals and the entire data is obtained purely in natural/ non-contrived settings hence the interference of the researcher is minimal. the study’s best suited time horizon is opted, i.e., cross-sectional design as the linkage of variables is only aimed to be demonstrated at a specific point of time. no second visit or any type of second observation is taken at any other point of time. data is collected in four working weeks and then entered in software for further analysis to come up with the reliable results. benevolent behavior positive psychological capital (psycap) psychological contract violation workplace bullying employee silence • defensive silence • relational silence • ineffectual silence namra jamshaid & sadia arshad 30 3.2 population and sample as respect and support for others especially patients and embracement of one’s powers is key to their survival while working under this context and being benevolent holds a strategical significance for all individuals working within it, any kind of mistreatment at workplaces will have a detrimental effect on their psychological state, compromising their positive emotions and confidence and where patients can be one to face its consequences. hence government administered hospitals represent all the study variables effectively. three main hospitals of lahore i.e. mayo hospital, ganga ram hospital and jinnah hospital are conveniently selected. 359 young junior doctors and staff nurses of three selected hospitals have been composed as an exact sample size of the current study through item response theory by olino et al. (2012). 3.3 measurement and scale silence (defensive, relational, ineffectual) behaviors are measured by using the 15items scale adapted from brinsfield (2013) representing defensive, relational and ineffectual silence. participants are scaled up to the level they get involved in numerous silence behaviors on a 5-point likert scale response format i.e. 1= strongly disagree till 5= strongly agree. sample item: “i feel it was risky to speak up”. workplace bullying is measured with a revised version of negative act questionnaire (naq-r) containing 22-items developed by einarsen et al (2009). consisting of three subscales as work-related, person-related and physically intimidating bullying, combined additively creating an overall workplace bullying scale. sample item: “being given task with unreasonable deadlines”. respondents are asked to indicate their responses on a 5 point likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). pcv is measured with 4-item scale adopted from robinson and morrison (2000). scale items are anchored on a 5 point likert scale providing range from 1 representing not at all to 5 representing very much. sample item include: “i feel betrayed by my organization”. benevolent behavior is measured with 4 items measuring the readiness of respondents in helping out and giving their time across a diverse set of situations, thus focusing more on the input side as compared to the preferred outputs. by adopting van der vegt et al (2003) helping scale, the sample item include: “i help others who have been absent”. respondents are asked to indicate their responses on a 5 point likert scale ranging from 1 (not at all) to 5 (very much). journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 31 psycap is measured with a 24-item scale adopted from luthans et al (2007). it is a higher order positive construct which comprises of four-facet constructs namely selfefficacy, hope, resilience and optimism. out of 24 items, 21 items are positively phrased and 3 items are reverse coded. sample items include: “i always look on the bright side of things regarding my job”. 5-point likert scale is provided to record the responses anchored at 1=strongly disagree and 5=strongly agree. 4. results 4.1 demographics of the study out of 359 participants involved in the survey, analysis represents with 52% of the female respondents and around 48% with males where a major chunk of the total sample falls below 30 years of age (72%), 25.6% lies between 31 to 35 years while the remaining falls in 36 and above. about 54.3% of the respondents hold their graduate degrees, 18% with their post graduate and around 28% under graduate category. fifty percent of the respondents have a working tenure below 1 year with their current supervisors whereas 44.3% have tenured between 1 to 3 years with their respective supervisors. five percent with 3-5 years and 0.3 percent have tenured for more than 5 years. 4.2 descriptive, correlations and reliabilities on a scale of 1 to 5, where 1 indicates strongly disagree and 5 indicates strongly agree, a value above 2.5 indicates most of the respondents rate themselves above the average level of mean. while initiating the statistical treatment of the raw data, reliabilities of the variable are calculated. reliabilities of all the variables are measured through calculating cronbach alpha. all variables showed above the acceptable values of reliabilities. table 4.1 correlation matrix and reliabilities variables mean sd 1 2 3 4 5 6 7 1.formal education 1.91 0.6 7 1 2.tenure with supervisor 1.61 0.5 9 .293* 3. workplace bullying 3.63 0.5 1 .069 .025 (0.80 ) 4. psychological contract violation 3.90 0.9 4 .056 -.014 .538* * (0.77) 5. benevolent behavior 3.80 0.9 8 .087 .020 .419* * .361** (0.81 ) namra jamshaid & sadia arshad 32 6. positive psychological capital 3.51 0.4 8 -.026 -.042 .536* * .653** .543* * (0.79 ) 7. employee silence 3.76 0.6 4 .038 -.016 .703* * .690** .457* .542* (0.83 ) n = 359, ** p < 0.01, * p < 0.05 analysis shows that predictor, mediator, moderators and outcome are significantly correlated to each other. specifically, employee silence has a significant and positive correlation with workplace bullying (r = 0.703, p<0.01). this means that increase in workplace bullying predicts greater employee silence. employee silence shows a positive and significant correlation with psychological contract violation (r = 0.690, p<0.01), whereas a negative and significant correlation with benevolent behavior and psycap (r = -.457, p<0.05) and (r = -.542, p<0.05) hence giving an initial support to all hypothesized statements. all the results of correlation matrix are summarized in table 4.1. 4.3 test of moderation, mediation and conditional indirect effect: simple linear regression is used to run test for the first hypothesis with x (workplace bullying) predicting y (employee silence). second, to test hypothesis 2, bootstrapping analysis, being non parametric simulations and a robust strategy in the estimation of indirect effects (preacher et al., 2007) is run to assess mediation (x predicts y via m). therefore obtaining the direct and indirect effects of mediation analysis, model 4 of hayes process macro (2018) is used. third, to examine the conditional effects of this dual stage moderated mediation, process model 1 is used, alongside simple slope representation. fourth, to run the conditional indirect effect of the dual stage moderated mediation, process model 21 is used. according to preacher et al. (2007), the strength of the independent variable (workplace bullying) on the outcome variable (employee silence) via mediator (pcv) differs across different levels of the moderators (benevolent behavior and psycap; +-1sd about mean respectively, respectively). table: 4.2 results of regression analysis mediator (pcv) outcome (employee silence) β b s.e(b) 95% cl β b s.e(b) 95% cl s.r.m (h1) wb 0.703 0.873** 0.047 [0.781-0.965] r2 0.494 simple mediation model (4) (h2) journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 33 wb 0.738 1.356** 0.066 [1.2271.484] 0.099 0.123** 0.044 [0.036-0.209] pcv 0.817 0.553** 0.023 [0.506-0.599] wb via pcv 0.603 0.750** 0.058 [1.040-1.351] simple moderation model (1) (h3) stage 1 moderation wb 0.151** 0.036 [0.0810.222] bb 0.815** 0.019 [0.7770.853] wb*bb 0.110** 0.018 [0.0740.145] r2 0.935 ∆r2 0.017 simple moderation model (1) (h4) stage 2 moderation pcv 0.531** 0.023 [0.489-0.574] psycap -0.112** 0.041 [0.030-0.193] pcv*psycap -0.083** 0.022 [0.040-0.126] r2 0.807 ∆r2 0.018 β: standardized regression coefficients; b: unstandardized regression coefficients; se (b): standard error of b; ci: confidence interval; pcv: psychological contract; psycap: positive psychological capital; **p < .01 4.3.1 workplace bullying-employee silence relationship employee silence shows a positive and statistically significant relationship with workplace bullying (b = 0.873, p<0.01) along with a 95% confidence interval of [llci=0.781, ulci=0.965], showing with every unit increase in workplace bullying, 0.873 units in employee silence is predicted. as values of the confidence interval are significantly different from zero, hence confirming the significance of the relationship. the results conclude that moving higher on bullying scale predicts greater level of employee silence and vice versa. the r2 value of 0.494 shows that workplace bullying alone explains 49.4% variance in employee silence. these results are consistent with the recent literature available on workplace bullying (i.e. rai & agarwal, 2018). hence the statistical analysis supports h1. 4.3.2 mediating role of pcv in workplace bullying-employee silence relationship in order to test the mediating effect of psychological contract violation in the bullying-silence relationship, the significance of independent variable on mediator is tested first. hence workplace bullying shows a positive and significant relationship with pcv (b = 1.356, p<0.01) along with a 95% confidence interval of [llci=1.227, ulci=1.484]. second, the significance of mediator on dependent variable is tested. pcv again shows a positive and significant relationship with employee silence when controlling for workplace bullying (b = 0.553, p<0.01) along with a 95% confidence namra jamshaid & sadia arshad 34 interval of [llci=0.506, ulci=0.599]. third, the significant indirect effect supported the mediating role of pcv on bullying-silence relationship (0.750, p<0.01) with a 95% confidence interval of [llci=1.040, ulci=1.351]. additionally, after the mediator being controlled, the association between bullying-silence (0.123, p<0.01) remain significant with a drop in beta value suggesting partial mediating effect of pcv in bullying-silence relationship. 4.3.3 moderating role of benevolent behavior in bullying-pcv relationship benevolent behavior shows a significant interaction term (b= 0.110, p< 0.01) with a 95% confidence interval of [llci=0.074, ulci=0.145]. the change in value of r2 (∆r2 = 0.017) also validate the significant predicting power of moderator variable. moreover, analysis shows that 93.5% variance in pcv is explained by the interaction term. the variation in the moderating effect of benevolent behavior in bullying-pcv relationship can be observed in figure 1. hence the coefficients of regression term along with graphical representation validate the moderating role of benevolent behavior in bullying-pcv relationship. alongside, its moderating role is higher at high levels workplace bullying and at lower level its role becomes insignificant. figure: 4.1: moderating role of benevolent behavior in bullying-pcv relationship 4.3.4 moderating role of psycap in pcv-silence relationship psycap shows a significant interaction term (b= -0.083, p< 0.01) with a 95% confidence interval of [llci=0.040, ulci=0.126]. the change in value of r2 (∆r2 = 0.018) also validate the significant predicting power of moderator variable. moreover, analysis shows that 80.7% variance in employee silence is explained by the interaction term. the variation in the moderating effect of psycap in pcv-silence relationship can be observed in figure 2. hence the coefficients of regression term along with graphical 1 1.5 2 2.5 3 3.5 4 4.5 5 low iv high iv p c v low moderator high moderator low wb high wb journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 35 representation validate the moderating role of psycap in pcv-silence relationship. alongside, its moderating role is lower at high levels of contract violation and at lower level its role becomes higher. figure: 4.2 moderating role of psycap in pcv-silence relationship 4.3.5 moderated mediation results for workplace bullying on employee silence via pcv across different levels of benevolent behavior (moderator 1) and psycap (moderator 2) to evaluate the dual-stage moderated mediation, following three conditions suggested by hayes (2018) are tested: (a) a significant indirect effect should be observed. (b) two significant interactions should be observed: first between the independent variable and moderator 1 in predicting mediator. second, mediator and the moderator 2 in predicting the outcome variable. (c) there should be a different conditional indirect effect from the independent variable to dependent variable via mediator at different levels (high and low) of moderator 1 and moderator 2, hence being the central condition for the existence of moderated mediation. it determines either mediator’s strength differs with different levels of moderators (hayes, 2018). the results of hypothesis 2 demonstrates the significant indirect relationship in bullying-silence via pcv and hence meeting the first condition. in observing second condition , results show that there is a significant interaction term between the independent variable and moderator 1 in predicting the mediator variable (b= 0.110, p< 0.01) and an interaction between the mediator and moderator 2 in predicting the 1 1.5 2 2.5 3 3.5 4 4.5 5 low iv high iv e m p lo y e e s il e n c e low moderator high moderator low pcv high pcv namra jamshaid & sadia arshad 36 dependent variable (b= -0.083, p< 0.01)is also found significant hence fulfilling condition 2. . next, to examine condition 3, hayes’ (2018) process macro for spss (model 21) is used. moderators are operationalized based on+-1sd about its mean score. results: looking at high levels of benevolent behavior, a significant indirect effect is found for those individuals who perceive psycap as being low or moderate, the indirect effect of workplace bullying on employee silence via pcv is significant. for those individuals who are low on benevolent behavior, no significant indirect relationship is found because regardless of the level of psychological capital in employees, zero is in the confidence interval of the bootstrap analysis. so if individuals are low on benevolent behavior, doesn’t matter the level of psychological capital. if individuals are high on benevolent behavior and observes psychological capital as low or moderate within themselves, the indirect effect of workplace bullying on employee silence via pcv is significant. table: 4.3: moderated mediation results for bullying-silence relationship via pcv across different levels of benevolent behavior (moderator 1) and psycap (moderator 2) moderator1 moderator2 conditional indirect effect bootstrap se bootstrap llci bootstrap ulci m1_pcv -0.9871 -0.4807 0.0177 .0157 -.0113 .0496 m1_pcv -0.9871 0.0000 0.0207 .0180 -.0131 .0572 m1_pcv -0.9871 0.4807 0.0237 .0205 -.0152 .0651 m1_pcv 0.0000 -0.4807 0.0620 .0185 .0313 .1028 m1_pcv 0.0000 0.0000 0.0727 .0200 .0390 .1166 m1_pcv 0.0000 0.4807 0.0833 .0224 .0453 .1331 m1_pcv 0.9871 -0.4807 0.1064 .0256 .0648 .1643 m1_pcv 0.9871 0.0000 0.1246 .0268 .0817 .1860 m1_pcv 0.9871 0.4807 0.1429 .0299 -.0953 .2132 5. discussion and conclusion anchored in the conservation of resource (cor) theory, this study’s aim is to analyze the passive coping strategies of employees in the face of workplace bullying. social exchange theory (set) theorizes the relationship to be mediated by pcv, and this empirical study lend support to the proposition. moreover, the results revealed that the process of mediation is stronger for individuals who report themselves at a higher level of benevolence and perceives psychological capital to be low. hence the cross level moderation is mediated. journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 37 taking a step beyond from previous literature examining bullying-turnover relationship (salin & notelaers, 2017), this research studies the establishment of passive coping strategies in response to workplace bullying and the mechanism to explain this relationship. it is proved that the perception of violation of psychological contract depicts a partial mediation effect between bullying and silence relationship. psychological contractthe perceived obligations in an employer-employee relationship based on its highly subjective interpretation and a sense of its violation strongly effect behaviors of employees and seems to partially explain some deleterious effects associated to bullying in this study. an escalated deleterious interaction with a wrongdoer result in resource loss, consequently developing feelings of pcv by increase psychological distress. hence aiming to conserve the remaining resources, such passive strategies are adopted by employees, hence indicating both the direct and indirect effect of workplace bullying on silence of employees. additionally, individual differences are also highlighted through the findings. among individuals reporting behaviors high on benevolence and perceive themselves low on psycap, the relation between workplace bullying and employee silence via pcv is stronger, hence providing with a more nuanced picture of employees’ reactions. typically, social relations have an involvement of reciprocity, congruous with equity theory’s perspective (adams, 1965) revolving around the concept that an impartial balance is seek by individuals between their own contributions and what are they receiving from the organization, which, in turn, is in line with research findings illustrating that the one’s highly committed to their work are sensitive to layoff processes (brockner et al., 1992). employees investing in prosocial behavior are highly expected of a respectful treatment in return, and pertain adverse negative reactions when that is not the case. benevolents have been proved to care less about pay and benefits or any external outcomes (kickul & lester, 2001), instead have been noticed that their strong reactions to breaches involves autonomy and control. hence it is supported in this study’s findings that for such employees, bullying behaviors strongly effect perception of psychological violations. though benevolent behavior moderated the relationship, the effect of moderation is minimal. the possible reason for this would be that when exposure to negative acts are reached to a certain threshold, negative behaviors of employees are equally high, regardless of the levels of benevolence reported by employees. further possible explanation of the findings is acknowledged. benevolent actions, generally endorsed positively both formally and informally by managers, not only reflect personal disposition, but is used as the response trying to offset or to cope with namra jamshaid & sadia arshad 38 bullying. although such interpersonal mistreatment results in deterioration of benevolence, few victims attempt to prove their worth and accordingly plan to stop the negative practices they are exposed toby being exceptionally responsive, friendly and cooperative. even being sick, they turn up to work to prove their loyalty and dedication and prevent themselves being marked as maligner (hoel et al., 2011). congruent with the experimental studies demonstrating that the one who experience ostracism tends to have a better performance in cognitive ability, tasks in an attempt to demonstrate their worth (jamieson et al., 2010). nevertheless, when such attempts to being helpful for offsetting bullying becomes ineffective, it may be that employees who have attempted this path becomes more cynical, resulting in a stronger sense of violation of the psychological contract and a stronger intention to remain silent. the moderation effects of psycap are significant but small. results demonstrates employees high on psycap depicts lower silent behaviors and vice versa as they have additional resources helping to overcome daily problems and are better able to interpret situations (chadwick & raver, 2013). this further suggests that positive psychological state of an individual can reduce the degree of employee adverse outcomes but cannot completely mitigate the consequences of such mistreatment. 6. limitations and directions first, a crosssectional design is used by which inferences are not drawn about workplace bullying-outcomes’ causal relationship. it is likely that the directions of reciprocal relations reported at a single time and different points in time are not always in the same directions. having benevolence and psycap are just as likely to modify an individual’s appraisal of the situation as an experience of mistreatment especially workplace bullying is likely to reduce the resilience of an individual with time. however, it can be concluded that positive psychological state have a tendency to reduce employee silence. second, the sample comprised mostly young junior doctors and staff nurses. in light of differences in experience of the sample, it is concluded that those having a higher experience and having higher job positions should also be incorporated in the study sample to yield different results. different government and private hospitals should too be incorporated for generalizing results to the entire health care sector. third, a full three-wave cross lagged longitudinal analysis should be designed among the study variables helping to identify a causal order making relationships more robust and decreasing the threat of cmv is another major limitation of the study. in journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 39 estimation 25% of the variance in any relationship is because of cmv (buckley et al., 1990). several promising future research directions are presented in this study. this study shows employees’ own exposure to negative acts results in violation of their psychological contracts and an increase in employee silence. a potential avenue for future research is to analyzing the extent to which witnessing others who are subjected to negative treatment results in their psychological contracts and their consequent passive reactions. as research shows that witnesses consequent in poor health and higher stress levels (salin, 2014). reason behind this could be speculated as witnessing bullying, too, shatters the beliefs of justice of the observer. as examination on layoffs demonstrates its impeding impacts not just on the ones legitimately influenced by it yet in addition who observes it (kets de vries & balazs, 1997). parallel to this, a researcher also hypothesized that occurrence of bullying followed by absence of organizational strides to stop is perceived as infringement of psychological contract from the bystanders’ perspective as well (parzefall & salin, 2010). this remains to have an empirical validation. future studies could also have an examination on other possible mediators in bullying-silence relationship. targets are plausible to not only encounter resentment, frustration and anger but other negative feelings in the form of fear, emotional exhaustion and lack of psychological safety (brinsfield, 2013; kiewitz et al., 2016) is also expected. provided the pakistan’s sociocultural milieu, analyzing the moderating roles of collectivism and power distance holds position for a very promising future research avenue. few studies, according to srivastava et al (2019) have been conducted to test underlying mechanism for workplace outcomes in response to silence of employees, hence an extensive research involving various central mechanisms and boundary conditions remains a valuable direction for future researchers. academic literature, surveys and reports all indicate the increase in the frequency of workplace bullying in recent years (e.g., 27%, workplace bullying institute survey, 2014). hence its detrimental effects largely costs to employees and organizations both, raising voices and speaking up for the concerns in an effort to reduce is a sensible approach to adopt. finding of the current study, however, suggests that employees, in response to mistreatment, prefer to remain silent, hence inhibiting mistreatments within work contexts and promotion of voice behaviors within employees serves as a key tool. previous research also proves the increased awareness on the nature, form, causes and consequences of bullying, clear imposition of policies such as “zero tolerance for mistreatment at work” and “managing with respect” and clear protocols in order to namra jamshaid & sadia arshad 40 report bullying serves as potential deterrents. management lax attitude and non supportive hr policies (noronha & d’cruz, 2009) are some prime reasons for the adoption of silence behaviors. a wellformulated and efficiently enforced antibullying strategies promote more active participation of hr in organizations. employees, particularly in high power distant cultures like pakistan are hesitant to straightforwardly report incidences. to counter this, reporting arrangements like hotlines and mobile applications like stop it should be adopted by organizations while maintaining the anonymity of the victim. for employees to share work related problems, organizational ombudsman should be established that operates outside the traditional hierarchy. through this, employees’ identities should also remain confidential hence protecting any sort of tension or damage either supervisor, helping employees combatting bullying by raising voices against it. in addition to fostering an intolerant environment for bullying through organizational practice, a close monitoring of supervisors’ behaviors followed by providing mechanisms to employees such as employee surveys and grievance procedures to report is a significant tool to reduce its severity. all except, workplace bullying is an organizational reality. hence efforts to make an entire bullying-free environment is next to impossible. so, in addition to make an effort in ending up this maltreatment, managers must limit its consequences by understanding its dynamics. as research shows that the relationship between bullyingsilence is preceded by particular reactions as frustration and anger (pcv) enhancing the victim’s psychological distress due to resource loss, it is important for managers and hr professionals to be sensitive for such cues and measures for relieving employees from these negative emotions should be timely fashioned. provision of psychological consultation services, where victims are heard, a supportive workplace environment which fosters interpersonal networks providing ample social support and opportunities for job enhancement in order to cope with negative situations will definitely serve the purpose. references adams, j. s. (1965). inequity in social exchange. advances in experimental social psychology, 2, 267-299. agarwal, u. a., & bhargava, s. (2014). the role of social exchange on work outcomes: a study of indian managers. the international journal of human resource management, 25(10), 1484-1504. journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 41 anderson, c., & brion, s. (2014). perspectives on power in organizations. annual review of organizational psychology and organizational behavior, 1(1), 67-97. aquino, k., & thau, s. (2009). workplace victimization: aggression from the target's perspective. annual review of psychology, 60. aquino, k., tripp, t. m., & bies, r. j. (2006). getting even or moving on? power, procedural justice, and types of offense as predictors of revenge, forgiveness, reconciliation, and avoidance in organizations. journal of applied psychology, 91(3), 653. aryee, s., chen, z. x., sun, l. y., & debrah, y. a. (2007). antecedents and outcomes of abusive supervision: test of a trickle-down model. journal of applied psychology, 92(1), 191. avey, j. b., reichard, r. j., luthans, f., & mhatre, k. h. (2011). meta‐analysis of the impact of positive psychological capital on employee attitudes, behaviors, and performance. human resource development quarterly, 22(2), 127-152. bandura, a. (1997). self-efficacy: the exercise of control. new york: freeman. bandura, a. (2007). an agentic perspective on positive psychology. in s. j. lopez (ed.), the science of human flourishing, new york: praeger. blau, p. m. (1964). exchange and power in social life wiley new york. brinsfield, c. t. (2013). employee silence motives: investigation of dimensionality and development of measures. journal of organizational behavior, 34(5), 671-697. brockner, j., tyler, t. r., & cooper-schneider, r. (1992). the influence of prior commitment to an institution on reactions to perceived unfairness: the higher they are, the harder they fall. administrative science quarterly, 241-261. brotheridge, c. m., & lee, r. t. (2010). restless and confused: emotional responses to workplace bullying in men and women. career development international, 15(7), 687-707. buckley, m. r., cote, j. a. and comstock, s. m. (1990). measurement errors in the behavioral sciences: the case of personality/attitude research. educational and psychological measurement, 50(3), 447-474. chadwick, i. c., & raver, j. l. (2013). continuously improving in tough times: overcoming resource constraints with psychological capital. academy of management proceedings, 1, 13895. namra jamshaid & sadia arshad 42 chung, h., & van oorschot, w. (2010). employment insecurity of european individuals during the financial crisis: a multi-level approach. reconciliation of work and welfare in europe working paper, 14. costa, s., & neves, p. (2017). job insecurity and work outcomes: the role of psychological contract breach and positive psychological capital. work & stress, 31(4), 375-394. cropanzano, r., & mitchell, m. s. (2005). social exchange theory: an interdisciplinary review. journal of management, 31(6), 874-900. d'cruz, p., noronha, e., & beale, d. (2014). the workplace bullying-organizational change interface: emerging challenges for human resource management. the international journal of human resource management, 25(10), 1434-1459. de vries, m. f. k., & balazs, k. (1997). the downside of downsizing. human relations, 50(1), 11-50. duan, j., bao, c., huang, c., & brinsfield, c. t. (2018). authoritarian leadership and employee silence in china. journal of management and organization, 24(1), 62. dyne, l. v., ang, s., & botero, i. c. (2003). conceptualizing employee silence and employee voice as multidimensional constructs. journal of management studies, 40(6), 1359-1392. einarsen, s. v., hoel, h., zapf, d., & cooper, c. l. (2020). bullying and harassment in the workplace: theory, research and practice. crc press. einarsen, s., hoel, h., & cooper, c. (eds.). (2003). bullying and emotional abuse in the workplace: international perspectives in research and practice. crc press. einarsen, s., hoel, h., & notelaers, g. (2009). measuring exposure to bullying and harassment at work: validity, factor structure and psychometric properties of the negative acts questionnaire-revised. work & stress, 23(1), 24-44. einarsen, s., hoel, h., zapf, d., & cooper, c. l. (2011). the concept of bullying and harassment at work: the european tradition. bullying and harassment in the workplace: developments in theory, research, and practice, 2, 3-40. fernández, i., carrera, p., sánchez, f., paez, d., & candia, l. (2000). differences between cultures in emotional verbal and non-verbal reactions. psicothema, 12(1), 83-92. journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 43 fox, s., & cowan, r. l. (2015). revision of the workplace bullying checklist: the importance of human resource management's role in defining and addressing workplace bullying. human resource management journal, 25(1), 116-130. fox, s., & stallworth, l. e. (2009). building a framework for two internal organizational approaches to resolving and preventing workplace bullying: alternative dispute resolution and training. consulting psychology journal: practice and research, 61(3), 220. ghayur, s. (2009). evolution of the industrial relations system in pakistan, ppc policy paper series. greenberg, j., & edwards, m. s. (eds.). (2009). voice and silence in organizations. emerald group publishing. guo, l., decoster, s., babalola, m. t., de schutter, l., garba, o. a., & riisla, k. (2018). authoritarian leadership and employee creativity: the moderating role of psychological capital and the mediating role of fear and defensive silence. journal of business research, 92, 219-230. halbesleben, j. r., neveu, j. p., paustian-underdahl, s. c., & westman, m. (2014). getting to the “cor” understanding the role of resources in conservation of resources theory. journal of management, 40(5), 1334-1364. harlos, k. (2016). employee silence in the context of unethical behavior at work: a commentary. german journal of human resource management, 30(3-4), 345355. harvey, p., stoner, j., hochwarter, w., & kacmar, c. (2007). coping with abusive supervision: the neutralizing effects of ingratiation and positive affect on negative employee outcomes. the leadership quarterly, 18(3), 264-280. hayes, a. f. (2018). introduction to mediation, moderation, and conditional process analysis: a regression based approach (2nd ed.). hershcovis, m. s. (2011). incivility, social undermining, bullying… oh my!: a call to reconcile constructs within workplace aggression research. journal of organizational behavior, 32(3), 499-519. hobfoll, s. e. (1989). conservation of resources: a new attempt at conceptualizing stress. american psychologist, 44(3), 513. hoel, h., & cooper, c. l. (2001). origins of bullying: theoretical frameworks for explaining workplace bullying. building a culture of respect, 21-38. crc press. namra jamshaid & sadia arshad 44 hoel, h., sheehan, m. j., cooper, c. l., & einarsen, s. (2011). organisational effects of workplace bullying. bullying and harassment in the workplace: developments in theory, research, and practice, 129-148. hom, p. w., mitchell, t. r., lee, t. w., & griffeth, r. w. (2012). reviewing employee turnover: focusing on proximal withdrawal states and an expanded criterion. psychological bulletin, 138(5), 831. huseman, r. c., hatfield, j. d., & miles, e. w. (1987). a new perspective on equity theory: the equity sensitivity construct. academy of management review, 12(2), 222-234. islam, n. (2004). sifarish, sycophants, power and collectivism: administrative culture in pakistan. international review of administrative sciences, 70(2), 311-330. jahanzeb, s., & fatima, t. (2018). how workplace ostracism influences interpersonal deviance: the mediating role of defensive silence and emotional exhaustion. journal of business and psychology, 33(6), 779-791. jamieson, j. p., harkins, s. g., & williams, k. d. (2010). need threat can motivate performance after ostracism. personality and social psychology bulletin, 36(5), 690-702. jiang, k., liu, d., mckay, p. f., lee, t. w., & mitchell, t. r. (2012). when and how is job embeddedness predictive of turnover? a meta-analytic investigation. journal of applied psychology, 97(5), 1077. keltner, d., gruenfeld, d. h., & anderson, c. (2003). power, approach, and inhibition. psychological review, 110(2), 265. kernan, m. c., racicot, b. m., & fisher, a. m. (2016). effects of abusive supervision, psychological climate, and felt violation on work outcomes: a moderated mediated model. journal of leadership & organizational studies, 23(3), 309321. khan, a. k., quratulain, s., & crawshaw, j. r. (2013). the mediating role of discrete emotions in the relationship between injustice and counterproductive work behaviors: a study in pakistan. journal of business and psychology, 28(1), 4961. kiazad, k., seibert, s. e., & kraimer, m. l. (2014). psychological contract breach and employee innovation: a conservation of resources perspective. journal of occupational and organizational psychology, 87(3), 535-556. journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 45 kickul, j., & lester, s. w. (2001). broken promises: equity sensitivity as a moderator between psychological contract breach and employee attitudes and behavior. journal of business and psychology, 16(2), 191-217. kiewitz, c., restubog, s. l. d., shoss, m. k., garcia, p. r. j. m., & tang, r. l. (2016). suffering in silence: investigating the role of fear in the relationship between abusive supervision and defensive silence. journal of applied psychology, 101(5), 731. kish-gephart, j. j., detert, j. r., treviño, l. k., & edmondson, a. c. (2009). silenced by fear: the nature, sources, and consequences of fear at work. research in organizational behavior, 29, 163-193. knoll, m., hall, r. j., & weigelt, o. (2019). a longitudinal study of the relationships between four differentially motivated forms of employee silence and burnout. journal of occupational health psychology, 24(5), 572. lapointe, é., vandenberghe, c., & boudrias, j. s. (2013). psychological contract breach, affective commitment to organization and supervisor, and newcomer adjustment: a three-wave moderated mediation model. journal of vocational behavior, 83(3), 528-538. leymann, h. (1996). the content and development of mobbing at work. european journal of work and organizational psychology, 5(2), 165-184. li, y., & sun, j. m. (2015). traditional chinese leadership and employee voice behavior: a cross-level examination. the leadership quarterly, 26(2), 172-189. liang, j., farh, c. i., & farh, j. l. (2012). psychological antecedents of promotive and prohibitive voice: a two-wave examination. academy of management journal, 55(1), 71-92. lim, s., & cortina, l. m. (2005). interpersonal mistreatment in the workplace: the interface and impact of general incivility and sexual harassment. journal of applied psychology, 90(3), 483. luthans, f. (2002). the need for and meaning of positive organizational behavior. journal of organizational behavior: the international journal of industrial, occupational and organizational psychology and behavior, 23(6), 695-706. luthans, f., avolio, b. j., avey, j. b., & norman, s. m. (2007). positive psychological capital: measurement and relationship with performance and satisfaction. personnel psychology, 60(3), 541-572. namra jamshaid & sadia arshad 46 luthans, f., norman, s. m., avolio, b. j., & avey, j. b. (2008). the mediating role of psychological capital in the supportive organizational climate—employee performance relationship. journal of organizational behavior, 29(2), 219-238. macintosh, j., wuest, j., gray, m. m., & cronkhite, m. (2010). workplace bullying in health care affects the meaning of work. qualitative health research, 20(8), 1128-1141. masten, a. s., & reed, m. g. j. (2002). resilience in development. handbook of positive psychology, 74, 88. mccormack, d., casimir, g., djurkovic, n., & yang, l. (2009). workplace bullying and intention to leave among schoolteachers in china: the mediating effect of affective commitment. journal of applied social psychology, 39(9), 21062127. merkin, r. s. (2006). power distance and facework strategies. journal of intercultural communication research, 35(2), 139-160. milliken, f. j., morrison, e. w., & hewlin, p. f. (2003). an exploratory study of employee silence: issues that employees don’t communicate upward and why. journal of management studies, 40(6), 1453-1476. morrison, e. w. (2014). employee voice and silence. annual review of organizational psychology and. organizational behavior, 1(1), 173-197. morrison, e. w., & rothman, n. b. (2009). silence and the dynamics of power. voice and silence in organizations, 6, 111-134. ng, t. w., & feldman, d. c. (2012). employee voice behavior: a meta‐analytic test of the conservation of resources framework. journal of organizational behavior, 33(2), 216-234. nifadkar, s., tsui, a. s., & ashforth, b. e. (2012). the way you make me feel and behave: supervisor-triggered newcomer affect and approach-avoidance behavior. academy of management journal, 55(5), 1146-1168. norman, s. m., avey, j. b., nimnicht, j. l., & graber pigeon, n. (2010). the interactive effects of psychological capital and organizational identity on employee organizational citizenship and deviance behaviors. journal of leadership & organizational studies, 17(4), 380-391. noronha, e., & d'cruz, p. (2009). employee identity in indian call centres: the notion of professionalism. sage publications india. journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 47 offe, c., & wiesenthal, h. (1980). two logics of collective action: theoretical notes on social class and organizational form. political power and social theory, 1(1), 67115. parzefall, m. r., & salin, d. m. (2010). perceptions of and reactions to workplace bullying: a social exchange perspective. human relations, 63(6), 761-780. peirce, e., smolinski, c. a., & rosen, b. (1998). why sexual harassment complaints fall on deaf ears. academy of management perspectives, 12(3), 41-54. peterson, c. (2000). the future of optimism. american psychologist, 55, 44–55. peterson, s. j., & luthans, f. (2003). the positive impact and development of hopeful leaders. leadership & organization development journal, 24(1), 26-31. piccoli, b., & de witte, h. (2015). job insecurity and emotional exhaustion: testing psychological contract breach versus distributive injustice as indicators of lack of reciprocity. work & stress, 29(3), 246-263. pinder, c. c., & harlos, k. p. (2001). employee silence: quiescence and acquiescence as responses to perceived injustice. personnel and human resources management, 331-369. preacher, k. j., rucker, d. d., & hayes, a. f. (2007). addressing moderated mediation hypotheses: theory, methods, and prescriptions. multivariate behavioral research, 42(1), 185-227. priesemuth, m., & taylor, r. m. (2016). the more i want, the less i have left to give: the moderating role of psychological entitlement on the relationship between psychological contract violation, depressive mood states, and citizenship behavior. journal of organizational behavior, 37(7), 967-982. rai, a., & agarwal, u. a. (2018). workplace bullying and employee silence: a moderated mediation model of psychological contract violation and workplace friendship. personnel review, 47(1), 226-256. raja, u., johns, g., & ntalianis, f. (2004). the impact of personality on psychological contracts. academy of management journal, 47(3), 350-367. robinson, s. l., & wolfe morrison, e. (2000). the development of psychological contract breach and violation: a longitudinal study. journal of organizational behavior, 21(5), 525-546. roscigno, v. j., lopez, s. h., & hodson, r. (2009). supervisory bullying, status inequalities and organizational context. social forces, 87(3), 1561-1589. namra jamshaid & sadia arshad 48 rosen, s., & tesser, a. (1970). on reluctance to communicate undesirable information: the mum effect. sociometry, 253-263. salin, d. (2013). bullying and well-being. handbook of unethical work behavior: implications for individual well-being: implications for individual well-being, 73. salin, d., & notelaers, g. (2017). the effect of exposure to bullying on turnover intentions: the role of perceived psychological contract violation and benevolent behaviour. work & stress, 31(4), 355-374. salin, d., & notelaers, g. (2020). the effects of workplace bullying on witnesses: violation of the psychological contract as an explanatory mechanism?. the international journal of human resource management, 31(18), 2319-2339. samnani, a. k. (2013). embracing new directions in workplace bullying research: a paradigmatic approach. journal of management inquiry, 22(1), 26-36. sauley, k. s., & bedeian, a. g. (2000). equity sensitivity: construction of a measure and examination of its psychometric properties. journal of management, 26(5), 885-910. schneider, s. l. (2001). in search of realistic optimism. american psychologist, 56, 250–263. seligman, m. e. p. (1998). learned optimism new york: pocket books. sheehan, m., mccabe, t. j., & garavan, t. n. (2020). workplace bullying and employee outcomes: a moderated mediated model. the international journal of human resource management, 31(11), 1379-1416. snyder, c. r., & lopez, s. j. (eds.). (2002). handbook of positive psychology. oxford university press. snyder, c. r., sympson, s. c., ybasco, f. c., borders, t. f., babyak, m. a., & higgins, r. l. (1996). development and validation of the state hope scale. journal of personality and social psychology, 70(2), 321. srivastava, s., jain, a. k., & sullivan, s. (2019). employee silence and burnout in india: the mediating role of emotional intelligence. personnel review, 48(4), 1045– 1060 stajkovic, a. d., & luthans, f. (1998). self-efficacy and work-related performance: a meta-analysis. psychological bulletin, 124(2), 240. sturm, r. e., & antonakis, j. (2015). interpersonal power: a review, critique, and research agenda. journal of management, 41(1), 136-163. journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 49 tepper, b. j. (2000). consequences of abusive supervision. academy of management journal, 43(2), 178-190. tepper, b. j., carr, j. c., breaux, d. m., geider, s., hu, c., & hua, w. (2009). abusive supervision, intentions to quit, and employees’ workplace deviance: a power/dependence analysis. organizational behavior and human decision processes, 109(2), 156-167. tepper, b. j., moss, s. e., lockhart, d. e., & carr, j. c. (2007). abusive supervision, upward maintenance communication, and subordinates' psychological distress. academy of management journal, 50(5), 1169-1180. timming, a. r., & johnstone, s. (2015). employee silence and the authoritarian personality: a political psychology of workplace democracy. international journal of organizational analysis, 23(1), 154-171. trépanier, s. g., fernet, c., & austin, s. (2013). workplace bullying and psychological health at work: the mediating role of satisfaction of needs for autonomy, competence and relatedness. work & stress, 27(2), 123-140. van der vegt, g. s., van de vliert, e., & oosterhof, a. (2003). informational dissimilarity and organizational citizenship behavior: the role of intrateam interdependence and team identification. academy of management journal, 46(6), 715-727. walumbwa, f. o., & schaubroeck, j. (2009). leader personality traits and employee voice behavior: mediating roles of ethical leadership and work group psychological safety. journal of applied psychology, 94(5), 1275. wang, y. d., & hsieh, h. h. (2013). organizational ethical climate, perceived organizational support, and employee silence: a cross-level investigation. human relations, 66(6), 783-802. whitman, m. v., halbesleben, j. r., & holmes iv, o. (2014). abusive supervision and feedback avoidance: the mediating role of emotional exhaustion. journal of organizational behavior, 35(1), 38-53. wu, t. y., & hu, c. (2013). abusive supervision and subordinate emotional labor: the moderating role of openness personality. journal of applied social psychology, 43(5), 956-970. xu, a. j., loi, r., & lam, l. w. (2015). the bad boss takes it all: how abusive supervision and leader–member exchange interact to influence employee silence. the leadership quarterly, 26(5), 763-774. namra jamshaid & sadia arshad 50 youssef, c. m., & luthans, f. (2007). positive organizational behavior in the workplace: the impact of hope, optimism, and resilience. journal of management, 33(5), 774-800. zabrodska, k., ellwood, c., zaeemdar, s., & mudrak, j. (2016). workplace bullying as sensemaking: an analysis of target and actor perspectives on initial hostile interactions. culture and organization, 22(2), 136-157. zapf, d. (1999). organisational, work group related and personal causes of mobbing/bullying at work. international journal of manpower, 20(1), 70-85. zapf, d. (2004). negative social behavior at work and workplace bullying. the fourth international conference on bullying and harassment in the workplace, 25. journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 51 survey questionnaire dear sir\madam, this survey is being conducted to analyze the impact of workplace bullying on employee silence. the results will have no effect what so ever upon your final grades. your anonymity will be strictly retained and results will be used for research purposes only. thank you for your time and cooperation! 1. general information: 2. please place a tick in one option for each statement to rate yourself on below given criteria at the state of workplace bullying. s tr o n g ly d is a g r e e d is a g r e e n e u tr a l a g r e e s tr o n g ly a g r e e wb1 someone withholding information which affects your performance. 1 2 3 4 5 wb2 being ordered to do work below your level of competence. 1 2 3 4 5 wb3 having your opinions ignored. 1 2 3 4 5 wb4 being given tasks with unreasonable deadlines. 1 2 3 4 5 wb5 excessive monitoring of your work. 1 2 3 4 5 wb6 pressure not to claim something to which by right you are entitled (e.g. sick leave, holiday entitlement, and travel expenses). 1 2 3 4 5 wb7 being exposed to an unmanageable workload. 1 2 3 4 5 wb8 being humiliated or ridiculed in connection with your work. 1 2 3 4 5 wb9 having key areas of responsibility removed or replaced with more trivial or unpleasant tasks. 1 2 3 4 5 wb1 0 spreading of gossip and rumors about you. 1 2 3 4 5 wb1 1 being ignored or excluded. 1 2 3 4 5 wb1 2 having insulting or offensive remarks made about your person, attitudes or your private life. 1 2 3 4 5 wb1 3 hints or signals from others that you should quit your job. 1 2 3 4 5 wb1 4 repeated reminders of your errors or mistakes. 1 2 3 4 5 wb1 5 being ignored or facing a hostile reaction when you approach. 1 2 3 4 5 wb1 6 persistent criticism of your errors or mistakes. 1 2 3 4 5 a) gender: a) male b) female b) formal education: (1) undergraduate (2) graduate (3) post graduate c) age: (1) below 30 (2) 31– 35 (3) 36 years or above d) working tenure under this supervisor (1) below 1 year (2) 1-3 years (3) 3-5 years (4) 5 years or more namra jamshaid & sadia arshad 52 wb1 7 practical jokes carried out by people you don’t get along with. 1 2 3 4 5 wb1 8 having allegations made against you. 1 2 3 4 5 wb1 9 being the subject of excessive teasing and sarcasm. 1 2 3 4 5 wb2 0 being shouted at or being the target of spontaneous anger. 1 2 3 4 5 wb2 1 intimidating behaviors such as finger-pointing, invasion of personal space, shoving, blocking your way. 1 2 3 4 5 wb2 2 threats of violence or physical abuse or actual abuse. 1 2 3 4 5 3. please place a tick in one option for each statement to rate yourself on below given criteria at the state of psychological contract violation. (please tick only one number for each of the following statements) pcv1 i feel betrayed by my organization. not at all 1 2 3 4 5 very much pcv2 i feel that the organization has violated the contract between us. not at all 1 2 3 4 5 very much pcv3 i have not received everything promised to me in exchange for my contributions. not at all 1 2 3 4 5 very much pcv4 almost all the promises made by my employer during recruitment have been kept thus far not at all 1 2 3 4 5 very much 4. please place a tick in one option for each statement to rate yourself on below given criteria at the state of benevolent behavior. (please tick only one number for each of the following statements) 5. please place a tick in one option for each statement to rate yourself on below given criteria at the state of positive psychological capital. s tr o n g ly d is a g r e e d is a g r e e n e u tr a l a g r e e s tr o n g ly a g r e e ppc1 i feel confident presenting information to a group of colleagues 1 2 3 4 5 ppc2 i feel confident in representing my work area in meetings with management. 1 2 3 4 5 bb1 i am always ready to help or to lend a helping hand to those around me not at all 1 2 3 4 5 very much bb2 i am willing to give of my time to help others who have work-related problems not at all 1 2 3 4 5 very much bb3 i help other team members with heavy work loads not at all 1 2 3 4 5 very much bb4 i help others who have been absent not at all 1 2 3 4 5 very much journal of applied economics and business studies, volume.4, issue 3 (2020) 15-54 https://doi.org/10.34260/jaebs.442 53 ppc3 i feel confident contributing to discussions about the company’s strategy 1 2 3 4 5 ppc4 i feel confident helping to set targets/goals in my work area. 1 2 3 4 5 ppc5 i feel confident contacting people outside the company (e.g., suppliers, customers) to discuss problems. 1 2 3 4 5 ppc6 i feel confident presenting information to a group of colleagues. 1 2 3 4 5 ppc7 if i should find myself in a jam at work, i could think of many ways to get out of it. 1 2 3 4 5 ppc8 at the present time, i am energetically pursuing my work goals 1 2 3 4 5 ppc9 there are lots of ways around any problem. 1 2 3 4 5 ppc10 right now i see myself as being pretty successful at work 1 2 3 4 5 ppc11 i can think of many ways to reach my current work goals. 1 2 3 4 5 ppc12 at this time, i am meeting the work goals that i have set for myself. 1 2 3 4 5 ppc13 when i have a setback at work, i have trouble recovering from it, moving on.(r) 1 2 3 4 5 ppc14 i usually manage difficulties one way or another at work. 1 2 3 4 5 ppc15 i can be “on my own,” so to speak, at work if i have to. 1 2 3 4 5 ppc16 i usually take stressful things at work in stride. 1 2 3 4 5 ppc17 i can get through difficult times at work because i’ve experienced difficulty before. 1 2 3 4 5 ppc18 i feel i can handle many things at a time at this job. 1 2 3 4 5 ppc19 when things are uncertain for me at work, i usually expect the best. 1 2 3 4 5 ppc20 if something can go wrong for me work-wise, it will.(r) 1 2 3 4 5 ppc21 i always look on the bright side of things regarding my job. 1 2 3 4 5 ppc22 i’m optimistic about what will happen to me in the future as it pertains to work. 1 2 3 4 5 ppc23 in this job, things never work out the way i want them to.(r) 1 2 3 4 5 ppc24 i approach this job as if “every cloud has a silver lining.” 1 2 3 4 5 6. please place a tick in one option for each statement to rate yourself on below given criteria at the state of employee silence s tr o n g ly d is a g r e e d is a g r e e n e u tr a l a g r e e s tr o n g ly a g r e e es1 (if) i did not believe my concerns would be addressed 1 2 3 4 5 es2 management did not appear interested in hearing about these types of issues 1 2 3 4 5 es3 no one is interested in taking appropriate action 1 2 3 4 5 es4 i did not feel i would be taken seriously 1 2 3 4 5 es5 i did not think it would do any good to speak up 1 2 3 4 5 es6 (d) i felt it is dangerous to speak up 1 2 3 4 5 es7 i remain silent to protect myself from harm 1 2 3 4 5 namra jamshaid & sadia arshad 54 es8 i feel it was risky to speak up 1 2 3 4 5 es9 i believe that speaking up may negatively impact my career 1 2 3 4 5 es10 i am afraid of adverse consequences (e.g., being criticized, losing my job) 1 2 3 4 5 es11 (r) i didn’t want to harm my relationship with another individual 1 2 3 4 5 es12 i did not want to create tension with co-worker 1 2 3 4 5 es13 i remain silent to avoid conflict with another individual 1 2 3 4 5 es14 i remain silent to protect my relationship with another individual 1 2 3 4 5 es15 i remain silent to avoid hurting someone’s feelings 1 2 3 4 5 journal of applied economics and business studies, volume. 5, issue 3 (2021) 77-104 https://doi.org/10.34260/jaebs.535 77 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x inflation dynamics in pakistan fauzia sohail1, ambreen fatima2 & sanam wagma khattak3 1 assistant professor, applied economics research centre (aerc) university of karachi 2 associate professor, applied economics research centre (aerc) university of karachi 3 lecturer, department of economics, university of peshawar (uop) abstract this study explains the correlation among consumer price behavior and inflation in pakistan by employing weekly micro data of retail price survey from october 2013 till september 2016, which is considered as the period of high inflation. inflation is a product of extensive margin (the frequency of price change) and intensive margin (magnitude of price change). the study decompose the variance of inflation into contribution from both margins thus to determine if the prices of various groups of commodities follows state dependent pricing model or time dependent pricing model. the extensive margin is a source of inflation in most of the state dependent pricing model while intensive margin is a basis of inflation in time dependent pricing model. the analysis reveals a greater correlation of inflation with the intensive margin of price change compared to the extensive margin. mounting inflation is although related more with greater frequencies of the price increase and lower frequency of price decrease, however, the size of the price increase and decrease also plays an important role. findings of the study reveals important role of price change behavior in explaining the inflation dynamics and are believed to assist policy makers in assessing the pricing mechanism and inflation dynamics in pakistan. keywords prices, inflation, commodities, monetary policy jel classification e31, e52, e12 1. introduction this study investigates new evidence on price setting mechanism in pakistan during the fauzia sohail, ambreen fatima & sanam wagma khattak 78 period of volatile inflation. we employ weekly data of retail price survey from october 2013 till september 2016. over the period under study, inflation rate mounted at 8.62 percent in 2013-14, while declined to less than 3 percent in 2016-17, which is thirteen years lowest in the country. the study significantly contributes to the literature by explaining the dynamics of pricing mechanism in various ways. firstly, the study endeavors to determine the most appropriate price-setting model for various product groups. price adjustments of various product groups whether follows the time-dependent pricing model, like, taylor (1980) and calvo (1983), state-dependent pricing model, like, caplin and spulber (1987) or a combination of both models like, dotsey et al., (1999) were left unexplored for the case of pakistan. further, the study significantly investigates the impact of intensive and extensive margins of inflation. aggregate inflation is a product of intensive margin, that is, the average magnitude of price change and an extensive margin, that is, the proportion of commodities changing prices. this identity is employed in the study for decomposing the variance of inflation into its components, that is, intensive and extensive margins of price change. thirdly, the study helps in identifying the relationship of inflation with a frequency of price increase and decrease. it is postulated that the magnitude of the price increase and decrease also plays an important role in inflation dynamics of pakistan. theoretically it has been explained in the literature that prices of some commodities respond swiftly to policy reform or shock to the economy, while stiffed prices of various other commodities remains tranquil. in micro founded price setting models, the degree of price stiffness or rigidness is explained by two models. first is the time dependent price model in which the instances of revising the price of any commodity are considered exogenous. second is the state dependent price model in which prices of the commodities are sensitive to the shocks and state of the economy, thus instances of price change were considered endogenous. however, literature supports that during volatile or high inflation periods, performance of time dependent pricing model becomes limited. internationally, the last two decades witnessed a significant contribution of studies that investigated whether price adjustments follows the time-dependent pricing mechanism or state-dependent pricing mechanism. klenow and kryvtsov (2005) considered pioneer in enlightening the inflation dynamic by developing the methodology that assists in determining the most suitable price setting assumption for any country. the study used the cpi database for the us and discerned the co-movement of the aggregated inflation with the intensive margin rather than the extensive margin. the empirical results, however, were not observed to follow any of the time-dependent but are found more closely associated with the statedependent models. the methodology was then followed by the number of other studies, for journal of applied economics and business studies, volume. 5, issue 3 (2021) 77-104 https://doi.org/10.34260/jaebs.535 79 instance, ikeda and nishioka (2007), kurri (2007) gagnon (2009), nchake, et al. (2014). gagnon (2009) found that in low inflation countries, extensive margin of price change have only weak correlation with inflation, whereas, intensive margin shows strong correlation with inflation. in case of pakistan, there exist two important studies on pricing behaviour. first one is malik et al., (2010) and the second is choudhary et al., (2011). both studies are based on the firm-level primary survey through field visits and one on one interview. the review of the two studies shows considerable differences in the findings, despite the fact that the first study, i-e., malik et al. (2010) is a subset of the second study, choudhary et al. (2011) in terms of sample coverage. this shows that studies based on a primary survey of a specific geographical region may not be generalized. hence another important contribution of the present study is that the results and analysis represent the whole country. the study thus contributes in terms of dataset as well as the rigorous methodology for determining the pricing mechanism in seventeen main cities of pakistan. more specifically study observes the inflation dynamic by employing the methodology developed by klenow and kryvtsov, (2005). the study is organized as follows. the next section illustrates the theoretical foundation of the study. section 3 presents brief review of existing literature, where the next section describes the micro price data employed in the study. section 5 elucidates methodological and technical aspect of the study. section 6 reveals the empirical results and analysis in detail. the last section concludes the study. 2. review of existing literature 2.1 theoretical review theoretically it has been explained in the literature that prices of some commodities respond swiftly to policy reform or shock to the economy, while stiffed prices of various other commodities remains tranquil. in micro founded price setting models, the degree of price stiffness or rigidness is explained by two models. first is the time dependent price model and the second is the state dependent price model. time dependent pricing models were originally recognized as simple staggered contract price model. taylor (1980) and calvo (1983) are considered pioneer in the development of these models. with the passage of time, extension in staggered pricing models proved useful in explaining the phenomena with viable policy decisions. the staggered or time dependent price model developed by taylor (1980) is also known as fixed duration model. the model was designed to have various key features, as explained by taylor (1999). for instance, (i) price setting mechanism is considered fauzia sohail, ambreen fatima & sanam wagma khattak 80 staggered or unsynchronized, which means that different firms or price setter alter their prices in different periods, (ii) there is a fixed period of time say a quarter, a month, a year or so on – usually known as contractual period between the buyer and seller, in which price remains unaltered and is independent of events or shocks transpire in the economy during the contractual interlude. hence, in this model, the instances of revising the price of any commodity are considered exogenous. the contractual agreements are unsynchronized and of varied length, hence these interludes overlap one another, thus the term “staggered contracts” was coined. (iii) price setters consider relevance of their decision of revising prices in other close markets as well. considered these foremost aspects, taylor (1980) postulated the deterministic price alteration mechanism by considering the firm to alter the prices after every tth period. however, on generalizing the model, taylor (1999) incorporated the series of contract types with varied lengths, say, t1, t2, t3 and so on. hence, in the time dependent pricing model, not all price setters change their price on a single point in time, so the aggregate price level adjusts slowly. the fixed duration model was further extended by blanchard (1983), blanchard (1987). the study introduces the stage-of-process model in which the input price was shown to affect the output price which continues to affect the input price of another firm and so on. this process thus engenders the dynamics in unsynchronized price setting mechanism. christiano (1985) extended the idea of staggered contract price model for cases in which contract lengths are shorter than the time period required for data collection. calvo (1983) developed the random duration version of the staggered contract model by assuming the contract lengths as stochastic rather than deterministic. the contract durations are supposed to be identically and independently distributed. according to this model contract finished randomly following a geometric or exponential distribution. the main difference between the taylor and calvo staggered contract price models is that in taylor model the contract length or the duration of fixed price is decided ex ante, whereas, in calvo model, the duration in which price remains fixed is not known. later on, time dependent models were criticized on the basis of exogeneity assumption. theories and models then begun to develop in which the instances of price change were considered endogenous rather than an exogenous matter. according to these new state dependent models, prices of the commodities are sensitive to the shocks and state of the economy. for instance, caplin and spulber (1987) accentuated the significance journal of applied economics and business studies, volume. 5, issue 3 (2021) 77-104 https://doi.org/10.34260/jaebs.535 81 of endogenous instances of price modification and eliminates the infrequent and large price change. the model strongly criticized the assumption of exogeneity and predetermined contractual agreements and assumed the (s, s) policies of sheshinski and weiss (1988) in modeling the endogenously determined price level. sheshinski and weiss (1988) emphasize the significance of shocks in reducing the interval of price adjustment. caplin and spulber, (1987) observed that on the whole, the monetary shocks lead the revision in prices by most of the price setters. the study also observed the money neutrality in the presence of relative price variability. an important modification in the state dependent pricing model of caplin and spulber (1987) is made by dotsey, king and wolman (1999). dotsey et. al. adapted the state dependent pricing rule to allow for random staggered contract model of calvo (1983). the model assumes that the fixed cost of changing price is random across firms thus modified the model for empirical application and policy investigations. according to this model, the proportion of firms that aimed to alter their prices in any particular period enhance with the boost in the inflation rate. according to this model price setter with small menu cost prefer to alter prices frequently as compared to the firms with soaring fixed cost. most recently, golosov and lucas (2007) extend the above described models by addressing the two key issues. first is the timings of price adjustment and the other is magnitude of adjustment. like other models, this model also assumed menu cost as a source of nominal rigidity. the distinctive facet of this model is that it consider impact of both idiosyncratic (i-e productivity shock like change in technology or preferences) as well an aggregate shock like general inflation on price adjustment for a typical firm. the model developed by dotsey et al. (1999) elucidates that the menu cost is affected by idiosyncratic shocks that in turn influence the firm’s decision of price adjustment. hence all firms that adjust their prices reached the same new price. whereas, golosov and lucas (2007) proved that idiosyncratic shocks alter the productivity and thus the price adjustment. this model arrests the observed heterogeneity in pricing mechanism. theoretical development of the price setting models led most of the international literature to explore the micro fundamentals of monetary and macroeconomic theories and their policy viability. hence, there is a growing empirical literature that aimed to determine the dynamic features of price setting behavior, which is described in the next section. fauzia sohail, ambreen fatima & sanam wagma khattak 82 2.2 empirical review klenow and kryvtsov (2005) considered pioneer in enlightening the inflation dynamic by developing the methodology that assists in determining the most suitable price setting assumption for any country. the study used cpi database for us from the period january 1988 to january 2005 and discerned the co movement of the aggregated inflation with the intensive margin (size of the price change) rather than the extensive margin (number of items reporting price change). the rationale was that the direction of price change (increase and decrease), which was strongly correlated with inflation counteracted each other. the empirical results, however, were not observed to follow any of the time dependent or first generation state dependent models but are found more closely associated with the second generation state dependent models. following the methodology developed by klenow and kryvtsov (2005), number of studies sought to decompose the variance of inflation into its components, that is, intensive and extensive margins of price change. the study ikeda and nishioka (2007) applied klenow and krytsov’s decomposition on japan’s cpi micro data for four years period from 2000 to 2004. the study observed that intensive margin explained most of the fluctuations in the japan’s inflation thus concluded in favour of time dependent pricing model, however, did not rule out the evidences of state dependent pricing model as well, considering the fact that the study period was relatively stable. kurri (2007) applied the similar klenow and krytsov decomposition to the finnish cpi micro data for the period february 1997 to december 2004 under the same ipn project. similar to the above results, the magnitude or the intensive margin of price change seemed to explain most of the variation in inflation. however, further estimation also indicates the presence of state dependent pricing rules for finnish economy. another valuable study, nchake et al. (2014) employed the lesotho cpi database over about seven year period from march 2002 to december 2009. the findings of the research study corroborate with the findings of the above mentioned studies as strong correlation of the intensive margin of price change was detected with the variance of inflation. the findings support the time-dependent price setting behavior for the small land-lock state. however, taking into account the gagnon (2009) critique, the study detected that the mounting inflation augmented the frequency of price increase and lowered the occurrence of price decrease that contradicts the assumptions of time dependent model. the results thus signify the importance of the direction of price change in driving the inflation variation. journal of applied economics and business studies, volume. 5, issue 3 (2021) 77-104 https://doi.org/10.34260/jaebs.535 83 in pakistan, mainly two approaches were used to investigate pricing related issues. first is by conducting a primary survey of the firms, which is relatively a costly procedure. whereas, the second approach is to use micro data underlying the consumer, sensitive, wholesale and producer price indices. adopting the first approach, malik et al. (2010) examine the price setting behavior of pakistani firms by conducting a primary survey from industrial cities of punjab namely faisalabad, gujrat, gujranwala and sialkot. the study determined that about 70 percent of firms follow either state dependent pricing strategies or blend of time and state dependent strategies. survey revealed that most of the firms usually alter their prices in the month of june and july. energy and raw material costs along with inflation were determined to increase prices of commodities while rivals’ pricing strategies, changes in demand and raw material cost were found to be the main sources of price decrease. indenture with customers and other fixed agreements of prices were believed to cause price rigidity in the country. choudhary et al. (2011) conducted 1086 structured interviews in sind and punjab provinces for examining the price-setting behavior of formal firms of the manufacturing and the service sectors in pakistan. the study found significantly high frequency of price change that resulted in diminishing the monetary policy impact in the country. exchange rate and cost shocks were revealed to impact significantly on price setting behaviors of the firms. in contrast to the above study, this study found that 52 percent of the firms follow time dependent policies compared to 25 percent firms that follow state dependent policies. however, only 10 percent firms follow both policies. to the best of our knowledge, the above two studies are the only research work exist till date that employed the first approach of conducting a questionnaire survey in pakistan on the subject. although few studies adopted the second approach of employing the micro level price dataset as well, however, aims and objectives of those studies are significantly different from the current study. for instance, hanif et al. (2016) employed monthly data for inflation from july 1992 to june 2014 aimed to determine the relationship between international commodities’ price volatility and inflation in pakistan. similarly, hanif (2012) also employed monthly micro price data of consumer price index (cpi) and wholesale price index (wpi) of 374 commodities from july 2001 to june 2011 to determine the impact of global oil and food prices on food inflation in pakistan. hence, this study is believed to be the first study that aims to examine the price setting mechanism in pakistan by analyzing the micro level price data. fauzia sohail, ambreen fatima & sanam wagma khattak 84 3. data description 3.1.theoretical model this study employs weekly data of retail price survey from third week of october 2013 till fourth week of september 2016, conducted by pakistan bureau of statistics (pbs) for computing the sensitive price index (spi). in the survey, prices of fifty three identical commodities prevailing in seventeen major cities of pakistan are surveyed from specified markets each week. the entire commodities of retail price survey of seventeen appraised cities are thus included in the study. overall there are 134,249 observations used in the study. the published data is the city average of individual price quote across markets. hence, the data employed is the average price statistics of each commodity per city per week. for example, the price of 1 kg of onion is surveyed in 4 different markets of islamabad. the bureau averaged four prices and published the average city price of a kg of onion. further, fifty three commodities are grouped into ten product categories based on the nature of the product. the commodities grouped in each category vary and listed in table a3. share of product categories in overall commodities is shown in table 1. table 1: share of product groups product groups share (%) staple food 18.9 perishable food 9.4 clothing & footwear 13.2 energy goods 13.2 other necessities 9.4 cooked food 7.5 meat & dairy 13.2 cooking oil & ghee 7.5 processed food 5.7 cigarette 1.9 total 100 source: calculated from pbs 4. methodology this section explains the methodology developed by klenow and kryvtsov (2005). inflation (πt) is the product of extensive margin (em) and intensive margin (im) of price change. extensive margin is computed as the proportion of products changing prices, journal of applied economics and business studies, volume. 5, issue 3 (2021) 77-104 https://doi.org/10.34260/jaebs.535 85 whereas, the intensive margin is the average magnitude of the price change of those products. mathematically, 𝜋𝑡 = 𝐸𝑀 ∗ 𝐼𝑀 (1) whereas, 𝑓𝑟𝑒𝑞𝑡 = 𝐸𝑀 = [ ∑ 𝑤𝑗𝑡𝑋𝑗𝑘𝑡 𝑛 𝑗,𝑘=1 ] and, 𝑠𝑖𝑧𝑒𝑡 = 𝐼𝑀 = [ ∑ 𝑤𝑗𝑡(𝑝𝑗𝑘𝑡 − 𝑝𝑗𝑘𝑡−1) 𝑛 𝑗,𝑘=1 ∑ 𝑤𝑗𝑡𝑋𝑗𝑘𝑡 𝑛 𝑗,𝑘=1 ] hence equation (1) becomes, 𝜋𝑡 = [∑ 𝑤𝑗𝑡𝑋𝑗𝑘𝑡 𝑛 𝑗,𝑘=1 ] ×[ ∑ 𝑤𝑗𝑡(𝑝𝑗𝑘𝑡−𝑝𝑗𝑘𝑡−1) 𝑛 𝑗,𝑘=1 ∑ 𝑤𝑗𝑡𝑋𝑗𝑘𝑡 𝑛 𝑗,𝑘=1 ] = ∑ 𝑤𝑗𝑡(𝑝𝑗𝑘𝑡 − 𝑝𝑗𝑘𝑡−1) 𝑛 𝑗,𝑘=1 (2) as previously defined, 𝑋𝑗𝑘𝑡 is the price change indicator of item j at time t and is defined as: 𝑋𝑗𝑘𝑡 = { . .0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 1 𝑖𝑓𝑃𝑗𝑘𝑡 ≠ 𝑃𝑗𝑘𝑡−1 , 𝑤𝑗𝑡is the cpi weight of the product and 𝑃𝑗𝑘𝑡 is the logarithm of the price. for determining the significance of em and im terms, klenow and kryvtsov (2005) decompose the variance of inflation [var (πt)] into the variance of size of price change [var(sizet)], the variance of frequency of price change [var(freqt)], and the covariance between the two terms. the required decomposition is obtained by the first order taylor series approximation of inflation (πt) around the average of freqtand sizet. 𝜋𝑡 ≅ 𝑓𝑟𝑒𝑞̅̅ ̅̅ ̅̅ .𝑠𝑖𝑧𝑒̅̅ ̅̅ ̅ + 𝑠𝑖𝑧𝑒̅̅ ̅̅ ̅(𝑓𝑟𝑒𝑞𝑡 − 𝑓𝑟𝑒𝑞̅̅ ̅̅ ̅̅ )+ 𝑓𝑟𝑒𝑞̅̅ ̅̅ ̅̅ (𝑠𝑖𝑧𝑒𝑡 − 𝑠𝑖𝑧𝑒̅̅ ̅̅ ̅) (3) where, 𝑓𝑟𝑒𝑞̅̅ ̅̅ ̅̅ and 𝑠𝑖𝑧𝑒̅̅ ̅̅ ̅ are the averages of 𝑓𝑟𝑒𝑞𝑡and 𝑠𝑖𝑧𝑒𝑡respectively. the variance of 𝜋𝑡 is then computed as: 𝑣𝑎𝑟 (𝜋𝑡) = 𝑓𝑟𝑒𝑞̅̅ ̅̅ ̅̅ 2𝑣𝑎𝑟(𝑠𝑖𝑧𝑒𝑡)⏟ 𝐼𝑀 𝑇𝑒𝑟𝑚 + 𝑠𝑖𝑧𝑒̅̅ ̅̅ ̅2𝑣𝑎𝑟 (𝑓𝑟𝑒𝑞𝑡)+ 2𝑓𝑟𝑒𝑞̅̅ ̅̅ ̅̅ .𝑠𝑖𝑧𝑒̅̅ ̅̅ ̅ 𝑐𝑜𝑣(𝑓𝑟𝑒𝑞𝑡,𝑠𝑖𝑧𝑒𝑡)⏟ 𝐸𝑀 𝑇𝑒𝑟𝑚 (4) equation 4 is very useful in determining the appropriate pricing models. it is shown that the variance of inflation [var (πt)] is composed of two terms. the first one is the im term and the second is em term. the im term in the above equation determine the variation in the inflation due to the variation in the magnitude or intensive margin of price fauzia sohail, ambreen fatima & sanam wagma khattak 86 change. however, the sdp term is also characterized by the extensive margin that captures the variation in inflation due to the variation in frequency and its covariance with size of price change. furthermore, klenow and kryvtsov (2005) found it significant to segregate inflation and the related components into terms with price increase (ijkt ↑ ) and price decrease (ijkt ↓ ). where; xjkt ↑ = { . . 0 otherwise 1 ifpjkt > pjkt−1 xjkt ↓ = { . . 0 otherwise 1 ifpjkt < pjkt−1 to decompose the inflation (πt) in to its positive and negative components, we can write; πt = πt + + πt − (5) where, πt + = freqt +. sizet + πt − = freqt −. (−sizet −) hence equation (5) can be written as, πt = freqt +. sizet + + freqt −. (−sizet −) (6) where, freqt + = [∑ wjtxjkt + n j,k=1 ] freqt − = [∑ wjtxjkt − n j,k=1 ] similarly, sizet + = [ ∑ wjtxjkt + (pjkt−pjkt−1) n j,k=1 ∑ wjt n j,k=1 xjkt + ] sizet − = [ ∑ wjtxjkt − (pjkt − pjkt−1) n j,k=1 ∑ wjt n j,k=1 xjkt − ] variance decomposition is then written as; var (πt) = {var(πt +)+ cov(πt +,πt −)}⏞ positive term + {var(πt −)+ cov(πt +,πt −)}⏞ negative term (7) journal of applied economics and business studies, volume. 5, issue 3 (2021) 77-104 https://doi.org/10.34260/jaebs.535 87 5. results and analysis it is clear from equation (1) that the changes in inflation can be determined either by the extensive margin (changes in the frequency of price) or the intensive margin (that is, the magnitude/size of the price change). the values of inflation〖(π〗_t), intensive margin 〖(size〗_t) and extensive margin 〖(freq〗_t) are estimated and reported in table 2 for all the commodities cumulatively and for nine identified groups individually. four weeks moving averages for cumulative data is plotted in figure 1 below, whereas, group wise figures are presented in annexure a4. table 2 reveals that the average weekly inflation is computed as 0.04 percent for entire commodities. the estimated values, however, varies considerably for various groups. for instance, energy group accounts for the lowest weekly average inflation rate of -0.42 percent while meat and dairy products and cooking oil & ghee exhibit the highest weekly average inflation rate of about 0.43 percent each group. the other low average inflation rate commodity groups are perishable food items’ group (-0.145 percent) and processed food items’ group (-0.11 percent).. in contrast, high average inflation rate commodity groups are identified as clothing & footwear (0.18 percent), other necessities (0.064 percent), staple food (0.053 percent) and cooked food items (0.15 percent). the weighted mean value of weekly frequency of price change or the extensive margin is 35 percent, whereas, the values are much volatile around this mean value with a standard deviation of 77 percent for entire commodities. the weighted average weekly size of price change or the intensive margin is diminutive at 0.03 percent with a standard deviation of 0.08 percent. table 2: average inflation and its components groups 𝝅𝒕 𝒇𝒓𝒆𝒒𝒕 𝒔𝒊𝒛𝒆 overall mean 0.04 (0.55) 35.01 (77.67) 0.03 (0.08) correlation with 𝜋𝑡 …. -0.22 0.35 group1: staple food items mean 0.053 (0.21) 12.85 (12.78) 0.014 (0.043) correlation with 𝜋𝑡 …. 0.076 0.45 group2: perishable food items mean -0.145 (0.33) 64.69 (39.106) 0.0009 (0.0076) correlation with 𝜋𝑡 …. -0.7 0.86 group3: clothing & mean 0.18 1.9 0.064 fauzia sohail, ambreen fatima & sanam wagma khattak 88 footwear (.2616) (2.45) (0.083) correlation with 𝜋𝑡 …. 0.75 0.66 group4: energy goods mean -0.42 (.53428) 29.57 (46.33) -0.004 (0.032) correlation with 𝜋𝑡 …. -0.62 0.54 group5: other necessities mean 0.06 (.1119) 1.29 (2.03) 0.085 (0.174) correlation with 𝜋𝑡 …. 0.32 0.48 group6: cooked food items mean 0.13 (.1204) 1.57 (1.75) 0.08 (0.066) correlation with 𝜋𝑡 …. 0.82 0.47 group7: meat & dairy products mean 0.43 (.8909) 77.51 (122.6) 0.03 (0.03) correlation with 𝜋𝑡 …. -0.24 0.42 group8: cooking oil & ghee mean 0.43 (0.26) 23.96 (35.98) -0.021 (0.023) correlation with 𝜋𝑡 …. 0.04 0.61 group9: processed food items mean -0.11 (0.56) 172.9 (170.26) -0.002 (0.005) correlation with 𝜋𝑡 …. -0.52 0.55 source: author’s calculations at the level of different groups, deviation of values is apparent in table 2. for instance, extensive margin for processed food items is as high as about 173 percent weekly, whereas, it is as low as 1.3 and 1.6 percent for groups 5 and 6. these results point towards the highly volatile behavior of wheat, sugar and tea prices during the period under study. it is noticed that commodity groups exhibited high frequency change provide evidences of small size of price change and vice versa. for instance, meat and dairy products’ group, processed food group, perishable food group and energy goods account for very small size of price change. in contrast, cooked food group, other necessities’ group and clothing & footwear group relatively large size of price change. journal of applied economics and business studies, volume. 5, issue 3 (2021) 77-104 https://doi.org/10.34260/jaebs.535 89 it is revealed that the intensive margin is strongly correlated with inflation in contrast to very weak association with the extensive margin. figure 1 clearly envisages this relationship where weekly inflation and weekly size of price change almost completely overlap each other during the entire period. the strong correlation between the two variables is also substantiated by table 2. for entire dataset the correlation between intensive margin and inflation is 35 percent, whereas extensive margin negatively correlates with inflation. this evidence indicates that more price decrease was found during the period under study compared to price increase. figure 1: intensive and extensive margin of inflation source: author’s illustration almost similar pattern is identified for most of the groups, for instance, staple food items, perishable food items, meat & dairy products, processed food items and other necessities group; where correlation of inflation with size of price change is strong compared to frequency of price change. however, few groups like clothing & footwear items and cooked food group shows the co-movement of inflation with both intensive and extensive margins. figures in the annexure a4 corroborate the findings of table 2. the variance decomposition of inflation for determining the significance of im and em term is obtained by employing equation (4). according to klenow and malin (2010), -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 1 6 1 1 1 6 2 1 2 6 3 1 3 6 4 1 4 6 5 1 5 6 6 1 6 6 7 1 7 6 8 1 8 6 9 1 9 6 1 0 1 1 0 6 1 1 1 1 1 6 1 2 1 1 2 6 1 3 1 1 3 6 1 4 1 1 4 6 inflation frequency of price change (divided by 10) size of price change (multiply by 10) fauzia sohail, ambreen fatima & sanam wagma khattak 90 intensive margin (im) drives most of the variation in inflation in the time dependent model explained by taylor (1980) and calvo (1983), whereas, extensive margin plays considerable part in various state dependent pricing models. product groups that follow tdp, var (freqt) [i-e, variance of the fraction of commodities experiencing price change] and cov(freqt, sizet) [covariance between fraction of commodities experiencing price change and magnitude of price alteration] become zero as the price alteration become invariable in tdp model. whereas, if any product group follows sdp model, the variance of inflation is swayed by the em term as price alters frequently in response to economic or political shocks. table 3 reveals the findings of variance decomposition as im and em terms. im term explains about 70 percent of the variance in inflation cumulatively, whereas, em term explain only about 30 percent of the variation. these findings corroborate with the above findings and explanation. the international literature also reveals the similar pattern of variance decomposition for us, like, klenow and kryvtsov (2008) and for lesotho, like, nchake et al. (2014), hence, validates the findings of this study. at the level of product groups, tdp term is responsible for more than 70 percent of variation in inflation in most of the groups, for instance, staple food group, perishable food group, energy group, other necessities group and processed food group. meat & dairy products group revealed more than 50 percent of variation in inflation because of the tdp term. whereas, three product groups, that is, clothing & footwear group, cooking oil & ghee group and cooked food group revealed sdp terms to cause variation in inflation more than fifty percent. table 3: variance decomposition groups time dependent pricing model (tdp) state dependent pricing model (sdp) overall 70.19 29.81 group1: staple food items 104.31 -4.31 group2: perishable food items 110.15 -10.15 group3: clothing & footwear 39.28 60.72 group4: energy goods 89.52 10.48 group5: other necessities 71.90 28.10 group6: cooked food items 33.45 66.55 journal of applied economics and business studies, volume. 5, issue 3 (2021) 77-104 https://doi.org/10.34260/jaebs.535 91 source: authors calculation klenow and kryvtsov (2005) found it useful to further estimate the value of inflation by taking into account price increase and decrease separately thus to determine the cause of strong correlation of inflation with the intensive margin and weaker relationship with extensive margin. the estimation also provides better justification in response to the argument raised by (gagnon, 2009) that the correlation between extensive margin and inflation is low as movements in the frequency of price increase and decrease partly offset each other nchake et.al. (2014). table 4 summarizes the values of inflation and its components separately for price increase and decrease. at the aggregated level, the weighted average weekly frequency of price increase is 17.76 percent, which is a little higher than weighted average weekly frequency of price decrease (17.24 percent). these fractions show weak correlation with inflation, which is also envisaged in figure 2 below. the size of price increase and decrease show positive correlation with inflation. however, magnitude of correlation results prove relatively strong association of freqt + and freqt − with inflation than sizet + and sizet −. almost similar pattern is identified at some of the group levels, for instance, staple food group, perishable food groups, energy groups, meat & dairy food group and processed food group. in clothing & footwear group frequency and size of price increase plays role in increasing inflation whereas, weak and zero correlation of frequency and size of price decrease shows no evidence of price decrease in this group during the period under study. group7: meat & dairy products 56.67 43.33 group8: cooking oil & ghee 46.07 53.93 group9: processed food items 106.12 -6.12 fauzia sohail, ambreen fatima & sanam wagma khattak 92 figure 2: frequency of price increase & decrease and inflation source: author’s illustration klenow and kryvtsov (2005) also determine the similar relationship of variables. the study regress inflation with price increase and decrease variables and although found statistically significant relationship of inflation with all four variables, that is, freqt +, freqt −, sizet + and sizet −, however, found more close association of inflation with freqt + and freqt − rather than sizet + and sizet −. similarly, (nakamura and steinsson, 2008) found strong correlation of freqt + and freqt − with inflation. this proves the above assertion that although intensive margin (im) is important in overall rise in inflation, however, frequency of price change in also significant. table 4: inflation and its components for price increase and decrease 𝝅𝒕 𝒇𝒓𝒆𝒒𝒕 + 𝒇𝒓𝒆𝒒𝒕 − 𝒔𝒊𝒛𝒆𝒕 + 𝒔𝒊𝒛𝒆𝒕 − 𝝅𝒕 + 𝝅𝒕 − overall mean 17.76 (38.08) 17.24 (40.82) 0.08 (0.114) -0.07 (0.113) 1.15 (2.55) -1.107 (2.67) correlation with 𝜋𝑡 -0.089 -0.34 0.065 0.03 -0.1 0.3 group1: staple food items mean 7.15 (7.00) 5.69 (6.37) 0.06 (0.05) -0.05 (0.035) 0.31 (0.3) -0.26 (0.31) correlation with 𝜋𝑡 0.29 -0.17 -0.05 0.35 0.32 0.38 group 2: perishable food items mean 30.75 (17.69) 33.94 (21.78) 0.14 (0.06) -0.13 (0.056) 4.8 (4.1) -4.95 (4.3) correlation with 𝜋𝑡 -0.64 -0.73 -0.45 0.4 -0.55 0.6 -1.5 -1 -0.5 0 0.5 1 1.5 2 1 6 1 1 1 6 2 1 2 6 3 1 3 6 4 1 4 6 5 1 5 6 6 1 6 6 7 1 7 6 8 1 8 6 9 1 9 6 1 0 1 1 0 6 1 1 1 1 1 6 1 2 1 1 2 6 1 3 1 1 3 6 1 4 1 1 4 6 frequency of price increase (divided by 10) frequency of price decrease (divided by 10) inflation journal of applied economics and business studies, volume. 5, issue 3 (2021) 77-104 https://doi.org/10.34260/jaebs.535 93 group 3: clothing & footwear mean 1.88 (2.44) 0.02 (0.207) 0.104 (0.08) -0.07 (0.00) 0.18 (0.26) -0.001 (0.02) correlation with 𝜋𝑡 0.76 -0.07 0.48 0.00 0.99 0.07 group 4: energy goods mean 1.63 (19.84) 18.94 (27.37) 0.05 (0.024) -0.05 (0.02) 0.63 (1.37) -1.05 (1.66) correlation with 𝜋𝑡 -0.46 -0.7 0.075 -0.013 -0.39 0.64 group5: other necessities mean 1.01 (1.64) 0.28 (0.6) 0.21 (0.36) -0.19 (0.52) 0.12 (0.4) -0.06 (0.41) correlation with 𝜋𝑡 0.44 -0.11 0.062 0.28 0.2 0.07 group 6: cooked food items mean 1.55 (1.75) 0.015 (0.105) 0.107 (0.05) -0.103 (0.08) 0.13 (0.12) -0.002 (0.02) correlation with 𝜋𝑡 0.82 0.085 0.106 -1.00 0.98 -0.11 group7: meat & dairy products mean 44.07 (59.99) 33.44 (65.02) 0.05 (0.03) -0.06 (0.02) 2.67 (4.3) -2.24 (4.5) correlation with 𝜋𝑡 0.07 -0.37 0.006 0.44 -0.17 0.36 group 8: cooking oil & ghee mean 8.50 (17.03) 15.46 (19.75) 0.03 (0..03) -0.04 (0.02) 0.16 (0.25) -0.45 (0.34) correlation with 𝜋𝑡 0.17 -0.082 -0.25 0.37 0.11 0.68 group 9: processed food items mean 87.22 (80.37) 85.68 (92.2) 0.04 (0.02) -0.07 (0.06) 1.96 (1.13) -2.07 (1.47) correlation with 𝜋𝑡 -0.45 -0.57 0.015 0.04 -0.47 0.74 source: author’s calculation table 5 below shows the variance decomposition of inflation incorporating the positive and negative aspects of inflation. table shows that, cumulatively, price increase contributes negatively by 80 percent of variance in inflation, whereas, 180 percent of variation is attributed to inflation contribution of price decrease. hence it is proved that negative component of variance in inflation contributes in reducing the overall inflation in pakistan from 2013 till 2016. table 5: variance decomposition groups positive term negative term overall -80.1 180.1 group1: staple food items 43.5 56.5 group2: perishable food items -328.0 428.0 group3: clothing & footwear 99.3 0.7 group4: energy goods -112.7 212.7 group5: other necessities -325.7 425.7 group6: cooked food items 101.5 -1.5 group7: meat & dairy products -116.7 216.7 group8: cooking oil & ghee 9.5 90.5 group9: processed food items -94.5 194.5 fauzia sohail, ambreen fatima & sanam wagma khattak 94 source: author’s calculation composition of positive and negative terms, however, varies for different groups. significant difference is observed for the clothing & footwear group and cooked food items’ group, where almost all the variation in inflation is explained by the price increase. it means that products included in this group seldom experienced any price decrease. however, variation in inflation in perishable food group, energy group, other necessities group, meat & dairy products, cooking oil & ghee group and processed food group is associated mostly by the negative term. 6. conclusion this study employs the methodology developed by klenow and kryvtsov (2005) firstly to determine the most appropriate price setting model for various product groups. secondly, the study determines the impact of intensive and extensive margins of price on inflation. the methodology also helps in identifying the relationship of inflation with frequency and magnitude of price increase and decrease. all these objectives are accomplished by employing the micro level price statistics the study describe the price setting behavior across seventeen large cities of pakistan by employing the micro data from retail price survey of fifty three commodities. the survey was conducted by pakistan bureau of statistics, for the purpose of constructing sensitive price index (spi). the weekly dataset employed is from 3rd week of october, 2013 till 4th week of september 2016. in this way a total of 134,249 observations are included in the study. further, these 53 commodities are then categorized into 10 product groups. the results of this chapter substantiate the findings of ikeda and nishioka (2007) empirically. in the case of pakistan at-least five product groups follows time dependent pricing model. these product groups are staple food group, perishable food group, energy group, processed food group, and other necessities group. whereas, three product groups, that is, clothing & footwear group, cooking oil & ghee group and cooked food group revealed sdp terms to cause variation in inflation more than fifty percent. the analysis reveals greater correlation of inflation with intensive margin of price change compared to extensive margin. it is also noticed that commodity groups exhibited high frequency provide evidences of small size of price change and vice versa. mounting inflation is although related more with greater frequencies of price increase and lower frequency of price decrease, however, size of price increase and decrease also plays an important role. cumulatively, price increase contributes negatively by 80 percent of variance in inflation, whereas, 180 percent of variation is attributed to inflation contribution of price decrease. group10: cigarette -80.1 180.1 journal of applied economics and business studies, volume. 5, issue 3 (2021) 77-104 https://doi.org/10.34260/jaebs.535 95 hence it is proved that negative component of variance in inflation contributes in reducing the overall inflation in pakistan from 2013 till 2016. overall, results reveal the important role of price change behaviour in explaining the inflation dynamics and are believed to assist policy makers in assessing the pricing mechanism and inflation dynamics in pakistan. the study found evidences of frequent price change for most of the product groups follows time dependent pricing model. according to choudhary et al. (2011), monetary policy significantly influence prices with longer duration compared to the smaller ones. however, policies to stabilize prices of these products depend on the level of persistency of price hike. for instance, in case of pakistan, prices of perishable commodities as well as meat & dairy products rise as a result of soaring demand on special occasions (like eid ul azha, moharram, ramadan etc.) or because of adverse weather conditions and floods. these shocks are mostly temporary in nature. the condition, however, points towards the imperfect marketing mechanism and failure of external policy. on the short run, the most viable policy option is to import perishable commodities like onion, tomato, potato, bananas and other fruits and vegetables, whose demands are expected to rise, from neighboring countries. otherwise, the spells of high prices persist in an economy. modest inflationary expectations and lack of customer resistance amid with hoarding and unsubstantiated rumors of crop failures also contributed towards the temporary rise and frequent adjustment in prices. the above policy option should amid with improved infrastructure, transportation system and modern storage mechanism. administrative control over prices could also be a good option only if proper check and balance would be ensured. in these cases of temporary price hike, monetary authorities are more concerned with underlying inflation rate thus to circumvent unnecessary instability in output and employment. in the case persistency observe in which food prices keeps on increasing for longer period of time, probably because of structural shift in demand for food products, the supply can only meet demand gradually by enhancing land and other inputs of farming. in the study, energy group although reveals relatively higher frequency of price decrease than price increase, however, magnitude of increase is relatively higher than decrease. the group follows both time dependent as well as state dependent model. according to sbp annual report 2014-15 and 2015-16 the decrease in oil price in international markets is transmitted into domestic market through domestic price fall of various energy goods, like, petrol, diesel etc. during most of the period under study but witnessed an increase of gas charges and electricity tariff. this policy action impacted directly to the production and transportation cost. rising price of energy slow down the fauzia sohail, ambreen fatima & sanam wagma khattak 96 productivity growth and increases inflation accelerating rate of unemployment and vice versa. results show that price setting mechanism for four product groups follow both time dependent as well as state dependent pricing models. this result suggests that macroeconomic model should incorporate both pricing strategies thus elucidate significant monetary transmission mechanism. it is proved that prices of some goods and services do not change instantaneously but remain fixed for longer period of time probably because of the cost involve in price alteration called the menu cost. literature also explains the customer’s anger over price increase as another significant explanation of price rigidity. by incorporating consumer anger in the macroeconomic model for the country, the study could significantly elucidate the impact of monetary tightening on inflation. limitations and scope for future research in this section, some limitations of the study are mentioned, as none can enfold all facets of research query. the dataset employed in this study is city average of individual prices across outlets, which is considered as the major limitation of this study. if the dataset at the outlet and product level would be available, the results and analysis are believed to be improved many folds. such a dataset would help in controlling the heterogeneity across outlets and cities. secondly, due to limited access of dataset, the study only employed most recent1 dataset, which was available from 3rd week of october, 2013 till 4th week of september, 2016. however, more comprehensive analysis could be done if the previous dataset would also make available for research purposes. availability of detailed dataset of prices would help in comparing various economic eras of the country, for instance, high and low inflation periods any particular period, when some economic or political shock was observed by an economy. this would help in analyzing the price setting behavior against the shock to an economy. this study is endowed with the promising scope for future research in countless dimensions depending on the curiosity and understanding of the researcher. however, our future plan is to extend the research by employing the producer’s price data. producer’s prices could be employed to investigate the price setting behavior at the firm level. producer’s prices are believed to be more rigid than consumer’s prices because of the long term contractual relationship between the agents. by extending the research in this direction, on one hand, this proposition could be tested and contribute to the literature on 1 this was most recent dataset, when the study was initiated journal of applied economics and business studies, volume. 5, issue 3 (2021) 77-104 https://doi.org/10.34260/jaebs.535 97 firm’s price setting behavior and on the other hand, a more comprehensive dsge modeling would be possible by employing classical maximum likelihood function with heterogeneous price setting behavior. this would also help in policy recommendations. conflict of interest all authors declare no conflicts of interest in this paper. references blanchard, o. j. (1983). price asynchronization and price level inertia, in dornbush, r. and simonson, m. author, inflation debt, and indexation. mit press, cambridge, 3-24. blanchard, o. j. (1987). aggregate and individual price adjustment. brookings papers on economic activity, 1: 57 122. calvo, g. (1983). staggered prices in a utility-maximizing framework. journal of monetary economics, 12(3), 383-398. caplin, a. s., and spulber, d. f. (1987). menu costs and the neutrality of money. the quarterly journal of economics, 102, 703-725. choudhary, m.a., naeem, s., faheem, a., hanif, n. and pasha, f. (2011). formal sector price discoveries: preliminary results from a developing countries. sbp working paper series, no. 42. christiano, l. j. (1985). a method for estimating the timing interval in a linear econometric model with application to taylor’s model of staggered contracts. journal of economic dynamics and control, 9, 363 404. dotsey, m., king, r. and wolman, a. (1999). state-dependent pricing and the general equilibrium dynamics of money and output. quarterly journal of economics, 114, 655-90. gagnon, e., (2009). price setting during low and high inflation: evidence from mexico. the quarterly journal of economics, 124(3), 1221-1263. golosov, m. and lucas, r., (2007). menu costs and phillips curves. journal of political economy . 115 (2) , 171-199. hanif m. n., (2012). a note on food inflation in pakistan. pakistan economic and social review, vol. 50, no. 2, 183-206 hanif, m. n., iqbal, j. and khan, i. n., (2016). global commodity prices and inflation in a small open economy. sbp working paper series, no. 76. ikeda, d. and nishioka, s. (2007). price setting behavior and hazard functions: evidence from japanese cpi micro data. bank of japan, working paper series, no.07-e-19. keynes, j. m. (1936). the general theory of employment, interest and money. london: macmillan (reprinted 2007). klenow, p. j., and kryvtsov, o. (2005). state-dependent or time-dependent pricing: does it matter for recent u.s. inflation? nber working paper. no. 11043. klenow, p. j. and kryvtsov o. (2008). state-dependent or time-dependent pricing: does it matter for recent us. inflation? the quarterly journal of economics. 123 (3), fauzia sohail, ambreen fatima & sanam wagma khattak 98 863–904. klenow, p. j. and malin b. a. (2010). microeconomic evidence on price-setting. nber working paper. no. 15826. kurri, s. (2007). price changes in finland: some evidence from micro cpi data. european central bank working paper series, no. 728. malik, w. s., satti, a. h. and saghir, g. (2010). price setting behavior of pakistani firms: evidence from four industrial cities of punjab. pide working paper, 65. nakamura, e., and steinsson, j. (2008). five facts about prices: a reevaluation of menu cost models. quarterly journal of economics 123, no.4, 1415-1464. nakamura, e. and steinsson, j. (2012). price rigidity: microeconomic evidence and macroeconomic implications. working paper, columbia university. nchake, m. a., edwards, l. and rankin, n. (2014). price setting behavior in lesotho: stylized facts from consumer retail prices. ersa working paper, no. 417 sheshinski, e. and weiss, y. (1983). optimum pricing policy under stochastic inflation. review of economic studies, 513-529. sheshinski, e. and weiss, y. (1988). staggered and synchronized price policies and multiproduct monopolies. tel-aviv, foerder institute for economic research, working paper, no. 24-78. taylor, j. (1980). staggered wage setting in a macro model. the american economic review, 108–113. taylor, j. b. (1999). the robustness and efficiency of monetary policy rules as guidelines for interest rate setting by the european central bank. journal of monetary economics, vol. 43(3), 655-679 journal of applied economics and business studies, volume. 5, issue 3 (2021) 77-104 https://doi.org/10.34260/jaebs.535 99 annexure a1: commodities included in survey with their units s. no. commodities units 1 wheat 10 kg 2 wheat flour (avg. quality) 10 kg 3 rice basmati (broken) kg 4 rice irri-6 kg 5 bread plain (medium size) each 6 beef kg 7 mutton kg 8 chicken (farm) kg 9 milk (fresh) ltr 10 curd kg 11 milk powder (nido) 400gm 12 egg (farm) doz 13 mustard oil kg 14 cooking oil (tin) 2.5 ltr. 15 vegetable ghee (tin) 2.5 kg 16 vegetable ghee (loose) kg 17 bananas doz 18 masoor pulse (washed) kg 19 moong pulse (washed) kg 20 mash pulse (washed) kg 21 gram pulse (washed) kg 22 potatoes kg 23 onions kg 24 tomatoes kg 25 sugar kg 26 gur kg 27 salt (powdered) kg 28 red chilies (powdered) kg 29 garlic kg 30 tea (packet) 200 gm. 31 cooked beef (plate) each 32 cooked daal (plate) each 33 tea (prepared) cup 34 cigarettes k-2 20's 35 long cloth mtr 36 shirting mtr 37 lawn mtr 38 georgette mtr 39 sandal gents (bata) pair 40 chappal sponge(bata) pair 41 sandal ladies (bata) pair 42 electricity chages (1 to 50) units 43 gas charges (upto 3.3719 mmbtu) mmbtu 44 kerosene ltr. 45 fire wood 40kg 46 energy saver each fauzia sohail, ambreen fatima & sanam wagma khattak 100 47 washing soap cake 48 match box each 49 petrol ltr 50 diesel ltr 51 l.p.g. each 52 telephone (local call) each 53 bath soap (life buoy) cake source: pbs a2: number of markets surveyed in each city cities markets 1. islamabad 4 2. rawalpindi 6 3. gujranwala 1 4. sialkot 1 5. lahore 7 6. faisalabad 2 7. sargodha 1 8. multan 3 9. bahawalpur 1 10. karachi 13 11. hyderabad 4 12. sukkur 2 13. larkana 1 14. peshawar 3 15. bannu 1 16. quetta 2 17. khuzdar 1 total 53 source: pbs journal of applied economics and business studies, volume. 5, issue 3 (2021) 77-104 https://doi.org/10.34260/jaebs.535 101 a3: commodities in each product group source: author’s illustration group1: staple food items group2: perishable food items group3: clothing & footwear group4: energy goods group5: other necessities group6: cooked food items group7: meat & dairy products group8: cooking oil & ghee group9: processed food items group10: cigarette wheat bananas long cloth electricity chages energy saver bread plain beef mustard oil wheat flour cigarettes rice basmati potatoes shirting gas charges washing soap cooked beef mutton cooking oil (tin) tea (packet) rice irri-6 onions lawn kerosene match box cooked daal chicken vegetable ghee (tin) sugar masoor pulse tomatoes georgette fire wood telephone tea milk vegetable ghee (loose) moong pulse garlic sandal gents petrol bath soap curd mash pulse chappal sponge diesel milk powder gram pulse sandal ladies l.p.g. egg gur salt red chilies fauzia sohail, ambreen fatima & sanam wagma khattak 102 a 4: intensive and extensive margins of inflation by product groups -0.100 -0.050 0.000 0.050 0.100 0.150 0.200 0.250 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 1 1 0 1 1 1 1 1 2 1 1 3 1 1 4 1 (a) staple food items inflation frequency of price change (divided by 10) size of price change -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 1 1 0 1 1 1 1 1 2 1 1 3 1 1 4 1 (b) perishable food items inflation frequency of price change (divided by 10) size of price change -0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50 1 1 0 1 9 2 8 3 7 4 6 5 5 6 4 7 3 8 2 9 1 1 0 0 1 0 9 1 1 8 1 2 7 1 3 6 1 4 5 (c) clothing & footwear frequency of price change (divided by 10) size of price change inflation -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 1 1 0 1 1 1 1 1 2 1 1 3 1 1 4 1 (d) energy goods inflation frequency of price change (divided by 10) size of price change journal of applied economics and business studies, volume. 5, issue 3 (2021) 77-104 https://doi.org/10.34260/jaebs.535 103 -0.40 -0.20 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1 1 0 1 9 2 8 3 7 4 6 5 5 6 4 7 3 8 2 9 1 1 0 0 1 0 9 1 1 8 1 2 7 1 3 6 1 4 5 (e) other necessities inflation frequency of price change size of price change 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 1 1 0 1 9 2 8 3 7 4 6 5 5 6 4 7 3 8 2 9 1 1 0 0 1 0 9 1 1 8 1 2 7 1 3 6 1 4 5 (f) cooked food inflation (multiplied by 100 frequency of price change size of price change -1.50 -1.00 -0.50 0.00 0.50 1.00 1.50 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 1 1 0 1 1 1 1 1 2 1 1 3 1 1 4 1 (g) meat & dairy products inflation frequency of price change (divided by 10) size of price change (multipled by 10) -0.14 -0.12 -0.10 -0.08 -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 1 1 3 2 5 3 7 4 9 6 1 7 3 8 5 9 7 1 0 9 1 2 1 1 3 3 1 4 5 (h) cooking oil & ghee inflation frequency of price change (divided by 100) size of price change fauzia sohail, ambreen fatima & sanam wagma khattak 104 source: author’s illustration -0.60 -0.50 -0.40 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 1 1 0 1 1 1 1 1 2 1 1 3 1 1 4 1 (i) processed food items inflation frequency of price change (divided by 100) size of price change 0.00 0.50 1.00 1.50 2.00 2.50 1 1 0 1 9 2 8 3 7 4 6 5 5 6 4 7 3 8 2 9 1 1 0 0 1 0 9 1 1 8 1 2 7 1 3 6 (j) cigarette inflation frequency of price change (divided by 10) size of price change (multiplied by 10) journal of applied economics and business studies, volume. 5, issue 4 (2021) 13-40 https://doi.org/10.34260/jaebs.542 13 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x socio-economic and mental health impacts of covid19 lockdown on students learning in pakistan mumtaz ahmed1, muhammad zeeshan younas2 & raja waqas shabbir3 1 assistant professor, department of economics, comsats university islamabad pakistan 2 phd scholar, school of economics, quaid-i-azam university islamabad pakistan 3 m.phil. scholar, school of economics, quaid-i-azam university islamabad pakistan abstract evaluation of the e-learning activities in this crisis period amid covid-19 has significant importance to propose effective mediations for the smooth running of learning and teaching. keeping this in mind, the present study takes a lead and identifies the key problems associated with online classes during the covid19 lockdown in pakistan. in addition, we examine the “stay at home” orders of government impacts on the mental health of students. the empirical analysis is based on a cross-sectional survey covering students from all educational groups across pakistan. a sample of 1,114 students is collected via a structured questionnaire constructed using google form and floated online to students in all educational institutions in pakistan. results indicate that students living in a city that has “stay at home” orders in place at the time of the survey is linked with a decrease in mental health by 0.525 standard deviations (p-value=0.0560), suggesting that “stay at home” orders have led to a significant reduction in the mental health of the students. students who report being stayed at home during lockdown have significantly better mental health scores. additionally, we provide a comprehensive descriptive statistical analysis of the online classes designed for lockdown on different socio-economic and mental health aspects. keywords covid-19, lockdown, online classes, mental health, pakistan jel classification a10, a11 mumtaz ahmed, muhammad zeeshan younas & raja waqas shabbir 14 1. introduction the covid-19 pandemic affected all spheres of life and sectors of the economy exacerbating the education crises in the developing world—where large disparities have already reduced the opportunities for vulnerable students. it is responsible for the largest disruption of education systems in modern history, affecting nearly 1.6 billion learners in around 200 countries across the globe. to slow the spread of the disease, lockdown measures have been put in place, limiting people’s ability to leave their homes and interact with other people. a major public health concern relates to how these measures impact the mental health of students. in this study, we attempt to examine the impact of lockdown measures on the mental health of students in pakistan. the closure of schools, colleges, universities, and other learning spaces has impacted nearly 94% of the world’s student population, and this figure goes up to 99% in low and lower-middle-income countries. there is an estimate that 23.8 million additional children may drop out permanently due to the harsh economic effects of the pandemic (un, 2020). unicef claims that at least 463 million school children have no access to distant learning during the strict lockdown. it is equivalent to one-third of the world's schoolchildren or 31% of the total global student population. pakistan lies in the south asian region—where 147 million students have no access to learning, making it the highest concentration of students without schools (unicef, 2020). pandemic compelled education institutions to end in-person instruction and switch to online teaching. cambridge university in the united kingdom was the first institution to move all face-to-face teaching online on march 24, 2020 (castle, 2020). in pakistan, the education institutes were also forced to replace the traditional approach with online teaching, using technology tools like google classroom, microsoft teams, whatsapp, zoom, etc. online education, by and large, remains a far distant reality for millions around the world and in pakistan in particular. the students in under-developed remote areas of pakistan such as balochistan, fata, khyber pakhtunkhwa, and gilgit-baltistan experienced difficulties due to unavailability and in some cases poor connectivity of internet facilities, with poor students having no access to gadgets necessary to acquire online educations (tahir, 2020). pakistan’s higher education commission (hec) encouraged education institutions to use the learning management system (lms) to carry out educational activities online across the country via an official notification. moreover, it instructed the institutions facing technological and spatial limitations to remain closed for any academic activities journal of applied economics and business studies, volume. 5, issue 4 (2021) 13-40 https://doi.org/10.34260/jaebs.542 15 till may 31, 2020 (hec, 2020). the roadmap is clear that online education is the only alternative in this pandemic and the foreseeable future. pakistan’s educational infrastructure lacks modern technology except for some reputed varsities. most of the topranked varsities are already working on online teaching programs which facilitated them to accelerate it during the pandemic. the government’s timely decision to lift the complete lockdown and to replace it with the selective smart lockdown carries enormous social and economic implications and possibly, it will have lasting effects on the plight of millions of students (children and youth). the un estimates that 23.8 million additional pre-primary to tertiary students will drop out and lose access to school next year due to pandemics' economic impact alone (un, 2020). the shutting of the learning institutions due to the covid-19 outbreak leads to an unprecedented impact on the educational sector. during the lockdowns (both strict and smart), educators are directed to teach through online learning platforms. abidah et al. (2020) highlight that, to overcome anxieties and mental stress during the lockdown and for the persistence of educational activities, there is a dire need to adopt innovative teaching. some other researchers (strielkowski, 2020; kumar, 2020) conclude that the covid-19 pandemic results in the digital revolution in the higher education system via interaction at the virtual environment, online examination, digital open books, teleconferencing, and online lectures. gonzalez et al. (2020) report a significantly positive influence of this pandemic on learning performances and efficiency by implementing online education schemes. however, the online learning strategies are often discriminatory to poor and marginalized students especially from remote areas in pakistan like balochistan, fata and gilgit-baltistan. during the lockdown, shut down of educational institutes badly damaged the teaching-learning process and education system of the country (sarwar et al., 2020; farooq et al., 2020). evaluation of the teaching-learning practices in this crisis period is extremely important to propose effective mediations for the smooth running of learning and teaching. keeping all this in mind, the fundamental objective of this research is to assess a) the awareness, knowledge, and attitude of the students regarding covid-19, b) learning status, platforms used for online classes and materials sharing during lockdown period, and c) impact of lockdown on economic, health and educational conditions of students during this lockdown amidst the covid-19 pandemic. the impact of covid19 on educational activities is analyzed by kapasia et al. (2020) for indian graduates and post-graduate students, bayham & fenichel, (2020) for the usa and cao et al. (2020) for chinese students. however, to the best of our knowledge, no prior study exists for pakistan on this issue covering wider dimensions of covid-19 lockdown and its impact mumtaz ahmed, muhammad zeeshan younas & raja waqas shabbir 16 on students including economic and mental health conditions during the lockdown. the empirical findings of the study are based on a sample of 1,114 students currently enrolled in different educational institutions across pakistan covering all educational groups, school-going, college students and university undergrad and grad students. the empirical findings yield interesting insights into the issue under study. the rest of the paper is organized as: section 2 provides the overall evolution of covid-19 and how it leads to the disruption of educational activities. in addition, some relevant literature is discussed as well. section 3 discusses the framework of the study along with data details and collection methods and the methodological approach. the empirical results and their discussion is provided in section 4 while section 5 concludes the study. 2. evolution of covid-19 and the educational disruption the unexpected and dramatic emergence of novel coronavirus disease (covid-19) took the world by storm and sent shock waves all around (iqbal & younas, 2021). it started from the chinese city of wuhan, in early december 2019, and later on december 31, wuhan municipal health commission, china, reported a cluster of cases of pneumonia in wuhan, hubei province (who, 2020). on january 7, 2020, the chinese authorities formally identified a new type of virus, later named covid-19. at first, it was declared an outbreak. the alarming level of spread and severity of the disease, having an unprecedented impact on human life in the post-war era, pushed the world health organization (who) to characterize the covid-19, as a pandemic, on march 11, 2020. the role of the government increases manifold to mitigate the disruptive impact of the disease in all spheres of life. the governments around the world followed the chinese model and locked down the countries, cities, and towns to bring life down to a standstill. after the confirmation from chinese authorities that the human-to-human transmission had taken place, in wuhan, the first epic center of the novel covid-19, went into an unprecedented lockdown on 23 january 2020 which was the beginning of a new norm for the world. from here on, the task was gigantic to stop the spread and restrict it to remain inside chinese borders. it was too big a task, and fighting an unknown enemy, with little clues, put experts and epidemiologists on the back foot. the world was dependent on china to learn about the disease. the nature of the disease and the inability to contain the disease in today’s globalized world helped it spread. standard operating procedures (sop’s), national and local measures, such as smart lockdowns, social distancing, restrictions on journal of applied economics and business studies, volume. 5, issue 4 (2021) 13-40 https://doi.org/10.34260/jaebs.542 17 immigration, use of sanitizers, and use of face-masks, and avoiding face-to-face meetings and teachings, helped avoid a large number of casualties. nevertheless, fatalities and spread are ceaseless, albeit at a lower rate. by and large, the pandemic has fundamentally changed the way people behave and the world function. it has forced people to adopt a new way of life in post-covid-19 times. nevertheless, the denial in some quarters remained high, citing the stories related to fake conspiracy theories and misconceptions. the knowledge, attitudes, and practices (kap) towards covid-19 is integral to assess the change in the behavior of the public, to avoid further complexities due to emotional panic and misconceptions about this deadly virus (zhong et al. 2020). to date, the infectious disease has affected 79.6 million. among those affected by the disease, 1,744,373 have succumbed to death, and 55.1 million have recovered. as of dec 24, 2020, pakistan has reported 465,000 confirmed cases, with recoveries of 417,000, and 9,668 people have died of covid-19 (who, 2020). pakistan confirmed its first two cases of covid-19 on february 26, 2020. the situation slowly deteriorated in the country, and at one point in time, pakistan was ranked among the thirdriskiest countries for covid-19 by the deep knowledge group on nations and their measures against the pandemic (dkg, 2020). in east asia and the pacific region, pakistan is ranked 33 among 36 countries, and only papua new guinea, solomon islands, and afghanistan were ranked lower to pakistan in safety scores. the rampage was real. pakistan recorded 5,830 cases on average in mid-june, and during the week of june 21, the fatality rate soared to 123 deaths per day (daily average by week). resilient pakistan, however, showed a steady decline in the new cases—which baffled analysts, infectious disease experts and remains a mystery to date. nevertheless, the establishment of the national command operation center (ncoc) was the game changer for pakistan—which helped to tackle down the surge in infections through awareness campaigns impacting behavioral patterns regarding precautionary measures. pakistan government’s smart-lockdown policy proved successful, and prompt response after some initial setbacks, the government handled the situation aptly. even then, it was nothing less than a miracle that was required to avoid a steep spike in infections. pakistan formulated a national action plan (nap) to serve as a policy document to ensure that all the guiding principles for preparedness, containment, and mitigation are followed to deal with the covid-19. as a developing country, pakistan lacks medical infrastructure and hospital admission facilities. according to the nap, all provinces are urged to identify and prepare the quarantine facilities to meet contingencies, as per the mumtaz ahmed, muhammad zeeshan younas & raja waqas shabbir 18 sop’s issued by the national institute of health (nih). pakistan government ordered to close all educational institutions across the country on march 13, 2020, after deliberation in a meeting of the country’s national security committee (nsc). the nap recommended using the vacant buildings of the educational institutions, including hostel facilities, to retain people for quarantine. to prevent the escalation of the disease through transmission, most of these places accommodated travelers. there are some conflicting views on the consequences of covid-19 on education. in a systematic review of school closures and management practices during covid-19, viner et al. (2020) argue that school closure did not contribute to control the epidemic during the sars outbreak in china, hong kong, and singapore. on the contrary, however, models based on covid-19 studies predict that school closures reduce the fatalities rate by 2–4%–which is much less than other sop’s adapted. for example, a study by kapasia et al. (2020) on the impact of lockdown on e-learning shows that 70% of the students who participated in online classes, faced multiple problems such as depression, anxiety, poor internet connectivity, and an unfavorable study environment at home. besides, students from unprivileged backgrounds and remote areas faced acute challenges during online learning. crawford et al. (2020) conduct a desktop analysis to provide intra-period higher education responses to covid-19 across 20-countries and find diverse responses across the spectrum. some education providers did not respond at all, while some redeveloped the online application education platform, and few provided isolation facilities to selected students. the findings show that how agile higher education is to prepare for a pandemic. some world-class institutions (for example, cambridge university) showed agility and were already prepared to transfer the whole academic work from the traditional system to online. in another study on the effect of covid-19 on education, edeh et al. (2020) conclude that the pandemic caused learning disruptions, decreased access to education, and increased student loans. the research finds that poor infrastructure hindered learning opportunities as network problems, power outages, poor digital skills, and unavailability of the internet adversely affected education. it recommends all stakeholders, including children, youth, and parents to adopt technology and improve their digital skills. burgess and seivertsen (2020) assess the impact of online education and studied the swedish case and argued that the “global lockdown of education institutions is going to cause a major (and likely unequal) interruption in students’ learning, disruptions in internal assessments, and the cancellation of public assessments for qualifications or their replacement by an journal of applied economics and business studies, volume. 5, issue 4 (2021) 13-40 https://doi.org/10.34260/jaebs.542 19 inferior alternative”. di-pietro et al. (2020) make use of existing literature as well as international data from sources such as eurostat, pisa, icils, pirls, and talis to assess the accurate impact of covid-19 on students learning by employing different methodologies. the study concludes that different methodologies can potentially undermine the children’s achievements and the study further highlights that students will suffer an unequal learning loss, with some students suffering more than others. secondly, the loss of learning time will influence both cognitive and non-cognitive skills acquisition and will have both shortterm and long-term consequences on the students learning. the authors acknowledge that it is a new terrain, and to assess the magnitude of damage, due to covid-19 interruptions, we can rely on pre-covid studies. carlsson et al. (2015) determine that even just ten days of extra schooling significantly raise scores on tests of the use of knowledge ‘crystallized intelligence’ by 1% of a standard deviation. the study conducted in sweden concludes that there is no significant impact on problem-solving skills (an example of ‘fluid intelligence). if we apply these results to an extraordinary situation in covid-19 times, there is a huge disparity among the skills of the pre-covid-19 students and post-covid-19 students. lavy (2015) studies the student groups from two countries, who were subjected to different hours of study in key subjects and concludes that one more hour per week over the school year in the main subjects increases test scores by around 6% of a standard deviation, which is significant differences in test scores. on a positive note, jena (2020) concludes that the emergence of covid-19 has worked as a catalyst for the educational institutions of india to grow, forcing the institutes to explore and opt for technology-rich platforms, not used before to ensure that education reaches remote areas. the education institutions in pakistan faced a complete (and in some rare cases a partial) lockdown and faced an unprecedented challenge to cope to readjust their academic year. in pakistan, the number of affected students stands at 46,803,407. the breakdown of this figure reveals that the number of students affected at the pre-primary level is 8,636,383, at a secondary level, it is 13,357,618 and at the tertiary level, it is about 1,878,101 (saeed, 2020). it is a critical juncture for a developing country like pakistan— which already has a very high number of students out of school. the lack of technology/digitalization, poor internet access (of only 36.8%), inability to use the internet, traditional educational policies and teaching methods, and a cultural hindrance to online learning made it difficult for all the population to have access to online teaching. the present study explores the different dimensions of e-learning in pakistan and assesses mumtaz ahmed, muhammad zeeshan younas & raja waqas shabbir 20 how pakistanis students coped with the impacts of covid-19 on education at schools, colleges and universities. all the researchers agree on the same point that physically attending the school raises the professional and social skills of students. from the very beginning, it increases the ability of the students, and missing school altogether can have profound consequences for the skill growth of the student. moreover, in the case of no exams, the inaccuracies in the results can bring forth life-changing differences. 3. data and methodology 3.1 study design, research instruments and participants a quantitative approach is utilized to evaluate the mode of learning, learning status, effectiveness, and problems related to online classes in pakistan during lockdown amidst the covid-19 pandemic. a cross-sectional survey is carried out among the students at different universities across pakistan. a self-administered questionnaire is designed based on previously published literature. the questionnaire comprises of five sections. the first section deals with the demographic information of the participants; the second section contains questions related to awareness, knowledge, and attitude regarding covid-19; the third section is related to the information about learning status during the lockdown, online classes material sharing platforms and modes of teaching; the fourth section describes the opinion of students regarding academic decisions; and the last section is associated with the impacts of lockdown on economic, health and educational conditions of students. this survey was conducted from august 20 to september 1, just before the two weeks of the scheduled reopening of educational institutions in pakistan by the government of pakistan. the key participants of the study include schools, colleges, undergraduate and graduate students studying in different educational institutions of pakistan. 3.2 data collection and analysis approach an online survey is recognized as the most suitable and convenient tool to collect statistics from a geographically diverse large sample. for that reason, data is collected online by using google forms, an online application provided by google inc. participants were called on different social media platforms to willingly participate in this study by using authors’ professional and personal networks, and endorsement by many public and private teachers, professors, student union presidents, and educators of higher education institutes from all over pakistan. participants who have pakistani nationality enrolled in any school, college and university, and have taken and/or taking online classes were asked journal of applied economics and business studies, volume. 5, issue 4 (2021) 13-40 https://doi.org/10.34260/jaebs.542 21 to fill in the survey by clicking on the online survey link. a total of 1,114 students from different educational institutions provided complete information related to the survey. the empirical analysis is exploratory in nature and several tools have been used to analyze the responses received based on the questionnaire. specifically, the impact of covid-19 lockdown on students’ learning is examined various frequency tables and percentages have been used to explore the pattern of the responses of the students who participated in the survey. the findings yield interesting results. 3.3 regression model to analyze the impact of “stay at home” orders on the mental health of the students, we define the following model; mental health = f (lockdown, individual characteristics, online classes characteristics) this baseline model can be specified in the regression form to make it estimate able with some econometrical technique: 〖 y〗_i= α+ βlockdown_i+ γx_i+ δz_i+ε_i (1) where 〖 y〗_i is the dependent variable of the model which represents the mental health of university student i. x_i is a vector that denotes the individual characteristics. z_(i )represents the online classes characteristics and other control variables. while ε_i is the disturbance term of the model. α is the intercept term and β,γ, and δ are the coefficients of the independent and control variables. 3.4 chi-square distribution the ratio of sample variance and population variance multiply by the degree of freedom is used to obtain the chi-square distribution. this happens when the population follows the normal distribution with population variance. the chi-square test uses adjusting chi-square significance to factor out the sample size. the following equation is a general representation of the chi-square test. x2 = [(n − 1) × s2] ∂2 the researcher widely uses the chi-square test for the association of attributes. based on chi-square testing we can accept or reject our pre-defined research hypothesis about the dependence between any two variables. the chi-square testing approach used in this study is based on the following steps. step 1: define the research hypothesis null hypothesis: h0: two variables are independent mumtaz ahmed, muhammad zeeshan younas & raja waqas shabbir 22 alternative hypothesis: h1: two variables are not independent step 2: construction of cross tabs the next step is to construct the cross tabs for each under consideration variable and then find the chi-square test value via using the following equation: 𝑋2 = ∑(∂𝑖 − 𝑒𝑖) 𝑒𝑖 step 3: find tabulation value after finding the calculated value of the chi-square test, the next step is to compare this value with the tabulated value which can find with the help of the following equation: 𝑋2tab = 𝑋2d. f at 𝛼 where d.f represents the degree of freedom which can be calculated with the following equation: 𝑑𝑓 = (𝑟 − 1)(𝑐 − 1) step 4: decision criteria reject h_0 if α≥p 4. empirical findings and discussion 4.1 descriptive statistical analysis 4.1.1 demographic characteristics the demographic analysis of the participants is provided in table 1. it can be seen that the majority of the respondents are male (64.4%), aged between 18 to 25 years (83.1%) and having a domicile of punjab province (58%). the number of students who belonged to urban areas is greater than in rural areas. additionally, the highest proportion of the respondents is of undergraduate students (65%), day scholars (69.6%), having major in science subjects (79.8%), and with a family income of between pkr 30,000 to pkr 60,000. table 1: demographic characteristics of participants variable name characteristics frequency percent student gender male 717 64.36 female 397 35.64 student age below 18 61 5.48 between 18 to 25 926 83.12 between 26 to 30 86 7.72 journal of applied economics and business studies, volume. 5, issue 4 (2021) 13-40 https://doi.org/10.34260/jaebs.542 23 above 30 41 3.68 student domicile azad jammu and kashmir 46 4.13 balochistan 54 4.85 fata 4 0.36 gilgit baltistan 26 2.33 islamabad (capital) 164 14.72 khyber pakhtunkhwa 126 11.31 punjab 646 57.99 sindh 48 4.31 region rural 360 32.32 urban 754 67.68 stream of study art 89 7.99 commerce 136 12.21 science 889 79.80 class level matric 38 3.41 inter & undergraduate 723 64.9 masters & above 353 31.69 residence day scholar 775 69.57 hosteller 339 30.43 student family income below pkr 30k 318 28.55 between pkr 30k to pkr 60k 441 39.59 above pkr 60k 355 31.87 total 1,114 100 table 2 presents the educational movement of students across different divisions of pakistan. the permanent residence of most students included in the survey is rawalpindi (21.36%), followed by islamabad (14.54%), lahore (10.59%), gujranwala (5.21%), sargodha (4.58%), gilgit-baltistan (4.04%), and peshawar (4.04%). furthermore, the highest concentration of students included in the survey is found in islamabad as an institutional region (49.73%), followed by lahore (13.64%), rawalpindi (6.73%), sargodha (5.66%), quetta (3.41%), and mardan (3.05%). islamabad and lahore are documented as the educational hubs in pakistan as first having the largest number of universities in pakistan and second is known as a city of colleges. a large number of educational institutes constitute a concentration of a substantial proportion of students in these cities. mumtaz ahmed, muhammad zeeshan younas & raja waqas shabbir 24 table 2: movement of students for the education permanent residence frequency percent institute location frequency percent rawalpindi 238 21.36 islamabad 554 49.73 islamabad 162 14.54 lahore 152 13.64 lahore 118 10.59 rawalpindi 75 6.73 gujranwala 58 5.21 sargodha 63 5.66 sargodha 51 4.58 quetta 38 3.41 gilgit-baltistan 45 4.04 mardan 34 3.05 peshawar 45 4.04 mirpur 27 2.42 mardan 41 3.68 peshawar 25 2.24 mirpur 36 3.23 karachi 24 2.15 quetta 36 3.23 multan 18 1.62 faisalabad 35 3.14 jhelum 17 1.53 hazara 33 2.96 gujranwala 16 1.44 larkana 31 2.78 faisalabad 15 1.35 kohat 30 2.69 larkana 12 1.08 multan 24 2.15 d g khan 8 0.72 dera ghazi khan 21 1.89 chakwal 5 0.45 karachi 20 1.8 bahawalpur 4 0.36 makran 17 1.53 gilgit-baltistan 4 0.36 muzaffarabad 12 1.08 abbottabad 3 0.27 bahawalpur 10 0.9 risalpur 3 0.27 bannu 9 0.81 sahiwal 3 0.27 sahiwal 9 0.81 taxila 3 0.27 kalat 6 0.54 attock 2 0.18 poonch 6 0.54 d i khan 2 0.18 hyderabad 5 0.45 hazara 2 0.18 swat 5 0.45 kohat 2 0.18 waziristan 5 0.45 sukkur 2 0.18 zhob 5 0.45 hyderabad 1 0.09 dera ismail khan 1 0.09 total 1,114 100 total 1,114 100 journal of applied economics and business studies, volume. 5, issue 4 (2021) 13-40 https://doi.org/10.34260/jaebs.542 25 4.1.2 awareness, knowledge, and attitude regarding covid-19 the second portion of this survey is designed to evaluate the awareness, knowledge, and attitude of the students regarding covid-19. table 3 reports the results. first, we asked the participants “when they heard about this novel covid-19 virus”. out of 1,114 respondents, 349 students (31.33%) heard about it in december 2019 and 283 students (25.4%) heard during march 2020. more than half of the students (60.41%) key source of information about covid-19 is social media. due to a rapid increase in the covid-19 reporting cases, the country was put under a nationwide lockdown from april 1 and stretched twice until 9 may. after 9th may 2020, the government of pakistan introduced an innovative method of lockdown which is named “smart lockdown”. it means all educational institutes would remain under strict lockdown till 15th september 2020 and partial lockdown would be implemented in different regions of the country. in other words, an area would be under strict lockdown if the number of newly reported covid19 cases acceded to the normally reporting number of cases. we asked the students about their place of residence during the lockdown. the highest proportion of the respondents (91.74%) reported that they are residing at their own home during this covid-19 compelling lockdown period. lastly, we asked the students about the difficulties they faced during the lockdown period. the majority of the participants (37.07%) reported that access to finance is the key problem during the lockdown, followed by access to transport (26.84%), access to health facilities (18.13%) and access to food difficulties (17.95%). table 3: awareness, knowledge, and attitude regarding covid-19 description frequency percent the time when heard about covid-19 during december 2019 349 31.33 during january 2020 247 22.17 during february 2020 235 21.1 during march 2020 283 25.4 source of information about covid-19 friends & family 112 10.05 social media 673 60.41 print media 20 1.8 television 309 27.74 place residing during the lockdown at own home 1022 91.74 at relatives or other places* 92 8.26 problems during lockdown access to finance 413 37.07 access to food 200 17.95 access to health facilities 202 18.13 access to transport 299 26.84 mumtaz ahmed, muhammad zeeshan younas & raja waqas shabbir 26 total 1,114 100 note: * other places including rented house, hostel, friends’ home, etc 4.1.3 learning status during the lockdown a series of questions were asked from the students to analyze the learning status and academic sphere during the lockdown period for instance, online lecture format, syllabus coverage, study time and availability of gadgets for online classes. table 4 reports our findings where 51.71% stated that the format of the online lecture was video in nature and the remaining reported it as an audio format. according to respondents, the higher education of pakistan failed to implement the plans about syllabus coverage during online classes. the highest number of students (57.81%) revealed that below 25% syllabus was covered during online classes while 27.2 % of students reported that 25% to 50% syllabus was covered during the lockdown period. the lockdown period also damaged the study hours as a high number of respondents (65.53%) reported that they did not study regularly during the lockdown period. among the surveyed students who were attending online classes, 61.85% used smart phones to attend the online classes followed by laptop users (35.19%) and desktop computer users (2.96%). when we asked about the possession of gadgets used for online classes, the majority of them (66.79%) claimed that they had their own gadget. 21.27% reported that they borrowed it from family or relatives and the rest of them borrowed from friends, neighbors, rented out, and other sources. to assess the learning status, we asked whether they had any experience of online class before the covid-19 lockdown where 80.25% answered with no option. table 4: learning status during the lockdown variable name description frequency percent format of lecture audio lecture 538 48.29 video lecture 576 51.71 syllabus covered below 25% 644 57.81 between 25% to 50% 303 27.2 above 50% 167 14.99 studied regularly yes 350 31.42 no 730 65.53 don't want to disclose 34 3.05 gadget for online classes desktop computer 33 2.96 journal of applied economics and business studies, volume. 5, issue 4 (2021) 13-40 https://doi.org/10.34260/jaebs.542 27 laptop 392 35.19 smart phone 689 61.85 possess of gadgets for online classes i have my own 744 66.79 family or relatives 237 21.27 friends 73 6.55 rented out 26 2.33 neighbours 20 1.8 others 14 1.26 experience of online class before covid-19 lockdown yes 220 19.75 no 894 80.25 total 1,114 100 4.1.4 platforms for online classes, material sharing, and evaluation reliable and quality platforms for online classes remained a key concern since the corona virus disease emerged. we asked several questions related to the platform used and/or recommended by the higher education commission for online classes in pakistan. it is found that students were using different platforms for online classes, e-lectures, material sharing, and learning evaluation such as youtube, google classroom, whatsapp, ms team, zoom app and other digital platforms like skype and face book. survey responses revealed that the majority of the students (39.68%) used zoom app for sharing material and attending online classes followed by ms team (31.87%), whatsapp group (7.01%), google classroom (3.32%) and youtube (1.97%). while 16.16% of students used other platforms including skype, face book and lms university software (see table 5). whatsapp groups remained convenient for the respondents to share material related to online classes as the majority of them had their own smart phones. as far as the evaluation of online classes is concerned, more than half of the students show disagreement about the regular evaluation of understanding by the respective educational institute. assignment and projects were the key criteria of course evaluation in the majority of pakistani educational institutes. when we asked the respondents about satisfaction with the evaluation method then almost 53% of students showed no satisfaction with this evaluation system. a sub-portion of our survey was related to the sensitivity of information and comfort ability with the online classes. a bigger portion of students (74.51%) did not feel comfortable with using online learning websites like edx, coursera and udemy etc. to cover the missing syllabus during the lockdown period. mumtaz ahmed, muhammad zeeshan younas & raja waqas shabbir 28 almost 52% of students showed fear related to misuse of personal information during online learning activities and profile making of online learning applications. when we asked about the overall satisfaction from online learning apps, 45.96% of students were not satisfied with the applications suggested by hec pakistan for online classes like zoom, team and google classroom. while 27.38% reported that they are just satisfied with these learning applications. table 5: platforms for online classes, materials sharing, and evaluation various platforms description frequency percent platforms of online classes zoom app 442 39.68 ms team 355 31.87 whatsapp 78 7 google classroom 37 3.32 youtube 22 1.97 others 180 16.16 platforms of materials sharing whatsapp groups 379 34.02 ms team 257 23.07 zoom app 137 12.31 email 132 11.65 google classroom 119 10.68 lms 25 2.44 others 65 5.84 regular evaluation of understanding by the institute yes 525 47.13 no 589 52.87 course evaluation no exam 125 11.22 online viva 177 15.89 quizzes 362 32.5 assignments/projects 450 40.39 satisfaction with the evaluation yes 403 36.18 no 589 52.87 don't know 122 10.95 comfortable with using online learning websites yes 284 25.49 no 830 74.51 fear related to the misuse of personal information yes 575 51.62 no 539 48.38 overall satisfaction from online learning apps not satisfied at all 512 45.96 just satisfied 305 27.38 journal of applied economics and business studies, volume. 5, issue 4 (2021) 13-40 https://doi.org/10.34260/jaebs.542 29 moderately satisfied 241 21.63 highly satisfied 28 2.51 don't know 28 2.51 total 1,114 100 4.1.5 opinion of students regarding academic decisions table 6 describes the opinion of respondents about different announcements and policies of government regarding academic activities. out of 1114 respondents, 80.61% thought that the lockdown decision and shut down of educational institutes by the government of pakistan was a good decision. while the highest number of students (81.24%) are in favor of the hec announcement about the opening of educational institutions on 15th sep 2020. when we asked about the ministry of education role played in online classes during the lockdown, they showed mixed opinions as 43.81% said yes and 43.9% said no to the question. table 6: opinion of students regarding academic decisions academic decisions and hec recommendations description frequency percent do you think, the lockdown decision by the government was good? yes 898 80.61 no 216 19.39 are you in favor of the opening of educational institutions on 15th sep 2020 as announced by hec? yes 905 81.24 no 140 12.57 don't know 69 6.19 do you think the ministry of education pakistan played its role well during a lockdown? yes 488 43.81 no 489 43.9 don't know 137 12.3 total 1,114 100 4.1.6 impact of covid-19 lockdown on students’ life the last segment of our study is associated with the impact of the covid-19 lockdown impact on student life. we divide these analyses into three sub-segments; economic impacts, health impacts, and educational impacts. table 7 reports that a bulk of students (71.01%) were suffering from psychological problems like stress, depression, and anxieties during the lockdown period. when we questioned “did you get any personal coaching to keep you calm during a lockdown?” then only 15.26% of students reported a yes answer. as far as economic impacts are concerned, 70.65% of students said that the lockdown policy badly damaged their family income and 44% reported that their financial mumtaz ahmed, muhammad zeeshan younas & raja waqas shabbir 30 needs were not fully met during the lockdown. while 40.75% revealed that they need financial support due to low family income effects during the lockdown. according to 47.85% of respondents, financial support in terms of 50% off in educational fees should be given to students due to no on-campus activities. while 26.66% asked for the full fee concession support, followed by 10 to 25% fee waive (13.82%) and no fee reduction (11.67%). a large number of students (47.94%) claimed that they were studying 3 to 6 hours daily before lockdown and 88.42% reported that the lockdown severely affected their study routines. 71.36% said that their study hours significantly decreased during the lockdown period. when we questioned about the enthusiasm level of study, 76.12% described that the lockdown negatively damaged their enthusiasm level of study. as mentioned earlier, internet connectivity was a major issue that students faced during online classes. out of 1114 students, only 46.59% had internet access. while 45.87% claimed that they had not a very good but an average kind of internet connectivity. separate room availability for online classes during the lockdown period was also a serious educational problem faced by the students as 50.54% of respondents said that they had no separate room available for online classes. table 7: impact of lockdown on economic, health, and educational conditions of students opinion description frequency percent economic impacts do you think the lockdown affected your family income? yes 787 70.65 no 244 21.9 don't know 83 7.45 your financial needs are fully met during the lockdown? yes 471 42.28 no 490 43.99 don't know 153 13.73 do you need financial support due to low family income effects during a lockdown? yes 454 40.75 no 499 44.79 don't know 161 14.45 what financial support should be given to students due to no on-campus activities? full fee waiver 297 26.66 half fee waiver 533 47.85 10-25% fee waiver 154 13.82 no fee reduction 130 11.67 journal of applied economics and business studies, volume. 5, issue 4 (2021) 13-40 https://doi.org/10.34260/jaebs.542 31 health impacts feeling of stress, depression, and anxieties during the lockdown yes 791 71.01 no 235 21.1 don't know 88 7.9 did you get any personal coaching to keep you calm during a lockdown? yes 170 15.26 no 873 78.37 don't know 71 6.37 educational impacts study hours before lockdown less than 3 hours 429 38.51 3 to 6 hours 534 47.94 more than 6 hours 151 13.55 did lockdown affect your study routine? yes 985 88.42 no 104 9.34 don't know 25 2.24 did your study hours decrease during a lockdown? yes 795 71.36 no 259 23.25 don't know 60 5.39 did lockdown damage your enthusiasm level of study? yes 848 76.12 no 170 15.26 don't know 96 8.62 internet access yes 519 46.59 no 595 53.41 internet connectivity average 511 45.87 excellent 103 9.25 good 277 24.87 no internet access 21 1.89 very poor 202 18.13 separate room available for classes during the lockdown yes 551 49.46 no 563 50.54 total 1,114 100 4.2 regression and chi-square analysis in regression analysis, we examine the impact of “stay at home” orders on the mental health of the students in pakistan. the dependent variable of our regression analysis is the mental health score which is measured by the use of the who 5-question module. it has been extensively used by researchers in different perspectives (krieger et al., 2014; downs et al., 2017). who 5-question module: “over the last two weeks ... [answers on mumtaz ahmed, muhammad zeeshan younas & raja waqas shabbir 32 a scale from 0 “at no time” to 5 “all of the time”.] 1) i have felt cheerful and in good spirits, 2) i have felt calm and relaxed, 3) i have felt active and vigorous, 4) i woke up feeling fresh and rested, and 5) my daily life has been filled with things that interest me”. an overall mental health score is obtained by summing answers to the 5 questions, with a higher score indicating better mental health. more explicitly, we regress mental health on a dummy variable indicating whether a lockdown was in place as well as a range of individual background characteristics of the students. cronbach’s alpha coefficient is used to check the internal consistency or reliability. the cronbach’s alpha of mental health is 0.725 which is in an acceptable range. the ols estimation-based results are reported in table 8. the lockdown coefficient is estimated to be negative and significant. it indicates that students living in a city that has “stay at home” orders in place at the time of the survey is linked with a decrease in mental health by 0.525 standard deviations (p-value=0.0560), suggesting that “stay at home” orders have led to a significant reduction in the mental health of the students. table 8: results of baseline model dependent variable= mental health score variables name coefficient s.e p-value lockdown -0.5257** 0.2746 0.0560 age 0.0088 0.0401 0.8260 gender male 0.0402* 0.3158 0.0990 urban region -0.1562 0.3108 0.6150 bs student 0.7013* 0.3983 0.0790 fa/fsc student 0.4974 0.8150 0.5420 master student 0.9428** 0.4533 0.0380 hostelite 0.3350 0.3167 0.2900 lockdown stay at home 1.0359** 0.5062 0.0410 online video lectures -0.1036 0.2856 0.7170 internet facility 0.1326 0.2849 0.6420 constant 5.8798*** 1.3684 0.0000 no. of observations 1114 ---- adjusted r-squared 0.15 ---- note: * indicates that variable is significant at 10% level of significance, ** at 5% level of significance, and *** at 1% level of significance. in table 8, several other patterns are worth mentioning. in line with previous literature (astbury, 2001; seedat et al., 2009; stevenson & wolfers, 2009; adams-prassl et al., journal of applied economics and business studies, volume. 5, issue 4 (2021) 13-40 https://doi.org/10.34260/jaebs.542 33 2020), our results reveal that being male is associated with significantly higher and better mental health. likewise, being a bs and master student of the university is allied with significantly better mental health. however, there is no significant association is observed between mental health and intermediate students. moreover, students who report being stayed at home during lockdown have significantly better mental health scores. it suggests that staying at home helps the students to improve mental health and cope up better with the covid-19 pandemic. the other characteristics like age, urban region, being a hostel student, online video lectures, and internet facilities have no statistically significant relationship with the mental health score. there is a strand of literature available (chang et al., 2013; dagher et al., 2015) which postulates that economic downturns or recessions can influence the mental health of males and females differently. in this study, we further examine whether the mental health impact of the “stay at home” orders varies by gender. for this objective, we re-estimate our baseline model of table 8 by introducing an additional interaction term between gender and dummy variable indicating whether the city was in lockdown. the ols-based estimated results are presented in table 9. again, the lockdown coefficient is estimated to be negative and significant. it indicates that students living in a city that has “stay at home” orders in place at the time of the survey is linked with a decrease in mental health by 0.8192 standard deviations (p-value=0.0160), suggesting that “stay at home” orders have led to a significant reduction in the mental health of the students. as indicated by the significant interaction coefficient, the gender gap is significantly higher in cities that introduced lockdown measures. the estimated gender gap is 0.860 standard deviations larger in cities that had a lockdown in place (p-value=0.0133). the rest of the results of table 9 are almost in line with the results of table 8. table 9: results after incorporating gender role dependent variable= mental health score variables name coefficient s.e p-value lockdown -0.8192** 0.3408 0.0160 age 0.0078 0.0399 0.8450 gender male 0.4255* 0.4261 0.0980 lockdown * male 0.8604** 0.5726 0.0133 urban region -0.1526 0.3098 0.6220 bs student 0.7038* 0.3972 0.0770 fa/fsc student 0.6260 0.8116 0.4410 master student 0.9477** 0.4521 0.0360 hostelite 0.3492 0.3161 0.2700 mumtaz ahmed, muhammad zeeshan younas & raja waqas shabbir 34 lockdown stay at home 0.9598** 0.5029 0.0570 online video lectures -0.0904 0.2846 0.7510 internet facility 0.1517 0.2830 0.5920 constant 6.3874*** 1.3167 0.0000 no. of observations 1114 ---- adjusted r-squared 0.15 ---- note: * indicates that variable is significant at 10% level of significance, ** at 5% level of significance, and *** at 1% level of significance. for the robustness checking of our baseline estimates, we introduced a family of dummy variables related to provinces i.e. punjab, sindh, kpk, and balochistan. table 10 presented the results of this incorporation. it reports that our baseline findings are not sensitive to the inclusion or exclusion of province-related variables except gender. in other words, all variables report the same results as reported in table 8 with the gender male variable as insignificant. table 10: results after incorporating province variables dependent variable= mental health variables name coefficient s.e p-value lockdown -0.5275** 0.2748 0.0550 age 0.0086 0.0401 0.8300 gender male 0.0272 0.3187 0.9320 urban region -0.1446 0.3132 0.6440 bs student 0.7076* 0.3990 0.0760 fa/fsc student 0.4947 0.8154 0.5440 master student 0.9535** 0.4548 0.0360 hostelite 0.3322 0.3169 0.2950 lockdown stay at home 1.0356** 0.5064 0.0410 online video lectures -0.0987 0.2862 0.7300 internet facility 0.1344 0.2851 0.6370 punjab 0.4018 0.3514 0.2530 sindh 0.6773 0.7367 0.3580 kpk 0.1606 0.5154 0.7550 balochistan -0.1118 0.7196 0.8770 constant 5.8485*** 1.3681 0.0000 no. of observations 1114 ---- adjusted r-squared 0.15 ---- note: * indicates that variable is significant at 10% level of significance, ** at 5% level of significance, journal of applied economics and business studies, volume. 5, issue 4 (2021) 13-40 https://doi.org/10.34260/jaebs.542 35 and *** at 1% level of significance. table 11 provides the chi-square analysis of our study. it reports the dependence of selective variables with lockdown variables. some independence-related null hypotheses are rejected and some are accepted. this table helps us to understand how the government decision “stay at home” influences the daily lives of students. table 11: chi-square analysis null hypothesis chisquare pvalue decision feeling of stress/depression and lockdown are independent 3.8765 0.099 null rejected family income and lockdown are independent 1.2804 0.572 null accepted difficulty in internet access and lockdown are independent 0.0806 0.776 null accepted study routine and lockdown are independent 5.7563 0.056 null rejected enthusiasm to study and lockdown are independent 3.2877 0.193 null accepted study environment at home and lockdown are independent 0.6029 0.437 null accepted access to transport and lockdown are independent 7.1759 0.028 null rejected access to food and lockdown are independent 3.8689 0.049 null rejected access to finance and lockdown are independent 3.4208 0.181 null accepted 5. concluding remarks and policy implications this study takes a lead and figures out the key challenges and issues that students across all over pakistan are facing. in particular, it digs deeps and explores the ways that how covid-19 lockdown impacted the economic, health and educational conditions of the students. the empirical analysis is carried out based on the survey data collected through a structured questionnaire distributed online to students across pakistan through different social media platforms. a total of 1,114 responses were received providing information on various aspects of online learning and demographics. in regression analysis, we examine the impact of “stay at home” orders on the mental health of the students where mental health is a dependent variable measured by the use of the who 5mumtaz ahmed, muhammad zeeshan younas & raja waqas shabbir 36 question module. the lockdown coefficient has a negative and statistically significant impact on the mental health of students. it indicates that students living in a city that has “stay at home” orders in place at the time of the survey is linked with a decrease in mental health which infers that lockdown orders have led to a substantial decline in the mental health of the students. our results also conclude that being a bs and master student of the university and being male is associated with significantly higher and better mental health. another important finding of the study is that students who report being stayed at home during lockdown have significantly better mental health scores. it proposes that staying at home during lockdown helps the students to improve mental health and cope up better with the covid-19 pandemic. robustness analysis indicates that these findings are not sensitive to the inclusion or exclusion of province related variables except gender similarly, the key findings of descriptive statistical analysis can be summarized as follows. in more than 48% of cases, the format of the lecture was audio rather than a video lecture, implying that the internet connectivity was poor and thus the audio mode of the lecture was preferred over the video. more than 65% of students said that they are no longer studying regularly. there can be several reasons, one of them can be that 33% of the students do not own the learning gadgets. they either rented out or borrowed from friends and family and since the family income of more than 68% of the students included in the sample is below 60,000 pkr, so it is highly unlikely to afford to buy a new one. another reason can be that more than 80% of the students don't have any prior experience of online learning. as far as economic impacts of covid-19 lockdown are concerned, more than 70% of the students reported that the lockdown affected their family income and 44% of them that their financial needs are not being fully met and more than 40% asked for financial support. around 88% of the respondents are in favor of providing a fee concession to the students with 48% recommending waving off half of the tuition fee while 27% are in favor of giving a full fee waiver to the students. the covid-19 lockdown has imparted severe health impacts on students’ lives with more than 71% responded said that they are in stress, anxiety and depression due to lockdown and the alarming situation is that only 15.26% received any personal coaching. the unprecedented scale and magnitude of online teaching is an experiment and a learning curve for educational institutions. the closure of educational institutions, schools, colleges, and universities not only interrupt the teaching but also creates massive issues of exams and assessments. the wrong assessment delays the recognition journal of applied economics and business studies, volume. 5, issue 4 (2021) 13-40 https://doi.org/10.34260/jaebs.542 37 of both the extraordinary potential and learning difficulties and its damaging consequences for the future of the students (burgess, 2020). as almost all universities and colleges replaced the traditional exams with online assessments or some other methods, there was a chance of measurement error. our study examines the impact of covid-19 on student education in pakistan further consolidates the facts established in the literature review section. it shows that educational institute closures caused by the pandemic exacerbated previously existing inequalities and those children who were already most at risk of being excluded from a quality education in pakistan have been most affected. the empirical results highlight the importance of immediate digital educational reforms in pakistan and targeted mediations to build a positive environment for online study among students from the vulnerable segment of the public. mumtaz ahmed, muhammad zeeshan younas & raja waqas shabbir 38 references adams-prassl, a., boneva, t., golin, m., & rauh, c. (2020). inequality in the impact of the coronavirus shock: evidence from real-time surveys. journal of public economics, 189, 104245. astbury, j. (2001). gender disparities in mental health. in: mental health. ministerial round tables 2001, 54th world health assembly, 2001, who, geneva, switzerland. bayham, j., & fenichel, e. p. (2020). impact of school closures for covid-19 on the us health-care workforce and net mortality: a modeling study. the lancet public health, 5(5), 271-278. cao, w., fang, z., hou, g., han, m., xu, x., dong, j., & zheng, j. (2020). the psychological impact of the covid-19 epidemic on college students in china. psychiatry research, 287(1), 112934. carlsson, m., dahl, g. b., öckert, b., & rooth, d. o. (2015). the effect of schooling on cognitive skills. review of economics and statistics, 97(3), 533-547. castle, s. (2020). cambridge university will hold its lectures online next year 2020. retrieved on 3 september 2020, from: https://www.nytimes.com/2020/05/19/world/europe/cambridge-universitycoronavirus.html chang, s. s., stuckler, d., yip, p., & gunnell, d. (2013). impact of 2008 global economic crisis on suicide: time trend study in 54 countries. bmj, 347. dagher, r. k., chen, j., & thomas, s. b. (2015). gender differences in mental health outcomes before, during, and after the great recession. plos one, 10(5), e0124103. davies, d. (2020). this year's a-level results are a fiasco – but the system was already broken | dan davies. retrieved on 3 september 2020, from: https://www.theguardian.com/commentisfree/2020/aug/15/a-level-results-systemofqual-england-exam-marking di pietro, g., biagi, f., dinis mota da costa, p., karpinski, z., and mazza, j. (2020). the likely impact of covid-19 on education: reflections based on the existing literature and recent international datasets, eur 30275 en, publications office of the european union, luxembourg, 2020, isbn 978-92-76-19937-3 (online), doi: 10.2760/126686 dkg. (2020). covid-19 regional safety assessment 250 countries, regions & territories retrieved on 3 september 2020, from: https://www.dkv.global/covidsafety-assessment-250-regions downs, a., laura a., duncan g., & anita, p. (2017). “using the who5 well-being index to identify college students at risk for mental health problems.” journal of college student development, 58(1): 113– 117. farooq, f., rathore, f. a., & mansoor, s. n. (2020). challenges of online medical journal of applied economics and business studies, volume. 5, issue 4 (2021) 13-40 https://doi.org/10.34260/jaebs.542 39 education in pakistan during covid-19 pandemic. j coll physicians surg pak, 30(1), 67-90. hec. (2020). hec announcements. retrieved on 3 september 2020, from: https://www.hec.gov.pk/english/hecannouncements/pages/why-onlineeducation.aspx hern, a. (2020). do the maths: why england's a-level grading system is unfair. retrieved 3 september 2020, from https://www.theguardian.com/education/2020/aug/14/dothe-maths-why-englands-a-level-grading-system-is-unfair hrw. (2020). impact of covid-19 on children’s education in africa. (2020). retrieved on 3 september 2020, from: https://www.hrw.org/news/2020/08/26/impact-covid19-childrens-education-africa iqbal, m. a., & younas, m. z. (2021). public knowledge, attitudes, and practices towards covid-19 in pakistan: a cross-sectional study. children and youth services review, 120, 105784. kapasia, n., paul, p., roy, a., saha, j., zaveri, a., mallick, r., & chouhan, p. (2020). impact of lockdown on the learning status of undergraduate and postgraduate students during the covid-19 pandemic in west bengal, india. children and youth services review, 116(20), 105194. krieger, t., zimmermann, j., huffziger, s., ubl, b., diener, c., kuehner, c., & holtforth, m. g. (2014). measuring depression with a well-being index: further evidence for the validity of the who well-being index (who-5) as a measure of the severity of depression. journal of affective disorders, 156, 240-244. kumar, d. n. s. (2020). impact of covid-19 on higher education. higher education digest. https://www.highereducationdigest.com/impact-of-covid-19-on-highereducation lavy, v. (2015). do differences in schools’ instruction time explain international achievement gaps? evidence from developed and developing countries. the economic journal, 125(588), f397-f424. onyema, e. m., eucheria, n. c., obafemi, f. a., sen, s., atonye, f. g., sharma, a., & alsayed, a. o. (2020). impact of coronavirus pandemic on education, 11(13), 108121. saeed, a. (2020). the impact of covid-19 on education in pakistan crss blog 2020. retrieved on 3 september 2020, from: https://crssblog.com/the-impact-of-covid19-on-education-in-pakistan/ sarwar, s., waheed, r., sarwar, s., & khan, a. (2020). covid-19 challenges to pakistan: is gis analysis useful to draw solutions? science of the total environment, 139089. seedat, s., scott, k. m., angermeyer, m. c., berglund, p., bromet, e. j., brugha, t. s., ... & kessler, r. c. (2009). cross-national associations between gender and mental disorders in the world health organization world mental health surveys. archives of general psychiatry, 66(7), 785-795. mumtaz ahmed, muhammad zeeshan younas & raja waqas shabbir 40 simon, b., & hans, h.s. (2020). the impact of covid-19 on education | vox, cepr policy portal. (2020). retrieved on 3 september 2020, from: https://voxeu.org/article/impact-covid-19-education stevenson, b., & wolfers, j. (2009). the paradox of declining female happiness. american economic journal: economic policy, 1(2), 190-225. strielkowski, w. (2020).covid-19 pandemic and the digital revolution in academia and higher education. preprints 2020, 2020040290. doi: 10.20944/preprints202004.0290.v1 tahir, p. (2020). education and covid-19. retrieved on 3 september 2020, from: https://www.dawn.com/news/1568714 unicef. (2020). at least a third of the world’s schoolchildren unable to access remote learning during covid-19 school closures. retrieved on 3 september 2020, from: https://www.unicef.org.au/about-us/media/august-2020/remote-learning-report viner, r. m., russell, s. j., croker, h., packer, j., ward, j., stansfield, c., ... & booy, r. (2020). school closure and management practices during coronavirus outbreaks including covid-19: a rapid systematic review. the lancet child & adolescent health, 4(5), 397-404. who timeline covid-19. (2020). retrieved on 3 september 2020, from: https://www.who.int/news-room/detail/27-04-2020-who-timeline---covid-19 zhong, b. l., luo, w., li, h. m., zhang, q. q., liu, x. g., li, w. t., & li, y. (2020). knowledge, attitudes, and practices towards covid-19 among chinese residents during the rapid rise period of the covid-19 outbreak: a quick online crosssectional survey. international journal of biological sciences, 16(10), 1745. journal of applied economics and business studies, volume. 3, issue 1 (2019) 1-14 https://doi.org/10.34260/jaebs.311 1 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x the role of institutions in the economic sustainability of global value chains: a transcendental phenomenological analysis of pakistani apparel industry ali asghar1 & rukhsana kalim 21 1 phd scholar, school of business and economics, university of management and technology, lahore, pakistan 2 professor of economics, school of business and economics, university of management and technology, lahore, pakistan abstract the geographical spread of the global value chains (gvc) not only makes them complex structures but also raises the challenges of governance for them. gereffi and lee (2014) argued that such governance issues may arise from either the vertical (within the structure of the gvc) or the horizontal (from outside) governance of a gvc; and at times, bring the sustainability of gvcs at stake. this study explores the phenomenon in the domain of the horizontal governance, outlining the role of the local institutions and the effects they cast upon the sustainability of the gvcs. this study incorporates the findings derived from transcendental phenomenological analysis (tpa) of the in-depth interviews of twenty-three export managers, working in the apparel firms producing for the gvcs on a long-term basis. the study delivers an explanation of the phenomenon and highlights the role played by the government institutions in it. the essence of this study elaborates the mechanisms of cooperation and coordination between the government institutions and the industry, and the difference in the expected and actual performance of the institutions. it highlights that the government institutions in pakistan, having rigid culture, are currently not fulfilling the requirements of the apparel production units linked with gvcs; and the behavior of government employees adds to it. the implications emerging from this study, directions for future research and limitations have also been discussed. besides, the application of nvivo software to perform the qualitative data analysis (qda) while performing the tpa adds to the methodological innovation of this study. keywords apparel, global value chains, institutions, moustakas, pakistan jel classification g20, q56, 1. introduction the pressure of competitiveness has pushed the production in fragments beyond national boundaries resulting in the emergence of the global value chains. global value chains (gvcs) are complex structures that offer challenges to the business as well as the 1*corresponding author: aliasghar123@gmail.com ali asghar & rukhsana kalim 2 institutions. literature published on gvcs reflects that the major challenges offered by them include governance, up-grading, and sustainability (kaplinsky & morris, 2000). this study focuses on the sustainability of the global value chains, especially about the local institutions, as they are prone to play an influential role in this regard (lee, gereffi, & beauvais, 2012; nadvi & halder, 2005; rodrik, 2018). the motivation for this study comes from a real-world incidence when a famous global apparel brand banned pakistan for apparel sourcing in 2013 (barrie, 2014; rana, 2013; reporter, 2014). the action of the global apparel brand was a consequence of the collapse of rana plaza in dhaka, bangladesh on april 24, 2013 (jacobs & singhal, 2017; reinecke & donaghey, 2015; siddiqui & uddin, 2016; sinkovics, hoque, & sinkovics, 2016). pakistani apparel industry anticipated an increase in their business as the brands starting leaving bangladesh (davis, 2013). at the same time, the previously mentioned brand demanded the governments of bangladesh and pakistan for improvement in the implementation of human rights conventions outlined by the international labor organization (ilo) through their local labour policy. in pakistan, an association of the industry directly connected with global apparel trade wrote a letter to the concerned government department and asked them to treat the matter on a serious note. the correspondence stopped with the unfortunate demise of the leader of the respective association and the delay in action against the brand’s demand resulted in the blacklisting of the country, which resulted in an initial loss of the global client who used to place annual orders between us$150-200 million (barrie, 2014; reporter, 2014). the industry representatives blamed the government, while the government kept its silence. the industrial units, once part of the global value chain, went into existential crisis resulting in local socio-economic chaos. the questions raised by this phenomenon had no clear explanation in the published literature, therefore, required for an in-depth exploration of the phenomena. the primary research questions outline by the study are: what role government institutions play in the economic sustainability of global value chains? and how? this study uses the agency theory to understand the type and nature of the relationship, interactions, and understanding between the stakeholders of the phenomena, i.e. the government institutions and the industry. the agency theory explains the agency relationships, their issues and solutions (arrow, 1971; panda & leepsa, 2017; ross, 1973; spence & zeckhauser, 1971). the agency relationships can be stated as an agreement between two parties stating one party – the agent – acting or making decisions on the behalf of the other – the principal (jensen & meckling, 1976; luhman & cuniffe, 2012; shapiro, 2005). the literature on agency theory highlights two main approaches to it: the positivist and the interpretivist. this research uses the interpretivist agency theory derived from the work of ross (1973), holmström (1979), and shavell (1979). as evident from the literature, journal of applied economics and business studies, volume. 3, issue 1 (2019) 1-14 https://doi.org/10.34260/jaebs.311 3 the interpretive agency theory has largely been used for studying the conflicts in the reallife situations, therefore, is popular among the researchers in the area of organizational and management studies (eisenhardt, 1989). the interpretive agency theory is a natural fit for this study as it aims at exploring the conflict among the industry (the apparel production units engaged with global value chains) and the government institutions engaged in facilitating the local firms for the global trade. on the theoretical praxis, the industry has been considered as the principal as it is running its business cycle and contributing to the national revenue through the payment of taxes and levies. on the other hand, the government institutions have been considered as the agents as they are dependent upon the revenue collected from the industry in terms of taxes and levies. 2. procedures: research design, methodology and data the researchers in the area of the global value chain have shown a recent inclination towards using phenomenology, for its strengths to explain the real-world complexities. for example, see the work of eller (2017), and cano-kollmann, cantwell, hannigan, mudambi, and song (2016). phenomenology is a qualitative research approach in which the researcher as a human makes sense of the world around him (cibangu & hepworth, 2016; saunders, lewis, & thornhill, 2009). it requires the researcher to understand the world with the point of view of those having lived experience of the phenomena or the “verstehen” (creswell, 2014; hennink, hutter, & bailey, 2014; moran, 2002). the literature on phenomenology refers to its certain types; and suggests the hermeneutic phenomenology and transcendental phenomenology as its two major types (laverty, 2003; moerer-urdahl & creswell, 2008). this constructivist-interpretivist study applies the transcendental phenomenology as propagated by husserl (2012) for its strength of acquiring and collecting data that explicates the essence of human experience (moerer-urdahl & creswell, 2008; moustakas, 1994). figure 1 exhibits the four steps of the transcendental phenomenological analysis as suggested by moustakas (1994) and adu (2015): ali asghar & rukhsana kalim 4 figure 1: transcendental phenomenological analysis process source: adu (2015), pp: 15 this study explains the role of institutions being played in the sustainability of the global value chains by incorporating the lived experience of the export managers, working in those active apparel production units in lahore, pakistan, engaged with global brands. the purposive sampling method was used to recruit 24 export managers having experience ranging from 10 to 25 years. the average experience of the sample is 16.91 years, which adds to the reliability of the qualitative data collected in the shape of their interviews. the interview guidelines were developed based on the agency theory with a focus on the subdimensions of “meaning” as exhibited in figure 2 below: figure 2: conceptual map of the sub-themes (generally focused in the interview guidelines) economic sustainability of gvcs meaning efficiency alignment of interest risk sharing continuation of existing job epoche phenomenological reduction imaginative variation synthesis of meanings bias-free state textual descriptions structural descriptions essence of the experience process product a systematic process of moving concrete data to abstract concepts and underlying meanings journal of applied economics and business studies, volume. 3, issue 1 (2019) 1-14 https://doi.org/10.34260/jaebs.311 5 during this study, the researcher observed that the saturation was achieved as he conducted the 20th interview; however, three more interviews were conducted for further assurance. the interview guidelines were generally referred to as a general plan for the interview, to initiate the discussion and as an approximate list to keep track that all relevant themes mentioned in it are touched in the discussion. the participants were interrupted through cross-questions only when necessary as the objective was to get their free and deep expressions. 2.1 step 1: the epoché the epoché, in denotative terms, is a greek word means “suspension of judgment”. epoche, in tpa, refers to explicating those presumptions to bring the researcher in a biasfree state of mind so that he can accept whatever information related to the phenomena comes to him in its original form (adu, 2015; moerer-urdahl & creswell, 2008; moustakas, 1994). since the principal researcher, conducting this research, has prior experience of working on the projects which included the global value chain stakeholders, therefore, this step was conducted with special care. in this regard, the personal statement by the principal researcher himself is given below: by clearing my mind through the epoche process, i recall my personal and professional experiences related to the industry, government institutions, global value chains, and their relationship. i remember my discussions with the seniors, colleagues and others, discussing the global value chains, industrial clusters, and porter’s thought. i recall my detailed discussions with the industry and institutional representatives, which were later compiled into the fact-finding reports. i also recall the discussions over global value chains, governance, and institutions with my university professors and phd class fellows. i recall the process of rediscovering my concepts of global value chains, governance, institutions, and economic sustainability through extensive reading of the relevant literature and putting them in the praxis with the real-world phenomena. i recall the blunt criticism from the board members before whom i presented my research proposal, pushing me to unlearn and relearn the concepts again and again. this iterative learning process helped me to learn how to think logically and understand the complexity of the real world. as i moved towards receptiveness, i was able to see, listen, and feel the people, incidences, and things as they are, without framing them into my perspective or the belief-system. after explicating the above thoughts through the reflective and iterative process, the researcher felt disconnected from the pre-conceptions and thoughts that could affect his receptiveness. now, he was able to fully concentrate on listening and hearing the participants’ experiences without filtering them through his process of thinking, feeling, and seeing. ali asghar & rukhsana kalim 6 2.2 step 2: phenomenological reduction the objective of the tpa is to provide a deep explanation of the phenomena from ‘the point of view’ of those who have experienced that phenomena (moustakas, 1994). since the participants of the study went through the same phenomena, therefore, the redundancy of statements is quite natural. the second step in tpa phenomenological reduction refers to the process of removing those repetitive and overlapping statements from the data so that the explanation of phenomena becomes rich and complete. this step helps the researcher to outline the textual descriptions (adu, 2015). moerer-urdahl and creswell (2008), in their study, have presented the textual descriptions as the list of significant statements. whereas, moustakas (1994) has stated this process as horizonalization. all interview transcripts were read multiple times and coded using nvivo by the researcher. the horizonalization process involved the identification of the significant statements about the experiences of the participants from the transcripts. the significant statements and the themes emerged therefrom acted as the basis of the textual descriptions explained in the following subheading. 2.2.1 textual descriptions: the role of institutions the textual descriptions, according to moustakas (1994) and moerer-urdahl and creswell (2008), are the outcomes of the phenomenological reduction which stand for “what” was experienced, as described by the researcher. a. coordination between the industry and the government institutions the coordination between the industry and the government institutions is very important for the sustainability of the global value chains expressed the industry representatives. the industry representatives doubted over the performance of the government institutions under their primary goals or purpose of the constitution. they also emphasized on increasing the coordination, especially for raising the awareness of the services, especially the facilitation schemes, which the relevant government institutions offer to the industry. the government institutions and the industry have to work in close coordination with each other for sustainable results. b. institutional efficacy the institutions are not much effective, as perceived the industry representatives. they were of the idea that the institutional policies significantly affect the business. according to them, these policies benefit the large-scale business only and their benefits do not reach to the grassroots level. they also expressed that this limited utility of the government institutions also raises from the complex institutional setup where the overlapping institutional roles create delusions. journal of applied economics and business studies, volume. 3, issue 1 (2019) 1-14 https://doi.org/10.34260/jaebs.311 7 the government institutions operate in a rigid policy framework and tight procedural control. this rigidity causes a delay in the provision of institutional services, incurring losses, to the business; especially in the realm of market dynamics. the institutional policies do not support the small and medium scale business which is an astonishing fact directly questioning the primary objectives for the constitution of the institutions. the existing institutional policy setup affects the sustainability of the global value chains, as does not support the local business to maintain long-term business relations with their foreign counterparts. c. competitiveness and government institutions the government plays a vital role in the competitiveness of the business in the global market. the government institutions also have limitations since they are the systems created by human and operated by a human. the industry representatives expressed their serious concerns over the low quality of services from the government institutions as it adds to the challenges of competitiveness to them. the industry representatives believe that ignoring the important elements of competitiveness may result in the eviction of an industry. d. global value chains and pakistani government institutions apparel manufacturing units are primarily the family-owned concerns in pakistan, expressed the industry representatives. according to them, they look forward towards the government to address human resource development at the national scale to enhance the competitiveness of the apparel industry. upgrading the human resource is very important to sustain the business with the global value chains, as it contributes to the national competitiveness. the local industry and the government has been vigilant in aligning the national industry with the global markets, as the global value chains are sensitive to the environmental changes, therefore, they continuously re-stratify the value chains to remain competitive. on the other hand, the government institutions in pakistan are rigid and have a uniform culture across the country. they have different thinking pattern as they strictly operate within their institutional scope and rules. such characteristics of the government institutions and their employees affect the business in different ways. e. weaknesses of the government institutions the industry representatives explained the weaknesses of the government institutions. according to them, government institutions do not consider environmental dynamics and have weak control over the local market dynamics. their employees do not have specialization in their respective jobs. the government institutions do not help the industry to prepare themselves for global trade, as the industry expects them to address the problems in the global trade, especially those related to the global value chains, holistically. according ali asghar & rukhsana kalim 8 to them, the government is not setting the right direction to earn competitiveness in the global markets. instead, the government policies are directly bringing negative effects to the cost of doing business and ease of doing business in pakistan. the complex government procedures constrain the industry, instead of facilitating it. 2.3 step 3: imaginative variations imaginative variations are the third step in applying the tpa to the qualitative data. this step refers to deleting the statements which may be non-repetitive or non-overlapping but are irrelevant to the study (moustakas, 1994). moerer-urdahl and creswell (2008), in their study, have referred to this step as meaning units or themes. adu (2015) advocates using the qualitative data analysis (qda) software – nvivo – and use the cluster analysis technique in it to perform this step. a few critics may discourage the application of software to perform analysis at this step, they may be right if their data can be managed otherwise. since this study incorporates 23 in-depth interviews, the data is enough to leave the chances of human errors, if handled without software. therefore, this study performed this step using the nvivo to enhance its reliability of the analysis. the essence of imaginative variations is to identify and bundle up those statements, which provide a meaningful description of the phenomena or a significant part of it. according to adu (2015), the researcher has to gather the logically relevant statements under the themes, therefore, requires the imagination of the researcher to test the variation in the meanings they provide when are arranged in different assortments. the researcher has to arrange them in a way that makes sense or explanation of the phenomena. this step was performed by loading the data spreadsheet to nvivo. the data, already gone through phenomenological reduction, was analyzed using the nvivo and was coded. in the next step, the cluster analysis was performed on the codes to identify and bundle up the relevant codes. the clusters suggested by the software were ignored, where there was no or weak logical connectivity among the grouped statements, and the statements were presented in such an assortment that gives a better explanation of the phenomena. 2.3.1 structural descriptions the structural descriptions; as guided by moustakas (1994), moerer-urdahl and creswell (2008), and adu (2015) are the outcomes of the imaginative variations which explains that “how” it was experienced. following structural descriptions evolved as an outcome of the imaginative variation: the coordination between the industry and the government institutions is the area, which plays a pivotal role in the phenomena of economic sustainability of global value chains. the institutional efficiency and effectiveness suffer when the coordination between both actors of the phenomena reduces. it was evident from the expressions of the industry journal of applied economics and business studies, volume. 3, issue 1 (2019) 1-14 https://doi.org/10.34260/jaebs.311 9 representatives that the industry cannot flourish without the support of the government institutions. when the government representatives design the policy for industry, they do not have sufficient knowledge, experience and exposure of the industry. they, therefore, need the input from the experts from industry who may not be the true representative of the industry, especially the small and medium enterprises. competitiveness seems to be, though not completely but, ignored. the lack of coordination here leads to the gap between government policies and actions, the market demand, and the industry’s requirements. the government has its vision and plans and its own pace of implementing those plans. on the other hand, the market, especially the global market, operates at its own pace. the industry representatives, as they are the actors in the market, therefore, have a different view of the gaps towards competitiveness than government representatives. it does not overshadow the importance or relevance of the government but explains that the government institutions at times may play a less effective role instead of leading the industry towards competitiveness. global value chains provide promising opportunities to the pakistani industry, but pakistani government institutions cannot understand and respond to their needs. as it was evident from the discussion of the industry representatives, they are unable to operate at their full capacity because of the scarcity of orders. it is not as if the global apparel demand is diminishing, but the government policies are not attractive for the global brands to enter into long-term production contracts with pakistani production houses, therefore they not only have to operate below their installed capacity but also have to bear the cost of attracting new customer more frequently. the government institutions do not understand the importance and the needs of the global value chains, nor their convention, slow and rigid systems sufficiently facilitate them. 2.4 step 4: synthesis of meanings synthesis of meanings is the final step of the transcendental phenomenological analysis (adu, 2015). this step comprises of synthesizing the textual description (the “what” part) and the structural descriptions (the “how” part), and express them as the essence of the experience. this essence of the experience can, therefore, be stated as a comprehensive description of the phenomena ‘from the point of view’ of those who have experienced it (moustakas, 1994). moerer-urdahl and creswell (2008) have described this step as “intuitive integration” of textual (what) and structure (how) descriptions. according to them, these descriptions capture the meaning ascribed to the experience or the “essence”. ali asghar & rukhsana kalim 10 2.4.1 the essence of the experience the phenomena of economic sustainability of global value chains are critical to the existence of the stakeholders of the global value chains. each stakeholder has different stakes and risks involved in it, which makes this phenomenon important for them. this finding also acknowledges the arguments of khan, lew, and akhtar (2016). the production units hired by the brands have a lot of pressure on them as they have to manage with their customer, their labour, and the government institutions (gereffi & lee, 2014; lund‐ thomsen, nadvi, chan, khara, & xue, 2012). if they fail to do so, their customer already has the option to shift its production to some other unit. the production units or the industry seeks high-quality support from the government institutions in return of the tax they pay and the legal requirements they fulfil, to help them in sustaining themselves as part of the gvcs. since the industry flourishes with the support of the government institutions; therefore, cooperation and coordination between them are very critical. the industrial policy plays a vital role in development and up-grading of a certain industry (khan et al., 2016; khan, mo yamin, lew, & sinkovics, 2015), but the participants of this study explained that it is not their case as the policy related to the apparel industry seems not been based upon the ground realities. on the other hand, the government organizations comprise of highly qualified and dignified civil servants. the coordination and cooperation with them become very tricky when the government institutions, being rigidly procedural, are unable to match their pace with the industrial demands. coinciding with the argument of yousaf, ihsan, and ellahi (2016), it was found that the lack of coordination and cooperation, added with the lower than expected efficacy of the government institutions affects the competitiveness, conduciveness and attractiveness of a country for the global buyers for long-term production agreements. 3 implications this transcendental phenomenological study explored the role of government institutions in the phenomena of economic sustainability of global value chains, which is a unique practice in the relevant literature. this study incorporated the qualitative data analysis (qda) software – nvivo – to perform the phenomenological reduction, which gives it an edge over the conventionally published phenomenological studies. the essence of this study highlights the importance of government institutions in the economic sustainability of the global value chains, especially in the context of pakistan. the transcendental phenomenological analysis (tpa) performed over the qualitative data gathered through interviewing 23 export managers from the apparel industry provides a good description of how the phenomenon takes shape in reality: what role the government institutions play in the economic sustainability of global value chains? and how? the essence of the study highlights the policy level implications, which must address the ways journal of applied economics and business studies, volume. 3, issue 1 (2019) 1-14 https://doi.org/10.34260/jaebs.311 11 to increase the cooperation between the industry and the government institutions. the policy must ensure their cooperation, information exchange, agreement on economic objectives, and updated knowledge of market dynamics. at an operational level, the industry and the government institutions can enhance their cooperation with each other by establishing the objective-based common forums, think tanks, and conducting series of events where both can better know and understand each other. 4 directions for future research this study guides to test the phenomenon of economic sustainability of the global value chains through other perspectives, thereby applying different methodologies and methods to bring more rigour in the knowledge of phenomena. besides, the researchers interested in this area can replicate this research to explore the role of other stakeholders of the global value chains, across the value chains other than apparel, or across different geographical locations. further, studies comparing the roles and contributions of different stakeholders in this phenomenon can also be conducted to expand the body of knowledge in this area. 5 limitations of the study: authors’ reflection transcendental phenomenological analysis (tpa) is a complex qualitative method in which the lived experience of the participants of a phenomenon are gathered, transformed into text, and analyzed through the four steps of analysis. each of these steps has its complexity and challenges during the execution, especially managing the thin margins of subjectivity. since this study has been extracted from a phd dissertation that has a bigger scope, therefore, it explains the phenomena from the perspective of industry representatives only. author’s note: i hereby declare that the given paper has been extracted from my phd thesis, and/or is an extension of my previous research work. references adu, p. (2015). using nvivo to conduct transcendental phenomenological analysis. https://www.slideshare.net/kontorphilip/using-nvivo-to-conduct-transcendentalphenomenological-analysis ali asghar & rukhsana kalim 12 arrow, k. (1971). essays in the theory of risk-bearing. amsterdam: north-holland publishing company. barrie, l. (2014, april 08, 2014). us: walt disney bans pakistan and bangladesh production. just-style. cano-kollmann, m., cantwell, j., hannigan, t. j., mudambi, r., & song, j. (2016). knowledge connectivity: an agenda for innovation research in international business: springer. cibangu, s. k., & hepworth, m. (2016). the uses of phenomenology and phenomenography: a critical review. library & information science research, 38, 148-160. doi: 10.1016/j.lisr2016.05.001 creswell, j. w. (2014). research design: qualitative, quantitative, and mixed methods approaches. usa: sage publications, inc. davis, m. (2013). pakistan exports expected to increase as apparel brands leave bangladesh. sourcing journal. eisenhardt, k. (1989). agency theory: an assessment and review. academy of management review, 14(1), 57-74. eller, k. h. (2017). private governance of global value chains from within: lessons from and for transnational law. transnational legal theory, 8(3), 296-329. gereffi, g., & lee, j. (2014). economic and social upgrading in global value chains and industrial clusters: why governance matters. journal of business ethics, 133(1), 25-38. doi: 10.1007/s10551-014-2373-7 hennink, m., hutter, i., & bailey, a. (2014). qualitative research methods. los angeles: sage publications inc. holmström, b. (1979). moral hazard and observability. the bell journal of economics, 10(1), 74-91. husserl, e. (2012). ideas: general introduction to pure phenomenology: routledge. jacobs, b. w., & singhal, v. r. (2017). the effect of the rana plaza disaster on shareholder wealth of retailers: implications for sourcing strategies and supply chain governance. journal of operations management, 49, 52-66. jensen, & meckling. (1976). theory of the firm: managerial behavior, agency costs, and ownership structure. journal of financial economics, 3(4), 305-360. kaplinsky, r., & morris, m. (2000). a handbook for value chain research (vol. 113): university of sussex, institute of development studies. khan, z., lew, y. k., & akhtar, p. (2016). the influence of industrial policy and national systems of innovation on emerging economy suppliers’ learning capability. industry and innovation, 23(6), 512-530. doi: 10.1080/13662716.2016.1189811 khan, z., mo yamin, p. r. r. s. d., lew, y. k., & sinkovics, r. r. (2015). the mirage of upgrading local automotive parts suppliers through the creation of vertical linkages journal of applied economics and business studies, volume. 3, issue 1 (2019) 1-14 https://doi.org/10.34260/jaebs.311 13 with mnes in developing economies. critical perspectives on international business, 11(3/4), 301-318. doi: 10.1108/cpoib-12-2012-0064 laverty, s. m. (2003). hermeneutic phenomenology and phenomenology: a comparison of historical and methodological considerations. international journal of qualitative methods, 2(3). lee, j., gereffi, g., & beauvais, j. (2012). global value chains and agrifood standards: challenges and possibilities for smallholders in developing countries. proceedings of the national academy of sciences, 109(31), 12326-12331. luhman, h., & cuniffe, e. (2012). agency theory: an assessment and review. academy of management review, 42, 12-43. lund‐thomsen, p., nadvi, k., chan, a., khara, n., & xue, h. (2012). labour in global value chains: work conditions in football manufacturing in china, india and pakistan. development and change, 43(6), 1211-1237. moerer-urdahl, t., & creswell, j. w. (2008). using transcendental phenomenology to explore the "ripple effect" in a leadership mentoring program. international journal of qualitative methods, 3(2), 19-35. moran, d. (2002). introduction to phenomenology. london and new york: routledge. moustakas, c. (1994). phenomenological research methods. california, usa: sage publications inc. nadvi, k., & halder, g. (2005). local clusters in global value chains: exploring dynamic linkages between germany and pakistan. entrepreneurship & regional development, 17(5), 339-363. doi: 10.1080/08985620500247785 panda, b., & leepsa, n. m. (2017). agency theory: review of theory and evidence on problems and perspectives. indian journal of corporate governance, 10(1), 74-95. rana, i. (2013, december 26, 2013). walt disney may stop sourcing goods from pakistan, the express tribune. retrieved from https://tribune.com.pk/story/650581/exportproducts-walt-disney-may-stop-sourcing-goods-from-pakistan/ reinecke, j., & donaghey, j. (2015). after rana plaza: building coalitional power for labour rights between unions and (consumption-based) social movement organisations. organization, 22(5), 720-740. reporter. (2014, 03 april 2014). walt disney stops imports from pakistan, trans asia news service. retrieved from http://transasianews.com/asia/53-pakistan/1338-waltdisney-stops-imports-frompakistan?tmpl=component&print=1&layout=default&page= rodrik, d. (2018). new technologies, global value chains, and developing economies: national bureau of economic research. ross, s. (1973). the economic theory of agency: the principal's problem. american economic review, 63(2), 134-139. ali asghar & rukhsana kalim 14 saunders, m., lewis, p., & thornhill, a. (2009). research methods for business students (5th ed.). england: pearson education ltd. shapiro, s. p. (2005). agency theory. annual review of sociology, 31, 263-284. shavell, s. (1979). risk sharing and incentives in the principal and agent relationship. the bell journal of economics, 10(1), 758-798. siddiqui, j., & uddin, s. (2016). human rights disasters, corporate accountability and the state: lessons learned from rana plaza. accounting, auditing & accountability journal, 29(4), 679-704. sinkovics, n., hoque, s. f., & sinkovics, r. r. (2016). rana plaza collapse aftermath: are csr compliance and auditing pressures effective? accounting, auditing & accountability journal, 29(4), 617-649. spence, m., & zeckhauser, r. (1971). insurance, information, and individual action. american economic review, 61(2), 380-387. yousaf, m., ihsan, f., & ellahi, a. (2016). exploring the impact of good governance on citizens' trust in pakistan. government information quarterly, 33(1), 200-209. doi: 10.1016/j.giq.2015.06.001 journal of applied economics and business studies, volume. 3, issue 1 (2019) 79-90 https://doi.org/10.34260/jaebs.315 79 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x an empirical analysis of the impact of working capital management on the firm performance in cement and ceramics industry of pakistana panel approach rashida bibi1, fayaz ahmad2 & muhammad aamir3* 1faculty of management sciences, international islamic university islamabad, pakistan 2department of statistics abdul wali khan university mardan, kp, pakistan 3department of statistics abdul wali khan university mardan, kp, pakistan abstract the real impact of working capital management is captured through its constituent policies such as inventory, receivable, and payable managements. the aim of this study is to analyze the effects of working capital management i.e. inventory turnover period, the cycle of cash conversion, accounts receivable and payable periods, on the firm performance of cement and ceramics industry of pakistan. the target population of the current research work is the whole cement and ceramics industry of pakistan. the fixed-effect model was used for empirical analysis and incorporated by using the stata software. the model parameters were tested at 1% and 5% level of significance. the outcomes of the fixed-effect model indicated that the account receivable period (ar) in days, inventory turnover (it) in days and cash conversion cycle (cc) has an inverse nexus with the firm value of cement and ceramics industry of pakistan. however, the account payable period has a positive relationship with the profitability of the cement and ceramics industry of pakistan. keywords cash conversion cycle, payable account period, fixed effect model jel classification g3, g30, g39, g32 1. introduction the social and economic development of the country depends on infrastructure development. therefore, the construction industry occupies an important place for the development of the infrastructure of the country, the economic growth of the country and more job opportunities. in this respect, the cement and ceramics industry occupy an important place because it acts as the barometer in the emerging countries of the world like pakistan. the number of plants of cement and ceramics in pakistan is increasing so that the increasing demand of the country for cement and ceramics can be met from this. the cement and ceramics industry currently have 49 units, and it has the 45 million annual production capacity to produce and meet the needs of the country. there are many * aamirkhan@awkum.edu.pk rashida bibi, fayaz ahmad & muhammad aamir 80 infrastructure developments projects which are in progress in pakistan, and the cement and ceramics are a basic need for their completion. the cpec project is one of its examples for making the economy strong, and there is a need to take measures so that these projects become successful. the success of these projects carries development for the overall economy of pakistan (ajilore, 2009). however, the projects’ success gets impact from the resources’ management. therefore, for the successful completion of projects, cement is needed for developing the ongoing projects on the healthy footing. in this way, the working of the industry is important for the economy of pakistan. therefore, the effective management of the resources is needed in this industry. the need for cement increases in pakistan, and therefore this sector is obtaining the loan (rs 75 billion) for meeting the increasing need for the cement and ceramics production in-country (arifin, 2013). the resources need to manage in this industry in an effective way like working capital management thus better results can be seen for the industry in the long run. the study sets the target to know about the impact of working capital management on the value of the cement and ceramics industry of pakistan. the study fills the gap left by previous studies, as it is taping the effect of working capital management (wcm) on the firm’s performance by using the accounting measure as well as the market measures. there are studies on this connection in pakistan but with accounting measures mainly and inclusion of market measure with accounting measure is the uniqueness of this work. therefore, through the help of this measure, the definite impact of working capital management (wcm) on the performance of wcm can be explored. the results of some of the past studies indicated that wcm influences the performance of the firms. the firms which can manage the working capital effectively can generate more profitability as inked to those who are unable to manage the wcm (ganesan, 2007). wcm capital helps to improve the value of the firms because managers keep an eye on every aspect of the wcm in the organizations (nwaobia and adedeji, 2014). 2. literature review the performance of the companies is important, because if companies are earning profitability, then they less rely on the loan from loan advancing institute. therefore, from this, it has inferred that the company’s operations go in the correct way and ultimately the profitability comes for the organizations in the long run. but the better performance also depends upon resources’ managing of the companies. working capital management is also one of the main resources of companies (richard, 2013). the current work is exploring the impact of the working capital management on the value of the cement and ceramics industry of pakistan. because effective management can lead to a better working journal of applied economics and business studies, volume. 3, issue 1 (2019) 79-90 https://doi.org/10.34260/jaebs.315 81 of the companies in the long run. the number of studies exists on the connection between the wcm and the value of the organizations. but the positive, as well as negative both relationships, exists between wcm and profitability of the organizations in the past literature (riyazahmed, 2016). the current liabilities, as well as current assets difference, is perceived as the wcm in the past literature (roselyn gakure, 2012). current assets denote account receivable, inventories, and cash and cash equivalent. the management of the cash and cash equivalent, account receivables and inventories, and payables is working capital management (ahmadi, arasi, and garajafary, 2014). the wcm is considered an important component for the better performance of the organizations. because resources are needed for the effective working of the companies. the shortage of wcm disturbs the workings of the companies in a negative way (banos-caballero, garcia-teruel, and martinez-solano, 2014). it is explored from the past literature that performance of the organizations and wcm has a positive relationship because the organizations that are effectively managing the working capital are noticing the long-run positive impact on the company’s value (suresh babu, 2014). the working capital management’s importance can be shown by the role that it plays in enhancing the performance of the organizations in a positive way (muneebi, 2012). the statistically significant link between the wcm and performance of the companies have denoted in past studies. it has applied the regression analysis approach on the 131 companies and data was taken from 2001 to 2004 for exploration of the intended relationship (lazaridis and tryfonidis, 2006). while in another study, it is explored that the performance of the organizations has a negative association with the wcm and illustrated in more details. the study was shown on the companies operating in the context of saudi arabia (eljelly, 2004). the influence of the wcm on the profitability of the companies has studied in detailed in past and highlighted that the impact of wcm on the performance of the organizations can be explored. the study was conducted in ghana and data was collected from the period 2005 to 2009, and with the application of the ordinary least squares (ols) technique, it has documented that wcm hurts the value of the firms. wcm comprises of the inventory turnover, account payable and receivable periods, and cash conversion cycle (cc) (vijayakumar, 2011). further, in pakistan context the results were different, and the study reveals that the performance of the organizations was hurt by the ratio of cash conversion cycle and inventory turnovers. while the account payable cycle has an encouraging influence on the value of the firms. work applies the panel approach, and along with this, the data over ten years has collected. the panel methodology was also used for the examination of the results of the study (rehman, afza, qayyumand, and bodla, 2010). the positive rashida bibi, fayaz ahmad & muhammad aamir 82 association between the performances of the companies and the inventory turnover period (it) has explored in the past study (usama, 2012). whereas on the opposite side, a negative relationship was explored among these two variables (zariyawati, diana and hirnissa, 2017). while, (almazari, 2013) has explored a significant negative relationship among the return of assets accounts receivable period (ar) variables. moreover, while in another study, the same relationship has explored between the ar and tobin (gakure, cheluget, onyango, and keraro, 2012). from the past literature (saptarshi, 2018) has also explored a positive relationship among the profitability of the companies and account payable period. the return on equity gets impact by the account payable period in a negative way for firms taken (mehtap, 2016). chashmsayadan and aghajan (2014) have explored a positive relationship between the performance of the companies and the cash conversion cycle. however, on the opposite side (ukaegbu, 2014; gulia, 2014; sivasankaran, 2018) has also explored a negative relationship among these two variables i.e. performance of the companies and cash conversion cycle. therefore, in this study, the impact of working capital management value on the organizations has explored by utilizing all these four factors including inventory turnover, account payable and receivable, and cash conversion cycle in the cement and ceramics industry of pakistan. 3. methodology the study uses the data of the companies falling under the cement and ceramics industry of pakistan. the data has taken from 2007 to 2016 for all those companies for which the data is available. the cement and ceramics industry of pakistan is the population for the current work. there is a total of 29 cement and ceramics companies listed at the pakistan stock exchange. therefore, the convenience sampling technique was used so that better results can come for the study. the 22 companies have involved for data collection, and total observations are 220. the data used in this study is secondary and panel. the study uses a correlation matrix, descriptive statistics, and a fixed-effect model. based on the hausman test results the application of the fixed-effect model was incorporated. the model for the study as follows: 𝑇𝑜𝑏𝑖𝑛 = 𝛼 + 𝛽1𝐴𝑅𝑖𝑡 + 𝛽2𝐴𝑃𝑖𝑡 + 𝛽3𝐼𝑇𝑖𝑡 + 𝛽4𝐶𝐶𝑖𝑡 + 𝛽5𝐴𝑔𝑒𝑖𝑡 + 𝛽6𝐿𝐸𝑉𝑖𝑡 + 𝑒𝑖𝑡 (1) here tobin is a proxy used for the firm performance, α is an intercept of the model, ap is the account payable period, ar denotes the period of account receivable, cc indicates a cycle of cash conversion, it is the ratio of stock turnover, and age is the firms’ age, and lev is the financial leverage, further coefficients of included variables are denoted by, “t” is a time and “i” is a number of firms, and finally error term is denoted by e. the four independent variables and one dependent variable of the study have denoted in this section also. tobin is a proxy used for firm value (performance). besides, it has journal of applied economics and business studies, volume. 3, issue 1 (2019) 79-90 https://doi.org/10.34260/jaebs.315 83 used in past studies too (mcnichols, rajan, and reichelstein, 2013). the detail on the variables of the study has denoted in the table given below. the independent variables have used in the past work of ntui, chrispina, tago, mkiibi, (2014). the size and leverage have used in the past study of coad, (2014). table 1: variables of the study no. variables proxy 1. firm value (or firm performance) • tobin (book value of leverage + equity’s market value) / total value of asset’s book • roa total net income / average assets of firms 2. working capital management • ar (total receivables / total sales)*365 • it (total inventory/ total cost of goods sold)*365 • ap (total payables /total purchase)*365 • cc [{(inventory minus account payable) divided by cost of goods sold} plus (total receivables / total sales)]*365 3. variable used as control • financial leverage total leverage / total assets of firms • age natural logarithm of age since firms establishment 4. results and discussion the descriptive statistics of the variables used in the study has denoted so that the properties of the variables can be determined from this. the average return on the assets of companies is 0.037, and some of the firms are noticing the negative return on assets (-0.270) while some of the firms have positive and more return on the assets of the companies. next is the tobin which is the other measure of firm performance, and the average value of the firm performance based on market information is 1.140, its maximum, and minimum value has also reported in the continuing section of this work. the average days of account receivable (ar) are 18.44, and some firms collect the account receivable in less than one day while some take 613.28 days for the collection of account receivables. on the other hand, the average days of account payable (ap) are 75.878 and minimum and maximum days denote the extent of days used in companies for payment credit purchase payment. the cash conversion cycle is 24.201 days on average. the inventory of the companies takes 45.69 days for the disposal. further, the control variables denote that the age of companies is 9.19 and lev is .559 on average. rashida bibi, fayaz ahmad & muhammad aamir 84 table 2: descriptive statistics variable mean standard deviation minimum maximum roa 0.037 0.102 -0.270 0.244 tobin 1.140 0.463 0.437 2.802 ar 18.448 45.347 0.022 613.286 ap 75.878 105.841 0.602 614.019 cc 24.201 64.916 -362.107 252.165 it 45.692 79.190 -683.671 645.055 age 9.198 0.288 8.539 9.907 lev 0.559 0.290 0.138 2.407 note: roa is the return on assets and tobin is market-based performance measures, ar is account receivable period, ap is period of account payable, cc is a cycle of cash conversion, itd is inventory turnover period, age denotes the age of firms, and lev is the debt of companies. table 3 presented the correlation matrix of this study. the correlation of the variables ar, ap, and it is negative with the roa of the study. but the correlation of cc in the present study is positive. the correlation of all the independent variables is negative with the tobin of the study, which is one measure of firm performance. the correlation among all variables is less than 70% in the present study and presented in table 3. therefore, from this, it has inferred that there is no problem of multicollinearity in the predictor variables of the study. table 3: correlation matrix variable roa tobin ar ap cc it age lev roa 1.000 tobin 0.128 1.000 ar -0.161 -0.028 1.000 ap -0.361 -0.027 -0.066 1.000 cc 0.169 -0.075 0.196 -0.668 1.000 it -0.002 -0.107 -0.530 0.093 0.230 1.000 age -0.168 -0.106 0.024 0.270 -0.117 0.103 1.000 lev -0.593 0.176 0.179 0.328 -0.121 -0.091 0.125 1.000 note: variables are the same as under table 2 the lm test of breuschgodfrey serial correlation is presented in table 4 which depicts that there is no serial correlation among the study variables as values are higher than 5% and the null hypothesis is accepted for this test, which favours the absence of serial correlation or there is no serial correlation. table.4: lm test of serial correlation model p-value roa .106 journal of applied economics and business studies, volume. 3, issue 1 (2019) 79-90 https://doi.org/10.34260/jaebs.315 85 tobin .109 note: the null hypothesis favours the absence of serial correlation. the p-value of the wald test for both the models depicts that endogeneity does not prevail in data, because the p-value is higher as related 5% for both the models (table 5). table.5: wald test of endogeneity model p-value roa .101 tobin .108 note: the null hypothesis favours the absence of endogeneity. the hausmann test (ht) was used for the exploration of the best model of the study. the p-values of the ht tests are less than 0.05 which reveals that the fixed-effect model is suitable for the current. the null hypothesis of the current study is that, that the random effect models are the best choice. however, the p-values of the ht test favours the fixedeffect model. in this way, the fixed-effect model has finalized as a final model of study because the presentation of the right model denotes the reliable results of the study. table 6: hausmann test model p-value roa <0.001 tobin <0.001 the fixed-effect model results of the current study are presented in table 7. the current study is applying to two measures of the firm performance, firstly is the market based while secondly is the accounting-based measures. the same results under the two models denoted the robustness and model exactness of the study. table 7: fixed effect results dependent variable tobin (1) roa (2) ar -.033** -.0004 * ap .053** .0005*** cc -.011* -.0002* it -3.541*** -7.591 age 1.299*** 0.165*** lev 0.292* 0.167*** constant -10.919*** -1.392*** r square 0.302 0.197 observations 220 220 note: ***, ** and * denote that variables are significant at 1%, 5% and 10% level the model results under proxy tobin denoted that the period of account receivables in days (ar) has an inverse effect on the firm performance with coefficient 0.033 at 5%. it rashida bibi, fayaz ahmad & muhammad aamir 86 depicts that increasing one unit in ar the firm performance decreases by – 0.033 units and vice versa. which means that an inverse relationship exists among these two variables of the study. consequently, it has inferred that the companies who are taking more days for the collection of the account receivables from the customers are decreasing their performance. therefore, the company’s value stays stable for the long survival of the companies. the companies can make the effective policies of an early collection of the account receivable by giving the discount to customers who are making early payments. the account receivables are created through credit sales from the companies. because all the sales cannot be made on a cash basis and therefore for enhancing sales, some of the sales have made on credit. the early collection ensures the more cash available, which assists in running the day to day operations of the business and in this way, firms’ value acquires impact. the results are in line with a study in the same field (ntui et al. (2014). the account payable period in days (ap) has a positive link with the firms with a coefficient .053 at 5%. therefore, from this, it has depicted that one unit increase in ap bringing 0.053 change in the firm performance. meaning that direct relationship exists among these two variables of the current study. hence, from these results, it has inferred that the companies who are taking more days for payments of the account payables to the customers are increasing their performance in the long run. the companies can make the effective policies of late payments of the account payables by giving the notice to suppliers who are waiting for the early payments. the image of the company also matters for the businesses and negotiation with the suppliers can help to achieve this end. the account payables are created through credit purchases made by the companies. because it is not possible for the organizations to purchase all the goods needed for companies operations running on the cash basis, the late payments ensure the more cash available in hand of organizations which they can invest according to the urgent need of the business and in this way ultimately the company’s performance gets affected. the results are in line with the study in the same field (sivasankaran, 2018). the cash conversion cycle (cc) in days has an inverse impact on the performance of the companies with coefficient -0.011. it indicates that the company takes additional days from getting things for doing the projects and then a collection of the cash from those projects the performance of the companies will affect negatively. hence, a negative relationship between these two variables of the study existed. the results are according to the study of ukaegbu (2014). it denotes the negative effect on the value of the companies at 1%. however, it denotes that the one unit increase in days for selling and converting the inventory into cash leads towards the decrease in the profitability of the firms. the companies who are taking care are consuming less time in converting the inventory into cash so that the sales of the business increase. also, the charges which are attached to the storage of inventory in the storeroom decreases and ultimately, the performance of the journal of applied economics and business studies, volume. 3, issue 1 (2019) 79-90 https://doi.org/10.34260/jaebs.315 87 organization increases. therefore, we can say that the more the days consume in selling the inventory, the less will be the profitability of the firms. the results are consistent with the (zariyawati et al., 2017) study in the same area. furthermore, tobin has a positive effect on the age of the companies because old firms have more resources and experience to diversify risk and increase performance. the other control variable leverage ratio (lev) denotes that the companies with leverage are experiencing more profitability as linked to those who are deprived of such a facility. further, itd is the strong predictor of this model, and in this way, we can perceive that more change in tobin is explained by the it. the 30.22% change in tobin has explained by the independent variables and control variables of the study. the results are the same as past work in this area (coad. 2014). model two of the study where roa is used as a measure of the firm performance is also denoting the same relationship. but one variable itd is insignificant under this model as related to model 1 of study. the predictive power of the model is also less under this model as linked to model one (table 4.4). therefore, the application of the two models signifies the robustness of the study, and in addition to this, it adds to the literature on wcm and companies’ value in the pakistani context. 5. conclusion and future work the aim of this study is to analyze the effects of working capital management i.e. inventory turnover period, the cycle of cash conversion, accounts receivable and payable periods, on the firm performance of cement and ceramics industry of pakistan. the whole industry of cement and ceramics is the population of the current work, but data has collected on those companies, which are easy to access and for which data was easily available. the stata software was used, and besides, the fixed-effect model has applied for exploring the results of the study. the results of the hausman have suggested the use of the fixed-effect model. for ensuring the robustness of the study, two measures of the firm performance have used. tobin is a market-based measure, and the other one is an accounting-based measure for the current study. the results signify that the period of account receivables (ar) in days (ar) has an inverse association with the firm performance at a 5% level. it means that an inverse relationship between these two variables of the study exists. hence, from these results, it is inferred that the companies who are taking more days for the gathering of the account receivables from the customers are decreasing their performance in the long run. the account payable period in days (ap) has a positive association with the firm value at a 5% level. therefore, from this, it has inferred that the companies who are taking more days for payments of the account payables to the customers are increasing their performance in the long run. the cash conversion cycle (cc) in days has a negative effect on the performance rashida bibi, fayaz ahmad & muhammad aamir 88 of the companies. it denotes that it's more presence lowers the value of the firms at 1%. therefore, it signifies that the one unit increase in more days for selling and converting the inventory into cash leads towards the decrease in the earning of the firms. the policy implication of the work is that the increase in the number of days required for making a payment could lead to enhance the earning of firms. however, negotiation is crucial in this context so that the overall image of the company cannot be affected. further, the concerned managers should reduce the days required to collect the cash from customers by giving the discount, and the inventory should be selling in a smaller number of days. besides the short cycle of cash conversion, there is a demand for increasing the earning of the firms. the limitation of the present work is that the data has collected for fewer years. so, in a future study, the long time frame can resolve this issue, and in this way, more deep insight as possible about the present phenomenon. references agha, h. (2014). impact of working capital management on profitability. european scientific journal, 10: 375 -381. ahmadi, m. arasi, i. s. and garajafary, m. (2014). studying the relationship between working capital management and profitability at tehran stock exchange: a case study of the food industry.research journal of applied sciences engineering & technology, 4(13): 1868-1874. ali, n. jaffar, m.a.a. and zaidi, s.m.r. (2015). the cement industry of pakistan: a swot analysis. journal for studies in management and planning, 1. almazari, a.a. (2013).the relationship between working capital management and profitability: evidence from saudi cement companies. british journal of economics, management, and trade, 4. akoto, r.k., awunyo-vitor, d., and angmor, p.l. (2013). working capital management and profitability: evidence from ghanaian listed manufacturing firms. journal of economics and international finance, 5: 373-379. banos-caballero, s., garcia-teruel, p. j., and martinez-solano, p. (2014). working capital management, corporate performance, and financial constraints. journal of business research, 67: 332-338. eljelly, a.m.a., (2004). liquidity-profitability tradeoff: an empirical investigation in an emerging market. int. j.commerce manage, 14: 48-61. journal of applied economics and business studies, volume. 3, issue 1 (2019) 79-90 https://doi.org/10.34260/jaebs.315 89 enqvist, j., graham, m., and nikkinen, j. (2014). the impact of working capital management on firm profitability in different business cycles: evidence from finland. research in international business and finance, 32: 36-49. ganesan, v. (2007). an analysis of working capital management efficiency in telecommunication equipment industry. rivier academic journal, 3(13). gulia, r. (2014). effects of working capital management on firms profitsevidence from the pharmaceutical sector. international journal of management and social sciences research, 3: 103-107. lamptey, l. l., frimpong, k., and morrison, a. b. (2017). empirical study on the influence of working capital management on performance of smes in a developing economy. british journal of economics, management & trade, 17: 1-10. lazaridis, i. and tryfonidis, d. (2006). ‘relationship between working capital management and profitability of listed companies in the athens stock. journal of financial management and analysis, 19. mathuva, d.m. (2010). influence of working capital management components on corporate profitability: a survey on kenyan listed firms. research journal of business management 3(1): 1-11. pollanen, r., abdel-maksoud, a., elbanna, s., and mahama, h. (2017). relationships between strategic performance measures, strategic decision-making, and organizational performance: empirical evidence from canadian public organizations. public management review, 19(5): 725-746. rehman, m. u., and anjum, n. (2013). determination of the impact of working capital management on profitability: an empirical study from the cement sector in pakistan. asian economic and financial review, 3(3): 319. saptarshi, d. (2018). analysing the impact of working capital management on profitability: a study on dse listed cement companies in bangladesh. journal of management and business research, 18:1-8. sarwat, s., iqbal, d., durrani, b. a., shaikh, k. h., and liaquat, f. (2017). impact of working capital management on the profitability of firms: case of pakistan’s cement sector. journal of advanced management science vol, 5(3). ukaegbu, b. (2014). the significance of working capital management in determining firm profitability: evidence from developing economies in africa. research in international business and finance, 31: 1-16. rashida bibi, fayaz ahmad & muhammad aamir 90 usama, m. (2012). working capital management and its effect on a firm's profitability and liquidity: in another food sector of (kse) karachi stock exchange. arabian journal of business and management review (oman chapter), 1(12): 62. vijayakumar, a. (2011). cash conversion cycle and corporate profitability – an empirical inquiry in indian. automobile firms international journal of research in commerce, it & management, 1(2): 84. yazdanfar, d., and ohman, p. (2014). the impact of the cash conversion cycle on firm profitability: an empirical study based on swedish data. international journal of managerial finance, 10(4): 442-452. zuberi, a.(2016). importance of cement industry in infrastructure development of pakistan, business recorder. https://fp.brecorder.com/2016/06/2016061456594/ https://fp.brecorder.com/2016/06/2016061456594/ journal of applied economics and business studies, volume. 3, issue 1 (2019) 61-78 https://doi.org/10.34260/jaebs.314 61 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x a spatiotemporal approach to measure impact of factors of production on innovations muhammad imran1*, muhammad rafiq2 & imranullah3 1assistant professor, department of economics, bacha khan university, charsadda. 2associate professor, institute of management sciences, peshawar. 3lecturer, department of economics, university of buner. abstract measuring innovations is momentous in research work on industrial location, regional economic planning and industrial profits. however, fewer studies have considered spatiotemporal factors for analysing the impact of factors of production on innovation till this date. present study is therefore an attempt to fill this void by developing a model relating to factors of production effects on industrial innovation with continuous spatial locations for pakistan. our methodology is consistent with the approach espoused by desmet and hansberg (2014). the model assumes that labour and capital affects the innovation at each available location during the production process. this model has also incorporated spill over-effects which declines with the upsurge in distance. migration of labour act as the basic element for the speculation of higher-level of innovation. successively, simulation of the model for a developing economy like pakistan, we conclude that the share of capital to innovation is much higher relative to the share of labour, but comparatively spill over-effects of capital is susceptible to the increase in distance. our study has theoretical contribution in the sense that we have improved the existing work of desmet and hansberg (2014), where labour was considered as the only source of innovation and capital, as an endogenous variable, was neglected. keywords industrial innovation, spatiotemporal approach, spillover effects, economy of pakistan. jel classification o14, e24, r12, r32 1. introduction economic literature has been systematically employed to evaluate different regionaleconomic problems pertaining to labour, capital, industrial output and economic structure. according to the complex economic system each region even with different level of development faces competition either from the neighbouring or comparatively developed regions. consequently, every region is trying to potentially utilize their resources to attain certain level of economic development by innovation to survive in the competitive neighbourhood. whereas, scarcity of one or more factors of production are susceptible to * imranecon@bkuc.edu.pk muhammad imran, muhammad rafiq & imranullah 62 the economic growth and competition with the contra regions. therefore, economic development mainly depends on the ability of regions to take advantage of their innovation capability by employing the available factors of production more efficiently. whilst, availability of different factors of production and proximity of concerned regions to the larger markets, adjoining developed regions are considerable makings of the region’s economic development through spillover-effects. previous literature in this field has either overlooked this important dimension or studies have scantly addressed this aspect. as an example, desmet and hansberg (2014), boucekkine et al. (2009), dunning and lundan (2008), caves (2007), and hilber & voicu (2009) had discussed effects of capital on innovation by connecting regional innovation to foreign direct investment (fdi) and concluded that the fdi recipient economies can utilize funds to accelerate their innovation and improve their fiscal and monetary position. however, ellison et al., (2010) and henderson et al., (1995) studied innovation effects on geographical locations of industries and found that spillover-effects determine the final location of industries. amidst these dynamic problems, this study has made an effort to elaborate the effects of factors of production on innovation. although, labour and capital are assumed to be the mainstays of producing value addition and innovation, however, incorporating both factors of production in a continuum of locations and their movement is of extreme importance to elicit the phenomenon of spatial innovation, where geographical locations are lined up in a continuous space (quah, 2002; brito, 2004). further, analysing factors of production and its effect on innovation, distance between different locations is an important factor that negatively affect diffusion of technology (spillover-effect) and supply of value-added. but, analysing innovation in a continuous location makes the model intractable for some reasons; for example, distribution of economic activities across space and time to clear both factors market and goods market.1 the only way to overcome this problem is the supposition put forwarded by brock and xepapadeas (2008, 2010); boucekkine, camacho: zou (2009) and quah (2002) that each location in the space perform as an isolated location. the term isolation still keeps positive trade costs and spillover-effect. the aforementioned studies have assessed either the share of labour or capital for analysing the magnitude of innovation, however, this study has provided a dynamic platform to deal with both the factors of production in a single model and derive the shares of each factor of production to innovation. the conceptual observation of innovation in terms of location through different factors of production need a dynamic model composed of related endogenous variables that explain patterns (migration and final destination to work) of factor of production which further determine their level of effectiveness on spatial innovation. observing labour agglomeration 1 as described by desmet and hansberg (2014) journal of applied economics and business studies, volume. 3, issue 1 (2019) 61-78 https://doi.org/10.34260/jaebs.314 63 through cost-led circular-casualty, any increase in the level of production of value-added in a location, provides more preferences of value-added at lower cost to the local consumers. while, rational consumers always try to maximize their level of utility by migrating to comparatively cheaper locations. therefore, low cost-of-living attracts more labours and, consequently, increase the supply of labour in the specific location which further attract producers of value-added to the concerned location, desmet and hansberg (2014). agglomeration of industries to a specific location raise the level of innovation both through investment in innovation and through spillover-effect. therefore, share of each factor of production to innovation is an essential ingredient for restructuring and development of an economy. our study is thus a contribution to the existing literature by incorporating some of the caveats of the previous research work. this analysis is based on time series data for the economy of pakistan from 1976-2014. ostensibly, service sector has higher share to the aggregate value-added production in the country than manufacturing sector, and reasonably service sector utilizes higher number of both factors of production. furthermore, share of labour to the aggregate growth in value-added production is comparatively higher than capital. the forthcoming sections are organized as, section 2 deals with the model deliberation to elucidate the effects of factors of production on innovation. section 3 while shedding lights on the structure of economy of pakistan as share of two factors of production to service sector and manufacturing sector, presents the results of ours spatiotemporal approach . final section entails concluding comments about this research work. 2. the model this study assumes that each of the isolated location, lined up in a continuous space, is composed of manufacturing sector or service sector; both sectors utilize labour and capital to produce value-added products. secondly, both factors of production are contributing to innovation in each location where the level of innovations, further, affect the surroundings through spillover-effect. third, transportation of value-added and spillover-effects are prone to remoteness, where impact of spillover effect will be higher with proximity. fourth, the innovation draws in each sector are random and are categorized by its level: higher the level of innovation shall spur greater spatial development. industries face problem of labours ‘wages corresponding to the location of their industry; as supply of labour determines the magnitude of industries at a particular location. the model presented by desmet and hansberg (2014) assumes that labour "𝐿" is the sole maker of value added products and innovation. nonetheless, our study is an improvement in the sense that we have endeavoured to incorporate capital along with labour as a determinant of innovation. this is evenly distributed across continuous space where the muhammad imran, muhammad rafiq & imranullah 64 number of locations are limited from 0 to 1. each location "ℓ1" is composed of manufacturing sector "𝑀" and service sector "𝑆" while each sector produces value-added commodities by utilizing labour, capital and technology at time"𝑡". model assumes the amount of land is fixed and immobile, but labour and capital are mobile. primarily, labour migrate in search of higher level of utility obtained by the consumption of differentiated goods and services. consumers order instantaneous utility function 𝑈(𝐶𝑖 ) where “c” denotes consumption and sectors are represented by the subscript “𝑖 ”(𝑖 = 𝑀, 𝑆). labor holds time. mobility of labour towards their desired location, directs capital movements. consequently, model assumes capital movement as the cause of labour migration. 2.1. consumer preferences consumer across locations are free to migrate in search of comparatively higher level of indirect utility. whilst, consumers also try to arrange their utility-preferences. labour supply their unit of time to the factor markets that weigh them relatively higher wages "𝑤" . consumer’ preferences are arranged as 𝑈(𝐶𝑀 , 𝐶𝑆 ) and they consume value-added products at cost 𝑃𝑖 𝐶𝑖 = 𝑝𝑀𝐶𝑀 (ℓ1, 𝑡) + 𝑝𝑆 𝐶𝑆(ℓ1, 𝑡) . cost function is assumed to follow constant elasticity of substitution (ces) with homogeneity of degree one. following market clearing conditions𝐶𝑖 = 𝑋𝑖, where aggregate production "𝑋𝑖 " in one of the two sectors (𝑖 = 𝑀, 𝑆) at location "ℓ1" requires 𝐸(𝑥𝑖 ) = 𝑤𝑖 𝐿𝑖 (ℓ1, 𝑡) + 𝜋𝑖 𝐾𝑖 (ℓ1, 𝑡) as production cost function. the second part of the function was missing in desmet and hansberg (2014). hence, we have reformulated the model. consumer preferences for the two sectors’ product in location "ℓ1" at time "𝑡" are organized as following: max {𝐶𝑖(ℓ1,𝑡)} ∞ 0 ∑ (𝜀𝑀𝐶𝑀 𝜇 + 𝜀𝑆 𝐶𝑆 𝜇 ) 1 𝜇⁄∞ 0 ….. (1) 𝑠. 𝑡. 𝐸(𝑋𝑖 (ℓ1, 𝑡)) = 𝑃𝑖 (ℓ1, 𝑡)𝐶𝑖 (ℓ1, 𝑡) consumers purchase whatever is produced by the two sectors. therefore, maximization process of eq.1 will result into consumers’ demand for the two sectors’ value-added products, wherever, consumption of goods and services is affected by real rewards to two factors of production. 𝐶𝑆 (ℓ1, 𝑡) = [ 𝐸(𝑋𝑖(ℓ1, 𝑡)) (𝑃𝑆 + 𝑃𝑀 𝜆 1 𝜇−1⁄ ) ⁄ ] , 𝐶𝑀(ℓ1, 𝑡) = 𝜆 1 𝜇−1⁄ [ 𝐸(𝑋𝑖 (ℓ1, 𝑡)) (𝑃𝑆 + 𝑃𝑀𝜆 1 𝜇−1⁄ ) ⁄ ] where 𝜆 = ( 𝑃𝑀 𝜀𝑀 𝑃𝑆 𝜀𝑆 ⁄ ) journal of applied economics and business studies, volume. 3, issue 1 (2019) 61-78 https://doi.org/10.34260/jaebs.314 65 the comparative price index"𝜆": that brings the two sectors’ prices to quotient, reflecting consumer behaviour of weighing the two sectors’ prices that helps consumer in organizing their preferences for both the sectors. if"𝜆 > 1", means it cost more to the consumer to purchase manufacturing sector produce than service sector, and vice-versa. function of consumer’s demand (as eq.1) can be rearranged to obtain the level of indirect utility. therefore, �̅� = ∑ (𝜀𝑀𝐶𝑀 𝜇 + 𝜀𝑆𝐶𝑆 𝜇 ) 1 𝜇⁄∞ 0 , indirect utility is composed of comparative price index and the cost that consumers pay to buy two sectors production, �̅� = [ℎ𝑀𝜆 𝜇 𝜇−1⁄ + ℎ𝑀] 1 𝜇⁄ [𝐸(𝑋𝑖) (𝑃𝑆 + 𝑃𝑀𝜆 1 𝜇−1⁄ )]. 2.2. producers this is pretty straight forward that an economy at each location is composed of two economic agents, that is, consumers and producers. consumers are the suppliers of labour and producers consist of owner of capital to produce value-added goods and services in one of the two sectors at each location. producers obtain the number of labours "𝐿" at the wage2 "𝑤" and add "𝐾" amount of capital at "𝜋" interest rate and start production in one of the two sectors at location "ℓ1". furthermore, both sectors follow imperfect competition, where, goods and services are produced and sold to consumers. consequently, our model has some prominent features, first, explanation of the role played by capital in the process of production; second, the share of capital to innovation and third, shifting of capital resources with labour migration. land is free of cost and available to all those who win the bid for labours. 2.2.1. technology furthermore, each sector "𝑖" try to innovate "0 < 𝜂 < 1" level of innovation at cost "𝛺(𝜂(ℓ1, 𝑡)) > 0" to stay in the market and produce, while the probability of success in innovation is denoted by "𝜃" whilst "1 − 𝜃" stands for failure. the cost of current innovation is paid out by the returns generated through the sale of value-added products at time"𝑡 − 1", i.e. 𝛺(𝜂(ℓ1, 𝑡)) 𝑃𝑖 (ℓ1, 𝑡)𝑋𝑖 (ℓ1, 𝑡 − 1) ⁄ . if the producer succeeded in utilizing their investment in innovation today, then the "𝑖𝑡ℎ" sector will add innovation to the existing level of technology "𝐴−(ℓ1, 𝑡)" through the innovation multiplier "𝜈" and start production. with the probability pr(𝜈 < 𝜈𝑖 ) = ( 1 𝜈⁄ ) 𝜃 firm will be able to add 𝐸(𝐴𝑖 +(ℓ1, 𝑡)│𝐴 −(ℓ1, 𝑡)) of technology: which is conditional to the existing level of technology. the probability of success should be pareto optimal. 𝐸(𝐴𝑖 +(ℓ1, 𝑡)|𝐴 −(ℓ1, 𝑡)) = [ 𝜂+𝜃−1 𝜃−1 ] ∙ 𝐴−(ℓ1, 𝑡) 𝑓𝑜𝑟 𝜃 > 1 𝑎𝑛𝑑 𝜂 > 0 … (2) 2 wage rate is determined by the free mobility of labor, but labor migrate before the realization of innovation. muhammad imran, muhammad rafiq & imranullah 66 it shows, that how much the particular sector will be able to achieve success in innovation, will add to its existing "𝐴−" level of technology. the negative " − " superscript stand for the level of technology before innovation realization, and the plus " + " denotes level of technology after innovation realization. 2.2.2. spillover-effects and timing3 as mentioned, the producers are located in a continuous locations(0 𝑡𝑜 1). where, distance negatively affect both spillover-effect and sales of production, while spillovereffect is positively correlated across locations. let 𝑐𝑟(ℓ1, ℓ2) > 0 represent correlation of spillover-effect, therefore, if the distance between ℓ1 𝑎𝑛𝑑 ℓ2 increases or decreases the correlation of spillover-effect is assumed to be positive but it declines with the level of distance. accordingly, the realization of spillover-effect follows𝑒−𝑑│ℓ1− ℓ2│𝐴𝑖 +(ℓ2, 𝑡 − 1), which means that the realization of spillover-effect is negatively affected by the distance " − 𝑑" between two locations ℓ1 𝑎𝑛𝑑 ℓ2. 𝐴𝑖 −(ℓ1, 𝑡) = max ℓ2𝜖│0,1│ 𝑒−𝑑│ℓ1− ℓ2│𝐴𝑖 + (ℓ2, 𝑡 − 1) …. (3) the spillover-effects of location ℓ1 is subject to the level of technology at ℓ2: means location ℓ2 at time 𝑡 − 1 already attained higher level of technology than ℓ1at time 𝑡. 2.2.3. production function amidst production of value-added goods and services in one of the two sectors, producers in the beginning try to acquire moderately higher probability of success in innovation; a unit of land is available to those producers who can compete for labour at the prevailing rate of labour wage. furthermore, firm add their unit of capital 4 to start production at location"ℓ1". combining all the variables in location "ℓ1" at time"𝑡", the resulting cobb-douglas production function for two sectors is as: 𝑋𝑖 (ℓ1, 𝑡) = 𝐴𝑖 𝛾 (ℓ1, 𝑡) 𝐾𝑖 𝛼 (ℓ1, 𝑡)𝐿𝑖 1−𝛼 (ℓ1, 𝑡) 𝑤ℎ𝑒𝑟𝑒 𝑖 = 𝑀, 𝑆 𝑎𝑛𝑑 𝛼 + 𝛾 = 1 …. (4) max ℓ1 𝐸(𝑋𝑖 (ℓ1, 𝑡)) = 𝑃𝑖 (ℓ1, 𝑡)𝐶𝑖 (ℓ1, 𝑡) 3 spillover-effect are indirect source of agglomeration and is effected by the length of distance. 4 ibid page 7 𝑒−𝑑│ℓ1− ℓ2│ ℓ2 𝑡 − 1 space ℓ ℓ1 𝑡 spillover-effect journal of applied economics and business studies, volume. 3, issue 1 (2019) 61-78 https://doi.org/10.34260/jaebs.314 67 production 𝑋𝑖 is being determined by the three factors at the right side of the eq. 5. whatever is produced at cost 𝐸(𝑋𝑖 (ℓ1, 𝑡)) are sold at price𝑃𝑖 . under the assumption of spillover-effect and probability of success in innovation, model will take "𝛺" to denote the difference in cost of innovation5. in order to guarantee difference in cost of innovation, we utilized eq.3, that the realization of technology in location "ℓ1" at time "𝑡" is negatively affected by the distance between the two locations ℓ1 𝑎𝑛𝑑 ℓ2 . assuming the cost of innovation as 𝛺(𝜂) = (𝑤(ℓ1, 𝑡) + 𝜋(ℓ1, 𝑡)) [𝛺1 + 𝛺2 ∙ 1 1−𝜂 ] , 𝑤ℎ𝑒𝑟𝑒 𝛺2 > 0, 𝛺2 ′ > 0 𝑎𝑛𝑑 𝛺2 ′′ ≤ 0. the capability of this innovation cost function is that the cost of innovation will increase with the increase in wages and profit. by putting cost of innovation function into the production function (eq.4) will result into 𝑃𝑖 (ℓ1, 𝑡). 𝑋𝑖 (ℓ1, 𝑡) ∙ [ 𝜂 𝜃 − 1⁄ + 1] − 𝑤 (ℓ1, 𝑡) ∙ 𝐿𝑖 (ℓ1, 𝑡) − 𝜋(ℓ1, 𝑡) ∙ 𝐾𝑖 (ℓ1, 𝑡) − 𝛺(𝜂(ℓ1, 𝑡)) …. (5) first order derivative with respect to "𝜂" of eq.5 will result into the share of capital and share of labour that affect the level of innovation. the advantage of eq. 6a and 6b is that these equations have their asymptote at 1. the derivation of one factor of production assumes another factor of production as given and constant. under fixed labour amount �̂�𝐾 = 1 − [( 𝛼(𝜃−1) 𝛾 ) ∙ 𝛺2 𝐴𝑖 −∙𝑃 𝑖 1 𝛾−1⁄ ∙ 𝜋 −1 𝛾⁄ �̅�𝑖 ] 1 2⁄ ….(6a) and fixed capital amount �̂�𝐿 = 1 − [( (𝜃−1)𝛾+1 𝛾𝛾+1∙(1−𝛼)2−𝛼 ) ∙ 𝛺2 𝐴 𝑖 −𝛾+1 ∙𝑃𝑖 2 ∙ 𝑤2−𝛼 �̅�𝑖 𝛼+1 ] 1 2⁄ ….(6b) 3. results and discussion this section is an illustrious elucidation of our reformulated model. however, this necessitates a discussion and depiction of the economy of pakistan. data for this study ranges from the period 1976 to 2014. the data is organized and calibrated as model to observe the outcomes that the economy has faced during this time period6. 5 the realization of innovation is still supposed to be the same, and correlation is 1. 6 main sources of the data are world bank and wto (world trade organization). muhammad imran, muhammad rafiq & imranullah 68 source: world bank annual census and wto for trade statistics. figure1 extricate the economic history of the country and summarize the relevant variables in percentage points. this is evident that value-added services and employment to goods sector are on the rise while other variables are facing decline7. at the outset from the year 1947, country’s economy embraced backward manufacturing and service sectors, but comparatively, after 1958 the economy progressed till early 1970’s, where growth was evident in almost all the sectors, including services and manufacturing, (see zaidi and saeed, 2013; mahmud, 2000; laporte, 1996). the designed model deliberated hitherto in section 2, is composed of two sectors8, therefore, each variable under discussion will be related to manufacturing or service sector. throughout the paper, the deep dark lines represent the service sector while the shaded one stands for manufacturing sector. to start our analyses, we need to find out the share of each factor of production to the valueadded sectors, as per eq.4. 7 the paper faces some shortcomings of unavailability of separate data on share of capital made to both sectors, therefore, we used gross capital and tried to assess the role played by capital in each sector at a time. 8 manufacturing refers to value-added net output of a sector after adding up all outputs and subtracting intermediate inputs. services include value-added in wholesale and retail trade (including hotels and restaurants), transport, and government, financial, professional, and personal services such as education, health care, and real estate services. services also included imputed bank service charges, import duties, and any statistical discrepancies noted by national compilers as well as discrepancies arising from rescaling. value-added is the net output of a sector after adding up all outputs and subtracting intermediate inputs. both manufacturing and services are calculated without making deductions for depreciation of fabricated assets and their origin is determined by the isic (international standard industrial classification). (source: world bank) 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% value added services value added manufacturing employment share to goods employment share to services capital share figure 1. summary of the economy of pakistan 1994-2014 1976-1994 journal of applied economics and business studies, volume. 3, issue 1 (2019) 61-78 https://doi.org/10.34260/jaebs.314 69 figure.2 depicts the share of labour to the aggregate value-added produced by each sector. evidently, share of labour to service sector has increased overtime in comparison to manufacturing sector. furthermore, this figure also includes share of labour to both the sectors, corrected for trade and otherwise. the first case as the share of labour to valueadded sectors when the country chooses not to trade is represented by the second-lap of curves (dashed curves) and third-lap or the middle curves represent the share of labour to value-added sectors as corrected for trade (the dotted curves)9. by observing the second lap dashed curves for both the sectors, it appears that the share of labour to service sector (the deep-coloured dashed curve) is higher than to manufacturing sector (the light-shaded dashed curve) but once the data is corrected for trade, the share of labour to manufacturing sector (the light-shaded dotted curve) exceeds the share of labour to service sector (the deep-coloured dotted curve), as is evident from 1977 to 1988. because of a larger trade volume in recent years, share of labour to manufacturing sector in present time is also greater than the service sector, represented by the dotted line. nonetheless, in aggregate terms the share of labour to service sector is more than the share of labour to the manufacturing sector, with and without trade. being an under developed economy, still service sector employees more labour on a comparative basis. the share of capital to the value-added production in each sector is portrayed in figure.3. service sector is utilizing higher amount of capital than the manufacturing sector10. at first, when country start trading, share of capital is equivalent for both the sectors, but with the 9 the second lap of curves supposes the gross trade and value added ratio as same as the gross output and value added ratio for the economy of pakistan, whereas the third lap of curves ignores intermediate inputs and trade in the concerned sector. (see desmet and hansberg (2014) page no.1224) 10 because of unavailability of the separate capital amount that is being utilized in both value-added sectors, we are left with gross capital amount that couldn’t be assessed for the factor “not corrected for trade”. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 figure 2. value added share of labor goods services services labor share(corrected for trade) goods labor share(corrected for trade) services labor share(not corrected for trade) goods labor share(not corrected for trade) muhammad imran, muhammad rafiq & imranullah 70 passage of times, service sector draw on more capital than the manufacturing sector, except for the year 2002. the commonalities between figure.2 and figure.3 are, the declining share of input to the goods sector and increased share of service sector. additionally, both the figures emulate that during the initial years shares of labour and capital were higher for the manufacturing sector11, while in the mid shares of both the factors were higher for the service sector, and at the end both sectors utilize almost similar shares of both factors12. nevertheless, the share of aggregate capital to innovation in each sector is explained through figure.5. 11 see also imran et al (2020) for detail study on factors productivity while having bilateral economic ties. 12 see (callen et al. 2016; shabbir et al., 2011; looney, 1997; seitz and licht, 1995) for further exposition 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 figure 3. value added share of capital goods services goods capital share(corrected for trade) services capital share(corrected for trade) 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 figure 4: productivit deflated price index (base year=2010) labor productivty in manufacturing sector labor productivity in service sector journal of applied economics and business studies, volume. 3, issue 1 (2019) 61-78 https://doi.org/10.34260/jaebs.314 71 as mentioned in section 2, the real cost of innovation is deflated by price index, therefore, figure.4 (for labour) and figure.5 (for both aggregate labour and aggregate capital), explicates the relevant factor-productivity deflated by consumer price index (cpi)13. figure.4 holds the share of labour to innovation in each sector, where shares are represented with different colours. although, value addition by manufacturing sector is comparatively lower (as per figure.2), but productivity of labour in the manufacturing sector is much higher than the service sector. during the times of economic instability, both curves tend to decline continuously, means the change in prices is much higher than change in labour-productivities, leading to a decline in innovation in both sectors, as innovation is represented by the slope of both the curves in figure.4. however, both the sectors demonstrate equal size of elasticity (still negatively sloped) to cpi, requiring some measures, for example, a decrease in the level of inflation or an increase in the factors productivity (investment in r&d etc.) to spur innovation in each sector. figure. 5 explains the aggregate productivity of each factor in both the sectors, where shaded-curve represent share of labour productivity in both the sectors and deep-black-curve represent share of capital productivity14. capital productivity – deflated by cpi – in both sectors is much higher than labour-productivity awaiting 2007. there are several reasons for small segment of labour to innovation; firstly, because of the low industrial base. besides, country lacks proper incentives and workable environment for new manufacturers, whereas country has abundant unemployed labour (see hassan, 2012 and iqbal, 1986). at the same times, although, capital-productivity is comparatively higher, but capital is further exposed to market fluctuations. eventually, this also establishes a fact that not only labour 13 factor-productivity denotes share of each factor of production to innovation which is, further, deflated by cpi to know the real factor-productivity at market price 14 data limitation concerning capital spent in both sectors separately, we are not able to assess capital productivity deflated to cpi. 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 figure 5. factor productivity deflated price index (base year=2010) labor productivity deflated price index capital productivity deflated price index muhammad imran, muhammad rafiq & imranullah 72 but capital has also an effect on innovation through the level of productivities. this is also evident as of eq.6a and eq.6b. subsequent to calculating the level of innovation backed by each factor of production to the concerned sectors, we have also explored the growth being contributed by each of the factors of production, since it is one of the assumption of our model that each location encompassed two sectors and the growth of these sectors represent growth and development of the particular location. henceforward, figure.6 presents share of labour and figure.7 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 figure 6: growth in value added per worker service sector manufacturing sector poly. (service sector) poly. (manufacturing sector) -0.3 -0.25 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2 figure 7. growth adjusted for gross capital service sector manufacturing sector poly. (service sector) poly. (manufacturing sector) journal of applied economics and business studies, volume. 3, issue 1 (2019) 61-78 https://doi.org/10.34260/jaebs.314 73 depicts share of capital to the value-added production. the share of labour to the growth of manufacturing sector plus service sector is evident in figure.6, where labour augments higher growth to service sector initially (explained in figure.4), amidst the declining trend of the curves. furthermore, figure.7 echoes the same pattern for capital, with a conspicuous occurrence of declining trend of capital towards the growth of service sector in year 2006. the polynomial curves (with degree 2) in both figures presents the summary of the direction of growth; growth stimulated by labour to both the sectors are greater (figure.6) than that of capital (figure.7). an alternative explanation can be that labour and capital both are comparatively utilized more in service sector than the manufacturing sector which reinforces our corollaries concerning figure.2 and figure.3. finally, concerning our assumption related to correlation of spillover-effects, (as per eq.6a and eq.6b) regarding share of each of the factor of production to innovation at any location, figure.8 presents the effect of distance on the spillover-effect and investment in innovation. impacts of spillover-effects and innovation are manifested by the three-dimensional diagram. data considered for this exercise is of 39 years. number of locations in a space are expanded to 500, where each location is supposed to be of the same size. therefore, through benchmark analyses the real cost of innovation is estimated as , 𝛺 = 0.0000385 , (see desmet and rossi-hansberg, 2012). additionally, the parameter of pareto distribution 𝑎 = 35 and to clearly understand the spillover-effects, distance-term is supposed to be 𝑑 = 0.004 𝑎𝑛𝑑 0.0015. distance term is measured as per 1000km, which separate the east from the west, (see ramondo and rodríguez-clare, 2013, desmet and rossi-hansberg, 2012). further, to decide shares of both factors of production to the value-addition in the two sectors, we have followed valentinyi and herrendorf (2008) and assumed 𝛼 = 0.6 𝑎𝑛𝑑 𝛾 = 0.4 to ensure that 𝛼 + 𝛾 = 1, while the elasticity of substitution is assumed to be stable at 𝜇 = −1.5, 𝑡ℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒 1 𝜇 − 1⁄ = 0.4 (see desmet and hansberg, 2014 and herrendorf, rogerson, and valentinyi, 2013) to ensure the incentive for producers to start production in one of the two sectors. for benchmark analysis through the specified and fixed variables mentioned in previous paragraph, we documented the innovation and spillover-effect in each location for the economy of pakistan throughout the specified time-period. intensity of yellow-colour represent higher share of concerned factor of production to innovation. intensity of yellow colour shows higher spillover-effect while accounting distance factor as d. to begin with the year 1976, higher distance costs have less effect on the share of capital to innovation (figure.8a) in comparison to the share of labour (figure.8b), given the distance effect of spillover-effect for the nearby locations. but with the passage of time both effects move in opposite directions and increase the share of labour to innovation visa vie the share of capital. whilst, lowering the distance effect (d = 0.0015) causes further decline in the share of capital (figure.8c) to spatial innovation than the share of labour (figure.8c). muhammad imran, muhammad rafiq & imranullah 74 figure 8. impact of capital and labour on spatial innovation in pakistan figure 8a (labour) and 8b (capital): under d = 0.004 journal of applied economics and business studies, volume. 3, issue 1 (2019) 61-78 https://doi.org/10.34260/jaebs.314 75 figure 8c (labour) and 8d (capital): d = 0.0015 muhammad imran, muhammad rafiq & imranullah 76 therefore, higher distance (d = 0.004) is comparatively more beneficial for labour to effect innovation as compared to capital. figure 8 (a, b, c, d) also represent the factorproductivity and their clustering and a measure of share of innovation that each factor of production adds to innovation. overall, the results concerning share of labour and share of capital to innovation proves that share of labour to innovation increases with time and spread out to a larger number of locations. while the share of capital declines and agglomerate in fewer locations with the increase in distance. the following three dimensional figures are worthwhile to identify the monopoly of locations where the particular factor of production is more helpful in innovation and formation of innovation clustering over time. 4. conclusion major objectives of our study is the assessment of the impact of factors of production on innovation in an underdeveloped economy like pakistan. our analysis is based upon the spatial model concerning the importance of the share of capital and share of labour to innovation. additionally, we have considered three inputs, that is, labour, capital and level of technology in a location, where migration of labour leads to the transfer of capital (as every producer owns a unit of capital). we infer that in any of the two sectors, producer needs to anticipate the number of labours as the foremost important criterion. besides, distance negatively affect spillover-effects. rendering to the designed model, capital-pulled innovations are positively related with profits (eq.6a) and labour-pulled innovations are positively correlated to wages (eq.6b). the application of the model is done for the economy of pakistan for the time period of 1976 to 2014. furthermore, it is established that labour and capital both are extensively utilized in service sector than manufacturing sector (as per figure.2 and figure.3). but correcting it for trade, the share of both the factors towards manufacturing has increased, while these shares decline for service sector. therefore, we found that trade is more beneficial for manufacturing sector for the purpose to increasing the shares of both factors of production (figure.2 and figure.3 after corrected for trade). moreover, share of capital to innovation in both sectors is comparatively higher than that of labour. this result is a contribution to the already existing work of desmet and hansberg (2014), where labour was only source of innovation and capital, as an endogenous variable, was neglected. therefore, capital is more important factor to innovate but more exposed to distance factor (figure.5), because distance leads to agglomeration of the capital resources to limited locations (figure.8a and figure.8c), while higher distance leads to higher rate of dispersion (figure.8b and figure.8d) of labour resources and increase the share of labour as spillover-effects to innovation. therefore, share of labour to spillover-effect is directly related to distance, as higher distance brings higher innovation, while share of capital is negatively related to increase in distance. journal of applied economics and business studies, volume. 3, issue 1 (2019) 61-78 https://doi.org/10.34260/jaebs.314 77 references boucekkine, r., camacho, c., & zou, b. (2009). bridging the gap between growth theory and the new economic geography: spatial ramsey model. ssrn electronic journal. doi:10.2139/ssrn.945077 brock, w., & xepapadeas, a. (2008). diffusion-induced instability and pattern formation in infinite horizon recursive optimal control. journal of economic dynamics and control, 32(9), 27452787. doi:10.1016/j.jedc.2007.08.005 brock, w., & xepapadeas, a. (2010). pattern formation, spatial externalities and regulation in coupled economic–ecological systems. journal of environmental economics and management, 59(2), 149-164. doi:10.1016/j.jeem.2009.07.003 callen, m., gulzar, s., hasanain, s. a., & khan, y. (2016). the political economy of public sector absence: experimental evidence from pakistan. doi:10.3386/w22340 caves, r. e. (2007). multinational enterprise and economic analysis. doi: 10.1017/cbo9780511619113 chow, g. c. (2010). interpreting china's economy. doi:10.1142/9789814317962. desmet, k., & rossi-hansberg, e. (2012). on the spatial economic impact of global warming. doi:10.3386/w18546. desmet, k., & rossi-hansberg, e. (2014). spatial development †. american economic review, 104(4), 1211-1243. doi:10.1257/aer.104.4.1211 de brito, m. p., & dekker, r. (2004). a framework for reverse logistics. reverse logistics, 3-27. doi:10.1007/978-3-540-24803-3_1 dunning, j. h., & lundan, s. m. (2008). institutions and the oli paradigm of the multinational enterprise. asia pac j manage, 25(4), 573-593. doi:10.1007/s10490-007-9074-z. ellison, g., glaeser, e. l., & kerr, w. r. (2010). what causes industry agglomeration? evidence from coagglomeration patterns. american economic review, 100(3), 1195-1213. doi:10.1257/aer.100.3.1195. herrendorf, b., rogerson, r., & valentinyi, á. (2013). two perspectives on preferences and structural transformation. american economic review, 103(7), 2752-2789. doi:10.1257/aer.103.7.2752 hassan ali. (2012). technology spillover impacts on total productivity of the manufacturing sector in pakistan. african journal of business management, 6(9). doi:10.5897/ajbm11.2352 henderson, v., kuncoro, a., & turner, m. (1995). industrial development in cities. journal of political economy, 103(5), 1067-1090. doi:10.1086/262013 hilber, c. a., & voicu, i. (2009). agglomeration economies and the location of foreign direct investment: empirical evidence from romania. regional studies, 44(3), 355-371. doi:10.1080/00343400902783230 imran, m., husen, a., kaleem, m., bangash, a. k., din, n. u., & sobia. (2020). effect of regional factor productivity on manufacturing sector: the case of sino-pak economic ties. financial innovation, 6(1). https://doi.org/10.1186/s40854-019-0163-z iqbal, m. (1986). substitution of labour, capital and energy in the manufacturing sector of pakistan. empirical economics, 11(2), 81-95. doi:10.1007/bf01987506 jenkins, & et al. (2000). embedded generation. doi:10.1049/pbpo031e muhammad imran, muhammad rafiq & imranullah 78 laporte,, r. (1996). pakistan in 1995: the continuing crises. asian survey, 36(2), 179-189. doi:10.2307/2645815 looney, r. e. (1997). infrastructure and private sector investment in pakistan. journal of asian economics, 8(3), 393-420. doi:10.1016/s1049-0078(97)90046-4 mahmud, s. f. (2000). the energy demand in the manufacturing sector of pakistan: some further results. energy economics, 22(6), 641-648. doi:10.1016/s0140-9883(99)00031-6 quah, d. (2002). spatial agglomeration dynamics. american economic review, 92(2), 247-252. doi:10.1257/000282802320189348. ramondo, n., & rodríguez-clare, a. (2013). trade, multinational production, and the gains from openness. journal of political economy, 121(2), 273-322. doi:10.1086/670136. rossi-hansberg, e. (2005). a spatial theory of trade. american economic review, 95(5), 14641491. doi:10.1257/000282805775014371. seitz, h., & licht, g. (1995). the impact of public infrastructure capital on regional manufacturing production cost. regional studies, 29(3), 231-240. doi:10.1080/00343409512331348923 shabbir, g., anwar, s., hussain, z., & imran, m. (2011). contribution of financial sector development in reducing unemployment in pakistan. international journal of economics and finance, 4(1). doi:10.5539/ijef.v4n1p260 triplett, j.e. and bosworth, b.p. (2004). productivity in the u.s. services sector, the brookings institutions: washington, d.c. valentinyi, á., & herrendorf, b. (2008). measuring factor income shares at the sectoral level. review of economic dynamics, 11(4), 820-835. doi:10.1016/j.red.2008.02.003 zaidi, m. r., & saeed, d. a. (2013). impact of electricity crises on sugarcane industry of pakistan. ssrn electronic journal. doi:10.2139/ssrn.2310508. journal of applied economics and business studies, volume. 2, issue 1 (2018) 75-90 https://doi.org/10.34260/jaebs.217 75 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online): 2663-693x estimating the role of money supply in determining aggregate investment in pakistan sohail alam khan1, muhammad tariq2*1, brekhna gul3 1 lecturer, department of economics, abdul wali khan university mardan 2assistant professor, department. of economics, abdul wali khan university mardan abstract this study investigates the role of money supply in determining the pakistan's aggregate investment for the period 1980-2015. hodric-prescott filter method has been applied for extracting the trend from the data. for estimating the results ordinary least squares method, granger causality test and vector autoregression has been used. the results revealed that money supply (m1 and m2) increases the aggregate investment in pakistan. other variables such as gdp growth rate and saving also showed a positive and significant association with aggregate investment. however, foreign direct investment remained insignificant. these findings recommended that the state bank of pakistan can used both m1 and m2 as an effective instrument of monetary policy for increasing the level of aggregate investment component of aggregate demand in the economy. keywords money supply, aggregate investment, granger causality test jel classification e51, e52, e12 copyright © 2018 saebr all rights reserved 1. introduction money supply is the total stock of money available in country at a specific time period. the central bank of a country usually adopts different tools i.e. open-market operations, reserve requirements and discount rate to control money supply in the country. the fluctuations in level of money stock in an economy have important implications for the performance of most of the macroeconomic indicators. on one side money supply expansion can put inflationary effects on an economy. while, on the other hand money supply increase may also affect the economic growth of a country positively. moreover, money supply as a determinant of aggregate demand plays a crucial role in the formulation and successful implantation of monetary policy for the stabilization of business cycles. during a period of recession, central banks use expansionary monetary policies. in contrast, 1 1*corresponding author: tariq_noor@awkum.edu.pk sohail alam, muhammad tariq, & brekhna gul 76 when the economy is in the expansion phase of business cycles, then the government can decrease the money supply for controlling inflation. the monetary policy works through different transmission channels in an economy. some of the channels are the interest rate channel, exchange rate channel, asset price channel, banking lending channel and balance sheet channel (lucky & kingsley (2017). through these channels the central bank’s monetary policy decisions affecting consumption and investment levels, employment and ultimately price level etc. (mishkin, 1996). it is a fact that all of these monetary transmission channels are having both positive and negative effects on all the major macroeconomic variables in an economy. however, the empirical results about all of these channels are still inconclusive. therefore, it is important to explore the effect of monetary policy on different components of aggregate demand for a more detail understanding about its role in the economy (khundrakpam, 2012; nasko, 2016; obadeyi et al, 2016). the different components of aggregate demand consist of aggregate consumption, aggregate investment, government expenditure and net exports. however, the fluctuations in all of these components of aggregate demand can result in a significant upward and downward shift in the output level in an economy. the aim of the present study is to investigate the impact money supply on aggregate investment in pakistan. 2. literature review gramm and nash (1971) investigated that change in the stock of money can affect income of agriculture sector and investment in united states. data were collected from 1919-1966 and the results showed the existence of a relationship between money stock and investment and income of agriculture sector. gertler and grinols (1982) also found that money supply affected investment. they also found that the investment depended on the capital holders demand for assets. growth in money supply also affect inflation level and thus demand for capital also changed. however, they reported that money supply growth decreased the investment. gaiotti and generale (2002) investigated the impact of monetary policy on behavior of investment and concluded that the role of money supply was an important determinant of the investment. majeed and khan (2008) investigated the determinants of private’s investment in pakistan utilizing a broader time period of 1970 to 2006. augmented dickey fuller test was used to identify the trend in data and then ordinary least square technique was used for regression. the results indicated that private production, net inflows and past stocks are the main determinants of private investment. ang (2010) investigated the key factors of investment in malaysia, using bank credit for investment as a one of the explanatory variables. using the time series data from 1960-2005 and applying the autoregressive distributed lag (ardl) and the error correction model (ecm), found that there was a journal of applied economics and business studies, volume. 2, issue 1 (2018) 75-90 https://doi.org/10.34260/jaebs.217 77 positive effect of monetary assets in economy on investment. furthermore, insecurity of macroeconomic variables affected private investment negatively. olweny and chiluwe (2012) investigated the effect of monetary policy on investment in kenya using the quarterly data from 1996-2009. they used the unit root and vector error correction model (vecm) to identify the short and long run relationship among the variables. the results indicated that public internal debt and treasury bill rate inversely affect private sector investment. further, increase of money supply and internal saving increased the investment. the results of study suggested that to boost up investment, deposit rates should be attractive and lending rates low. ayeni (2014) examined the effect of major macroeconomic variables on investment in nigeria using annual time series data from 1979-2012. for the analysis of the data, augmented dickey fuller test, johansen co-integration and ardl (for examining the long and short run association) were utilized. the results depicted that the exchange rate, gdp, financing private sector, rate of real interest and rate of inflation were important determinates of the investment in nigeria. eshun, adu and buabeng (2014) using time series annual data from 1970-2010 and applying the ardl bound test found that the high real rate of interest decreased investment in shorter run as well as in longer run and that unavailability of credit placed restrictions on the investment level. imoughele and ismaila (2014) examined the determinant of private saving in nigeria utilizing the time series data from 1981-2012. they found that the determinants of the private investment in nigeria are per capita income, inflation rate, financial deepening and term of trade. ali and shaheen (2016) examined the determinants of private investment in pakistan using the time series data from 1980-2011. they found a negative association between inflation and investment. while a positive relationship of gross domestic product, savings and credit with the investment was found. umoru and ohiomu (2017) explored the link between money supply and investment instability in nigeria. they used restricted var model for analysis. the result of the analysis indicated the existence of co-integration in variables and shock of money supply was the reason of investment instability. they suggested that in nigeria the policy makers need to focus on the money supply (m2) in implementing the monetary policy. additionally, it is necessary that government should focus on interest rate management and investment in order to boost real sector activities. ajayi and kolapo (2018) examined the impact of the gross domestic product, money supply, exchange rate, interest rate and inflation rate on investment of nigeria. the result indicated that in short run, gross domestic product and exchange rate are positively related to private investment, while money supply was negative related to the domestic private sohail alam, muhammad tariq, & brekhna gul 78 investment. in long run, money supply negatively affects the investment and gross domestic product positively affect the investment. 3. research methods this section presents the details about the data and the methods used to study the relationship between investment and money supply. first, in section 3.1, data, data sources and variables definitions are presented. after that the theoretical framework of the study has been given in section 3.2. finally, section 3.3 outlines the empirical model. 3.1 data and definitions of the variables of the study the main objective of the study is to investigate the impact of money supply on the aggregate investment of pakistan. for this purpose, time series annual data over the period of 1980 to 2015 have been used for the analysis. all the variables along with its definitions and symbols are given in table.1 as follows. table 1: definitions of variables of the study variables definition symbols sources money supply money supply i.e. m1c and m2c 𝑀𝑆𝐶 economic surveys of pakistan various issues narrow money supply paper notes and coins in circulation in million rupees in pakistan m1c broad money supply narrow money + checkable deposits + money market instruments in million rupees in pakistan m2c gross domestic product growth rate annual percentage changes in the gross domestic product of pakistan gdprc aggregate investment pakistan annual total public and private investment in million rupees ic saving rate pakistan annual saving rate in percentage sc foreign direct investment total foreign direct investment in an year in million rupees fdic the superscript “c” on the symbols of variables indicates the deviated form. for converting the data into deviation form and extracting only the cyclical component, hodrick prescott (hp) filter method is used. the purpose of the application of the hp filter method to data is to remove trend from data and making it stationary. 3.2 theoretical framework the relationship between money supply and aggregate investment can be understood in the framework of the keynesian is-lm approach. according to keynesian approach, the money supply influences the national income through the aggregate demand channel. more journal of applied economics and business studies, volume. 2, issue 1 (2018) 75-90 https://doi.org/10.34260/jaebs.217 79 specifically, when there is a rise or fall in the money supply it will influence the various components of aggregate demand which will in turn affect the national income. whereas, the components of the aggregate demand are aggregate consumption, aggregate investment, government expenditures and net exports. the relationship between money supply and aggregate investment can be understood by focusing on figure.1 as below. figure 1: effects of money supply on aggregate demand and national income figure 1 shows the relationship between the money supply and national income. in the figure interest rate (r) has been given on the vertical axis. whereas, national income (y) has been represented on horizontal axis. the graph shows simultaneous equilibrium in both the goods and money market at point “a” where the lm (m0) is intersected with the is0. at this point the level of interest rate is (r0) and income (y0). now suppose the money supply is increased by the central bank. this increase in the money supply will bring decrease in the interest rate. this fall in the interest rate from r0 to r1 will increase the aggregate demand in the economy through its different components i.e. aggregate consumption, business investment, government spending and net exports. with the increase in the aggregate demand, the lm curve will shift towards the right side from lm (m0) to lm (m1) i.e. from point “a” to point “b”. this will in turn shift the income level from y0 to y1. in light of figure.1, another framework has been developed which is showing the relationship between the money supply and aggregate investment. this relationship between the money supply and aggregate investment given in figure.2 provides the basis for the development of the empirical model of the study given in equation (1). lm(m1) lm(m0) is0 y r r1 0 a y0 y1 b r0 income in te re st r a te sohail alam, muhammad tariq, & brekhna gul 80 figure 2: relationship between money supply and aggregate investment 3.3 empirical model for examining the relationship between money supply and aggregate investment, equation (1) has been formulated. 𝐼𝐶 = 𝛼0 + 𝛼1 𝑀𝑠 𝐶 + 𝛼2𝐺𝐷𝑃𝑟 𝐶 + 𝛼3𝑆 𝐶 + 𝛼4𝐹𝑑𝑖 𝐶 + 𝑒𝑖 (1) equation (1) shows the aggregate investment function in which the aggregate investment 𝐼𝐶 is dependent variable while 𝑀𝑠𝐶 , 𝐺𝐷𝑃𝑟𝐶 , 𝑆𝐶 , 𝐹𝑑𝑖𝐶 are independent variables. in addition, 𝛼0, 𝛼1 , 𝛼2, 𝛼3, 𝛼4 are the coefficients and 𝑒𝑖 is the error term. for the computation of the results of the study different econometric techniques have been used. then granger causality test has been used for checking the interrelationship between the study variables. then var test has been applied for checking the relationship between the variables through a system equation. 4. results and discussion the present section shows the empirical results of the study. first, section 4.1 shows the regression results. then, section 4.2 shows the granger causality test results. finally, the vector autoregression results have been presented in section 4.3. aggregate demand money supply aggregate consumptio n aggregate investment governme nt expenditur e net exports journal of applied economics and business studies, volume. 2, issue 1 (2018) 75-90 https://doi.org/10.34260/jaebs.217 81 4.1 regression results for the investment model with m1 and m2 to check the relationship between the money supply and aggregate consumption, ordinary least squares method has been applied. the regression results are given in table 2 as follows. table 2: regression results of investment model with m1 and m2 𝐼𝐶 = 𝛼0 + 𝛼1 𝑀𝑠 𝐶 + 𝛼2𝐺𝐷𝑃𝑟 𝐶 + 𝛼3𝑆 𝐶 + 𝛼4𝐹𝑑𝑖 𝐶 + 𝑒𝑖 𝑀1𝐶 𝑀2𝐶 independent variables coefficient st. error independent variables coefficient st. error 𝑀1𝐶 0.168226** 0.261576 𝑀2𝐶 0.010450** 0.029145 𝐺𝐷𝑃𝑟𝐶 20360.02** 9159.402 𝐺𝐷𝑃𝑟𝐶 21128.15** 10846.15 𝑆𝐶 16157.39 7837.185 𝑆𝐶 16273.18** 7968.465 𝐹𝑑𝑖𝐶 1.160148 0.267463 𝐹𝑑𝑖𝐶 0.961514 0.344151 intercept 1.19e-08 15130.31 intercept 1.33e-08 15614.12 rsquare : 0.52 adj. rsquare : 0.48 dw statistics : 1.71 rsquare : 0.51 adj. rsquare : 0.49 dw statistics : 1.73 • asterisk *, ** and *** shows 1%, 5% and 10% level of significance. table 2 shows the regression results of the study with m1 and m2 separately in the model. the results show that m1 and gdp growth rate are statistically significant and showing a positive impact on aggregate investment. whereas, saving and foreign direct investment remained insignificant. in the second model on the right side, m1 has been replaced with m2. the purpose of putting m2 instead of m1 in the model is to know whether it can make any difference. the results show that m2, gdp growth rate and saving were significant and showing a positive impact on aggregate investment. 4.2 results of granger causality test for the investment model. the granger causality test results have also been computed for the investment model which is placed in table 3 as follows. table 3: granger causality test results for investment model with m1 null hypothesis observation fstatistic probability 𝑀1𝐶 does not affect𝐼𝐶 𝐼𝐶 does not affect𝑀1𝐶 33 0.07890 0.35265 0.0243 0.0059 𝐺𝐷𝑃𝑟𝐶 does not affect 𝐼𝐶 𝐼𝐶 does not affect 𝐺𝐷𝑃𝑟𝐶 33 8.79289 6.68508 0.0011 0.0042 𝑆𝐶 does not affect 𝐼𝐶 𝐼𝐶 does not affect 𝑆𝐶 33 0.91793 2.28466 0.0110 0.1205 𝐹𝑑𝑖𝐶 does not affect 𝐼𝐶 𝐼𝐶 does not affect𝐹𝑑𝑖𝐶 33 10.8625 3.33284 0.5003 0.6503 sohail alam, muhammad tariq, & brekhna gul 82 𝐺𝐷𝑃𝑟𝐶 does not affect 𝑀1𝐶 𝑀1𝐶 does not affect 𝐺𝐷𝑃𝑟𝐶 33 0.31758 0.60938 0.7305 0.0507 𝑆𝐶 does not affect 𝑀1𝐶 𝑀1𝐶 does not affect𝑆𝐶 33 0.85775 1.85947 0.4350 0.1745 𝐹𝑑𝑖𝐶 does not affect 𝑀1𝐶 𝑀1𝐶 does not affect 𝐹𝑑𝑖𝐶 33 4.05798 4.21268 0.0283 0.7252 𝑆𝐶 does not affect 𝐺𝐷𝑃𝑟𝐶 𝐺𝐷𝑃𝑟𝐶 does not affect𝑆𝐶 33 6.21502 2.33090 0.0058 0.1158 𝐹𝑑𝑖𝐶 does not affect 𝐺𝐷𝑃𝑟𝐶 𝐺𝐷𝑃𝑟𝐶 does not affect 𝐹𝑑𝑖𝐶 33 4.24238 5.76688 0.0246 0.0080 𝐹𝑑𝑖𝐶 does not affect 𝑆𝐶 𝑆𝐶 does not affect 𝐹𝑑𝑖𝐶 33 3.05324 0.87792 0.6632 0.4268 first, granger causality test results have been computed for the investment model with m1. it has been noted that money supply and investment affect each other. similarly, gdp growth rate and investment also affects each other. moreover, two-way relationships have been also found between foreign direct investment and gdp growth rate. furthermore, oneway relationship has been found between the other variables except fdi and saving rate which does not show any relationship with each other. after that m1 has been replaced with m2, and granger causality test has been applied to data again. the results computed are given in table 4, as below. table 4: granger causality test results for investment model with m2 null hypothesis observation fstatistic probability 𝑀2𝐶 does not affect𝐼𝐶 𝐼𝐶 does not affect𝑀2𝐶 33 3.07070 3.71399 0.0623 0.0371 𝐺𝐷𝑃𝑟𝐶 does not affect 𝐼𝐶 𝐼𝐶 does not affect 𝐺𝐷𝑃𝑟𝐶 33 8.79289 6.68508 0.0011 0.0042 𝑆𝐶 does not affect 𝐼𝐶 𝐼𝐶 does not affect 𝑆𝐶 33 0.91793 2.28466 0.4110 0.1205 𝐹𝑑𝑖𝐶 does not affect 𝐼𝐶 𝐼𝐶 does not affect 𝐹𝑑𝑖𝐶 33 10.8625 3.33284 0.0003 0.0503 𝐺𝐷𝑃𝐺𝑟𝐶 does not affect 𝑀2𝐶 𝑀2𝐶 does not affect 𝐺𝐷𝑃𝐺𝑟𝐶 33 4.18739 1.02536 0.0256 0.3717 𝑆𝐶 does not affect 𝑀2𝐶 𝑀2𝐶 does not affect𝑆𝐶 33 1.85767 0.52824 0.1748 0.5954 𝐹𝑑𝑖𝐶 does not affect 𝑀2𝐶 𝑀2𝐶 does not affect 𝐹𝑑𝑖𝐶 33 11.4724 10.6518 0.0002 0.0004 𝑆𝐶 does not affect 𝐺𝐷𝑃𝑟𝐶 𝐺𝐷𝑃𝑟𝐶 does not affect𝑆𝐶 33 6.21502 2.33090 0.0058 0.1158 𝐹𝑑𝑖𝐶 does not affect 𝐺𝐷𝑃𝑟𝐶 𝐺𝐷𝑃𝑟𝐶 does not affect 𝐹𝑑𝑖𝐶 33 4.24238 5.76688 0.0246 0.0080 𝐹𝑑𝑖𝐶 does not affect 𝑆𝐶 𝑆𝐶 does not affect 𝐹𝑑𝑖𝐶 33 3.05324 0.87792 0.0632 0.4268 journal of applied economics and business studies, volume. 2, issue 1 (2018) 75-90 https://doi.org/10.34260/jaebs.217 83 first, granger causality test results have been computed for the investment model with m2 too. the results show that two way relationship has been found between money supply and investment, gdp growth rate and investment, fdi and investment, fdi and money supply (m2) and fdi and gdp growth rate. however, one-way relationship has been noted between other variables except saving rate and money supply. 4.3 var results for the investment model investment model results of var with m1 and m2 have been shown in this section. first the individual significance of the variables for each model of the whole system equation has been checked. after that the joint significance of all the variables of different models has been computed through wald test. finally, the cholesky decomposition test is used for the computation of impulse response functions. 4.3.1 individual significance of the variables of var model with m1. for showing the individual significance of the variables, var test has been applied. the results are given in table 5. in the table the individual significance of the explanatory variables has been checked for the five models of the system i.e. investment, money supply (m1), gdp growth rate, saving rate and foreign direct investment. first, the investment model results show that two lags for investment, two lags for money supply (m1), second lag for gdp growth, first lag of saving rate turned significant, whereas, foreign direct investment turned insignificant. similarly, in the money supply model only money supply and gdp growth rate turned significant. moreover, in the gdp growth rate mode, investment, money supply, lag gdp growth rate, saving rate and foreign direct investment became significant. similarly, in the saving rate model, only investment turned significant. however, in the foreign direct investment model all variables turned insignificant. table 5: results of var for investment model with m1 𝑰𝑪 𝑴𝟏𝑪 𝑮𝑫𝑷𝒓𝑪 𝑺𝑪 𝑭𝒅𝒊𝑪 𝐼𝐶 (-1) 0.414714** (0.20934) [ 1.98102] 0.264508 (0.30228) [ 0.87503] 6.31e-06** (6.6e-06) [0.95107] 1.20e-05 (1.1e-05) [1.13843] 0.304287 (0.13453) [ 2.26185] 𝐼𝐶 (-2) 0.285718** (0.18617) [ 1.53474] -0.361579 (0.26882) [-1.34507] 3.48e-06** (5.9e-06) [0.58985] 2.05e-06** (9.4e-06) [0.21886] -0.171599 (0.11964) [-1.43434] 𝑀1𝐶 (-1) 0.296313** (0.15948) [ 1.85798] 0.155850** (0.23028) [0.67677] 1.05e-06 (5.1e-06) [0.20738] 2.71e-07 (8.0e-06) [0.03371] 0.183899 (0.10249) [ 1.79437] 𝑀1𝐶 (-2) 0.120552** (0.17040) [ 0.70746] 0.025937 (0.24605) [0.10541] 8.21e-06** (5.4e-06) [1.52084] 1.16e-05 (8.6e-06) [ 1.35234] -0.231122 (0.10950) [-2.11063] sohail alam, muhammad tariq, & brekhna gul 84 𝐺𝐷𝑃𝑟𝐶 (-1) 19488.11 (6680.26) [ 2.91727] 1396.679** (9646.02) [0.14479] 0.468930 (0.21169) [2.21519] 0.503346 (0.33666) [1.49511] 4482.225 (4292.92) [ 1.04410] 𝐺𝐷𝑃𝑟𝐶 (-2) 14977.32** (7235.23) [ 2.07006] -5812.181 (10447.4) [-0.55633] 0.210520** (0.22927) [0.91820] 0.237622 (0.36463) [ 0.65168] 6191.032 (4649.56) [ 1.33153] 𝑆𝐶 (-1) 1624.993** (4015.74) [0.40466] -5634.751 (5798.56) [-0.97175] 0.084712** (0.12725) [ 0.66569] 0.297915 (0.20238) [ 1.47207] 3528.701 (2580.63) [ 1.36738] 𝑆 (-2) 202.1481 (4126.05) [ 0.04899] 142.8277 (5957.85) [ 0.02397] 0.146605 (0.13075) [ 1.12127] -0.107262 (0.20794) [-0.51584] 412.6367 (2651.52) [ 0.15562] 𝐹𝑑𝑖𝐶 (-1) 1.248696 (0.27408) [ 4.55591] -0.978436 (0.39576) [-2.47227] 1.94e-05** (8.7e-06) [ 2.23483] -1.34e-05 (1.4e-05) [-0.97073] 0.714889 (0.17613) [ 4.05879] 𝐹𝑑𝑖𝐶 (-2) -0.343613 (0.38112) [-0.90160] 0.607276 (0.55032) [ 1.10350] -8.82e-06 (1.2e-05) [-0.73016] 1.64e-05 (1.9e-05) [ 0.85237] -0.903702 (0.24492) [-3.68984] c 2257.226 (7550.39) [ 0.29895] -898.8537 (10902.4) [-0.08245] -0.017303 (0.23926) [-0.07232] 0.084542 (0.38051) [ 0.22218] 869.4860 (4852.09) [ 0.17920] r-squared 0.690508 0.593550 0.563488 0.489189 0.366058 adj. r-squared 0.676194 0.580618 0.545074 0.477002 0.359720 sum sq. residuals 4.08e+10 8.50e+10 40.93773 103.5419 1.68e+10 s.e. equation 43047.45 62158.69 1.364114 2.169435 27663.50 f-statistic 16.21135 1.101072 2.839958 2.106878 7.204032 4.3.2 joint significance of the variables of var model with m1 the joint significance of the variables has been checked by using the wald test. the overall significance of all the four models of the system i.e. investment model, money supply model, gdp growth rate model, saving rate model and foreign direct investment model has been checked. table 6 shows that for all the models, no restriction has been put on the explanatory variables with the hypothesis that all these explanatory variables do not influence the dependent variables. however, all of these hypotheses have been rejected and it is confirmed that all the explanatory variables in all the models do influence the explained variables. from the investment model it is clear that money supply (m1) does play role in the determination of the aggregate investment. journal of applied economics and business studies, volume. 2, issue 1 (2018) 75-90 https://doi.org/10.34260/jaebs.217 85 table 6: wald test results for joint significance var model with m1 null hypotheses of all models: the joint effect of all the independent variables is equal to zero investment model test stat. value df prob. chisquare 162.1173 11 0.0000 money supply(m2) model test stat. value df prob. chisquare 11.01552 11 0.0420 gdp growth rate model test stat. value df prob. chisquare 0.0028 11 0.0028 saving model test stat. value df prob. chisquare 21.09896 11 0.0324 foreign direct investment model test stat, value df prob. chisquare 72.04038 11 0.0000 4.3.3 impulse response function results of var model with m1 cholesky decomposition is used for examining the response of the dependent variables to independent variables in all the seven models. the results are given in figure 3 as follows. figure 3: impulse response function results of var model for m1 the results show the shocks received from the explanatory variables by dependent variables in the entire model. 80 ,0 00 40 ,0 00 0 40 ,0 00 80 ,0 00 12 0,00 0 2 4 6 8 10 response of ic to ic 80 ,0 00 40 ,0 00 0 40 ,0 00 80 ,0 00 12 0,00 0 2 4 6 8 10 response of ic to m1c 80 ,0 00 40 ,0 00 0 40 ,0 00 80 ,0 00 12 0,00 0 2 4 6 8 10 response of ic to gdpgrc 80 ,0 00 40 ,0 00 0 40 ,0 00 80 ,0 00 12 0,00 0 2 4 6 8 10 response of ic to sc 80 ,0 00 40 ,0 00 0 40 ,0 00 80 ,0 00 12 0,00 0 2 4 6 8 10 response of ic to fdic 80 ,0 00 40 ,0 00 0 40 ,0 00 80 ,0 00 2 4 6 8 10 response of m1c to ic 80 ,0 00 40 ,0 00 0 40 ,0 00 80 ,0 00 2 4 6 8 10 response of m1c to m1c 80 ,0 00 40 ,0 00 0 40 ,0 00 80 ,0 00 2 4 6 8 10 response of m1c to gdpgrc 80 ,0 00 40 ,0 00 0 40 ,0 00 80 ,0 00 2 4 6 8 10 response of m1c to sc 80 ,0 00 40 ,0 00 0 40 ,0 00 80 ,0 00 2 4 6 8 10 response of m1c to fdic 2 1 0 1 2 2 4 6 8 10 response of gdpgrc to ic 2 1 0 1 2 2 4 6 8 10 response of gdpgrc to m1c 2 1 0 1 2 2 4 6 8 10 response of gdpgrc to gdpgrc 2 1 0 1 2 2 4 6 8 10 response of gdpgrc to sc 2 1 0 1 2 2 4 6 8 10 response of gdpgrc to fdic 2 1 0 1 2 3 2 4 6 8 10 response of sc to ic 2 1 0 1 2 3 2 4 6 8 10 response of sc to m1c 2 1 0 1 2 3 2 4 6 8 10 response of sc to gdpgrc 2 1 0 1 2 3 2 4 6 8 10 response of sc to sc 2 1 0 1 2 3 2 4 6 8 10 response of sc to fdic 40 ,0 00 20 ,0 00 0 20 ,0 00 40 ,0 00 2 4 6 8 10 response of fdic to ic 40 ,0 00 20 ,0 00 0 20 ,0 00 40 ,0 00 2 4 6 8 10 response of fdic to m1c 40 ,0 00 20 ,0 00 0 20 ,0 00 40 ,0 00 2 4 6 8 10 response of fdic to gdpgrc 40 ,0 00 20 ,0 00 0 20 ,0 00 40 ,0 00 2 4 6 8 10 response of fdic to sc 40 ,0 00 20 ,0 00 0 20 ,0 00 40 ,0 00 2 4 6 8 10 response of fdic to fdic response to cholesky one s.d. innovations ± 2 s.e. sohail alam, muhammad tariq, & brekhna gul 86 4.4. var results for the investment model with m2 the present sections show the var results with m2, first, var results has been computed for investigating the individual significance of the variables through a system analysis. then wald test has been applied for showing the over significance of the variables. finally, cholesky decomposition test is used for the computation of impulse response functions. 4.4.2 individual significance of the variables of var model with m2 for showing the individual significance of the variable’s var test has been applied. the results are given in table 7. in the table the individual significance of the explanatory variables has been checked for the five models of the system i.e. investment, money supply (m2), gdp growth rate, saving rate and foreign direct investment. first the investment model results show that lag investment, lag money supply, lag gdp growth rate, and lag saving rate became significant in the investment model. similarly, in the money supply model only money supply and foreign direct investment turned significant. moreover, in the gdp growth rate model investment, money supply, lag gdp growth rate and saving rate tureen significant. similarly, in the saving rate model only investment, money supply and saving rate became significant. finally, in the foreign direct investment model lag values of investment, gdp growth rate turned significant, whereas, all other variables remained insignificant. table 7: results of var model with m2 𝐼𝐶 𝑀2𝐶 𝐺𝐷𝑃𝑟𝐶 𝑆𝐶 𝐹𝑑𝑖𝐶 𝐼𝐶 (-1) 0.477987** (0.20133) [ 2.37415] 5.914224 (1.64977) [3.58488] 1.31e-05** (6.7e-06) [1.96621] 4.56e-06** (1.1e-05) [0.42143] 0.026602** (0.11329) [ 0.23481] 𝐼𝐶 (-2) 0.472664 (0.17158) [ 2.75470] 0.719550 (1.40603) [0.51176] 3.47e-06 (5.7e-06) [0.61061] 2.24e-06 (9.2e-06) [0.24298] -0.043393 (0.09655) [-0.44943] 𝑀2𝐶 (-1) 0.046599** (0.02280) [ 2.04413] 0.528302** (0.18680) [2.82811] 9.61e-07** (7.6e-07) [1.27085] 1.13e-06 (1.2e-06) [ 0.92042] -0.017484 (0.01283) [-1.36303] 𝑀2𝐶 (-2) 0.021799** (0.01622) [ 1.34374] 0.024041 (0.13294) [ 0.18085] 4.23e-07** (5.4e-07) [ 0.78606] 1.26e-07** (8.7e-07) [ 0.14505] 0.040432 (0.00913) [ 4.42920] 𝐺𝐷𝑃𝑟𝐶 (-1) 21613.38** (6125.84) [ 3.52823] 29156.34 (50197.4) [ 0.58083] 0.406120** (0.20316) [1.99905] -0.596378 (0.32920) [-1.81159] 8898.768** (3446.98) [ 2.58161] 𝐺𝐷𝑃𝑟𝐶 (-2) 14945.21 (6750.81) [ 2.21384 56390.25 (55318.5) [ 1.01937] 0.196829 (0.22388) [0.87916] 0.111858 (0.36279) [ 0.30833] 6360.126** (3798.64) [ 1.67432] 𝑆𝐶 (-1) 1156.336 (3991.91) [0.28967] 20449.35 (32711.2) [ 0.62515] 0.034245** (0.13239) [ 0.25867] 0.323421** (0.21452) [ 1.50762] 1086.915 (2246.23) [ 0.48388] journal of applied economics and business studies, volume. 2, issue 1 (2018) 75-90 https://doi.org/10.34260/jaebs.217 87 𝑆𝐶 (-2) 1159.180** (3926.46) [0.29522] -27164.13 (32174.8) [-0.84427] 0.179044** (0.13022) [ 1.37497] -0.122269 (0.21101) [-0.57945 1043.100 (2209.40) [ 0.47212] 𝐹𝑑𝑖𝐶 (-1) 1.037172 (0.25692) [ 4.03697 13.76025** (2.10528) [ 6.53605] 2.24e-05 (8.5e-06) [ 2.62332] -1.21e-05 (1.4e-05) [-0.87890] 0.751103 (0.14457) [ 5.19555] 𝐹𝑑𝑖𝐶 (-2) -1.344520 (0.43468) [-3.09312] 9.841367 (3.56194) [ 2.76292] 1.13e-05 (1.4e-05) [ 0.78294] -1.47e-05 (2.3e-05) [-0.62924] -0.737086 (0.24459) [-3.01352] c 941.5429 (7188.24) [ 0.13098] 1351.896 (58903.1) [ 0.02295] 0.011165 (0.23839) [ 0.04684] 0.049431 (0.38629) [ 0.12796] 949.9758 (4044.79) [ 0.23486] r-squared 0.500399 0.412214 0.431473 0.367243 0.275483 adj. r-squared 0.490581 0.405947 0.412143 0.355081 0.260703 sum sq. residuals 3.74e+10 2.51e+12 41.12674 107.9904 1.18e+10 s.e. equation 41227.34 337831.7 1.367259 2.215548 23198.41 f-statistic 17.87288 10.91192 2.816796 1.929463 11.17248 4.4.3 joint significance of the variables of var model with m2 the joint significance of the variables has been checked by using the wald test. the overall significance of all the four models of the system i.e. investment model, money supply model, gdp growth rate model, saving rate model and foreign direct investment model has been checked. the results computed are given in table 8. for all the models no restriction has been put on the explanatory variables with the hypothesis that these all these explanatory variables do not influence the dependent variables. however, all of these hypotheses have been rejected and it is confirmed that all the explanatory variables in all the models do influence the explained variables. from the investment model it is clear that money supply (m2) does play role in the determination of the aggregate investment. table 8: wald test results for joint significance var model with m2 null hypotheses of all models: the overall effect of the independent variables is equal to zero investment model test stat. value df probability chisquare 178.732 11 0.0000 money supply(m2) model test stat. value df probability chisquare 109.120 11 0.0000 gdp growth rate model test stat. value df probability chisquare 28.171 11 0.0030 saving model test stat. value df probability chisquare 19.323 11 0.0455 foreign direct investment model test stat. value df probability chisquare 111.724 11 0.000 sohail alam, muhammad tariq, & brekhna gul 88 4.4.4 impulse response function results of var model with m2 cholesky decomposition is used for examining the response of the dependent variables to independent variables in all the seven models. the results are given in figure 4 as follows. figure 4: impulse response function results of var model for m2 the results show that aggregate investment in this model is affected by shocks from all the explanatory variables. similarly, money supply, gdp growth rate and inflation have also been affected by shocks from the explanatory variables. 5. conclusion the study estimated the role of money supply (i.e. m1 and m2) in the determination of the aggregate investment component of aggregate demand for pakistan during the period 1980 2015. first, hodric prescott filter method was applied to the data for separating the cyclical components from the trend components of the data for making the data stationary. after that ordinary least squares method was applied for the regression results. granger causality test has also been applied for checking the causal relationship between the variables. vector auto regression for the robustness of the results also applied. the results showed that monetary aggregates i.e. m1 and m2 positively affect the aggregate investment in pakistan. furthermore, the granger causality test showed a one directional relationship 8 0 ,0 0 0 4 0 ,0 0 0 0 4 0 ,0 0 0 8 0 ,0 0 0 2 4 6 8 1 0 response of ic to ic 8 0 ,0 0 0 4 0 ,0 0 0 0 4 0 ,0 0 0 8 0 ,0 0 0 2 4 6 8 1 0 response of ic to m2c 8 0 ,0 0 0 4 0 ,0 0 0 0 4 0 ,0 0 0 8 0 ,0 0 0 2 4 6 8 1 0 response of ic to gdpgrc 8 0 ,0 0 0 4 0 ,0 0 0 0 4 0 ,0 0 0 8 0 ,0 0 0 2 4 6 8 1 0 response of ic to sc 8 0 ,0 0 0 4 0 ,0 0 0 0 4 0 ,0 0 0 8 0 ,0 0 0 2 4 6 8 1 0 response of ic to fdic 4 0 0 ,0 0 0 0 4 0 0 ,0 0 0 2 4 6 8 1 0 response of m2c to ic 4 0 0 ,0 0 0 0 4 0 0 ,0 0 0 2 4 6 8 1 0 response of m2c to m2c 4 0 0 ,0 0 0 0 4 0 0 ,0 0 0 2 4 6 8 1 0 response of m2c to gdpgrc 4 0 0 ,0 0 0 0 4 0 0 ,0 0 0 2 4 6 8 1 0 response of m2c to sc 4 0 0 ,0 0 0 0 4 0 0 ,0 0 0 2 4 6 8 1 0 response of m2c to fdic 2 1 0 1 2 2 4 6 8 1 0 response of gdpgrc to ic 2 1 0 1 2 2 4 6 8 1 0 response of gdpgrc to m2c 2 1 0 1 2 2 4 6 8 1 0 response of gdpgrc to gdpgrc 2 1 0 1 2 2 4 6 8 1 0 response of gdpgrc to sc 2 1 0 1 2 2 4 6 8 1 0 response of gdpgrc to fdic 2 1 0 1 2 3 2 4 6 8 1 0 response of sc to ic 2 1 0 1 2 3 2 4 6 8 1 0 response of sc to m2c 2 1 0 1 2 3 2 4 6 8 1 0 response of sc to gdpgrc 2 1 0 1 2 3 2 4 6 8 1 0 response of sc to sc 2 1 0 1 2 3 2 4 6 8 1 0 response of sc to fdic 4 0 ,0 0 0 2 0 ,0 0 0 0 2 0 ,0 0 0 4 0 ,0 0 0 2 4 6 8 1 0 response of fdic to ic 4 0 ,0 0 0 2 0 ,0 0 0 0 2 0 ,0 0 0 4 0 ,0 0 0 2 4 6 8 1 0 response of fdic to m2c 4 0 ,0 0 0 2 0 ,0 0 0 0 2 0 ,0 0 0 4 0 ,0 0 0 2 4 6 8 1 0 response of fdic to gdpgrc 4 0 ,0 0 0 2 0 ,0 0 0 0 2 0 ,0 0 0 4 0 ,0 0 0 2 4 6 8 1 0 response of fdic to sc 4 0 ,0 0 0 2 0 ,0 0 0 0 2 0 ,0 0 0 4 0 ,0 0 0 2 4 6 8 1 0 response of fdic to fdic response to cholesky one s.d. innovations ± 2 s.e. journal of applied economics and business studies, volume. 2, issue 1 (2018) 75-90 https://doi.org/10.34260/jaebs.217 89 from m1 and m2 towards aggregate investment. similarly, the vector auto regression also supported these results and revealed a positive relationship between money supply and aggregate investment. furthermore, gross domestic product growth rate and saving also turned positively significant. however, foreign direct investment remained insignificant. it is concluded on the basis of that monetary policy aggregates both m1 and m2 play an important role in the determination of the aggregate investment in pakistan. references ajayi, l. b., & kolapo, f. t., (2018). is domestic private investment sensitive to macroeconomic indicators? further evidence from nigeria. world journal of economics and finance. 4(2),100-105. ali, m. m., &shaheen, s., (2016). an analysis of determinants of private investment in pakistan. international interdisciplinary journal of scholarly research (iijsr). 2(2). ang, j. b. (2010). determinants of private investment in malaysia: what causes the post crisis slumps?. contemporary economic policy, 28(3), 378-391. ayeni, r. k. (2014). macroeconomic determinants of private sector investment: an ardl approach-evidence from nigeria. global advance research journal of management and business studies, 3(2), 82-89. eshun, m. e., adu, g., &buabeng, e. (2014). the financial determinants of private investment in ghana. munich personal repec archive. paper no. 57570. http://mpra.ub.uni-muenchen.de/57570. gaiotti, e., & generale, a. (2002). does monetary policy have asymmetric effects? a look at the investment decisions of italian firms. giornaledeglieconomisti e annali di economia, 29-59. gertler, m., & grinols, e. (1982). monetary randomness and investment. journal of monetary economics, 10(2), 239–258. gramm, w. p., & nash, r. t. (1971). the impact of changes in the stock of money of agricultural income and investment: comment. journal of money, credit and banking, 3(3), 712. hodrick, r. j. and prescott, e. c. (1981). postwar u. s. business cycles: an empirical investigation. nber working paper no. 451. imoughele, l. e., & ismaila, m. (2014). an econometric analysis of the determinants of private domestic savings in nigeria (1981-2012). international journal of humanities and social science, 4(5), 12-21. khundrakpam, j. k. (2012). estimating impacts of monetary policy on aggregate demand in india. reserve bank of india mpra paper no. 50902 sohail alam, muhammad tariq, & brekhna gul 90 lucky & kingsley, (2017). monetary policy transmission mechanisms and domestic real investment in nigeria: a time series study 1981-2015. international journal of economics and financial management,2(2). majeed, m. t., & khan, s. (2008). the determinants of private investment and the relationship between public and private investment in pakistan. munich personal repec archive, 49301.http://mpra.ub.uni-muenchen.de/49301/ mishkin, f. s. (1996). the channels of monetary transmission: lessons for monetary policy. nber working paper no. 5464. nasko, a. m. (2016). impact of monetary policy on the economy of nigeria. pyrex journal of business and finance management research,2(10),163-179. obadeyi, j. a., okhiria, a. o and afolabi, v. k. (2016). evaluating the impact of monetary policy on the growth of emerging economy: nigerian experience. american journal of economics,6(5), 241-249. olweny, t., &chiluwe, m. (2012). the effect of monetary policy on private sector investment in kenya. journal of applied finance and banking, 2(2), 239. umoru, d., &ohiomu, s. (2017). econometric analysis of investment volatility and the money in circulation in nigeria. russian federation sochi journal of economy.11(4), 341-351. journal of applied economics and business studies, volume. 5, issue 1 (2021) 67-82 https://doi.org/10.34260/jaebs.514 67 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x workplace deviance in public sector organizations: evidence from pakistan syed arif hussain shah1, tazeem ali shah2, abrar ullah3 and muhammad yasir4* 1 department of management, international islamic university, islamabad, pakistan 2 department of management, international islamic university, islamabad, pakistan 3 department of management sciences, university of swabi, swabi, pakistan 4 department of management sciences and commerce, bacha khan university, charsadda, pakistan abstract deviant workplace behaviors have become an important area of research due to the recent revelation of high-profile corporate scandals. scholars view that deviant workplace behaviors can be controlled when the factors that affect workplace deviance are properly understood. therefore, this study is having two sections. section one identifies the level of workplace deviance prevailing in the public sector hospitals of pakistan, for which data was collected from 219 respondents in the understudy sector. findings of this study show that workplace deviance exists in the understudy sector at a moderate level. section two describes the factors that have the ability to influence the emergence of deviant workplace behaviors. thus, this study searched for workplace deviance related articles available at the academic research databases such as scopus and web of science. the keywords that were used for searching articles were “workplace deviance”, “organizational deviance”, “deviance” and “deviant behaviors”. this study outlines twenty-five factors that affect workplace deviance, thereby highlighting how workplace deviance can be minimized. lastly, implications and suggestions for further research and practice are highlighted. keywords workplace deviance public sector pakistan jel classification i1, i12, i19 1. introduction after the revelation of several high-profile corporate scandals like enron, worldcom, and lehman brothers, workplace deviance has become an important issue to address. according to the report of ethics resource center, fifty-two percent of employees have observed deviant workplace behavior (erc, 2012) and previous literature shows that deviant workplace behavior lead towards diminishing stakeholders’ return, augmented psychological distress, diminishing psychological * dryasir@bkuc.edu.pk syed arif hussain shah, tazeem ali shah, abrar ullah and muhammad yasir 68 well-being, higher turnover rates, and lower level of employees’ job satisfaction (penney & spector, 2005; appelbaum, iaconi, & matousek, 2007; lu & lin, 2014). moreover, organizations lose around 5% of their revenues every year because of deviant workplace behaviors (acfe, 2014). thus, workplace deviance may cause significant organizational costs. the previous literature also reports several forms of workplace deviance that exists in pakistan’s public sector organizations, such as, stealing official belongings, taking longer lunch breaks, employees’ leaving office early, and frequent late arrival (bashir, nasir, qayyum, & bashir, 2012; yasir & khan, 2020; shaheen, abrar, saleem, shabbir, & zulfiqar, 2021). in addition, corruption has almost become a norm in this sector. matters like getting a gas connection, getting a contract to build a road, or reporting an issue to the police may not be entertained without bribing the officials and if one fails to do so, the case may remain pending for an unidentified period of time (bashir, khattak, hanif, & chohan, 2011). this shows that workplace deviance exists in this sector. therefore, scholars and practitioners argue that it is important to examine the factors which cause the emergence of deviant workplace behavior and how it can be controlled (nasir & bashir, 2012; ahmed, kiyani, & hashmi, 2013; shahzad & malik, 2014). because of its universal value, ethical issues are addressed in both non-western and western countries (yasir & mohamad, 2016; khan, yasir, yusof, bhatti, & umar, 2017). but, in the developing countries, the demand for organizational ethics is not as high as compared to the developed countries (lyon & maher, 2005; blackburn, bose, & haque, 2006; yasir, rasli, & qureshi, 2017). in addition, ahmed, shad, mumtaz, and tanveer (2012) identified that the organizational ethical values of the west can be applied in a non-western society, therefore, top-management of the pakistani organizations needs to assess their work environment based on ethical values already developed in the west as unethical behaviors like falsifying documents, theft, harassment, abuse, and several other deviant behaviors are common practices at the workplace all over the world leading towards negatively effecting organizational reputation and performance. therefore, scholars and practitioners are compelled to find solutions for workplace deviance in order to limit its occurrence (peng, tseng, & lee, 2011; ferris, spence, brown, & heller, 2012; neves & story, 2015; yasir & rasli, 2018; moon, morais, de moura, & uskul, 2020; shaheen et al., 2021). thus, it is necessary to study the factors that contribute towards workplace deviance. therefore, objectives of the current research are to identify the level of workplace deviance that exists in the understudy sector and to further identify the factors that affect (negatively and positively) workplace deviance, thereby showing those factors that cause the emergence of workplace deviance and also highlighting those factors through which it can be reduced and controlled. journal of applied economics and business studies, volume. 5, issue 1 (2021) 67-82 https://doi.org/10.34260/jaebs.514 69 2. literature review increasing attention in the harmful effects of deviant workplace behaviors is due to the massive costs that organizations incur when employees are involved in bad behaviors, for instance purposefully neglecting supervisor instructions, theft, or falsification of documents (biron, 2010; guay, choi, oh, mitchell, mount, & shin, 2016). the notion of deviance originates from previous studies on group norms (sherif & sherif, 1953). but, from the last two decades, the concept of deviant behaviors has attracted much more attention, than before (bennett & robinson, 2000; biron, 2010; thau & mitchell, 2010; de clercq, bouckenooghe, raja, & matsyborska, 2014; lian, ferris, morrison, & brown, 2014; zagenczyk, restubog, kiewitz, kiazad, & tang, 2014; guay et al., 2016; yasir, jamil, adil, hamayun, & irum, 2017; yasir & khan, 2020). scholars define and term employees negative workplace behaviors in several different ways, such as workplace deviance or deviant workplace behavior (bennett & robinson, 2003), antisocial behavior (giacalone & greenberg, 1997), workplace violence (neuman & baron, 1998), counterproductive behavior (spector & fox, 2005), bad behavior (griffin & lopez, 2005), workplace aggression (neuman & baron, 1998, 2005), workplace incivility (pearson, andersson, & porath, 2005), mobbing/bullying (zapf, 1999; einarsen, hoel, zapf, & cooper, 2011), organizational misbehavior (vardi & weitz, 2003), dysfunctional behavior (griffin, o'leary-kelly, & collins, 1998), corruption (lange, 2008), and unethical behavior (kish-gephart, harrison, & treviño, 2010). however, the main theme of these notions is the same, as they are harmful to the employees and the organization. 2.1 workplace deviance workplace deviance has been categorized into four distinct quadrants; (a) production deviance, (b) political deviance, (c) property deviance, and (d) personal aggression (robinson & bennett, 1995). syed arif hussain shah, tazeem ali shah, abrar ullah and muhammad yasir 70 figure 1: dimensions of workplace deviance source: robinson and bennett (1995) figure 1 shows that organizational deviance includes those actions that are against the organization, for instance, sabotaging equipment’s and/or wasting resources, however, interpersonal deviance includes such behaviors that cause harm to employees, for instance, verbal harassment and/or assault (robinson & bennett, 1995). initially, it was hollenger and clark who grouped employees’ deviant behaviors into two broader behavioral kinds (hollinger & clark, 1982a; hollinger & clark, 1982b, 1983; hollinger, 1986). the first was labeled as property deviance, this includes misuse of employer assets, for instance, theft, and property damage. the second category was labeled as production deviance, this involves behaviors that detract from production during working hours (e.g., sloppy work and drug use) and not being on the work as scheduled. sackett (2002) study suggested a hierarchical model, having general deviance factor at the top, with group factors namely organizational and interpersonal deviance below this general factor, and specific behavior domains, for instance, drug usage, absence, and theft, below these group factors. lastly, berry, ones, and sackett (2007) identified that the most widely used conceptualization of workplace deviance is the model proposed by bennett and robinson (2000). journal of applied economics and business studies, volume. 5, issue 1 (2021) 67-82 https://doi.org/10.34260/jaebs.514 71 2.2 public healthcare sector in pakistan, the public healthcare sector is primarily overlooked due to inadequate budgetary allocation, corruption, political instability, and the absence of the will of the political leadership in order to improve the current condition of this sector (khan, nawaz, & khan, 2015). the pakistan economic survey (pes) 2018-19 shows that there are 1,279 public sector hospitals, whereas, basic health units are 5,527, dispensaries are 5,671, and rural health centers are 686 in the country. in addition, there are 220,829 doctors, 22,595 dentists, and 108,474 nurses, hence, providing a health facilities ratio of 963 individuals per doctor, 9413 individuals per dentist, and 1,608 individuals per hospital bed (pes, 2018). the pes (2019) shows that the government had allocated rs. 421.8 billion for the expenditure on the public healthcare sector (see figure 2), thereby making it 1.1% of the gdp. despite the government efforts and funds allocation, the desired health outcomes have not been achieved in the country due to several socio-economic factors like growing population, unhygienic environmental conditions, uneven distribution of health benefits, and poverty. figure 2: expenditure on public healthcare sector source: pes (2019) moreover, pakistan medical and dental council (pmdc) and pakistan nursing council (pnc) are the registration and statutory regulatory authority for medical and nursing practitioners in the country. the higher education commission (hec) of pakistan also plays an important role, for instance, issuing licenses to educational institutions and verification of degrees in the country. furthermore, the national bioethics committee (nbc) was established in 2004 in pakistan for the purpose of promoting ethical practices in the healthcare sector. despite having such regulatory bodies in pakistan, the healthcare sector of the country is still hampered by the nonexistence of hospital ethics committees and the absence of courses related to ethics in the curriculum (jafree, zakar, fischer, & zakar, 2015). moreover, pnc and pmdc have made ethics education compulsory but several nursing and medical institutions in the country do not teach compulsory courses in ethics (shaikh & humayun, 2012). syed arif hussain shah, tazeem ali shah, abrar ullah and muhammad yasir 72 thus, a large number (57%) of doctors are reported to have no knowledge of the code of ethics of pmdc (imran, haider, jawaid, & mazhar, 2015). furthermore, corruption and poor governance are prevailing in the management of equipment and drugs (naz, khan, daraz, hussain, & khan, 2012), and ethical compliance during clinical practices is not monitored in the public healthcare sector of pakistan (jafree et al., 2015). moreover, doctors and nurses are reported to be absent for several days in the public sector hospitals of pakistan (saeed & ibrahim, 2005). as, callen, gulzar, hasanain, and khan (2016) also found that 68.5% of doctors were absent during normal working hours in the understudy sector. in addition, ghost workers, violent work environment, conflicts, protests, poor governance, corruption, understaffed hospitals, and political interference are prevalent problems that exist in the public sector hospitals of pakistan (saeed & ibrahim, 2005; naz et al., 2012; yousafzai, 2015; hamid, kanwal, bajwa, khalid, & muba, 2016; hussain, yusoff, banoori, khan, & khan, 2016; jafree, 2017). thus, the level of workplace deviance is high in the public sector hospitals, however, only a few private sector hospitals follow a zero-tolerance policy towards workplace deviance (shahzad & malik, 2014). hence, the overall public sector hospitals of pakistan present an unimpressive picture, however, some developments are taking place during the last few years but still, the situation is not satisfactory. 3. methodology 3.1 section one sampling in this study, data was collected through a convenient sampling technique from five public sector hospitals in pakistan through questionnaires. 500 respondents (doctors and nurses) were given the questionnaire, and usable completed questionnaires were 219. measurement instruments the workplace deviance construct is assessed by using bennett and robinson (2000) measurement scale for deviant workplace behavior having nineteen items, with twelve items for organizational deviance and seven items for interpersonal deviance (bennett & robinson, 2000). but, in this study two of the items were removed from the scale based on the validity issue, that is (a) “dragged out work in order to get overtime” and (b) “falsified a receipt to get reimbursed for more money than you spent on hospital expenses”. thus, the final questionnaire had seventeen items. moreover, the workplace deviance scale was anchored at “1= never”, “2= rarely”, “3= sometimes”, “4= often”, “5= always”. journal of applied economics and business studies, volume. 5, issue 1 (2021) 67-82 https://doi.org/10.34260/jaebs.514 73 section two this study highlights the factors that have been previously investigated in relationship with workplace deviance and these factors are having the ability to influence workplace deviance. this study searched for workplace deviance related articles published between 2005 and 2019 at the academic research databases namely web of science and scopus. the keywords that were used for searching articles were “workplace deviance”, “organizational deviance”, “deviance” or “deviant behaviors”. 4. analysis section one initially, reliability analysis was conducted for the items of workplace deviance and the value of cronbach alpha was .864, which is in the acceptable range (sekaran, 2006). furthermore, to answer the first research question. the means of the respondents’ responses were calculated to determine the extent to which workplace deviance is prevailing in the understudy sector. table 1 depicts the results of the descriptive analysis which addresses research question one. table 1: descriptive statistics items mean std. deviation skewness kurtosis interpersonal d 01 3.2009 .88608 -.406 -1.243 interpersonal d 02 3.0913 .86251 -.091 -1.061 interpersonal d 03 3.3288 .76124 -.382 -.776 interpersonal d 04 3.3607 .82535 -.708 -.589 interpersonal d 05 3.2557 .69626 -.312 -.777 interpersonal d 06 3.2694 .81021 -.530 -.753 interpersonal d 07 3.1781 .72332 -.210 -.091 organizational d 08 3.2100 .83044 -.362 -.886 organizational d 09 3.3973 .86321 -.823 -.669 organizational d 10 3.3470 .93756 -.577 -.754 organizational d 11 3.0228 .95995 -.171 -1.379 organizational d 12 3.3425 .83313 -.522 -.802 organizational d 13 3.3014 .91398 -.380 -1.120 organizational d 14 3.3744 .90182 -.850 -.445 organizational d 15 3.2420 .67107 -.419 -.385 organizational d 16 2.4338 .63459 .414 -.063 organizational d 17 3.2237 .69743 -.336 -.916 variable mean std. deviation level interpersonal deviance 3.2407 .53112 moderate organizational deviance 3.1895 .47393 moderate workplace deviance 3.2106 .45907 moderate note: n=219 syed arif hussain shah, tazeem ali shah, abrar ullah and muhammad yasir 74 to determine the level of workplace deviance prevailing in the understudy organizations, mean scores based on five points likert scale were categorized into five levels as very low level, low level, moderate level, high level, and very high level. this method has been used in the previous literature (alanazi & abbod, 2014; nazari, pihie, akmaliah, idris, & basri, 2014; alshurman & alkhateeb, 2015) in identifying levels of a different phenomenon which was based on the following equation. category length = highest scale value − lowest scale value number of categories = (5-1)/5 = 0.8 very low = 1 + 0.8 = 1.8 low = 1.8 + 0.8 = 2.6 moderate = 2.6 + 0.8 = 3.4 high = 3.4 + 0.8 = 4.2 very higher = more than 4.2 hence, the mean value of 1 to 1.8 indicating a very low level, 1.8 to 2.6 indicating a low level, 2.6 to 3.4 indicating moderate level, 3.4 to 4.2 indicating the high level, and more than 4.2 indicating a very high level of existence of workplace deviance. moreover, results from table 1 indicate that the level of workplace deviance and its components exists at a moderate level in the understudy sector. section two table 2 outlines the twenty-five factors that were extracted from the previous literature and are having the ability to influence workplace deviance. table 2: factors affecting workplace deviance s/n antecedent relationship dependent variable source 1. abusive supervision + workplace deviance mitchell and ambrose (2007) 2. authentic leadership erkutlu and chafra (2013) 3. basic need satisfaction lian, ferris, and brown (2012) 4. co-workers’ solidarity itzkovich and heilbrunn (2016) 5. depression + zhu, lyu, and ye (2019) 6. emotional exhaustion + mulki, jaramillo, and locander (2006) 7. employees’ trust in leader mo and shi (2015) 8. ethical climate yasir and rasli (2018) journal of applied economics and business studies, volume. 5, issue 1 (2021) 67-82 https://doi.org/10.34260/jaebs.514 75 9. ethical ideology henle, giacalone, and jurkiewicz (2005) 10. ethical leadership mo and shi (2015) 11. goal congruence de clercq et al. (2014) 12. injustice + ferris et al. (2012) 13. job satisfaction darrat, amyx, and bennett (2010) 14. organizationbased self esteem ferris, brown, and heller (2009) 15. organizational commitment tepper, henle, lambert, giacalone, and duffy (2008) 16. organizational support liu and ding (2012) 17. perceived organizational ethical values biron (2010) 18. personality guay et al. (2016) 19. psychological capital norman, avey, nimnicht, and pigeon (2010) 20. psychological contract breach + chiu and peng (2008) 21. self-control bordia, restubog, and tang (2008) 22. socialized charismatic leadership brown and treviño (2006) 23. work engagement de clercq et al. (2014) 24. work-family conflict + darrat et al. (2010) 25. workplace sexual harassment + zhu et al. (2019) 5. discussion the findings of this study are consistent with the prior literature which highlights that workplace deviance exists in the public sector organizations of pakistan, for instance, theft, purposely ignoring supervisor’s instruction, harassment, gossiping about the supervisor, corruption, blaming co-workers, and intentionally arriving late at work (saeed & ibrahim, 2005; bashir et al., 2012; nasir & bashir, 2012; shahzad & malik, 2014; jafree et al., 2015; shaheen et al., 2021). specifically, deviant behaviors reported in the understudy sector are; verbal and physical violence (shahzad & malik, 2014; jafree, 2017), bullying and mobbing behavior (gadit & mugford, 2008; bano & malik, 2013; somani, karmaliani, farlane, asad, & hirani, 2015), corruption and bribery (naz syed arif hussain shah, tazeem ali shah, abrar ullah and muhammad yasir 76 et al., 2012; haroon, 2014; yousafzai, 2015), protests (abbasi, 2014), non-reportage of errors and student nurses used as adjunct staff (jafree et al., 2015), sexual harassment (shaikh, 2000; somani, karmaliani, mc farlane, asad, & hirani, 2015; jafree, 2017), and absenteeism (saeed & ibrahim, 2005; naz et al., 2012). thus, many forms of deviant workplace behaviors (interpersonal and organizational) exist in the understudy sector which needs to be minimized and controlled. this study further identified several factors that have the ability to affect the occurrence of deviant workplace behaviors. the current study adds value to the existing body of knowledge and provides a literature review regarding the construct of workplace deviance. however, this study has some limitations such as the selection bias; as this study provided a single reference of a study for each factor that affects workplace deviance identified in table 2. though several other studies might be available on each factor, this study aimed to provide a single recent study available on the given factor. moreover, it is possible that the authors may have overlooked relevant studies while reviewing the prior literature. thus, there is a need for further research in order to highlight the factors that influence workplace deviance and how it can be controlled. this study further implies that organizations must conduct ethics training and development programs for employees in general and supervisors in specific. the focus of these training programs should include themes like exhibiting concern for others, disciplining bad behavior, and communicating the significance of ethical behavior, thus emphasizing employees expected behavior. moreover, top management of the organizations should start using social media campaigns as an awareness tool, specifically to create facebook pages/groups and whatsapp groups for their respective organizations in which ethical standards are consistently communicated, thereby emphasizing the importance and utilization of ethical conduct. thus, establishing an ethical work climate in which individuals are more likely to refrain from deviant behavior. hence, this study shows that organizations must consider promoting ethical workplace behavior and address the factors that contribute to the occurrence of workplace deviance, thus minimizing the occurrence of workplace deviance. 6. conclusion the findings of this study show that workplace deviance exists in the public sector hospitals of pakistan at a moderate level. thus, the level of workplace deviance in this sector needs to be controlled and limit. this study further identified several factors (see table 2) through which the occurrence of workplace deviance can be minimized for instance ethical leadership, authentic leadership, and ethical climate. lastly, this study also identified several factors that cause the emergence of workplace deviance, for instance, abusive supervision and injustice. journal of applied economics and business studies, volume. 5, issue 1 (2021) 67-82 https://doi.org/10.34260/jaebs.514 77 references abbasi, i. n. (2014). protest of doctors: a basic human right or an ethical dilemma. bmc medical ethics., 15(1), 24. ahmed, b., shad, i., mumtaz, r., and tanveer, z. (2012). organizational ethics and job satisfaction: evidence from pakistan. african journal of business management, 6(8), 2966. ahmed, w., kiyani, a. a., and hashmi, s. h. (2013). the study on organizational cynicism, organizational injustice & breach of psychological contract as the determinants of deviant work behavior. актуальні проблеми економіки.(2), 145-154. alanazi, a., and abbod, m. (2014). e-learning repository system for sharing learning resources among saudi universities. in g. vincenti, a. bucciero and c. vaz de carvalho (eds.), e-learning, e-education, and online training: first international conference, eleot 2014, bethesda, md, usa, september 18-20, 2014, revised selected papers (pp. 10-18): springer international publishing. alshurman, w. m., and alkhateeb, a. a. (2015). the extent to which teachers of talented and creative student posses knowledge and skills according to cec-nagc criteria in jordan. european scientific journal, 11(28). appelbaum, s. h., iaconi, g. d., and matousek, a. (2007). positive and negative deviant workplace behaviors: causes, impacts, and solutions. corporate governance: the international journal of business in society., 7(5), 586-598. association of certified fraud examiners. (2014). 2014 report to the nation on occupational fraud and abuse. retrieved from http://www.acfe.com/rttn/docs/2014-report-tonations.pdf bano, s., and malik, s. (2013). impact of workplace bullying on organizational outcome. pakistan journal of commerce and social science, 7(3), 618-627. bashir, s., khattak, h. r., hanif, a., and chohan, s. n. (2011). whistle-blowing in public sector organizations: evidence from pakistan. the american review of public administration, 41(3), 285-296. bashir, s., nasir, m., qayyum, s., and bashir, a. (2012). dimensionality of counterproductive work behaviors in public sector organizations of pakistan. public organization review, 12(4), 357-366. bennett, r. j., and robinson, s. l. (2000). development of a measure of workplace deviance. journal of applied psychology., 85(3), 349. bennett, r. j., and robinson, s. l. (2003). the past, present, and future of workplace deviance research organizational behavior: the state of the science, 2nd ed (pp. 247-281). mahwah, nj, us: lawrence erlbaum associates publishers. berry, c. m., ones, d. s., and sackett, p. r. (2007). interpersonal deviance, organizational deviance, and their common correlates: a review and meta-analysis. journal of applied psychology., 92(2), 410. biron, m. (2010). negative reciprocity and the association between perceived organizational ethical values and organizational deviance. human relations, 63(6), 875-897. blackburn, k., bose, n., and haque, m. e. (2006). the incidence and persistence of corruption in economic development. journal of economic dynamics and control, 30(12), 24472467. http://www.acfe.com/rttn/docs/2014-report-to-nations.pdf http://www.acfe.com/rttn/docs/2014-report-to-nations.pdf syed arif hussain shah, tazeem ali shah, abrar ullah and muhammad yasir 78 bordia, p., restubog, s. l. d., and tang, r. l. (2008). when employees strike back: investigating mediating mechanisms between psychological contract breach and workplace deviance. journal of applied psychology., 93(5), 1104. brown, m. e., and treviño, l. k. (2006). socialized charismatic leadership, values congruence, and deviance in work groups. journal of applied psychology, 91(4), 954. callen, m., gulzar, s., hasanain, s. a., and khan, y. (2016). the political economy of public sector absence: experimental evidence from pakistan. national bureau of economic research. cambridge. chiu, s.-f., and peng, j.-c. (2008). the relationship between psychological contract breach and employee deviance: the moderating role of hostile attributional style. journal of vocational behavior, 73(3), 426-433. darrat, m., amyx, d., and bennett, r. (2010). an investigation into the effects of work–family conflict and job satisfaction on salesperson deviance. journal of personal selling & sales management, 30(3), 239-251. de clercq, d., bouckenooghe, d., raja, u., and matsyborska, g. (2014). unpacking the goal congruence–organizational deviance relationship: the roles of work engagement and emotional intelligence. journal of business ethics, 124(4), 695-711. einarsen, s., hoel, h., zapf, d., and cooper, c. l. (2011). the concept of bullying and harassment at work: the european tradition. bullying and harassment in the workplace: developments in theory, research, and practice, 2, 3-40. erkutlu, h., and chafra, j. (2013). effects of trust and psychological contract violation on authentic leadership and organizational deviance. management research review, 36(9), 828-848. ethics resource center. (2012). national business ethics survey of fortune 500 employees, washington, dc. ferris, d. l., brown, d. j., and heller, d. (2009). organizational supports and organizational deviance: the mediating role of organization-based self-esteem. organizational behavior and human decision processes, 108(2), 279-286. ferris, d. l., spence, j. r., brown, d. j., and heller, d. (2012). interpersonal injustice and workplace deviance: the role of esteem threat. journal of management, 38(6), 17881811. doi: 10.1177/0149206310372259 gadit, a., and mugford, g. (2008). a pilot study of bullying and harassment among medical professionals in pakistan, focussing on psychiatry: need for a medical ombudsman. journal of medical ethics, 34(6), 463-466. giacalone, r. a., and greenberg, j. e. (1997). antisocial behavior in organizations: sage publications, inc. griffin, r. w., and lopez, y. p. (2005). “bad behavior” in organizations: a review and typology for future research. journal of management, 31(6), 988-1005. griffin, r. w., o'leary-kelly, a., and collins, j. (1998). dysfunctional work behaviors in organizations trends in organizational behavior (pp. 65-82). new york, ny, us: john wiley & sons ltd. guay, r. p., choi, d., oh, i. s., mitchell, m. s., mount, m., and shin, k.-h. (2016). why people harm the organization and its members: relationships among personality, organizational commitment, and workplace deviance. human performance, forthcoming, 29(1), 1-15. journal of applied economics and business studies, volume. 5, issue 1 (2021) 67-82 https://doi.org/10.34260/jaebs.514 79 hamid, s., kanwal, r., bajwa, m. h., khalid, s., and muba, h. (2016). ethical issues faced by nurses during nursing practice in district layyah, pakistan. diversity & equality in health and care, 13(4), 302-308. haroon, m. z. (2014). corruption in healthcare system: an obstacle in achieving optimal outcome. journal of ayub medical college abbottabad, 26(2). henle, c. a., giacalone, r. a., and jurkiewicz, c. l. (2005). the role of ethical ideology in workplace deviance. journal of business ethics, 56(3), 219-230. hollinger, r., and clark, j. (1982a). employee deviance a response to the perceived quality of the work experience. work and occupations, 9(1), 97-114. hollinger, r. c. (1986). acts against the workplace: social bonding and employee deviance. deviant behavior, 7(1), 53-75. hollinger, r. c., and clark, j. p. (1982b). formal and informal social controls of employee deviance*. the sociological quarterly, 23(3), 333-343. hollinger, r. c., and clark, j. p. (1983). theft by employees: lexington books lexington, ma. hussain, a., yusoff, r. m., banoori, s. a., khan, a., and khan, m. a. (2016). enhancing effectiveness of employees through training and development in the health care department of khyber pakhtunkhwa pakistan: a literature review. international review of management and marketing, 6(4), 731-737. imran, n., haider, i. i., jawaid, m., and mazhar, n. (2015). health ethics education: knowledge, attitudes and practice of healthcare ethics among interns and residents in pakistan. journal of postgraduate medical institute (peshawar-pakistan), 28(4). itzkovich, y., and heilbrunn, s. (2016). the role of co-workers’ solidarity as an antecedent of incivility and deviant behavior in organizations. deviant behavior, 1-16. jafree, s. r. (2017). workplace violence against women nurses working in two public sector hospitals of lahore, pakistan. nursing outlook. doi: 10.1016/j.outlook.2017.01.00 jafree, s. r., zakar, r., fischer, f., and zakar, m. z. (2015). ethical violations in the clinical setting: the hidden curriculum learning experience of pakistani nurses. bmc medical ethics., 16(1), 1. khan, h., yasir, m., yusof, h. m., bhatti, m. n., and umar, a. (2017). the relationship between ethical leadership and organizational citizenship behavior: evidence from pakistan. city university research journal, 45-62. khan, z. a., nawaz, d., and khan, i. (2015). the impact of leadership styles on innovation in health services. gomal university journal of research, 31(1). kish-gephart, j. j., harrison, d. a., and treviño, l. k. (2010). bad apples, bad cases, and bad barrels: meta-analytic evidence about sources of unethical decisions at work. journal of applied psychology., 95(1), 1. lange, d. (2008). a multidimensional conceptualization of organizational corruption control. academy of management review, 33(3), 710-729. lian, h., ferris, d. l., and brown, d. j. (2012). does taking the good with the bad make things worse? how abusive supervision and leader–member exchange interact to impact need satisfaction and organizational deviance. organizational behavior and human decision processes, 117(1), 41-52. lian, h., ferris, d. l., morrison, r., and brown, d. j. (2014). blame it on the supervisor or the subordinate? reciprocal relations between abusive supervision and organizational deviance. journal of applied psychology., 99(4), 651. syed arif hussain shah, tazeem ali shah, abrar ullah and muhammad yasir 80 liu, n.-t., and ding, c. g. (2012). general ethical judgments, perceived organizational support, interactional justice, and workplace deviance. the international journal of human resource management, 23(13), 2712-2735. lu, c.-s., and lin, c.-c. (2014). the effects of ethical leadership and ethical climate on employee ethical behavior in the international port context. journal of business ethics, 124(2), 209-223. lyon, j. d., and maher, m. w. (2005). the importance of business risk in setting audit fees: evidence from cases of client misconduct. journal of accounting research, 43(1), 133151. mitchell, m. s., and ambrose, m. l. (2007). abusive supervision and workplace deviance and the moderating effects of negative reciprocity beliefs. journal of applied psychology., 92(4), 1159. mo, s., and shi, j. (2015). linking ethical leadership to employee burnout, workplace deviance and performance: testing the mediating roles of trust in leader and surface acting. journal of business ethics, 1-11. doi: doi:10.1007/s10551-015-2821-z moon, c., morais, c., de moura, g. r., and uskul, a. k. (2020). the role of organizational structure and deviant status in employees’ reactions to and acceptance of workplace deviance. international journal of conflict management. doi: 10.1108/ijcma-032020-0036 mulki, j. p., jaramillo, f., and locander, w. b. (2006). emotional exhaustion and organizational deviance: can the right job and a leader's style make a difference? journal of business research, 59(12), 1222-1230. nasir, m., and bashir, a. (2012). examining workplace deviance in public sector organizations of pakistan. international journal of social economics, 39(4), 240-253. doi: doi:10.1108/03068291211205677 naz, a., khan, w., daraz, u., hussain, m., and khan, t. (2012). an analytical study of patients’ health problems in public hospitals of khyber pakhtunkhwa pakistan. international journal of business and social science, 3(5). nazari, k., pihie, l., akmaliah, z., idris, k., and basri, r. (2014). exploring lecturers’ perception on learning organization dimensions and demographic variables in technical and vocational colleges. pertanika journal of social sciences & humanities, 22(1), 119-136. neuman, j. h., and baron, r. a. (1998). workplace violence and workplace aggression: evidence concerning specific forms, potential causes, and preferred targets. journal of management, 24(3), 391-419. neuman, j. h., and baron, r. a. (2005). aggression in the workplace: a social-psychological perspective counterproductive work behavior: investigations of actors and targets (pp. 13-40). washington, dc: american psychological association. neves, p., and story, j. (2015). ethical leadership and reputation: combined indirect effects on organizational deviance. journal of business ethics, 127(1), 165-176. norman, s. m., avey, j. b., nimnicht, j. l., and pigeon, n. g. (2010). the interactive effects of psychological capital and organizational identity on employee citizenship and deviance behaviors. journal of leadership & organizational studies, 17(4), 380-391. pakistan economic survey. (2018). health and nutrition. retrieved from http://www.finance.gov.pk/survey/chapters_19/economic_survey_2018_19.pdf http://www.finance.gov.pk/survey/chapters_19/economic_survey_2018_19.pdf journal of applied economics and business studies, volume. 5, issue 1 (2021) 67-82 https://doi.org/10.34260/jaebs.514 81 pakistan economic survey. (2019). health and nutrition. retrieved from http://www.finance.gov.pk/survey/chapter_20/11_health_and_nutrition.pdf pearson, c. m., andersson, l. m., and porath, c. l. (2005). workplace incivility counterproductive work behavior: investigations of actors and targets (pp. 177-200). washington, dc, us: american psychological association. peng, j.-c., tseng, m.-m., and lee, y.-l. (2011). relationships among supervisor feedback environment, work-related stressors, and employee deviance. journal of nursing research, 19(1), 13-24. penney, l. m., and spector, p. e. (2005). job stress, incivility, and counterproductive work behavior (cwb): the moderating role of negative affectivity. journal of organizational behavior, 26(7), 777-796. robinson, s. l., and bennett, r. j. (1995). a typology of deviant workplace behaviors: a multidimensional scaling study. academy of management journal, 38(2), 555-572. sackett, p. r. (2002). the structure of counterproductive work behaviors: dimensionality and relationships with facets of job performance. international journal of selection and assessment., 10(1-2), 5-11. saeed, a., and ibrahim, h. (2005). reasons for the problems faced by patients in government hospitals: results of a survey in a government hospital in karachi, pakistan. the journal of the pakistan medical association, 55(1), 45-47. sekaran, u. (2006). research methods for business: a skill building approach: john wiley & sons. shaheen, s., abrar, m., saleem, s., shabbir, r., and zulfiqar, s. (2021). linking organizational cronyism to deviant workplace behavior: testing the mediating role of employee negligence in pakistani higher education institutions. international journal of leadership in education, 1-23. doi: 10.1080/13603124.2020.1869313 shahzad, a., and malik, r. (2014). workplace violence: an extensive issue for nurses in pakistan-: a qualitative investigation. journal of interpersonal violence., 29(11), 20212034. shaikh, a., and humayun, n. (2012). medical ethics in undergraduate medical education in pakistan: towards a curricular change: intech open access publisher. shaikh, m. a. (2000). sexual harassment in medical profession-perspective from pakistan. journal-pakistan medical association., 50(4), 130-131. sherif, m., and sherif, c. w. (1953). groups in harmony and tension; an integration of studies of intergroup relations. oxford, england: harper & brothers. somani, r., karmaliani, r., farlane, j. m., asad, n., and hirani, s. (2015). prevalence of bullying/mobbing behaviour among nurses of private and public hospitals in karachi, pakistan. international journal of nursing education, 7(2), 235-239. somani, r., karmaliani, r., mc farlane, j., asad, n., and hirani, s. (2015). sexual harassment towards nurses in pakistan: are we safe? international journal of nursing education, 7(2), 286-289. spector, p. e., and fox, s. (2005). the stressor-emotion model of counterproductive work behavior counterproductive work behavior: investigations of actors and targets (pp. 151-174). washington, dc, us: american psychological association. http://www.finance.gov.pk/survey/chapter_20/11_health_and_nutrition.pdf syed arif hussain shah, tazeem ali shah, abrar ullah and muhammad yasir 82 tepper, b. j., henle, c. a., lambert, l. s., giacalone, r. a., and duffy, m. k. (2008). abusive supervision and subordinates' organization deviance. journal of applied psychology., 93(4), 721. thau, s., and mitchell, m. s. (2010). self-gain or self-regulation impairment? tests of competing explanations of the supervisor abuse and employee deviance relationship through perceptions of distributive justice. journal of applied psychology., 95(6), 1009. vardi, y., and weitz, e. (2003). misbehavior in organizations: theory, research, and management: psychology press. yasir, m., jamil, f., adil, m., hamayun, m., and irum, s. (2017). direct and indirect effects of transformational leadership on workplace deviance: evidence from non-profit organizations of pakistan. journal of managerial sciences, 11(4), 121-141. yasir, m., and khan, n. (2020). mediating role of employees’ trust in the relationship between ethical leadership and workplace deviance in the public sector hospitals of khyber pakhtunkhwa. leadership, education, personality: an interdisciplinary journal, 2(2), 113-123. doi: 10.1365/s42681-020-00010-5 yasir, m., and mohamad, n. a. (2016). ethics and morality: comparing ethical leadership with servant, authentic and transformational leadership styles. international review of management and marketing, 6(4s), 310-316. yasir, m., and rasli, a. (2018). direct and indirect effects of ethical leadership on workplace deviance in public healthcare sector of pakistan. journal of advances in management research, 15(4), 558-574. doi: https://doi.org/10.1108/jamr-11-2017-0109 yasir, m., rasli, a., and qureshi, m. i. (2017). investigation of the factors that affect and gets affected by organizational ethical climate. advanced science letters, 23(9), 93519355. doi: https://doi.org/10.1166/asl.2017.10479 yousafzai, a. w. (2015). corruption in medical practice: where do we stand? journal of ayub medical college abbottabad, 27(3), 515-516. zagenczyk, t. j., restubog, s. l. d., kiewitz, c., kiazad, k., and tang, r. l. (2014). psychological contracts as a mediator between machiavellianism and employee citizenship and deviant behaviors. journal of management, 40(4), 1098-1122. zapf, d. (1999). organisational, work group related and personal causes of mobbing/bullying at work. international journal of manpower., 20(1/2), 70-85. zhu, h., lyu, y., and ye, y. (2019). workplace sexual harassment, workplace deviance, and family undermining. international journal of contemporary hospitality management, 31(2), 594-614. https://doi.org/10.1108/jamr-11-2017-0109 https://doi.org/10.1166/asl.2017.10479 journal of applied economics and business studies, volume. 5, issue 1 (2021) 175-190 https://doi.org/10.34260/jaebs.519 175 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x an analysis of food insecurity in pakistan: prevalence of undernourishment (pou) and food insecurity experience scale (fies) ghulam sadiq afridi1*, abdul jabbar2, shahzad khan 3 and nadeem akmal4 1. member social sciences division, pakistan agricultural research council, pakistan. 2. assistant professor, international institute of islamic economic, international islamic university, islamabad, pakistan. 3. assistant professor, institute of development studies, the university of agriculture, peshawar, pakistan 4. principal scientific officer, social sciences division, pakistan agricultural research council, pakistan abstract this study covers only sdg target 2.1 (2.1.1-prevalance of undernourishment and 2.1.2-food insecurity experience scale). though fao is the custodian organization for estimating these targets across the globe, however, it is the first ever attempt for estimating these targets by parc-mnfs&r. hies data for the year 2018-19 has been used for estimation of these targets and compared with the results of hies-2015-16 estimated by fao. according to the results 18.38 percent households are undernourished in pakistan and this situation is worse in urban areas (23.43%) compared with rural areas (16.61%). punjab has highest proportion of undernourished individuals/households with 21.48 percent followed by sindh province with 17.40 percent households. khyber pakhtunkhwa has the lowest proportion of 12.67 percent and baluchistan with 16.95 percent households. national level results of fies supports the results of pou except urban/rural order. according to fies results, about 16 percent of the households (individuals) are moderate and/or severe food insecure with more than 02 percent as severe food insecure in pakistan. sindh province shows highest proportion with more than 19 percent followed by kp province with nearly 17 percent keywords pou, fies, sdg, food insecurity, pakistan jel classification c02, c31, c63 * sadiqafridi@gmail.com ghulam sadiq afridi, abdul jabbar, shahzad khan and nadeem akmal 176 households as moderate and severe food insecure. however, highest proportions of more than 03 percent households were found as severe food insecure in punjab province. in conclusion pakistan has shown tremendous achievements towards the zero hunger targets by 2030, however, more efforts are needed for sustainable agriculture and food system in order to address the food insecurity level through better access and availability of food. awareness campaign about healthy and nutritious food intake, and measures for adoption of dietary guidelines are recommended for preventing undernourishment. 1. introduction food security, as defined by the committee on world food security of united nations’ as, “all people, at all times, have physical, social, and economic access to sufficient, safe, and nutritious food that meets their food preferences and dietary needs for an active and healthy life”. for last three decades, the concepts of food security have appeared to imitate modifications in official policy thinking (clay, 2002; heidhues et al., 2004). the definition of food security was confined to the availability of food along with stability of price of the essential food item that was initiated in the mid-1970s, during the world food conference in 1974. this definition was “availability at all times of adequate world food supplies of basic foodstuffs to sustain a steady expansion of food consumption and to offset fluctuations in production and prices”. later on food and agriculture organization (fao) paying attention towards balancing the equation of food demand and supply and new definition of food security was raised as “ensuring that all people at all times have both physical and economic access to the basic food that they need” (fao, 1983). fao revised this definition in food security analysis further by including individual and household level, in addition to regional and national level of aggregation. the highly significant report on poverty and hunger by world bank report (world bank, 1986) focused on temporal dynamics of food insecurity (clay, 2002). the world bank report is pioneered in introducing the differences among chronic food insecurity that is caused by low income and/or structural poverty and transitory food insecurity. this was supplemented by sen’s theory of famine (1981) who stressed on the effect of personal entitlements to food access i.e., labor, production, , trade and transfer based resources. the most acknowledgeable definition of food security of world food summit (1996) fortifies on multidimensional nature of food security that includes all pillars of food security like, food availability, access to food, utilization and stability. this definition enabled policy makers to focus on the promotion and recovery of livelihood choices, that primary introduced by chambers and conway (1992), an academician. developmental programs of the international organizations now revolve around journal of applied economics and business studies, volume. 5, issue 1 (2021) 175-190 https://doi.org/10.34260/jaebs.519 177 livelihood approaches. according to devereux (devereux 2000), food insecurity as a social and political construct has textured the present for analysis leaving the link among food security, starvation and crop failure in past. many different indicators have been used during the last 30 years to provide information on household food insecurity, but none existed that combined the properties of validity, reliability and cross-country and over-time comparability. in 1995, the household food security survey module (hfssm) was applied on a nationally representative sample of the us population in the continued population survey and the responses to the 18 questions it comprised, analyzed through the lenses of the rasch measurement model firstly to obtain a proper measure of the severity of the food insecurity condition, defined as inability of households or individuals to access food, owing to money shortage or other resources commonly devoted to it. however, climate change, increasing population pressure, souring the prices of food and environmental stresses will not have insignificant impact on food security in the coming decades. many adaptation and mitigations strategies and policy measures are needed to respond global climate changes like optimal allocation of water and land resources, managing food losses and wastage, trade of food, processing of food and prices of food. in this regards international food policy research institute (ifpri) has worked on different dimensions of food security covering, promotion of sustainable agricultural technologies analysis of cash transfers, building resilience to shocks, and trade-offs management in food security. 1.1 prevalence of undernourishmentpou (sdg indicator 2.1.1): pou is an estimate and refers to that population’s proportion that have habitual food consumption providing insufficient dietary energy levels required for maintaining a normal active and healthy life. sukhatme (sukhatme, 1961) originally proposed to monitor food security by estimating the prevalence of undernourishment in the population based on two new ideas: i) part of the variability observed in food consumption has nothing to do with food insecurity, as it reflects the normal variability due to differences in food requirements; ii) it is not just a matter of availability but also of access to food. application of sukhatme method becomes possible for most countries where household surveys including food consumption module. 1.2 food insecurity experience scale – fies (sdg indicator 2.1.2): on the other hand, fies is food insecurity measure centering experience-base and mainly encompasses direct responses of the people to questions about their experiences in the past one year who face problems in access to food. food and agriculture organization (fao) under voice for hungry project developed analytical techniques to measure experience-based food security measurements. these techniques made it viable to compare the rate of food insecurity prevalence not only across countries but even also ghulam sadiq afridi, abdul jabbar, shahzad khan and nadeem akmal 178 at sub-national level. these techniques produce: meaningful and reliable information on the adequacy of food access at the individual or household level, is direct measure of food insecurity experienced by people and households, easily applicable to less expenditure within any individual or household survey, and provides actionable information that policy makers can use to identify vulnerable population groups and guide policy interventions when applied to large population survey. 2. methodology this section describes sources of data, techniques and methods used for analysis of the data. data and sources: hies data for the year 2018-19 was used for this study. beside this, numerous data set have been used for the analysis. hies data covers only age but no height. for calculating body mass index (bmi), height data is pre-requisite and this data was drawn from pakistan demographic and health survey, 2017-18. for calories and other nutrients distribution pakistan dietary guidelines for better nutrition developed by planning commission of pakistan-2018 were used. pou estimation: estimation of pou is a cumbersome process and involved a lot of complicated analysis and large dataset. it is not only handling the large data set of more than 25000 households but also handling nutritional data, reference value for dietary energy requirements of all classes of age simultaneously. for calculation of all these required figures following steps and assumptions are generated. dietary energy intakes and requirements: the energy amount required for living a long-term active and healthy life is called dietary energy requirements. any individual is considered to be undernourished if the energy intake level of his/her habitual dietary is less than the minimum dietary energy requirement assumed to be appropriated by nutritionists. similarly for a group of individuals there is a range of energy requirements that are compatible with long-term good health and also there are so many other factors like, age, physiological status, gender, body weight and lifestyle that affect energy requirements. probability distribution framework: there is difference between energy requirements and actual intake. the estimation of the undernourished in a population cannot be viewed as a simple accounting exercise involving the comparison of the observed household dietary energy consumption (dec) with the estimated household dietary energy requirement (der), and counting the individuals in the households with such that dietary energy consumption is less than dietary energy requirement (dec < der). journal of applied economics and business studies, volume. 5, issue 1 (2021) 175-190 https://doi.org/10.34260/jaebs.519 179 probability distribution of der: there are two types of probability distribution. one, for individual within population group – that how the energy requirement of individuals is distributed within a population that is represented by a probability distribution where the values of energy requirement are associated to a probability of occurrence. and second, each population group has its own probability distribution of dietary energy requirements. probability distribution of dietary energy intake: now again individual within population group for dietary energy intake that how the dietary energy intake (dei) of individuals is distributed within population represented by a probability distribution where values of dietary energy intake are associated to a probability of occurrence is different from that each population group has its own probability distribution of dietary energy intake. prevalence of undernourishment (pou): now pou can be represented by the following equation and figure-1. 𝑃𝑜𝑈 = ∫ 𝑓(𝑥|𝐷𝐸𝐶; 𝐶𝑉; 𝑆𝐾)𝑑𝑥 𝑥<𝑀𝐷𝐸𝑅 (1) figure-1: distribution of pou it is depicted from the above figure that within a probability distribution framework, the prevalence of undernourishment (pou) is the probability that a randomly selected individual from a population has habitual access to food which does not provide the dietary energy necessary to cover his/her habitual energy requirements. it is just an estimate and not a direct measure but shows individuals proportion in a population that ghulam sadiq afridi, abdul jabbar, shahzad khan and nadeem akmal 180 suffering from chronic hunger (a state, lasting for at least one year, of inability to acquire enough food to satisfy the energy requirements). fies estimation: for the first time pakistan bureau of statistics incorporated a separate section (section 5b) in hies data for the year 2018-19 to capture data on food insecurity experience scale (fies). this section covers all the basic 8 questions of the fies survey module based on respondent reply with dichotomous (yes/no) responses on food insecurity experience during last 12 months. this report is based on the hies 2018-19 data set. the 8 questions included at section 5b of hies 2018-19 are: during the last 12 months, was there a time when: 1. you were worried you would run out of food because of a lack of money or other resources? 2. you were unable to eat healthy and nutritious food because of a lack of money or other resources? 3. you ate only a few kinds of foods because of a lack of money or other resources? 4. you had to skip a meal because there was not enough money or other resources to get food? 5. you ate less than you thought you should because of a lack of money or other resources? 6. your household ran out of food because of a lack of money or other resources? 7. you were hungry but did not eat because there was not enough money or other resources for food? 8. you went without eating for a whole day because of a lack of money or other resources? the model: rasch (rasch, 1960) developed a model based on the theory of latent response. this model states to a family of mathematical models that attempt to describe the association between latent traits (unobservable characteristic or attribute) and their manifestations (i.e., observed outcomes, responses or performance). this model is said to be foundation of the item response theory and has been used to capture the food insecurity experience based on dichotomous response of yes and no. rasch model is an inimitable method of mathematical modeling centered upon a latent trait and accomplishes stochastic (probabilistic) conjoint additivity (conjoint means measurement of persons and items on the same scale and additivity is the equal-interval property of the scale). mathematically, rasch model is equal to single parameter logistic model, though, rasch model constrains the item discrimination to 1, while the 1-parameter logistic model strives to fit the data as much as possible and does not limit the discrimination factor to 1. the rasch model is considered superior due to its more concerning with journal of applied economics and business studies, volume. 5, issue 1 (2021) 175-190 https://doi.org/10.34260/jaebs.519 181 developing the variable which is being used for measuring the dimension of interest. therefore, when constructing an instrument fitting, the rasch model would be best, improving the precision of the items. rasch model implies that raw score (i.e., simple sum of affirmative responses) has adequate statistic to measure severity of respondents’ as both affirming an item and denying it convey information individual measures of severity depend on the number of affirmed items, not on which particular set of items have been affirmed. unexpected patterns, however, contribute to determine measures of mis-fit, used to test the empirical validity of the model use of raw score for classification greatly facilitates use of the method, but it is only legitimate if data conform to the model’s assumptions of equal discrimination of all item and conditional independence of the responses to each item. conditional maximum likelihood (cml) can be used to estimate parameters. it imposes no assumption on the shape of the distribution of the latent trait in the population (+) provides consistent estimates of standard errors under the rasch model assumptions (+) it only uses non-extreme response patterns, as severity parameters for zero or maximum raw score cannot be estimated. if items are not adequately chosen, this may limit the size of effective samples. 𝑃𝑟𝑜𝑏(𝑋𝑖,𝑗 = 1) = exp(𝑎𝑖−𝑏𝑗) 1−exp(𝑎𝑖−𝑏𝑗) (2) • it is the foundation of item-response theory • 𝑋𝑖,𝑗 ∈ {0,1} is the “response” of the i-th respondent to the j-th “item”. • the probability that a respondent whose position on a scale is 𝑎𝑖 might respond to an item positioned at 𝑏𝑗on the same scale is a (logistic) function of the difference (𝑎𝑖 − 𝑏𝑗) • the model provides the probabilistic basis for estimating the parameters associated with both items and respondents and for conducting statistical tests of the strength of association of the responses to the latent trait and of goodness of fit rasch model has has unique properties i.e., (i) raw score is a sufficient statistic for the latent trait measure. it means that two respondents with the same raw score but different response patterns will be assigned the same measure (even though the absolute error around the measure may differ), (ii) respondent parameters form an interval scale on the latent trait metrics. these properties are subject to assumptions of (i) infit statistic in the range 0.7 – 1.3, (ii) high rasch reliability measures and (iii) no correlation among “residuals”. ghulam sadiq afridi, abdul jabbar, shahzad khan and nadeem akmal 182 3. results and discussion: prevalence of undernourishment: estimation of pou is a cumbersome process and fao is custodian for its estimation across the globe and rarely it is estimated by local experts. table-1, as following divulges the results of the prevalence of undernourishment in pakistan. according to the results 18.38 percent households are undernourished in pakistan and this situation is worse in urban areas (23.43%) compared with rural areas (16.61%). the provincial picture is interesting as depicted from table-1 that punjab has highest proportion of undernourished individuals/households with 21.48 percent followed by sindh province with 17.40 percent households. khyber pakhtunkhwa has the lowest proportion of 12.67 percent and baluchistan with 16.95 percent households. the order of provincial and nation rural/urban undernourished households though is not too much different from the results of 2015-16, however, the country has shown commendable achievements in catering poverty and undernourishment. during the past three years pakistan has made improvement achieving food security by lowering the proportion of undernourished households from 22.12 percent (fao results based on 2015-16 data) to 18.38 percent. similarly, punjab has shown improvement by lowering the proportion of undernourished from 23.05 percent to 21.48 percent, sindh from 26.12 percent to 17.40 percent. the food insecurity status of the country was on increasing trend since 2008 as reflected from fao results. pou was 21.8 in 2008, increased to 21.9 during 2010 to 2013 and further boosted to 22.12 during 2014 to 2016 (fao, 2016). however, the score dropped to 18.38 as per this report results, indicates that pakistan has done nicely to overcome the issue of food insecurity by reducing the proportion of undernourished people. table-1: national and provincial estimates of pou province region dec min. dietary energy requirement avg. dietary energy requirement max. dietary energy requirement pou 2018-19 punjab overall 2141 1723 2220 2637 21.48 sindh overall 2191 1732 2233 2650 17.40 kp overall 2216 1672 2149 2550 12.67 baluchistan overall 2084 1701 2190 2595 16.95 pakistan overall 2162 1715 2209 2623 18.38 pakistan urban 2142 1754 2266 2696 23.43 pakistan rural 2174 1692 2176 2581 16.61 source: author’s own estimation food insecurity experience scale: sustainable development goal (sdg)-2 related to zero hunger and 2.1.2 being its sub indicator is based on food insecurity experience scale (fies). data set on fies journal of applied economics and business studies, volume. 5, issue 1 (2021) 175-190 https://doi.org/10.34260/jaebs.519 183 indicators was not available so far in pakistan, however, pakistan bureau of statistics included these indicators in its hies data set for the year 2018-19. based on hies data set fies was estimated and the results are presented in the following sections. reliability of residual correlation: all 08 questions as mentioned in section 2.2 were analyses and were checked for their reliability over residual correlation. the results are given in table-2, which depicts that residual correlations are in normal range and no additional dimension is captured by the fies data. this is visible by the constant descending slope of the curve, without any sudden change of slope. it confirms the results of residual correlation matrix about the uni-dimensionality of the data. rasch model reliability was found to be 79 percent. table-2: reliability and residual correlation: rasch reliability = 0.79 residual correlation healthy fewfood skipped ateless runout hungry wholeday worried 0.18 -0.05 -0.02 -0.14 -0.04 0.00 0.03 healthy 0.26 -0.19 -0.09 -0.14 -0.11 -0.05 fewfood -0.07 -0.01 -0.11 -0.10 -0.07 skipped 0.03 0.09 0.01 -0.05 ateless 0.00 -0.00 -0.12 runout 0.42 0.42 hungry 0.51 source: author’s own estimation item severity parameters and infit statistics: item severity parameters show that the experience with the lower severity is accessing food considered “healthy”, followed by having to eat only “few kinds of food” and being “worried” the household would runout of food. the severity is gradually increasing up to the most severe experience, having not to eat for a “whole day” owing to deficiency of money or other resources. the severity pattern makes sense and describes the process followed in pakistan towards food insecurity. infit statistics suggest that all items have equal weight in the measure of food insecurity and therefore they all can be kept in the scale for a proper measure of food insecurity. ghulam sadiq afridi, abdul jabbar, shahzad khan and nadeem akmal 184 table-3: item statistics: severity s.e. infit s.e. infit outfit worried -0.861 0.029 1.009 0.014 1.174 healthy -3.056 0.038 0.910 0.024 2.001 fewfood -2.715 0.035 0.995 0.021 2.392 skipped 0.719 0.035 1.081 0.021 1.375 ateless -0.168 0.030 1.101 0.016 1.520 runout 1.735 0.043 0.757 0.029 0.998 hungry 1.918 0.045 0.751 0.031 0.652 wholeday 2.827 0.058 0.814 0.040 0.294 total number of cases 24809 source: author’s own estimation prevalence of food insecurity: table-4 to table-6 divulges prevalence of food insecurity in pakistan by household and individual level across urban and rural territories as well provincial break up. according to results about 16 percent of the households (individuals) are moderate and/or severe food insecure with more than 02 percent as severe food insecure. sindh province shows highest proportion with more than 19 percent followed by kp province with nearly 17 percent households as moderate and severe food insecure. however, highest proportions of more than 03 percent households were found as severe food insecure in punjab province. table-4: prevalence of food insecurity by household and individual in pakistan (%) prevalence rates of food insecurity (% of households) moderate or severe moe severe moe 15.92 0.84 2.37 0.32 prevalence rates of food insecurity (% of individuals) moderate or severe moe severe moe 15.98 0.91 2.40 0.35 source: author’s own estimation table-5: prevalence of food insecurity by urban and rural household/individual in pakistan (%) prevalence rates of food insecurity by urban/rural (% of households) moderate or severe moe severe moe urban 9.22 1.12 1.24 0.40 rural 19.96 1.14 3.05 0.46 prevalence rates of food insecurity by urban/rural (% of individuals) journal of applied economics and business studies, volume. 5, issue 1 (2021) 175-190 https://doi.org/10.34260/jaebs.519 185 moderate or severe moe severe moe urban 9.33 1.21 1.28 0.44 rural 19.72 1.23 3.04 0.48 source: author’s own estimation table-6: prevalence of food insecurity by provincial household/individual in pakistan (%) prevalence rates of food insecurity by province (% of households) moderate or severe moe severe moe punjab 14.43 1.16 3.06 0.50 sindh 19.51 1.77 1.55 0.47 khyber pakhtunkhwa 16.73 1.90 0.90 0.48 baluchistan 15.18 2.72 2.05 0.95 prevalence rates of food insecurity by urban/rural (% of individuals) moderate or severe moe severe moe punjab 14.55 1.27 3.13 0.55 sindh 19.36 1.94 1.75 0.55 khyber pakhtunkhwa 15.95 2.03 0.96 0.54 baluchistan 16.47 3.18 2.31 1.14 source: author’s own estimation raw and diagnostic statistics given infit statistics are in the acceptable range, the raw score can be considered an ordinal measure of food insecurity (the higher the raw score, the higher the severity). however, the estimate is associated to a measurement error that needs to be taken into account in the estimate of prevalence of food insecurity. the two most discrepant items compared to the global standard are worried and healthy. one possibility to calibrate the pakistani scale against the fies global standard is to consider these two items “unique” to the metric of pakistan, and not to use them in order to compute cross-country comparable thresholds of food insecurity. table-7: raw score rs severity error 0 -4.177 1.576 1 -3.210 1.249 2 -1.906 1.096 3 -0.794 1.013 4 0.176 0.959 5 1.078 0.943 6 1.999 0.983 7 3.058 1.164 8 4.269 1.576 source: author’s own estimation ghulam sadiq afridi, abdul jabbar, shahzad khan and nadeem akmal 186 table-8: absolute difference between country and global standard items worried 0.77 healthy 0.69 fewfood 0.26 skipped 0.02 ateless 0.20 runout 0.33 hungry 0.17 wholeday 0.50 correlation between common items = 90.2 % source: author’s own estimation table-9: infit statistics by urban and rural urban rural worried 0.97 1.02 healthy 0.87 0.93 fewfood 1.03 0.97 skipped 1.01 1.10 ateless 1.08 1.11 runout 0.78 0.75 hungry 0.75 0.75 wholeday 0.89 0.79 number of cases 8873.00 15936.00 source: author’s own estimation differential item functioning: differential item functioning (dif) studies the possible different functioning of the scale across population characteristics. this would not be a good property of the scale, as the “measurement invariance” property asserts that the scale should function in the same way independently from the population sampled. in this case, we can conclude that there is no dif across urban/rural contexts and the same scale can be used to measure food insecurity in these areas. table-10: infit statistics and dif by province punjab sindh khyber pakhtunkhwa baluchistan worried 1.04 0.89 1.18 0.86 healthy 0.84 0.92 0.99 0.84 fewfood 0.89 1.15 0.97 1.19 skipped 0.90 1.41 1.05 1.11 ateless 1.21 1.06 0.93 0.87 runout 0.77 0.57 0.90 0.77 hungry 0.72 0.68 0.78 0.87 wholeday 0.89 0.75 0.71 1.01 number of cases 11781.00 6216.00 4485.00 2327.00 source: author’s own estimation journal of applied economics and business studies, volume. 5, issue 1 (2021) 175-190 https://doi.org/10.34260/jaebs.519 187 conclusions and recommendations agricultural profile of pakistan is pretty well. pakistan is ranked 2nd in production of indigenous oilseed, milk of buffalo and meat of buffalo, and, 3rd in cottonseed production and chilies , 4th in cotton lint, pulses, mango, goat milk, , roots and tubers, 5th in chickpeas and spices , 6th in wheat, spinach, sugarcane, apricot and dates, 7th in broccoli and cauliflower , 8th in onion, mandarin, tangerines and tobacco , 11th in oranges, pistachio and wools, 13th in rice, 14th in peas, eggs and birds’ and and 15th in lentil production. large population of the country exerts tremendous pressure on the available stock of food and fiber. production or availability of food is not a serious issue in pakistan, but its accessibility both physically (good example is milk; 50 percent surplus production) and economically (low income and high prices) is a matter of concern. pou refers to physical access of food and its utilization. on the other hand, fies scores insinuates economic access to food. in pakistan about 75 percent of the calories are taken from three main sources (wheat/rice about 50 percent, oil/ghee 15 percent and sugar about 10 percent). consumption of these three commodities is not only unhealthy food but also deficit in required balanced food resultantly which causes prevalence of undernourishment and other diet related issues. the disaggregated side analysis of the food commodities divulge that international commodity prices did not affect domestic prices especially wheat and sugar, but supply demand gap causes prices to rise. the demand side causes of food inflation cannot be ignored but supply also pushed inflation through increased cost of production of major and stable foods. government policy regarding productivity enhancement of important crops like wheat, rice, sugarcane and pulses is good initiative towards more food availability but enhancing purchasing power of the consumers through lowering cost of production and enhancing per capita income may incentivize the economic access to food and nutrition. furthermore, government should launch awareness campaign for balanced and healthy food habits. it is also recommended that government should take measurement to undertake dietary guidelines for consumers and strictly enforce labeling of processed food for ingredients and health hazards. ghulam sadiq afridi, abdul jabbar, shahzad khan and nadeem akmal 188 figures on equating plot and differential item functioning (dif): figure-2: equating plot figure-3: dif by region (1=urban, 2 = rural) figure-4: dif by province (1=punjab, 2=sindh, 3= kp, 4=balochistan) references: chambers, r. and r. conway, (1992). sustainable rural livelihoods: practical concepts for the 21st century. ids discussion paper, no. 296. pp.127-130. clay, e. (2002). food security: concepts and measurement, paper for fao expert consultation on trade and food security: conceptualising the linkages rome, 11-12 july 2002. ● ● ● ● ● ● ● ● −4 −2 0 2 4 − 4 − 2 0 2 4 dif by urban/rural 1 national 1 worried healthy fewfood skipped ateless runouthungry wholeday ● ● ● ● ● ● ● ● −4 −2 0 2 4 − 4 − 2 0 2 4 dif by urban/rural 2 national 2 worried healthy fewfood skipped ateless runouthungry wholeday ● ● ● ● ● ● ● ● −4 −2 0 2 4 −4 −2 0 2 4 dif by region 1 national 1 worried healthy fewfood skipped ateless runout hungry wholeday ● ● ● ● ● ● ● ● −4 −2 0 2 4 −4 −2 0 2 4 dif by region 2 national 2 worried healthy fewfood skipped ateless runout hungry wholeday ● ● ● ● ● ● ● ● −4 −2 0 2 4 −4 −2 0 2 4 dif by region 3 national 3 worried healthy fewfood skipped ateless runout hungry wholeday ● ● ● ● ● ● ● ● −4 −2 0 2 4 −4 −2 0 2 4 dif by region 4 national 4 worried healthy fewfood skipped ateless runout hungry wholeday journal of applied economics and business studies, volume. 5, issue 1 (2021) 175-190 https://doi.org/10.34260/jaebs.519 189 devereux s (2000). famine in the twentieth century, ids working paper, 105. institute of development studies, brighton. fao. (1983). world food security: a reappraisal of the concepts and approaches. director general’s report. rome. fao. (1996). world food summit. report of the world food summit, 13-17 november, 1996. food and agriculture organization, rome, italy. http://www.fao.org/3/w3548e/w3548e00.htm. fao. (2016). country fact sheet on food and agriculture policy trends: socioeconomic context and role of agriculture. food and agriculture organization of the united nations, rome, italy, 2016. heidhues, f., a. atsain, h. nyangito, m. padilla, g. ghersi and j. le vallée (2004). development strategies and food and nutrition security in africa: an assessment. 2020 discussion paper no. 38. rasch, g. (1961). on general laws and the meaning of measurement in psychology. in j. neyman (ed.), proceedings of the fourth symposium on mathematical statistics and probability, vol. 4. berkeley: university of california press, 321– 333. sen, a. k. (1981). "ingredients of famine analysis: availability and entitlements", quarterly journal of economics, vol. 95. shukhatme, p.v. (1961). the world hunger and future needs in food supplies. journal of royal statistics society, 124: 463, 1961. world bank. (1986). “the world bank’s study for reducing poverty and hunger”. a report to the development community. environmentallysustainable developmentstudiesand monographsseries no. 4. theworldbank,washington, d.c. http://www.fao.org/3/w3548e/w3548e00.htm ghulam sadiq afridi, abdul jabbar, shahzad khan and nadeem akmal 190 acknowledgement introduction and methodology of this article is based on the established literature and methodology developed by fao for estimation of pou and fies. i am highly indebted to pay gratitude to dr. viviani sara and dr. meghan miller from fao, rome for their valuable comments on the analysis of fies. i would also like to express my sincere gratefulness to dr. sattar, fao for his continuous support and help in estimation of pou. i am also thankful to my colleagues dr. jabbar, dr. shahzad khan and dr. nadeem akmal for their support and technical assistance. journal of applied economics and business studies, volume. 3, issue 1 (2019) 41-60 https://doi.org/10.34260/jaebs.313 41 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x efficiency assessment of public education & health sector in selected middle-income countries with special reference to millennium development goals (mdgs) sami ullah khan1*, zahoor khan2 & gulnaz hameed3 1 phd scholar at institute of management sciences, peshawar and lecturer, department of economics, university of swabi. 2 assistant professor, institute of management sciences, peshawar 3 assistant professor, department of economics, faculty of social sciences, pir mehir ali shah arid agriculture university rawalpindi. abstract this study investigates relative efficiency of public education and health sector in selected middle income countries with special reference to millennium development goals (mdgs). the study uses data for two reference years; 2000 (implementation year of mdgs for developing countries) and 2015 (the final year of mdgs). data envelopment analysis (dea) and malmquist productivity index (mpi) are used to calculate relative efficiency, operating scale of the countries (dmus) and productivity change in relative efficiency over time respectively. the paper conceptualizes relative efficiency of the countries in discretionary, multicriteria input-output variables context to investigate efficiency differences among the countries and deduce important takeaways. educational expenditure, teachers at primary level, health expenditure, birth attended by skilled staff are used as input variables while enrollment at primary level, completion of primary level education, infant survival per annum and child survival per annum are used as output variables. the dea results show that all countries could not operate at efficient level to target mdgs. the level of efficiency was not same under different dea specifications in both the periods. some countries were inefficient because of their size; either having too large size or too small size of operation. the sources of change in efficiency over the time were either because of real change in efficiency or change in technology frontier or both. the study identified a set of institutional and individuals factors which contribute to the efficiency and inefficiency of dmus under investigation. keywords efficiency, data envelopment analysis (dea), public education, health, malmquist productivity index (mpi) jel classification c14 h51 h52 i18 i21 i28 1. introduction scarcity of resources is supposed to be the main economic problem faced by all economic agents while making various economic choices4. hence, every individual wants to get maximum benefits from the given resources by utilizing these scares resources * corresponding author: samiullahkhan@uoswabi.edu.pk 4 robbins (1932). sami ullah khan, zahoor khan & gulnaz hameed 42 efficiently. equally important, the same idea is required at societal and macro level to ensure efficient allocation of resources for the welfare of the general masses. health and education are the basic needs as well as fundamental rights of every individual and thus considered main factors of human development (ul-haq, 1995). it is the responsibility of every government to provide basic education and health facility to their people. in every budget, government allocate sufficient amount of their revenue to the education and health sector. the percentage of expenditure allocated to these sectors varies across the countries. mostly, in developed countries the percentage expenditure allocated to these two sectors are comparatively high than developing countries. therefore; the education level, quality and health performance indicators of developing countries are mostly unsatisfactory comparatively. the member countries of united nation (un), during their millennium meeting in 2000, signed an agreement and set time-bounded goals, known as millennium development goals (mdgs). these goals are related to poverty, education and health5. after this agreement, all member countries focused on these sectors and allocated resources to these sectors while keeping in mind the goals. in the un report of 2015, most of the countries have not achieved health and education related targets of mdgs. in developing countries, some countries performed better in achieving those targets while some of developing countries did not perform well and could not achieve mdgs targets. the purpose of this study is to answer the basic yet important question why all the developing countries could not perform equally despite of assigning equal targets? there might be issues in the targets set by the un or might be different possible reasons that why majority of the countries could not achieve mdgs. this study has important lessons learnt for developing countries while now pursuing sustainable development goals; an extension of mdgs. studies are available that examined efficiency in health or/and education sector across the countries however; there exist a significant gap in literature related to efficiency in health and education sector with reference to mdgs goals and targets. most of the studies focused on calculating efficiency scores by dea or frontier disposal hull (fdh) methods (asandului et al., 2014; gupta & verhoeven, 2001; pang & herrera, 2005). some studies focus comparing efficiency in education and health sectors among oecd and developed countries (afonso & aubyn, 2006; afonso et al., 2010; aristovnik, 2012) with few exceptions which compared efficiency in these two sectors in developing countries context (gupta & verhoeven, 2001; pang & herrera, 2005; wang & alvi, 2011; jimenez and lockheed, 1995). however; to the best of our knowledge we could not find any study which specifically focuses on calculating efficiency in health and education sector by considering 5https://www.afdb.org/fileadmin/uploads/afdb/documents/publications/annexe%20mdg2011.pdf https://www.afdb.org/fileadmin/uploads/afdb/documents/publications/annexe%20mdg2011.pdf journal of applied economics and business studies, volume. 3, issue 1 (2019) 41-60 https://doi.org/10.34260/jaebs.313 43 the mdgs targets and goals. efficiency in health and education sector (combined) among developing countries is not studied extensively. most of the studies have selected countries within specific geographical region. the selection of dmus is very critical and efficiency scores are very sensitive to numbers and types of dmus. so, by comparing efficiency among countries based on region is very critical. most of the studies have only calculated efficiency scores for specific time or year and did not considered changes in efficiency over a period of time. majority of the studies did not consider the returns to scale in health and education sector and ignored the scale of country at which they are currently operating. in light of the above discussion, the main objective of this study is to examine health and educational performance of selected middle income countries in context of mdgs by using multi-criteria input and output approach. data envelopment analysis (dea), a nonparametric multi-criteria technical efficiency assessment method is used for two different periods; 2000 and 2015. the paper uses an interesting idea by taking into consideration countries as productive units which use certain resources as inputs and process it into desirable outputs such as health and education outcomes. in this context, each country is considered as a decision making unit (dmu). we select these two time periods because 2000 was the starting year of mdgs and 2015 was the end year of mdgs targets. the selection of countries (dmus) is based on their per capita income level and considered only selected middle income countries6 as dmus. the study selects only middle income countries in order to ensure homogeneity among dmus as dea analysis is sensitive to selection of dmus. variables (specifically output variables) are selected based on mdgs targets and goals related to education and health sectors. the study calculated overall technical efficiency, pure technical efficiency, scale efficiency and supper efficiency. returns to scale for each dmu and for both periods are also calculated to observe operating level of each country and the size of each country. to study that how dmus are performing overtime and how their productivity changing overtime, malmquist productivity index (mpi) is also calculated. 2. literature review the literature on relative efficiency in health and education sector, across the countries, is mostly related to developed, european, oecd and emerging countries. afonso et al. (2005) calculated efficiency and performance of public sector expenditure in seven sectors7 as a whole and separately as well for 23 industrialized oecd countries during 1990 and 2000. they calculated public sector performance (psp) index and used free disposable hull (fdh) for calculating efficiency. another study by afonso & aubyn (2006) examined the efficiency in educational expenditure among 25 oecd countries. they compare the 6 the classification of countries is based on world bank’s countries classification 7 administrative, education, health, public infrastructure, distribution, stability and economic performance sami ullah khan, zahoor khan & gulnaz hameed 44 efficiency in educational output and its determinants by carrying dea and tobit analysis. wang & alvi (2011) examined the relative efficiency and determinants of efficiency in ten oecd and seven asian countries by implementing dea and tobit model. the results show that in oecd countries us, germany and new zealand shown highest relative efficiency while in asian countries japan was relative efficient country in public expenditure. further they concluded that by increasing the share of private spending, government could increase the efficiency. asandului et al. (2014) calculated the efficiency of healthcare system in 30 european countries for the year 2010. they used non-parametric dea method for the analysis. out of 30 countries only six countries were efficient. similarly in another study samut & cafri (2016) evaluated the efficiency in health sector by calculating efficiency of hospitals across 29 oecd countries during 2000 and 2010. afonso & kazemi (2017) analysed both input and output oriented efficiency in public spending among 20 oecd countries by using data for the period 2009-2013. mandl et al. (2008) examined the variation in public spending specially in education and research & development (r&d) sector and efficiency in these spending among 27 countries. similarly, aristovnik (2012) analysed the impact of information and communication technology (ict) on performance of education sector and their efficiency in european and oecd countries. by implementing dea analysis, the findings reveal that there is a significant variation across countries in ict efficiency. studies are also available that examined efficiency in health and/or education sector in developing countries. jimenez & lockheed (1995) compared the relative efficiency in public and private education at secondary level in five selected developing countries. using dea analysis, the results confirmed that efficiency of private schools is higher than public schools in relative terms. pang & herrera (2005) also examined the efficiency in public expenditure with special focus on health and education across 140 developing countries as a whole and region wise separately for the time period 1996 to 2002. androniceanu & ohanyan (2016) analysed the effect of international monetary fund (imf) financial assistance program on comparative performance of education and healthcare system in two developing countries i.e. romania and bulgaria. grigoli & kapsoli (2018) calculated the efficiency in health expenditure in 80 developing and emerging countries during 2001-2010. a comprehensive study on efficiency in health and education sectors along with infrastructure sector was conducted by herrera & ouedraogo (2018) for 175 countries for the period 2006-2016 by using non-parametric techniques. some studies are also available that compared the cross countries efficiency in higher education sector. agasisti & pérez-esparrells (2010) compared efficiency of 57 italian and 46 spanish universities in two time periods 2000-01 and 2004-05. for the comparison of efficiency in higher education, they used dea. in overall, the efficiency of spanish universities is higher than italian universities. agasisti & johnes (2009) evaluated the journal of applied economics and business studies, volume. 3, issue 1 (2019) 41-60 https://doi.org/10.34260/jaebs.313 45 efficiency in higher education across italian and english universities during four year from 2001 to 2004. joumady & ris (2005) examined the performance of higher educational institution in eight european countries during 1994/95 by using dea analysis. they considered 209 institutions of different characteristics and examined their technical efficiency. 3. material and methods 3.1 theoretical framework the pure theory of public expenditure presented by samuelson (1954) described that there are two types of economic quantities; one is called input and the other is called output. everyone always wants to maximize the output while inputs/factors are those economic quantities which everyone always wants to minimize. furthermore, the theory also stated that there is lack of efficiency in providing the public goods to the people due to not existing of market price for public goods. however, the efficiency in public expenditure could be measured that how much benefit government provided to their masses by using the revenue resources. we compare the cost and welfare of activities to assess how beneficial is government intervention while providing public goods. a state is said to be more efficient when it provides potentially possible maximum welfares to their residents from the use of given inputs (i.e. taxes) or if the public cost is lower than the welfare of an activity for the general masses. the link between government input(s) and output(s) and the efficiency in this production process is given in figure-1. a government uses different inputs (monetary and non-monetary) to produce an output. the choice of input and output variable is based on mdgs targets and past literature (afonso & aubyn, 2006; asandului et al., 2014; mandl et al., 2008; gupta & verhoeven, 2001; jimenez & lockheed, 1995 etc.). the inputs for this study are public expenditures for basic education and health care services and physical capital in health and education sectors while the output is any facility or welfare gained by the masses i.e. improvement in education and health. here, in this process the efficiency is measured by the input/output ratio. small value of the ratio means that less amount of a set of inputs is required to produce given level of output. following figure 1 offer interconnection between factor inputs and outputs. sami ullah khan, zahoor khan & gulnaz hameed 46 figure 1: theoretical framework for the efficiency in production process8 3.2 dea framework the data envelopment analysis (dea) model was first proposed and developed by farrell (1957) and charnes et al. (1978). the dea model is used to measure the efficiency (relative efficiency) of different organizations or production units. the organizations or production units are called decision making units (dmus). the main function of the dmus is to employ a specific combination of inputs to produce an efficient output or a set of efficient outputs. the mathematical form of the dea model can be explained for each dmu and presented in output-input ratio by giving specific weight. 𝐸𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑐𝑦 𝑆𝑐𝑜𝑟𝑒 = 𝑊𝑒𝑖𝑔ℎ𝑡𝑒𝑑 𝑂𝑢𝑡𝑝𝑢𝑡(𝑠) 𝑊𝑒𝑖𝑔ℎ𝑡𝑒𝑑 𝐼𝑛𝑝𝑢𝑡(𝑠) ⁄ (1) a dmu is said to be relatively efficient if the efficiency score in equal to one, and relative inefficient if the score is otherwise. 8 see khan and sulaiman, 2016. education related inputs (public expenditure on education and teachers) health related inputs (public expenditure on health and skilled staff) education related outputs (enrollment and survival at primary) health related outputs (infant survival and child survival per annum) government policies and regulations and other exogenous factors type of dmu production process (scale of operation, type of technology, managerial and technical proficiency) journal of applied economics and business studies, volume. 3, issue 1 (2019) 41-60 https://doi.org/10.34260/jaebs.313 47 3.2.1 the input oriented dea model the dea analysis is based on linear programming. for n dmus, 𝐷1, 𝐷2, 𝐷3, 𝐷4, … … … 𝐷𝑛, the input oriented efficiency under crs and vrs specification is defined by the given minimization problem as under; 𝑴𝒊𝒏 ∅ − 𝜺(∑ 𝒔𝒍𝒋 − +𝒎𝒋=𝟏 ∑ 𝒔𝒍𝒌 +𝒍 𝒌=𝟏 ) (2) 𝑪𝒐𝒏𝒔𝒕𝒓𝒂𝒊𝒏 𝑭𝒖𝒏𝒄𝒕𝒊𝒐𝒏 𝒇𝒐𝒓 𝑪𝑹𝑺 { ∑ 𝜌𝑖 𝐼𝑡𝑖𝑗 + 𝑠𝑙𝑗 −𝑛 𝑖=1 = ∅𝐼𝑇 𝑗 = 1,2, … . , 𝑚 ∑ 𝜌𝑖 𝑂𝑡𝑘𝑗 − 𝑠𝑙𝑘 +𝑛 𝑖=1 = 𝑂𝑇 𝑘 = 1,2, … . , 𝑙 𝜌𝑖 ≥ 0 𝑖 = 1,2, … . , 𝑛 (3) 𝑪𝒐𝒏𝒔𝒕𝒓𝒂𝒊𝒏 𝑭𝒖𝒏𝒄𝒕𝒊𝒐𝒏 𝒇𝒐𝒓 𝑽𝑹𝑺 { ∑ 𝜌𝑖 𝐼𝑡𝑖𝑗 + 𝑠𝑙𝑗 −𝑛 𝑖=1 = ∅𝐼𝑇 𝑗 = 1,2, … . , 𝑚 ∑ 𝜌𝑖 𝑂𝑡𝑘𝑗 − 𝑠𝑙𝑘 +𝑛 𝑖=1 = 𝑂𝑇 𝑘 = 1,2, … . , 𝑙 ∑ 𝜌𝑖 = 𝑛 𝑖=1 1 𝑖 = 1,2, … . , 𝑛 (4) in equation (2), (3) and (4), ∅is the efficiency score measure the technical efficiency of ith dmu, 𝐼𝑇 is the vector of inputs and 𝑂𝑇 is the vector of outputs, 𝑠𝑙𝑗 − and 𝑠𝑙𝑘 + are the slacks of inputs and outputs respectively. in case of input oriented efficiency, ∅ ≤ 1. a dmu is said to be inefficient and lie inside the frontier, if the value of ∅ is less than one. alternatively, a dmu is said to be efficient if the value of ∅ is equal to one and hence the dmu will be lie on production frontier curve. 𝜌is the measure of weight given to ith dmu and give location to ith dmu according to its efficiency score. in input oriented efficiency model the objective is to achieve the fixed level of output by utilizing the possible minimum level of inputs. 3.2.2 supper efficiency after calculating overall technical, pure technical and scale efficiency for the year 2000 and 2015, we also calculated super efficiency by relaxing the assumption of maximum score of 100 percent (equal to one) for efficient dmus. super efficiency score enable us to rank dmus according to their efficiency score. the study calculated input-oriented crs supper efficiency model which is first time proposed by andersen & petersen (1993). this model rank efficient dmus under crs specification of dea model. the study also calculates supper efficiency for input oriented vrs dea model. this model allows us to rank pure technically efficient dmus after allowing the constraint of maximum score beyond one (100 percent). the objective and constraint functions under crs and vrs based input oriented supper efficiency is calculated by the following problem. 𝑴𝒊𝒏 ∅𝑺 (5) sami ullah khan, zahoor khan & gulnaz hameed 48 𝑪𝒐𝒏𝒔𝒕𝒓𝒂𝒊𝒏 𝑭𝒖𝒏𝒄𝒕𝒊𝒐𝒏 𝒇𝒐𝒓 𝑪𝑹𝑺 { ∑ 𝜌𝑖 𝐼𝑡𝑖𝑗 𝑛 𝑖=1 ≤ ∅ 𝑠𝐼𝑇 𝑗 = 1,2, … . , 𝑚 ∑ 𝜌𝑖 𝑂𝑡𝑘𝑗 𝑛 𝑗=1 ≥ 𝑂𝑇 𝑘 = 1,2, … . , 𝑙 𝜌𝑖 ≥ 0 𝑖 = 1,2, … . , 𝑛 (6) 𝑪𝒐𝒏𝒔𝒕𝒓𝒂𝒊𝒏 𝑭𝒖𝒏𝒄𝒕𝒊𝒐𝒏 𝒇𝒐𝒓 𝑽𝑹𝑺 { ∑ 𝜌𝑖 𝐼𝑡𝑖𝑗 𝑛 𝑖=1 ≤ ∅ 𝑠𝐼𝑇 𝑗 = 1,2, … . , 𝑚 ∑ 𝜌𝑖 𝑂𝑡𝑘𝑗 𝑛 𝑗=1 ≥ 𝑂𝑇 𝑘 = 1,2, … . , 𝑙 ∑ 𝜌𝑖 = 𝑛 𝑖=1 1 𝑖 = 1,2, … . , 𝑛 (7) in equations (5), (6) and (7) ∅𝑠 are the supper efficiency scores associated to each dmu. 3.2.3 returns to scale next, to investigate returns to scale (rts) in production process of each dmu, the study compare the efficiency score under crs and vrs specifications. the returns to scale is calculated by taking the ratio of crs based efficiency score and vrs based efficiency score for each dmu and the score are them compared with score obtained from nonincreasing return to scale. finally, the following conclusion is drawn; i. first, crs/vrs efficiency scores are obtained. if the ratio is equal to one, the dmu will follow crts and the dmu size is optimal. ii. if the ratio is less than one, the dmu may face a disadvantage of its operation (the dmu size are either too small or too big). 3.3 malmquist productivity index (mpi) malmquist productivity index (mpi) is an important technique in dea framework which has been used for productivity measurement of a dmu over time based on technology of base year. the total productivity change of a dmu from one period to other period can be decomposed into two parts; firstly, it may be due to change in efficiency (real change in productivity) and secondly, it may be due to shift in efficient frontier because of change in technology. let us consider to calculate mpi across two time periods i.e. 𝑡 & 𝑡 + 1. 𝑀𝑃𝐼𝑖 𝑡 = 𝐷𝑖 𝑡(𝐼𝑡𝑡+1,𝑂𝑡𝑡+1) 𝐷𝑖 𝑡(𝐼𝑡𝑡,𝑂𝑡𝑡) (8) 𝑀𝑃𝐼𝑖 𝑡+1 = 𝐷𝑖 𝑡+1(𝐼𝑡𝑡+1,𝑂𝑡𝑡+1) 𝐷𝑖 𝑡+1(𝐼𝑡𝑡,𝑂𝑡𝑡) (9) 𝑀𝑃𝐼𝑖 𝐺.𝑀𝑒𝑎𝑛 = { 𝐷𝑖 𝑡+1(𝐼𝑡𝑡+1,𝑂𝑡𝑡+1) 𝐷𝑖 𝑡(𝐼𝑡𝑡,𝑂𝑡𝑡) } × [{ 𝐷𝑖 𝑡(𝐼𝑡𝑡,𝑂𝑡𝑡) 𝐷𝑖 𝑡+1(𝐼𝑡𝑡,𝑂𝑡𝑡) } . { 𝐷𝑖 𝑡(𝐼𝑡𝑡+1,𝑂𝑡𝑡+1) 𝐷𝑖 𝑡+1(𝐼𝑡𝑡+1,𝑂𝑡𝑡+1) }] 1/2 (10) here, the subscript 𝑖 shows that we are calculating mpi based on input oriented dea. 𝐼𝑡 𝑎𝑛𝑑 𝑂𝑡 are inputs and outputs vectors respectively while 𝐷𝑖 𝑡 𝑎𝑛𝑑 𝐷𝑖 𝑡+1 show the distance function at two different time periods. in the above equation (10), the first part is called journal of applied economics and business studies, volume. 3, issue 1 (2019) 41-60 https://doi.org/10.34260/jaebs.313 49 change in efficiency across time periods while the second part is called change in technology or shift in frontier. 3.4 data the study used secondary data for analysis for two different time periods; 2000 and 2015. the selection of these two time periods is due to mdgs as these goals were set in 2000 and completed in 2015. the countries (dmus) were selected by the world bank on the basis of income per capita. the main sources of data are world development indicator (wdi), millennium development goals, the world bank; united nation mdgs database. table 1: input and output variables with name, definition and unit of measurement and reference source of selection input/ output variable description/operation definition unit reason for selection input educational expenditure total public educational expenditure measured in current us $. usd gupta & verhoeven (2001) input teachers at primary level total teachers employed at primary level. numbers afonso & aubyn (2005) input health expenditure total public sector expenditure measured in current us $. usd gupta & verhoeven (2001) input birth attended by skilled staff total number of children attended by skilled staff during birth. numbers mdgs target 5.2 output net enrollment rate at primary total number of children enrolled at primary level numbers mdgs goal 2.1, gupta & verhoeven, 2001 output completed primary total number of enrolled students who completed primary level education numbers mdgs target 2.2, hauner & kyobe (2010) output infant survival total number of children survived under the age of one year. numbers mdgs target 4.2, afonso & aubyn (2005) output child survival total number of children survived under the age of five years. numbers mdgs target 4.1, afonso & aubyn (2005) 4. results and discussion 4.1 efficiency analysis for the year 2000 the results of dea model for the year 2000 is presented in table-2 below. the first column of the table describes name of a country (dmu). column 2 and 3 present overall technical efficiency scores of countries and their ranking on the basis of crs specification sami ullah khan, zahoor khan & gulnaz hameed 50 respectively. the crs results show that out of 64 countries, only ten (10) countries are relative efficient having score one and ranked as one in order (shown as bold). the ten efficient dmus in public health and education sector under crs are albania, argentina, belarus, bhutan, cameroon, ghana, macedonia, pakistan, south africa and vanuatu. in the remaining inefficient dmus, costa and cuba are only 2 percent inefficient (score=0.98) while iran and peru are the most inefficient dmus having efficiency scores 0.30 and 0.28 respectively. it means that for costa and cuba there is a need of only two percent improvement to reach to the desirable efficient level while iran and peru need almost 70 percent improvement (70 percent inefficient) in the usage of input resources to become efficient while targeting selected mdgs. the crs model shows technical efficiency mixed with scale efficiency. to calculate the pure technical efficiency, we run the dea model under variable returns to scale (vrs).the efficiency score and raking of countries based on vrs specification is presented in column 4 and five. the vrs efficiency gives the pure technical efficiency of a dmu that how efficiently they convert inputs into desirable outputs. in our case, out of 64 countries 29 countries are relative efficient having score equal to one and ranked as one. it means that almost half of dmus are efficient technically. panama and cuba are only one and two percent technically inefficient (99 and 98 percent efficient) respectively. technically, most inefficient dmus in education and health sector for the year 2000 are vietnam, indonesia and kiribati. column six gives the returns to scale (rts) information of countries in health and education sector. rts specification gives information about the operating size of dmu and show that the inefficiency is either due to small size or large size. a country will be operating at optimal size if rts is constant and will be operating at small size if rts is increasing and will be operating at large size if rts is decreasing. the results show that ten (10) relative efficient dmus are operating at optimal level showing the rts as constant. only two countries have shown increasing rts which means that they are inefficient in term of scale efficiency due to small size of health and education sector and can achieve the optimal size of operation and become efficient by increasing the size of health and education sector. the other remaining dmus are inefficient in term of scale efficiency because they are operating above the optimal level and shown decreasing rts. the crs and vrs efficiency model has the assumption that a dmu can get a maximum score of one (100 percent) and is bound between 0 and 1. now by relaxing this assumption, we can rank the efficient dmus as it allows the efficient dmus to get score above one. for this, the study also calculates supper efficiency under crs and vrs specification. supper efficiency is used to rank the efficient dmus. the scores of inefficient dmus remains the same as they are already getting score below one. under crs specification supper efficiency score indicate that within efficient dmus, vanuatu is ranked at number one because the journal of applied economics and business studies, volume. 3, issue 1 (2019) 41-60 https://doi.org/10.34260/jaebs.313 51 supper efficiency score is 3.44 which is maximum among all the dmus’ score. bhutan get second highest score under crs supper efficiency and ranked at number two while ghana is at number 10 and get minimum score among efficient dmus. under vrs supper efficiency specification, we see that some countries are ranked as infeasible and ranked as one. the meaning of infeasible score under vrs supper efficiency is that the scores of these dmus are extremely large in number and they are the most efficient dmus. therefore, we give top ranking of one to all those five dmus which marked as “infeasible” under vrs supper efficiency model. 4.2 efficiency analysis for the year 2015 the results of relative efficiency calculated by dea model under different specifications for the year 2015 is presented in table-3. total numbers of selected dmus for the year 2015 are 42 selected middle income countries. the result of crs model is presented in column two. crs efficiency score gives the overall relative efficiency (both technical and scale combined) of a dmu under the assumption that all dmus are operating at efficient frontier. the crs results show that out of 42 countries, eight (08) countries i.e. belarus, domenici, ecuador, pakistan, panama, sao tome, vincent & grenadines and fiji are relative efficient having score one and ranked as number one in order. in the remaining inefficient dmus, maldives, serbia and peru are only one percent inefficient (99 percent efficient) and grenada, macedonia, tajikistan and romania are only two percent inefficient (98 percent efficient). total 28 countries get score above or equal to 0.90 and show 90 percent or above efficiency level. the most inefficient dmus are mexico, china and brazil respectively. to calculate the pure technical efficiency, we run the dea model under variable returns to scale (vrs) specification. the efficiency score and raking of countries based on vrs specification is presented in column 4 and 5. the vrs efficiency gives the pure technical efficiency of a dmu that how efficiently they convert inputs into outputs. in our case, out of 42 countries 17 countries are relative efficient in education and health having score equal to one and ranked as one. a total of 38 dmus scored above or equal to 90 percent. the most inefficient dmu under vrs in health and education sector is also brazil having score of 0.71. returns to scale (rts) information of countries in health and education sector are presented in column 6. rts specification gives information about the operating size of dmu and show that inefficiency is either due to small size or large size of a dmu. the results show that seven (07) relative efficient dmus are operating at optimal level showing the rts as constant. fourteen (14) countries have shown increasing rts which means that they are inefficient in term of scale efficiency due to small size of health and education sector and can achieve the optimal size of operation and become efficient by increasing the size of sami ullah khan, zahoor khan & gulnaz hameed 52 health and education sector. the remaining dmus are inefficient in term of scale efficiency because they are operating above the optimal level and shown decreasing rts. sao tome is ranked at number one because the supper efficiency score is 4.47 which is maximum among all dmus’ scores. belarus and domenici get second and third highest score under crs supper efficiency and ranked as number two and three respectively while fiji is at number 08 and get minimum score among efficient dmus. under vrs supper efficiency specification, we see that some countries are marked as infeasible and ranked as one. the meaning of infeasible score under vrs supper efficiency is that the scores of these dmus are extremely large and they are most efficient dmus. therefore, we give top ranking of one to all these six dmus which marked as “infeasible” under vrs supper efficiency model. the minimum scores of efficient dmus under vrs supper efficiency specification are assigned to indonesia, panama and tajikistan. 4.3 malmquist productivity index (mpi) analysis change in productivity, over the time from 2000 (base period) to 2015 (current period), in health and education sector has been assessed via malmquist productivity index (mpi). the mpi shows change in efficiency from one period to other period along with sources of change in efficiency. thirty three (33) middle income countries based on their classification as middle income countries in 2000 and 2015 were selected for mpi analysis and the results are presented in table-4. the first column present the names of dmus, the score of mpi based on crs dea specification is presented in column two while column three and four presents the sources of change in efficiency i.e. change in efficiency and change in frontier respectively. the malmquist index score show that out of 33 dmus, sixteen have scored less than one while seventeen dmus reported score greater than 1. it means that sixteen dmus reported decline in change in efficiency from period 2000 to 2015 while seventeen dmus reported improvement in efficiency from base period to the current period. the efficiency from one period to another period changed because of two reasons; improvement in technical efficiency of dmus or because of change in technology (frontier shift). if the score of change in efficiency is less than, equal to or greater than one it means that real efficiency increased, remains constant or decreased respectively from 2000 to 2015. the analysis shows that for twenty five dmus ‘change in efficiency’ score is observed less than one which indicate that real technical productivity of these dmus has improved over time. the change in efficiency score of three dmus are equal to one it means that the real efficiency of these dmus did not change over time. the remaining five dmus efficiency has been decreased over time (see column 3). the result of shift in frontier indicates that twenty two dmus reported a positive shift in dea frontier across the period while eleven dmus reported decline in dea frontier across the periods. for albania, azerbaijan, cuba kyrgyz, macedonia, pakistan and south africa the source of increase in productivity was technological change during the period. brazil, bulgaria, indonesia, iran, other countries journal of applied economics and business studies, volume. 3, issue 1 (2019) 41-60 https://doi.org/10.34260/jaebs.313 53 like mexico, panama, russia, thailand, uzbekistan and vietnam reported positive change in productivity due to real technical efficiency change. the reason of decline in productivity in belarus, belize, domenici, dominican republic, ecuador, fiji, georgia, kazakhstan, mongolia, romania and sri lanka was change in efficiency frontier (technological change) across the periods. sami ullah khan, zahoor khan & gulnaz hameed 54 table 2: results of dea model under different specifications for the year 2000 country crs specification vrs specification returns to scale (rts) supper efficiency under crs supper efficiency under vrs score rank score rank score rank score rank albania 1.00 1 1.00 1 constant 1.29 7 20.46 3 algeria 0.93 4 1.00 1 decreasing 0.93 21.27 2 argentina 1.00 1 1.00 1 constant 1.64 5 1.67 11 armenia 0.92 5 0.94 4 decreasing 0.92 --0.94 -- azerbaijan 0.95 3 1.00 1 decreasing 0.95 --6.49 4 belarus 1.00 1 1.00 1 constant 3.20 3 3.20 7 belize 0.70 14 0.86 12 decreasing 0.70 --0.86 -- bhutan 1.00 1 1.00 1 constant 3.27 2 5.07 5 botswana 0.91 6 0.91 7 increasing 0.91 --0.91 -- brazil 0.65 17 0.87 11 decreasing 0.65 --0.87 -- bulgaria 0.64 18 0.86 12 decreasing 0.64 --0.86 -- cabo 0.73 12 0.74 18 decreasing 0.73 --0.74 -- cambodia 0.43 34 1.00 1 decreasing 0.43 --1.18 19 cameroon 1.00 1 1.00 1 constant 1.20 9 1.21 17 colombia 0.67 16 0.79 16 decreasing 0.67 --0.79 -- costa 0.98 2 1.00 1 decreasing 0.98 --1.04 22 cote 0.46 31 0.80 15 decreasing 0.46 --0.80 -- cuba 0.98 2 0.98 3 increasing 0.98 --0.98 -- dominica 0.61 21 0.89 9 decreasing 0.61 --0.89 -- dominican republic 0.45 32 0.87 11 decreasing 0.45 --0.87 -- ecuador 0.34 40 0.90 8 decreasing 0.34 --0.90 -- egypt 0.67 16 1.00 1 decreasing 0.67 --1.01 23 fiji 0.48 29 1.00 1 decreasing 0.48 --2.60 9 georgia 0.57 24 1.00 1 decreasing 0.57 --3.25 6 ghana 1.00 1 1.00 1 constant 1.09 10 1.16 20 guatemala 0.40 36 0.89 9 decreasing 0.40 --0.89 -- guyana 0.45 32 0.82 13 decreasing 0.45 --0.82 -- india 0.87 8 0.89 9 decreasing 0.87 --0.89 -- indonesia 0.61 21 0.65 20 decreasing 0.61 --0.65 -- iran 0.30 42 0.93 5 decreasing 0.30 --0.93 -- jordan 0.71 13 1.00 1 decreasing 0.71 --1.21 18 kazakhstan 0.47 30 0.92 6 decreasing 0.47 --0.92 -- kiribati 0.52 27 0.54 21 decreasing 0.52 --0.54 -- kyrgyz 0.84 9 1.00 1 decreasing 0.84 --1.01 24 lao 0.50 28 1.00 1 decreasing 0.50 --1.51 13 lebanon 0.33 41 0.94 4 decreasing 0.33 --0.94 -- lesotho 0.58 23 0.75 17 decreasing 0.58 --0.75 -- macedonia 1.00 1 1.00 1 constant 2.22 4 2.87 8 journal of applied economics and business studies, volume. 3, issue 1 (2019) 41-60 https://doi.org/10.34260/jaebs.313 55 mauritania 0.48 29 0.91 7 decreasing 0.48 --0.91 -- mauritius 0.38 37 0.80 15 decreasing 0.38 --0.80 -- mexico 0.48 29 1.00 1 decreasing 0.48 --1.43 14 moldova 0.36 39 1.00 1 decreasing 0.36 --infeasible 1 mongolia 0.68 15 1.00 1 decreasing 0.68 --1.36 15 myanmar 0.37 38 0.88 10 decreasing 0.37 --0.88 -- namibia 0.45 32 0.82 13 decreasing 0.45 --0.82 -- nicaragua 0.41 35 0.91 7 decreasing 0.41 --0.91 -- nigeria 0.45 32 1.00 1 decreasing 0.45 --1.10 21 pakistan 1.00 1 1.00 1 constant 1.27 8 1.28 16 panama 0.63 19 0.99 2 decreasing 0.63 --0.99 -- peru 0.28 43 1.00 1 decreasing 0.28 --infeasible 1 philippines 0.59 22 0.86 12 decreasing 0.59 --0.86 -- romania 0.89 7 1.00 1 decreasing 0.89 --1.93 10 russia 0.81 10 1.00 1 decreasing 0.81 --1.01 25 south africa 1.00 1 1.00 1 constant 1.49 6 1.55 12 sri lanka 0.56 25 0.86 12 decreasing 0.56 --0.86 -- tajikistan 0.58 23 0.90 8 decreasing 0.58 --0.90 -- thailand 0.55 26 1.00 1 decreasing 0.55 --infeasible 1 tonga 0.46 31 0.82 13 decreasing 0.46 --0.82 -- tunisia 0.75 11 0.82 13 decreasing 0.75 --0.82 -- ukraine 0.40 36 1.00 1 decreasing 0.40 --infeasible 1 uzbekistan 0.44 33 0.80 15 decreasing 0.44 --0.80 -- vanuatu 1.00 1 1.00 1 constant 3.44 1 infeasible 1 vietnam 0.57 24 0.68 19 decreasing 0.57 --0.68 -- zambia 0.62 20 0.81 14 decreasing 0.62 --0.81 -- sami ullah khan, zahoor khan & gulnaz hameed 56 table 3: results of dea model under different specifications for the year 2015 country crs specification vrs specification returns to scale (rts) supper efficiency under crs supper efficiency under vrs score rank score rank score rank score rank albania 0.87 14 0.92 8 increasing 0.87 --0.92 -- azerbaijan 0.86 15 0.90 10 increasing 0.86 --0.90 -- belarus 1.00 1 1.00 1 constant 1.61 2 2.16 5 belize 0.92 8 0.95 5 increasing 0.92 --0.95 -- brazil 0.66 22 0.71 13 decreasing 0.66 --0.71 -- bulgaria 0.89 13 0.92 8 increasing 0.89 --0.92 -- china 0.72 21 0.84 12 decreasing 0.72 --0.84 -- colombia 0.84 16 0.84 12 increasing 0.84 --0.84 -- cuba 0.96 4 0.99 2 decreasing 0.96 --0.99 -- dominica 1.00 1 1.00 1 constant 1.59 3 1.60 7 dominican republic 0.92 8 0.93 7 increasing 0.92 --0.93 -- ecuador 1.00 1 1.00 1 constant 1.48 5 infeasible 1 el salvador 0.90 10 0.91 9 increasing 0.90 --0.91 -- fiji 1.00 1 1.00 1 decreasing 1.00 8 infeasible 1 georgia 0.94 6 0.95 5 increasing 0.94 --0.95 -- grenada 0.98 3 0.99 2 increasing 0.98 --0.99 -- indonesia 0.82 18 1.00 1 decreasing 0.82 --1.01 10 iran 0.79 19 0.88 11 decreasing 0.79 --0.88 -- kazakhstan 0.94 6 0.94 6 decreasing 0.94 --0.94 -- kyrgyz 0.95 5 0.95 5 increasing 0.95 --0.95 -- macedonia 0.98 3 0.98 3 increasing 0.98 --0.98 -- malaysia 0.83 17 0.90 10 decreasing 0.83 --0.90 -- maldives 0.99 2 0.99 2 increasing 0.99 --0.99 -- mexico 0.78 20 0.91 9 decreasing 0.78 --0.91 -- moldova 0.96 4 0.96 4 increasing 0.96 --0.96 -- mongolia 0.96 4 0.96 4 increasing 0.96 --0.96 -- pakistan 1.00 1 1.00 1 constant 1.49 4 1.56 8 panama 1.00 1 1.00 1 constant 1.01 7 1.01 10 peru 0.99 2 1.00 1 decreasing 0.99 --1.04 9 romania 0.98 3 1.00 1 decreasing 0.98 --infeasible 1 russia 0.84 16 0.91 9 decreasing 0.84 --0.91 sao tome 1.00 1 1.00 1 constant 4.74 1 63.07 3 serbia 0.99 2 0.99 2 decreasing 0.99 --0.99 south africa 0.93 7 1.00 1 decreasing 0.93 --infeasible 1 sri lanka 0.91 9 1.00 1 decreasing 0.91 --1.70 6 tajikistan 0.98 3 1.00 1 decreasing 0.98 --1.01 10 thailand 0.82 18 0.94 6 decreasing 0.82 --0.94 turkey 0.89 12 1.00 1 decreasing 0.89 --infeasible 1 ukraine 0.87 14 0.98 3 decreasing 0.87 --0.98 uzbekistan 0.93 7 1.00 1 decreasing 0.93 --4.15 4 vietnam 0.90 11 1.00 1 decreasing 0.90 --infeasible 1 vincent & grenadines 1.00 1 1.00 1 constant 1.03 6 562.47 2 journal of applied economics and business studies, volume. 3, issue 1 (2019) 41-60 https://doi.org/10.34260/jaebs.313 57 table 4: malmquist productivity index (mpi) analysis (2000 is base period and 2015 is current period) country (dmus) based on input-oriented crs change in productivity change in efficiency shift in frontier malmquist index efficiency change frontier shift albania 2.71699 1.09325 2.48523 increased decreased upward azerbaijan 3.40103 1.11239 3.05742 increased decreased upward belarus 0.45250 1.00000 0.45250 decreased constant downward belize 0.38226 0.73439 0.52051 decreased increased downward brazil 1.97190 0.99079 1.99024 increased increased upward bulgaria 1.68722 0.78787 2.14148 increased increased upward colombia 0.92634 0.82876 1.11775 decreased increased upward cuba 2.59667 1.03569 2.50719 increased decreased upward dominica 0.59918 0.64349 0.93113 decreased increased downward dominican republic 0.32780 0.49657 0.66013 decreased increased downward ecuador 0.21486 0.38152 0.56317 decreased increased downward fiji 0.25136 0.50007 0.50266 decreased increased downward georgia 0.46556 0.59935 0.77677 decreased increased downward indonesia 1.79807 0.81386 2.20931 increased increased upward iran 1.26497 0.40580 3.11722 increased increased upward kazakhstan 0.49445 0.57549 0.85918 decreased increased downward kyrgyz 2.29275 1.02357 2.23996 increased decreased upward macedonia 2.33637 1.00918 2.31512 increased constant upward mexico 1.57580 0.69200 2.27719 increased increased upward moldova 0.76436 0.41929 1.82297 decreased increased upward mongolia 0.36647 0.68751 0.53304 decreased increased downward pakistan 1.75203 1.00000 1.75203 increased constant upward panama 1.79264 0.66671 2.68879 increased increased upward peru 0.54764 0.28947 1.89189 decreased increased upward romania 0.51748 0.89894 0.57566 decreased increased downward russia 2.28680 0.98098 2.33115 increased increased upward south africa 4.21227 1.07322 3.92488 increased decreased upward sri lanka 0.65025 0.65482 0.99302 decreased increased downward tajikistan 0.68073 0.60609 1.12314 decreased increased upward thailand 1.78594 0.65546 2.72470 increased increased upward ukraine 0.58997 0.43848 1.34547 decreased increased upward uzbekistan 1.30986 0.51579 2.53950 increased increased upward vietnam 1.18014 0.65504 1.80164 increased increased upward sami ullah khan, zahoor khan & gulnaz hameed 58 5. conclusion and recommendations this study examined the relative efficiency in public education and health sector in selected middle income countries regarding mdgs targets for two different time periods; 2000 and 2015. educational expenditure, teachers at primary level, heath expenditure, birth attended by skilled staff are used as inputs; enrollment at primary level, survival at primary level, infant and child survival per annum are used as outputs. dea technique has been utilized to measure efficiency under crs, vrs and supper efficiency specification to calculate overall relative efficiency, pure technical efficiency and ranking of efficient dmus. returns to scale of dmus are also calculated to observe the size of dmu while mpi technique is used to investigate the change in productivity and sources of change in productivity across the two periods. the results of the study indicate that the level of efficiency is different among dmus under different dea specification in both time periods. the result is mixed; most of the countries show technical inefficiency in achieving the mdgs targets related to health and education sector. some countries are inefficient because there size of operation is too big or too small. some countries shown improvement in productivity and efficiency over time either because of improvement in technical efficiency or technological improvement. the results of this study suggest that there is a vast scope for further improvement both in education and health sectors. countries could increase their efficiency both technically and scale wise. to achieve the targets of mdgs, particularly health and education related targets; countries need to increase resources for health and education sectors. there is also need to utilize these resources efficiently and optimally. there is a great potential for inefficient countries to improve their performance in health and education sectors with given resources. the results suggest that inefficiency is observed in both; technical and scale related aspects. the current study has some limitations and there exist space for further detail investigation of phenomenon in future. the study evaluated efficiency for two time periods only and one could study it for multi time periods and analyze it through panel data approach. various economic, social, environmental determinants could be investigated for the source of efficiency/inefficiency in health and education sector. there is also need of efficiency analysis of both sectors separately. the study may also be conducted for a large sample of data by including lower income countries. journal of applied economics and business studies, volume. 3, issue 1 (2019) 41-60 https://doi.org/10.34260/jaebs.313 59 references afonso, a., & aubyn, m. s. (2006). cross-country efficiency of secondary education provision: a semi-parametric analysis with non-discretionary inputs. economic modelling, 23(3), 476-491. afonso, a., & aubyn, m. s. (2005). non-parametric approaches to education and health efficiency in oecd countries. journal of applied economics, 8(2), 227-246. afonso, a., & kazemi, m. (2017). assessing public spending efficiency in 20 oecd countries.inequality and finance in macrodynamics, 23(2017), 7-42. afonso, a., schuknecht, l., & tanzi, v. (2005). public sector efficiency: an international comparison. public choice, 123(3-4), 321-347. afonso, a., schuknecht, l., & tanzi, v. (2010). public sector efficiency: evidence for new eu member states and emerging markets. applied economics, 42(17), 2147-2164. agasisti, t., & johnes, g. (2009). beyond frontiers: comparing the efficiency of higher education decision‐making units across more than one country. education economics, 17(1), 59-79. agasisti, t., & pérez-esparrells, c. (2010). comparing efficiency in a cross-country perspective: the case of italian and spanish state universities. higher education, 59(1), 85-103. andersen, p., & petersen, n. c. (1993). a procedure for ranking efficient units in data envelopment analysis. management science, 39(10), 1261-1264. androniceanu, a., & ohanyan, g. (2016). comparative approach on education and healthcare in romania and bulgaria as beneficiaries of the imf financial assistance. administratie si management public, (26), 25-48. aristovnik, a. (2012). the impact of ict on educational performance and its efficiency in selected eu and oecd countries: a non-parametric analysis. turkish online journal of educational technology, 11(3), 144-152. asandului, l., roman, m., & fatulescu, p. (2014). the efficiency of healthcare systems in europe: a data envelopment analysis approach. procedia economics and finance, 10, 261-268. charnes, a., cooper, w. w., & rhodes, e. (1978). measuring the efficiency of decision making units. european journal of operational research, 2(6), 429-444. farrell, m. j. (1957). the measurement of productive efficiency. journal of the royal statistical society. series a (general), 120(3), 253-290. sami ullah khan, zahoor khan & gulnaz hameed 60 grigoli, f., & kapsoli, j. (2018). waste not, want not: the efficiency of health expenditure in emerging and developing economies. review of development economics, 22(1), 384-403. gupta, s., & verhoeven, m. (2001). the efficiency of government expenditure: experiences from africa. journal of policy modeling, 23(4), 433-467. herrera, s., & ouedraogo, a. (2018). efficiency of public spending in education, health, and infrastructure: an international benchmarking exercise: the world bank. hauner, d., & kyobe, a. (2010). determinants of government efficiency. world development, 38(11), 1527-1542. jimenez, e., & lockheed, m. e. (1995). public and private secondary education in developing countries: a comparative study (vol. 309): world bank publications. joumady, o., & ris, c. (2005). performance in european higher education: a non‐ parametric production frontier approach. education economics, 13(2), 189-205. khan, z., & sulaiman, j. (2015). social and financial efficiency of microfinance institutions in pakistan. the pakistan development review, 54(4i-ii), 389-403. mandl, u., dierx, a., & ilzkovitz, f. (2008). the effectiveness and efficiency of public spending: directorate general economic and financial affairs (dg ecfin), european commission. nations, u. (2000). resolution adopted by the general assembly [without reference to a main committee (a/55/l. 2)], 55/2. united nations millennium declaration. pakistan economic survey. (2015-16). ministry of finance; islamabad, pakistan. pang, g., & herrera, s. (2005). efficiency of public spending in developing countries: an efficiency frontier approach. world bank policy research working paper(3645). robbins, l. (1932). essay on the nature and significance of economic science. samuelson, p. a. (1954). the pure theory of public expenditure. the review of economics and statistics, 387-389. samut, p. k., & cafrı, r. (2016). analysis of the efficiency determinants of health systems in oecd countries by dea and panel tobit. social indicators research, 129(1), 113-132. ul haq, m. (1995). reflections on human development: oxford university press. wang, e. c., & alvi, e. (2011). relative efficiency of government spending and its determinants: evidence from east asian countries. eurasian economic review, 1(1), 3-28. journal of applied economics and business studies, volume. 5, issue 2 (2021) 17-34 https://doi.org/10.34260/jaebs.522 17 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x impact of institutional quality on bilateral exports: exploring the role of development muhammad zubair chishti1*, babar hussain2, and muhammad aqib khursheed3 1school of business, zhengzhou university, henan, china 2assistant professor, school of economics, international islamic university, islamabad, pakistan 3school of economics, international islamic university islamabad. abstract this study uses the gravity model to analyze the homogeneous and heterogeneous effect of institutional quality and development on bilateral exports. we use the panel data of 61countries for the period 2000 to 2016 and employ the poisson pseudo maximum liklihood (ppml) econometric technique with a high-dimensional fixed effect (hdfe) for an estimation that allows the analysis in the presence of high dimensional fixed effects. the findings reveal that the direct effect of institutional quality and level of development on bilateral exports is positive and significant. further, the institutional quality and the level of development of the exporter country have more impact on bilateral exports than that of the importer country. our estimation results of homogeneity of institutions show that when both trading countries share the same level of institutional quality, it boosts the bilateral exports. the major finding of this study reveals that the interaction effect of institutional quality and level of development on bilateral exports is positive and significant. high value of interaction term of exporter economy and low value of importer country suggest that interaction effect of institutional quality and level of development on bilateral exports of exporter country have a greater impact than the interaction effect of institutional quality and level of development of importer country due to having the more production and exports facilities in exporter country. based on the findings, some essential policies are also recommended, followed by some future research gaps. keywords institutional quality, development, gravity model, poisson pseudo maximum liklihood, highdimensional fixed effect jel classification f1, f63, o43 * chishtimz9@gmail.com; mzchishti@eco.qau.edu.pk mailto:chishtimz9@gmail.com mailto:mzchishti@eco.qau.edu.pk muhammad zubair chishti, babar hussain, and muhammad aqib khursheed 18 1. introduction the economists have a mutual consensus that both trade and institutions are the crucial determinants of growth and income levels. further, based on the scholars’ viewpoint on institutions, trade, and growth, the available literature can be classified into two bunches. first, the recent and past literature exhibits the positive and significant impact of growth on trade (donaldson, 2015; bernhofen and brown, 2005; frankel and romer, 1999). second, the other part of the literature on this topic shows that highquality institutions are the key determinant of economic growth and development. despite the direct impacts of institutions on growth, domestic institutions are also intermediaries in explaining the linkage between trade and growth (shah et al., 2020; pascali and luigi, 2017; florensa et al., 2015; nunn and trefler, 2014). the dramatic rise in the size of trade flows in several emerging market economies raises the question that what factor is essential in determining the international trade flows. however, there is an abundant evidence in the literature that indicates that institutional quality is responsible for this dramatic increase (angkinand and chiu, 2011). if the quality of institutions is strong, it reduces the uncertainty in the international transaction thus, it reduces the transaction cost of trade (hou et al., 2021). if the quality of institution is poor it leads political instability, attract more corruption and the flawed rule of law which increase the uncertainty about expected gain from international transaction thus discourage international trade (bandyopadhayay and roy, 2016 institutional homogeneities are a more important factor than the simple quality of institutions for bilateral trade; thus, the homogeneous quality of institutions positively affects the bilateral trade (islam and reshef 2006; de groot et al. 2004). there is a fascinating reason why we study institutional homogeneity rather than institutional quality; evidently, we want to know that under what conditions institutional homogeneity matters rather than the institutional quality. we extensively knew that there are many countries in the world where institutional quality is simultaneously high, then how can we check the effects of institutional distinctions of these nations on trade flows. however, if the institutional quality of both exporter and importer countries is high yet differs between nations, it is hard to expand the institutional quality more; hence, we use institutional homogeneity. we consider institutional homogeneity and institutional similarity the “same legal origin” from la porta et al. (1997 and 1998). if both exporter and importer countries share the same legal origin, then it is considerd that both countries have the same contractual environment (acemoglu johnson and robinson, 2001; acemoglu and johnson, 2005). the basic idea about legal origins is that countries have distinct1 legal 1 see glaeser and shleifer (2002) for more details and explaination about this distinction of legal origins and their families journal of applied economics and business studies, volume. 5, issue 2 (2021) 17-34 https://doi.org/10.34260/jaebs.522 19 origins, and this legal origin widely matters for financial and economic outcomes (la porta et al., 1997; la porta et al.,1998). beverelli et al. (2018) and bandyopadhayay and roy (2016) argue that heterogeneous institutional quality positively affect bilateral exports. however, dutt and traca (2010) explore that the heterogeneous institutional quality negatively affects the bilateral exports. in general, the effect of institutional quality on bilateral exports varies across different countries. hence, the prime focus of the current study is to check the role of development in homogeneous and heterogeneous impact of institutional quality on bilateral exports. however, this study explains the interaction effect of institutions and development, indicating that domestic institutions are the main source of comparative advantage. within gravity structure and for a complete review on the importace of gravity model in bilateral exports this study is methodologically related to beverelli et al., (2018). in present period exchange of goods and services within counties and across different countries is necessary, without it survival of a country is not possible. institutional quality is also important factor to determine the international trade. in the literature some studies show that quality of institutions affects the bilateral trade positively and significantly, other studies show that there is no linear relationship between institutions and bilateral exports because regional trade agreements and development also affect the bilateral exports (naanwaab and diarrassouba, 2013; bandyopadhayay et al., 2016). however, existing literature does not account for the role of development in the homogeneous effect of institutional quality on bilateral exports. therefore our focus is to check the role of development in homogeneous and heterogeneous impact of institutional quality on bilateral exports. the remainder of the study is organized as follows. in section 2, the literature review is discussed, while section 3 presents the methodology and data. section 4 provides the results and discussion, and section 5 explains the conclusion and policy recommendations. 2. literature review in recent years, the role of institutions has received much attention from not only on the economic perspective to explain the variations in per capita income across different countries but also their impact on international trade flows (kuncic, 2013; francois and manchin, 2007). however institutional quality and trade have positive and significant relationship (do and levchenko, 2009). literature on institutions, development and bilateral trade suggests that institutional quality and development positively affect the bilateral trade (bandyopadhayay and roy, 2016; mattew et al., 2017). moreover large numbers of studies found that there are positive characteristics of innovation; introduction of new varieties, knowledge and cost-decreasing firms, but other side also explains that huge studies recommend that national institutions are key determinant in international trade (beverelli et al., 2018). institutional differences are muhammad zubair chishti, babar hussain, and muhammad aqib khursheed 20 also source of comparative advantage: employ other things the poor countries may not gain from international trade and factor prices may actually deviate as a result of trade. thus institutional changes are more essential determinant of trade flows (levchenko, 2007). the nexus of homogeneous institutions and bilateral trade flows also has been discussed widely in literature during recent times. anderson and marcouillier (2002) are amongst the first contributors to show that bilateral trade flows are positively affected by trading countries which have same quality of institutions. some studies show the negative impact of institutional quality on trade (mendonca et al., 2014), (gilpareja and llorca-vivero, 2017). there are also some other studies which show that there is positive effect of institutional quality on trade (araujo et al. 2016, beverelli et al., 2018) but dutt and traca (2010) show the inconclusive impact of institutions on trade. in the context of relationship between institutions and trade it has shown that there is no linear relationship between institutions and bilateral exports because regional trade agreements and development also affect the bilateral exports (naanwaab and diarrassouba, 2013; bandyopadhayay et al., 2016). so the literature on this topic can be divided into three parts on the basis of connection among these variables. 2.1 institutions and trade abundant seminal and recent literature suggests that there are significant relationship between institutions and trade (nguyen et al., 2018; arif and chishti, 2020; chishti, 2020; chishti, 2021; chishti et al., 2021; ullah, chishti, and majeed, 2020). there are many cross-country variations in which economic and political lifecycle is planned. a huge literature shows the large cross-country deviations in economic institutions and there is robust relationship between these institutions and trade, as mendonca et al. (2014) explore that the higher is the variations in the institutional quality among countries the larger its obstructive impacts on trade. bilateral trade plays very important role in international trade therefore the question is which domestic quality of institution affects the bilateral sectorial trade however, alvarez et al. (2018) answerd that both the institutional situations at destination and the institutional distance between exporting and importing countries are appropriate elements for bilateral trade flows. a pioneer study on institutions and trade by north (1993) find that the better quality of institutions reduces the uncertainty in exchange due to incomplete information and imperfect insight. however better quality of institutions play important role by reducing the transaction costs. 2.2 institutions, development and trade the quality of institutions and development widely matter in trade. institutions with high quality determine the large trade flows and institutions with low quality determine the less trade flows. domestic institutions are main source of high trade flows and high development. however institution affects the trade and development such as beverelli et al. (2018) find that stronger institutions stimulate the trade and journal of applied economics and business studies, volume. 5, issue 2 (2021) 17-34 https://doi.org/10.34260/jaebs.522 21 development. similarly mattew et al. (2017) examine that stronger institutions enhance the trade. however dollar and kraay (2003) investigate that countries with welldeveloped institutes trade more. bojnec et al. (2009) investigate that the effect of level of development on the patterns of bilateral trade is depend on the institutional determinants. institutional quality also measured by corruption so bandyopadhayay and roy (2016) use the interection between corruption and development previously this interection is used by marjit et al. (2014) and discover that the exports of specific product by the rich countries are negatively and significantly influenced by higher corruption of these nations. markusen (2013) use the gdp per-capita as a development variable and examines the role of gdp per-capita in determining trade flows and theoratically and empirically conclude that international trade positively and significantly depends on per-capita income of the exporter country but aggregate income remain constant. 2.3 homogeneous and hetrogeneous instituions and trade in literature seminal studies use the same legal origin as homogeniety of institutions (de groot et al., 2004; 2005). the basic idea about legal origins proposd by (la porta et al., 1997; la porta et al.,1998) is that countries have distinict legal origins and this legal origins widely matter for finacial and economic outcomes. however la porta et al. (1998); la porta et al. (1999) examine the legal rule, origins of this rule and quality of their enforcement covering the protection of corporate creditors and shareholders and finds that civil law nations are weakest and common law nations have strongest legal defense of investor. de groot et al., (2004) examine the impact of quality of institutions on the bilateral trade flows using gravity the model, also extend their analysis to the institutional homogeneity between exporter and importer country and find that institutional quality positively affect the bilateral exports and having homogenous institutional quality between pair of countries boosts the bilateral trade faster. samilary, islam and reshef (2006) explore the institutional quality and institutional similarity and find that quality of institutions and similarity of institutions design promote bilateral trade by decreasing the transaction cost. likewise miura and takechi (2014) examine that if the quality of both trading partners is low then they have to improve their institutional quality for international transaction. on the other side, if the institutional quality of two trading countries is equally high, then institutional homogeneity matter for international trade. in the framework of institutional heterogeneity, bandyopadhayay and roy (2016) used the intrection term between corruption and development to investigate the role of development to explain the hetrogeneous quality of institutions on bilateral exports and find that the exports of specific good by the high income countries are negatively influenced by greater domestic corruption and the corruption of importer country also decreases its exports of these goods but imports are less affected by corruption than exports. marjit et al.(2014) muhammad zubair chishti, babar hussain, and muhammad aqib khursheed 22 also use the interection between the corruption and level of development to explore the affect of isntitutions on trade openess and find that the impact of corruption in trade depends on relative factor abundants. 3. methodology 3.1 theoretical background and empirical model the traditional form of gravity model forecasts bilateral trade flows with respect to the distance and economic sizes between two nations. tinbergen (1962) was the 1st to recommend a gravity equation for international trade flows. this model suggests that trade between two nations is directly related to the gdp and inversely proportional to the distance between these countries. deardorff (1998) derive the traditional heckscher–ohlin model within gravity framework, and show that gravity equation could be constant with numerous trade models. more recent study on gravity model developed by anderson and van wincoop (2003) which explains the gravity model on anderson (1979) to add a applied way for the estimation of gravity equations. the basic point of anderson and van wincoop (2003) is mrts. if we do not account mrts in a gravity model it can leads biased results. hence the traditional structural gravity model can be derived as follow. 𝑋𝑖𝑗 = 𝑇𝑖𝑗 yi 𝐸𝑗 πi 𝑃𝑗 (1) where xi j is bilateral trade from i country (exporting) to j (importing) country, ti j shows any determinant of bilateral trade costs between country i and j (like distance, rta, country-specific trade related drivers like institutions and etc.). yi is total value of production of country i (total sale at home and abroad). e j is expenditure in country j. π i & p j are structural mrts coefficients of the determinants of trade flows. we can also denote multilateral resistance terms as follow: 𝜋𝑖 = ∑ tij ej pj 𝑃𝑗 = ∑ tij ej 𝜋𝑖 (2) the above multilateral resistances have numerous interesting properties: mrts are used to capture the evidence that bilateral trade between two nations are not relies only on the size and distance between them but also show that how these nations are remote from rest of the world. hence anderson and van wincoop (2003) discover that if other things remain same, more multilaterally isolated countries will trade more and more with each other’s. multilateral resistance terms (mrts) is an index which decay the frequency of trade costs on traceable goods if they sell to and on the consumers if they purchase from world market of goods (anderson and yotov, 2010; yotov et al., 2016). however on the basis of gravity model there are three objectives in this section. first we check the impact of homogenous institutional quality on bilateral exports. journal of applied economics and business studies, volume. 5, issue 2 (2021) 17-34 https://doi.org/10.34260/jaebs.522 23 second we study the effect of heterogeneous institutional quality on bilateral exports. third we demonstrate the effect of institutional quality on bilateral exports exploring the role of development. we develop our analysis sequentially. we start our analysis from standard gravity model then we show the impact of institutional quality and development on bilateral trade within gravity model. to achieve our first objective about role of development in homogeneous effects of institutional quality on bilateral exports, we introduce a dummy variable lawij from (miura and takechi, 2014) which takes the value 1 if two countries share the same legal origin and zero for others which captures the homogeniety and non-homogeniety of institutions respectively. 𝑙𝑛𝑥𝑖𝑗𝑡 = 𝐺𝑅𝐴𝑉𝑖𝑗 𝛽 + 𝛽1𝐼𝑄𝑖𝑡 + 𝛽2𝐼𝑄𝑗𝑡 + 𝛽3𝑙𝑛𝑦𝑖𝑡 + 𝛽4𝑙𝑛𝑦𝑗𝑡 + 𝛽5𝐿𝑎𝑤𝑖𝑗 + 𝛽6𝐸𝑅𝑖𝑡 + 𝛽7𝐼𝑄𝑖𝑡 ∗ 𝑙𝑛𝑦𝑖𝑡 + 𝛽8𝐼𝑄𝑗𝑡 ∗ 𝑙𝑛𝑦𝑖𝑡 + 𝜂𝑖𝑗 + 𝜇𝑡 + 𝜀𝑖𝑗𝑡 (3) in order to achieve our second objective of the impact of heterogeneous impact of institutions on bilateral exports by recognizing role of development, we use the following gravity type specification, 𝑙𝑛𝑥𝑖𝑗𝑡 = 𝐺𝑅𝐴𝑉𝑖𝑗 𝛽 + 𝛽1𝐼𝑄𝑖𝑡 + 𝛽2𝐼𝑄𝑗𝑡 + 𝛽3𝑙𝑛𝑦𝑖𝑡 + 𝛽4𝑙𝑛𝑦𝑗𝑡 + 𝛽5𝐸𝑅𝑖𝑡 + 𝛽6𝐼𝑄𝑖𝑡 ∗ 𝑙𝑛𝑦𝑖𝑡 + 𝛽7𝐼𝑄𝑗𝑡 ∗ 𝑙𝑛𝑦𝑖𝑡 + 𝜂𝑖𝑗 + 𝜇𝑡 + 𝜀𝑖𝑗𝑡 (4) equation (3) obtained from equation (1) after log-linearazing and replacing the bilateral trade cost varibale tij with gravij which include all determinant of trade cost. and also included two main variables of our interest institutional qulity and gdp percapita. additionaly we also included law varibale for homogenity of institutions and excahnge rate varibale. however in equation (3) xijt is bilateral trade flow from country i (exporting) to country j (importing) in time t. grav is including all graveity variables: lndistij is logrithem distance in kilomaters between country i and country j, this distance is measured from capital of exporter country to capital of importer country. cntgij is that two countries are sharing the common border or not. if courty i and j speak same officail language(langij). if two countries share any colonial relationships(clnyij). wheter the i and j countries have any rta. iqit is institutional quality of exporting in time t. iqjt is institutional quality of importing country in tim t. lnyit and lnyjt is gdp per capita in us dollar of exporter and importer country respectively, gdp per capita is proxy for economic development. erit is real exchange rate of exporter country in time t which is measurd from local currency to us dollar. ηij is fixed effect which is used to capture the country-pair effects which control the mrts in the gravity model and μt is year fixed effect capturing time-varying effects. the main terms in equation (4) are the interaction term effects between institutional quality and development. however, this approach allow to explore the indirect impact of institutional quality on bilateral exports relies on the level of development (bandyopadhay et al., 2016). muhammad zubair chishti, babar hussain, and muhammad aqib khursheed 24 𝜕 𝑙𝑛 𝑋𝑖𝑗𝑡 𝜕𝐼𝑄𝑖𝑡 = 𝛽1 + 𝛽3𝑙𝑛𝑦𝑖𝑡 (5) 𝜕 𝑙𝑛 𝑋𝑖𝑗𝑡 𝜕𝐼𝑄𝑗𝑡 = 𝛽2 + 𝛽4𝑙𝑛𝑦𝑖𝑡 (6) the effect of institutional quality on bilateral exports of exporter country is determined by β1 likewise, the effect of importer country’s institutional quality on bilateral exports can be determined as shown in above equation (6). on the other hand for importer countries the term β2 dominate and β2 determined the impact of institutional quality on bilateral exports. 3.2 description of variables our estimates are from panel data set of 61 poor and rich countries2 , time period range from 2000 to 2016. our selection of countries and time period is depend upon the availability of consistent data of bilateral exports and institutional quality. due to data availability limitations imposed by the united nations' commodity trade statistics database (uncomtrade) and economic freedom of the world (efw) database for many developing countries in our sample, we choose to deploy the data for the time period 2000 to 2016. hence, we cannot extend our time period other than 2016 due to data availability restriction. further, we select countries on the bases of following criteria: 1. country as a context of study, 2. the countries involved in this sample denotes a significant portion of the whole world. 3. maximum intra-trading partners to avoid zero trade. 4. countries from different regions. for example, intra-trade among european union is round about 70 percent, 52 percent for asian countries, 50 percent for north american regions, 26 percent for south american regions and 18% for african countries (wto, 2015). but we are restricted by data availability of some countries to select accurate percentage of intra-trade flows. our dependent variable is xijt bilateral exports from i exporting country to j importing country which includes the bilateral exports of all commodities from sitc revision 2 (standard international trade classification). the source of bilateral exports data is united nations' commodity trade statistics (uncomtrade) database. our first explanatory variable is gravity variables inclusing: logrithem of bilateral distance (lndistij) which is proxy for bilateral trade cost between two countries (beverelli et al., 2018), if both countries share the same border (cont ij), if the both countries share the same official language (langij), if two countries share any colonial relationships (clnyij). the data source for all gravity variables is cepii’ geo dist database. we also include rta (regional trade agreement) in gravity variable. our rta data is from mario larch's rta database on the basis of egger and larch (2008), finally this data set is based on wto rta. our key interested variable is quality of institution of 2 there are total 61 countries, 20 poor countries and 41 rich countries. poor countries mean those countries that are presented as low income or lower middle income countries. rich countries mean those countries that are presented as upper middle income or high income countries (this classification is made by the world bank). journal of applied economics and business studies, volume. 5, issue 2 (2021) 17-34 https://doi.org/10.34260/jaebs.522 25 exporting country i and importing country j. quality of institution varibale is economic freedom of the world index which is based on “fraser institute for institutions”. the source of this variable is economic freedom of the world (efw) database. efw index is based on five sub-indices which capture the degree of economic freedom. these five indices are as follow: (i) size of government (ii) legal structure and protection of property rights (iii) access to sound money (iv) freedom to trade internationally (v) regulation of business, credit and labor. the range of each index is between 0-10, lower value indicates bad and greater value indicates good quality. our second main variable of interest is development. the proxy for development is real gdp per-capita (bojenec et al., 2009). this variable is taken in us dollars, constant 2010. real exchange rate is also a control variable in this regression. the source of real exchange rate is united nations conferences on trade and development (unctad) database .the real exchange rate is defined here as one local currency unit is equal to one us dollar. however we are unable to use bilateral exchage rate data due to large data set and availability of data. so we use local currency to dollar $ exchange rate data. then we move on interection term between institutional quality and development to check the role of development in effect of instituional quality of bilateral exports, as bandyopadhayay and roy, (2016) use interection term between corrpution index and development in their study. this interection term is construct for exporter country by using institutional quality of exporter and development of exporter, and for importer by using institutional quality of importer and development of importer. we use homogeneous and hetrogeneous institutional quality in this study. to construct this variaable we follow miura and takechi, (2014) they use dummy variable which takes the value 1 for homogeneous institutional quality if two nations are sharing the same legal origin and zero for hetrogeneous institutional quality if both countries are not share the same legel origin. the information of legal law origin provided to us by andrei shleifer’s website3 3.3 method of estimation by following the, beverelli et al. (2018) we use the ppml estimation method because this method became famous for structural gravity model due to following striking approaches: (1) ppml method deals the problems of heteroskedasticity which often presents in bilateral trade dataset (santos silva and tenreyro, 2006). (2) it deals with zero trade data which is also present in bilateral trade dataset because every country does not trade with all countries. (3) it can be use easily in stata to estimate the gravity model (anderson et al., 2016). we knew in basic gravity model that if we do not control mrts (which is fixed effects in ppml) then it leads biased estimation as beverelli et al. (2018) observed that if we do not control mrts properly in ppml it leads biased estimaton. in stata 3 https://scholar.harvard.edu/shleifer/publications/quality-government muhammad zubair chishti, babar hussain, and muhammad aqib khursheed 26 simple ppml command is used to estimate the poisson pseudo maximum likelihood but according to zylkin (2018) in simple ppml command we treat fixed effects as exogenous variable which leads biased results. because in simple ppml command we create dummies for years, exporters and importers then we grouped these dummies to make a single variable of year-fixed-effects, exporter-fixed-effects and importer-fixedeffects then we regress these fixed effects as exogenous variable. in this way these fixed effect dummies correlate with main variable dummies and yield biased result. to avoid the biased estimation, zylkin (2018) introduce a new command ppml_panel_sg (ppml panel standard gravity) to deal fixed effects separately but there is also a problem if we have big data set and large number of variables because it is not too fast command. however a new ppml command ppml_hdfe (ppml with multiple high dimensional fixed effects) is introduced by correia et al. (2019) having all properties of simple ppml in addition it deal fixed effects saperately and faster command to estimate the ppml estimates. ppml_hdfe is implements a novel and robust method to check the existence of poisson pseudo maximum likelihood estimates4. recently ppml high dimesnsional command is used by larch et al. (2019) to test the impact of currency unions on trade for big data set having 200 countries for 65 years and find that this command facilitate the larg data set having fixed effects. 4. results and discussion 4.1 the impact of institutions and development on bilateral exports: in this section we demonstrate the strength of our methods by attaining the partial estimates of the direct effect of domestic institutions and level of developments on bilateral exports. in additions we also develop the institutional homogeneity and institutional heterogeneity affecting the bilateral exports. we start with baseline panel estimation because this is the most popular and standard approach to investigate the effect of institutional quality on bilateral exports by following the previous literature: tinbergen, 1962; anderson and van wincoop, 2003; beverelli et al., 2018. the estimation from column (1) of table (1) is attained with ppml_hdfe established on standard gravity model which is not included institutional quality variables as hou et al., (2021) and beverelli (2018) use these variable in their baseline model. in this column we use standard variables which capture the trade cost, including; bilateral distance, contagious border, common official language, colonial relationships and bilateral rta. in addition we also use the exporter-fixed effect, importer-fixed effect and time-fixed effects in our estimation. the baseline standard gravity estimates from column 1 of table (1) reveals the time-invariant effect of gravity variables on bilateral exporter which shows that the distance and contagious border are barriers to bilateral trade while having the common 4 more information about ppml_panel_sg and ppml_hdfe codes, examples and its help files can be found at website of authors of t hese command thomas zylkin: http://www.tomzylkin.com/ journal of applied economics and business studies, volume. 5, issue 2 (2021) 17-34 https://doi.org/10.34260/jaebs.522 27 official language, sharing the colonial relations and sharing the bilateral trade agreements encourages bilateral trade. the estimate impact of ln_dist is negative and significant. the impact of contagious border, common-official language, colonial relationships and rta is positive and significant; these results are consistent with hou et al., (2021). but we are more interested in time-varying effects of specific variables on bilateral exports, however our baseline ppml_hdfe estimates is robust to estimate the ppml_hdfe with institutions and development variables. in order to obtain the various determinants of bilateral exports and the impact of quality of institutions and level of development, the results from column (2) of table (1) are attained with the same baseline estimation. additionally we extend it and add gdp per capita of exporter and importer country which is use for level of development. here the gdp per capita of exporter and importer is positively and significantly affect the bilateral exports. the coefficient of gdp per capita of exporter is larger than importer country. hence the level of development of exporter matter more than importer’s, this results are consistent with florensa et al., (2015). the estimates from column 3 of table (1) are attained with same as column 2; in additions to the standard gravity model we introduce here the institutional quality variable efw for the exporter as well as for importer country which captures the effect of institutional quality on bilateral exports. mainly we discuss the results from column 4 and column 5 of table (1) because in column 3 of table (1) show the average impact of institutional quality and development on bilateral exports. further we extended our gravity model and finally we reach our main estimate results where we also included exchange rate as a control variable and our main interested variable homogeneity of institutions along with institutional quality and level of development variables. three main findings are obtained in column 4 and 5 of table (1): first our results are enable us to obtain the impact of institutional quality on the bilateral exports in the existence of complete set of exporter-fixed effects, importerfixed effects and time-fixed effects, this confirms the idea of beverelli et al. (2018) that if we use fixed effects in our estimation then we will not face any problem of collinearity. if the institutional quality of exporter as well as thee institutional quality of importer country is equally low or relatively low then the institutional quality is imperative in explaining the impact of quality of institutions on bilateral exports. hence, our results reveal that institutional quality of exporter as well as the institutional quality of importer countries is positive and significant but the institutional quality of exporter country has larger significant value than importer country; however quality of institutions of exporter country is more important than the quality of institutions of importer country to raise the bilateral exports flow. second we obtain positive and significant impact of level of development on bilateral exports which leads us that high level of development boosts the bilateral exports. in case of development we also obatain the higher positive signifant value of development of exporter country than importer country which confirm that if a country acts as a exporter then the level of muhammad zubair chishti, babar hussain, and muhammad aqib khursheed 28 development more important rather than it is importer. these are consistent with daniel et al. (2006) that the quality of institutions of exporter country is more important for different transactions(exporte and imports). third the positive and statistically significant results of law variable (which is used to measure the institutional homogeneity) in column 5 of table (1) leads that if the both countries share the same level of the institutional characteristics, it promotes the bilateral exports. as suggested by miura and takechi (2014) that this results have essential implication for homogeniety of institutional quality, especially for high institutional quality countries. institutional homogeniety is very important if we have high institutional quality countries in our sample, i.e if the institutional quality of both countries are equally low or relatively low then we can measure the effect of institutional differences on bilateral exports and able to say that improvent in institutional quality will increase the bilateral exports but if the both countries already have high institutional quality at maximum level but different then it is not possible to improve further the quality of institutions because it not reveals that in which path the rule would be improved. however in this situation institutional harmonization will facilitate the bilateral exports (khan et al., 2020). table 1: the impact of institutions and development on bilateral exports (1) (2) (3) (4) (5) variables ppml baseline ppml baseline gdp ppml baseline efw ppml hetro main ppml homo main ln dist ij -0.701*** -0.709*** -0.710*** -0.710*** -0.720*** (0.0127) (0.0136) (0.0133) (0.0133) (0.0134) cont ij 0.594*** 0.601*** 0.598*** 0.598*** 0.516*** (0.0259) (0.0260) (0.0259) (0.0259) (0.0282) lang ij 0.270*** 0.274*** 0.271*** 0.270*** 0.0772*** (0.0290) (0.0285) (0.0286) (0.0286) (0.0285) col ij 0.254*** 0.247*** 0.245*** 0.245*** 0.266*** (0.0296) (0.0287) (0.0285) (0.0286) (0.0276) rta ijt 0.444*** 0.409*** 0.408*** 0.410*** 0.434*** (0.0242) (0.0245) (0.0243) (0.0243) (0.0248) ln p. gdp it 1.001*** 0.895*** 0.896*** 0.896*** (0.0774) (0.0873) (0.0871) (0.0858) ln p. gdp jt 0.328*** 0.199* 0.195 0.201* (0.107) (0.118) (0.119) (0.117) efw it 0.131*** 0.137*** 0.136*** (0.0430) (0.0434) (0.0439) efw jt 0.231*** 0.233*** 0.229*** (0.0451) (0.0453) (0.0451) er it 0.138* 0.139* (0.0816) (0.0812) law ijt 0.244*** (0.0169) journal of applied economics and business studies, volume. 5, issue 2 (2021) 17-34 https://doi.org/10.34260/jaebs.522 29 constant 28.79*** 15.61*** 15.32*** 15.21*** 15.23*** (0.115) (1.252) (1.193) (1.206) (1.194) observations 56,352 56,250 52,773 52,605 52,605 r-squared 0.932 0.936 0.936 0.936 0.937 exporter fe yes yes yes yes yes importer fe yes yes yes yes yes year fe yes yes yes yes yes note: *** p<0.01, ** p<0.05, * p<0.1. robust standard errors in parentheses. note: this table shows the estimation results of ppml high dimensional fixed effects. all estimation is obtained by absorbing the exporter, importer and time fixed effects which vary over time in panel specification. mrts are properly controlled by ppml hdfe. 4.2 interaction effect of institutions and development on bilateral exports: we also obtain the differential estimate of the interaction impact of quality of institutions and level of development on bilateral exports. moving on interaction effect of institutional quality and gdp per-capita, we eventually discover robust evidence of asymmetric impacts of institutions and level of development on bilateral exports that the quality of institutions and level of development of exporter country increases the bilateral exports, we also explore the similar distorted impacts of importer country’s quality of institutions and level of development on bilateral exports. the ppml with hdfe regression from column 1 of table (2) reveals the positive and significant impact of quality of institutions and level of development on bilateral exports by controlling the bilateral distance, contagious border, common official language, colonial relationship and exchange rate and negative effect of interactive terms, moreover including the individual terms and interaction terms simultaneously in the same regression. here interaction terms become negative due to multicollinearity between interactive terms and individual effects as, adefabi (2011) observed that if we include individual effects and interaction effects simulataneously in the same regression the coefficients change their signs or become insignificant due to enterance of multicollinearity in regression between individual and interacation effects. further, lamsiraroj (2016), chishti et al., (2020) and chishti et al., (2021) find that the multicollinearity in regression is due to interaction term and it is constructed because the correlation of two variables that’s why this multicollinearity is unavoidable if we include both interaction and individual effects simulateously in the same regression. the positive and significant impact of quality of institutions and level of development due to the result of direct effect of both variables on bilateral exports. therefore, it is essential to include the institutional quality and level of development individually alongwith their products to test jointly whether these variables affect the bilateral exportes by themselves or through the interactive terms. however, we find that muhammad zubair chishti, babar hussain, and muhammad aqib khursheed 30 institutional qualtiy and level of development dirctly affect the bilateral exports as well as they indirectly affect the bilateral exports by interacting each others. similar specification adopted in the column 2 of table (2) where we include only one individual effect gdp per-capita along with interaction terms which shows that the coefficient of both interaction terms become positive and significant but the coefficient of gdp per capita of importer become insignificant because of multicollinearity. in column 2 of table (2) it is heterogeneous effects but in column 3 of table (2) we also check the homogeneous effects, in this column we also find the same results with insignificant effects of individual term. after the discussion of interaction and individual terms and problem of multicollinearity our main results are presented in column 4 and column 5 of table (2) where only interaction terms are included while the individual terms are omitted to avoid the problem of multicollinearity which is constructed in interactions. in column 4 of table (2) heterogeneous interaction impact of institutional quality and level of development on bilateral exports is shown. the coefficient of interaction terms is positive and significant which shows that the interaction effect of institutional quality and level of development is positive and statistically significant these results are correlated with marjit et al.(2014). in our results the value of interaction term of exporter is high and value of importer country is low which indicates that interaction effect of institutional quality and level of development on bilateral exports of exporter country have greater impact than the interaction effect of institutional quality and level of development of importer country this happens because most countries want to increase exports to maximize their comparative advantage and government also encourages exports and discourage imports by using trade protection , tariff on imports and subsidy on exports, because exports bring jobs, higher wages and increase the foreign reserves. however the institutional quality and level of development of exporter country supports to increase the exports rather than the imports. in column 5 of table (2) our main result of homogenous institutional quality is shown, here all variables are same as column 4 of table (2) additionally law variables is add to check the homogeneity of institutions. the coefficient of law is positive and significant which show that the homogeneous interaction effect of institutional quality and level of developments on bilateral exports is positive and significant. however if the quality of institutions and level of development interact with each and they also have same level of institutional quality, it will boost the bilateral exports these results are consistent with bandyopadhay et al., (2016). table 2: interaction effect of institutions and development on bilateral exports (1) (2) (3) (4) (5) variables ppml int full ppml int +dev hetro ppml int +dev homo ppml int main hetro ppml int main homo ln dist ij -0.721*** -0.710*** -0.720*** -0.708*** -0.718*** journal of applied economics and business studies, volume. 5, issue 2 (2021) 17-34 https://doi.org/10.34260/jaebs.522 31 (0.0134) (0.0133) (0.0134) (0.0127) (0.0127) cont ij 0.515*** 0.598*** 0.516*** 0.599*** 0.517*** (0.0281) (0.0260) (0.0282) (0.0258) (0.0280) lang ij 0.0783*** 0.270*** 0.0769*** 0.270*** 0.0781*** (0.0284) (0.0286) (0.0286) (0.0289) (0.0290) col ij 0.266*** 0.245*** 0.266*** 0.246*** 0.267*** (0.0276) (0.0286) (0.0276) (0.0287) (0.0277) rta ijt 0.433*** 0.410*** 0.435*** 0.417*** 0.441*** (0.0249) (0.0244) (0.0249) (0.0238) (0.0244) ln p. gdp it 1.176*** 0.825*** 0.825*** (0.214) (0.103) (0.102) ln p. gdp jt 0.648*** 0.0626 0.0706 (0.232) (0.141) (0.139) efw it 0.594** (0.303) efw jt 0.907*** (0.269) int it -0.0478 0.0132*** 0.0132*** 0.0461*** 0.0461*** (0.0320) (0.00457) (0.00462) (0.00356) (0.00356) int jt -0.0704** 0.0225*** 0.0222*** 0.0267*** 0.0268*** (0.0292) (0.00482) (0.00482) (0.00422) (0.00422) er it 0.146* 0.136* 0.137* 0.135* 0.134* (0.0819) (0.0819) (0.0814) (0.0810) (0.0901 law ijt 0.243*** 0.244*** 0.243*** (0.0169) (0.0170) (0.0171) constant 8.370*** 17.30*** 17.30*** 23.48*** 23.54*** (2.812) (1.420) (1.398) (0.396) (0.396) observations 52,605 52,605 52,605 52,773 52,773 r-squared 0.937 0.936 0.937 0.935 0.935 exporter fe yes yes yes yes yes importer fe yes yes yes yes yes year fe yes yes yes yes yes note: *** p<0.01, ** p<0.05, * p<0.1. robust standard errors in parentheses. conclusion institutions are fundamental determinants of economic exchange within border as well as across the border. numerous previous studies on the impact of institutions on bilateral trade flows exists, though these studies do not deal satisfactorily with mrts and interaction effects of homogenous institutions and development on bilateral exports. so we evaluate the causal effect of institutional quality, institutional homogeneity and development on bilateral exports. this study apply a novel methodology which allows to investigate the effect of quality of institutions and homogeneity of institutions as a driver of informal trade barriers on bilateral export flows with proper set of fixedeffects. we employ both institutional quality and level of development for bilateral exports so our approach is consistent with theoretical gravity model. muhammad zubair chishti, babar hussain, and muhammad aqib khursheed 32 however, the impact of quality of institutions and development on bilateral export flow that we have estimated is positive and significant. the result of homogenous institutional quality is also positive and significant which reveals that homogenous institutional quality encourages the bilateral export flows. enrich with consistent estimates of the impact of quality of institutions and level of development on bilateral exports, we also check the interaction impact of quality of institutions and level of development on bilateral exports. we show that quality of institutions and level of development also affects the bilateral exports indirectly; both are interact with each other and affect the bilateral exports significantly and positively. our homogenous impact of quality of institutions and level of development suggest that if both countries share the same level of institutional quality then the interaction effect of institutional quality and level of development boosts the bilateral exports. therefore, government should improve the institutional quality because it leads certainty for international trade which helps the foreign exporters and importers to increase the trade flows. government should also pay attention to improve the level of development because it facilitates the exporters and importers to boosts the bilateral trade flows. besides the future research can be done by applying our methodology and estimators, to measure behind the border such as sales taxes, value-added taxes, and trade facilitation for extended period, as well as environmental measures like consumption based carbon emission exports and imports. references acemoglu, d., & robinson, j. (2010). the role of institutions in growth and development. review of economics and institutions, 1(2), 1-33. arif, m., chishti, m. (2020). analyzing the effectiveness of fiscal decentralization in economic growth: the role of institutions. iranian economic review, (), -. doi: 10.22059/ier.2020.78144 acemoglu, d., johnson, s., & robinson, j. (2001). the colonial origins of comparative development: an empirical investigation. american economic review, 5(1), 1369-1401. alvarez, i., barbero, j., rodriguez-pose, a., & zofio, j. (2018). does institutional quality matter for trade? institutional conditions in a sectoral trade framework. world development, 103(1), 72-87. anderson, j. e. (1979). a theoratical foundation for the gravity equation. the american economic review, 69(1), 106-116. anderson, j., & marcouiler, d. (2002). insecurity and the pattern of trade: an empirical investigation. review of economic and statistics, 84(2), 342-352. anderson, j., & van wincoop, e. (2003). gravity with gravitas: a solution to the border puzzle. the american economic review, 93(1), 170-192. angkinand, a. p., & chiu, m. e. (2011). will institutional reform enhance bilateral trade flows? analyses from different reform aspects. journal of economic policy reform, 14(3), 243258. journal of applied economics and business studies, volume. 5, issue 2 (2021) 17-34 https://doi.org/10.34260/jaebs.522 33 bandyopadhayay, s., & roy, s. (2016). the effects of corruption on trade flows: a disaggregated analysis. international trade and international finance, 1(1), 97-116. bernhofen, d. m., & brown, j. c. (2005). an empirical assessment of the comparative advantage gains from trade: evidence from japan. american economic review, 95(1), 208-225. beverelli, c., larch, m., keck, a., & v.yotov, y. (2018). institions, trade and development: a quantitative analysis. cesifo working ppaers, 6920. bojnec, s., imre, f., & kti, m. (2009). the institutional determinants of bilateral agricultural and food trade. applied studies in agribusiness and commerce, 1(3), 1-15. correia, s., guimarães, p., & zylkin, t. (2019). ppmlhdfe: fast poisson estimation with highdimensional fixed effects. zrxiv preprint arxiv: 1903.01690. chishti, m. (2020). analysis of the nexus between demographic changes and economic growth in pakistan: role of capital stock. iranian economic review, (), -. doi: 10.22059/ier.2020.76529 chishti, m. z., ahmad, m., rehman, a., & khan, m. k. (2021). mitigations pathways towards sustainable development: assessing the influence of fiscal and monetary policies on carbon emissions in brics economies. journal of cleaner production, 292, 126035.. chishti, m. z., ullah, s., ozturk, i., & usman, a. (2020). examining the asymmetric effects of globalization and tourism on pollution emissions in south asia. environmental science and pollution research, 27, 27721-27737.. chishti, m. z., iqbal, j., mahmood, f., & azeem, h. s. m. (2020). the implication of the oscillations in exchange rate for the commodity-wise trade flows between pakistan and china: an evidence from ardl approach. review of pacific basin financial markets and policies, 2050030. chishti, m. (2021). assessing the dynamic effects of oscillations in the exchange rate on commodity-wise trade flows between pakistan and saudi arabia: evidence from ardl approach. iranian economic review, (), -. doi: 10.22059/ier.2021.79457 chishti, m., hussain, b., jehangir, f. (2021). do foreign direct investment and trade openness influence poverty in pakistan? evidence from nardl approach. iranian economic review, (), -. doi: 10.22059/ier.2021.81556 de groot, h. l., rietveld, p., subramanian, u., & linders, g.-j. (2004). the institutional determinants of bilateral trade patterns. kyklos, 57(1), 103-123. deardorff, a. v. (1998). determinants of bilateral trade: does gravity work in a neoclassical world? in the regionalization of the world economy, 1(1), 7-32. donaldson, d. (2015). the gains rom market integration. nnual rewiew of economics, 7(1), 619647. florensa, l. m., márquez-ramos, l., & recalde, m. l. (2015). the effect of economic integration and institutional quality of trade agreements on trade margins: evidence for latin america. review of world economics, 151(2), 329-351. francois, j., & manchin, m. (2007). institutions, infrastructure, and trade. world development, 46(1), 165-175. frankel, j. a., & romer, d. h. (1999). does trade cause growth? american economic review, 89(3), 379-399. muhammad zubair chishti, babar hussain, and muhammad aqib khursheed 34 hou, y., wang, y., & xue, w. (2021). what explains trade costs? institutional quality and other determinants. review of development economics, 25(1), 478-499. islam, r., & reshef, a. (2006). if your institutions are good, does it matter if they are differenen? the world bank, 1-61. khan, m. a., popp, j., talib, m. n. a., lakner, z., khan, m. a., & oláh, j. (2020). asymmetric impact of institutional quality on tourism inflows among selected asian pacific countries. sustainability, 12(3), 1223. la porta, r., lopez-de-ailanes, f., shleifer, a., & vishny, r. w. (1997). legal determinant of external finance. the jounal of finance, 52(3), 1131-1150. la porta, r., lopez-de-silanes, f., shleifer, a., & vishny, r. w. (1998). law and finance. journal of political economy, 106(6), 1113-1155. levchenko, a. (2007). institutional quality and international trade. review of economic studies, 74(3), 791-819. marjit, s., mandal, b., & roy, s. (2014). trade openess, corruption and factor abundace: evidence from a dynamic panel. review of development economics, 18(1), 45-58. mendonca, t. g., lirio, v. s., braga, m. j., & da silva, o. m. (2014). institutions and bilateral agricultural trade. economics and finance, 14(1), 164-172. miura, k., & takechi, k. (2014). institutional quality and homogeneity, and types of international transactions. scientific research, 1(1), 1-21. nguyen, c. p., su, t. d., & nguyen, t. v. h. (2018). institutional quality and economic growth: the case of emerging economies. theoretical economics letters, 8(11), 1943. north, d. c. (1993). institutions and credible commitment. journal of institutional and theoratical economics, 1(1), 11-23. nunn, n., & trefler, d. (2014). domestic institutions as a source of comparative advantage. handbook of international economics, 4(1), 263-315. pascali, l. (2017). the wind of change: maritime technology, trade, and economic development. american economic review, 107(9), 2821-2854. shah, q., zubair, s., & hussain, s. (2020). the impact of institutions on economic growth in selected developing countries: an analysis based on bayesian panel estimation. journal of applied economics and bussiness studies (jaebs), 4(4), 251-272. silva, j. s., & tenreyro, s. (2006). the log of gravity. the review of economics and statistics, 88(4), 641-658. silva, j. s., & tenreyro, s. (2010). on the existence of the maximum likelihood estimates in poisson regression. economic letters, 107(2), 310-312. tinbergen, n. (1962). foot-paddling in gulls. british birds, 55, pp. 117-120. ullah, s., chishti, m.z. & majeed, m.t. the asymmetric effects of oil price changes on environmental pollution: evidence from the top ten carbon emitters. environ sci pollut res 27, 29623–29635 (2020). https://doi.org/10.1007/s11356-020-09264-4 zylkin, t. (2018). ppml_panel_sg: stata module to estimate "structural gravity" models via poisson pml. journal of applied economics and business studies, volume. 5, issue 2 (2021) 87-102 https://doi.org/10.34260/jaebs.526 87 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x university teachers’ perspective on hec’s faculty appointment criteria: an interpretive phenomenological analysis shoaib irshad1*, sadia irshad2 and sadaf kashif3 1ph.d. scholar, iqra university, islamabad campus & deputy director, quality assurance agency, higher education commission, islamabad 2assistant professor, air university, islamabad 3assistant professor, iqra university, islamabad campus abstract system devised by the higher education commission (hec) for higher education institutions (hei) aims to promote innovation and develop human capital. the efficiency of the system is ascertained if it could help in recruitment and retention of faculty members. therefore, the study of faculty members' perspective on hec’s faculty appointment criteria (fac) is needed to determine whether it motivates and facilitates them. this study reports the faculty’s perspectives on fac. this qualitative interpretive phenomenological study gathers data using a semistructured questionnaire for interviewing. the analysis reveals that the criteria do not cover the overall performance of faculty members and is only based upon minimum qualification, duration of service and number of publications. there is a dire need of reviewing the current appointment criteria and for that the involvement of all stakeholders is suggested to devise a profound scheme for better human resource development at heis of pakistan. keywords faculty appointment criteria, job design, job satisfaction, interpretive phenomenologi cal analysis, discrimination jel classification i2, i23, i28, j28, j71, m51 1. introduction higher education system is highly important in the modern rapidly growing educational paradigm globally. it is not only fundamental in promoting innovation but also in developing human capital and knowledge economy (dill & van vught, 2010). the success and development of the higher education system is primarily dependent * sirshad@hec.gov.pk shoaib irshad, sadia irshad and sadaf kashif 88 upon the performance of faculty members. unfortunately, in pakistan, the faculty underperforms or low-performs in all areas including research due to many tangible and intangible issues. the faculty members view it as dissatisfaction due to poor compensations in terms of recognition of their performances (centra, 1977). the important achievement and hallmark of success for faculty in academia is to get promoted, for that purpose different criteria have been devised in the higher education system for promotion of faculty (wiley, wallingford, monllor-tormos, & konyufogel, 2016). the higher education commission (hereafter hec) of pakistan has devised faculty appointment criteria (hereafter fac) for appointments and promotions of faculty members from the position of lecturer to full professor. the criteria is primarily based upon minimum qualification, length of service and research publications; leaving out the other very important parameters of performance of faculty members (khuram, bhutto, & jabeen, 2017). therefore, it is imperative to study the faculty's view on the criteria for recruitment in heis of pakistan. this study presumes that hec’s faculty appointment criteria affect faculty negatively. it is a source of dissatisfaction, disappointment and demotivation of those faculty members who are unable to meet these criteria due to multiple reasons. the objectively stated stance brings forth uni-dimensionality of these criteria. the faculty’s performance, efficiency and competence is judged against one yardstick: research publications. resultantly, the impact apparently can be seen in terms of faculty pursuing one task only, leaving aside the other important jobs expected from them. in this way, these set criteria become a hindrance and obstacle for faculty to move on to the next scale. addressing faculty’s dissatisfaction with faculty appointment criteria, the present study investigates the faculty’s perspectives to unearth the problems they face in their professional growth. the appointment criteria devised by any human resource department ensures merit, justice and employee satisfaction. however, the faculty members of higher education institutions (hereafter heis) of pakistan feel deprived when they map their performances with the higher education commission’s faculty appointment criteria. they consider the current criteria as uni-dimensional, as the only performance parameter of criteria is of research output in the form of research publications and that too in hec’s recognized journals only. these discriminatory criteria adversely affect the overall performance of faculty members and their focus towards other key assignments, as they start only focusing on getting their research publications to meet the defined criteria. in order to address this issue it is important to gauge hei’s faculty’s perspective. in order to study the identified problem, the present study intends to investigate employees’ views on hec’s fac. therefore, for the purpose of this study the researcher explored the perspectives of faculty members of higher education institutions regarding hec’s fac and also located the problems and reasons due to which faculty cannot meet set criteria. journal of applied economics and business studies, volume. 5, issue 2 (2021) 87-102 https://doi.org/10.34260/jaebs.526 89 the basic assumption for conducting this research is that in general the faculty members who serve in heis of pakistan consider the hec’s fac does not cover the overall performance evaluation of faculty members and is only based upon minimum qualification, duration of service and number of publications. moreover, the study postulates that there are a number of problems which heis’ social sciences faculty face when they aim to fulfill it. 2 literature review this study incorporates the thematic critical review of the existing literature on hec’s faculty appointment criteria, since this study intends to explore the perspectives of faculty members of higher education institutions (heis) of pakistan regarding hec’s faculty appointment criteria and its impact on the overall performance of faculty members. in this respect there are very few researches that have been conducted to these sorts of criteria; however the review includes those previous researches which address a variety of related perspectives. these perspectives would, nonetheless, help conceptualize the phenomenon of employees’ job dissatisfaction due to the appointment criteria. therefore, this study will be focused to review the different themes associated with faculty appointment criteria, perspectives of faculty members associated with these criteria and the problems which the faculty members are facing owing to these criteria. globally, the organizations are striving hard to implant internationally acclaimed best practices of human resource management (hrm) to better compensate their employees, and in order to get desired outcomes (kang & lee, 2021). the compensations are directly linked with the performances of the employees, and mechanisms have been devised to better gauge the performances in order to implement compensations accordingly. the understanding of performance measurement tools for any kind of job is pivotal in order to better compensate the employees, as per their satisfaction level. the higher education institutions (heis), are not different from other organizations and are committed in taking steps to implant best hrm practices. as globally, the faculty members are considered the pivotal and integral part of heis, therefore special efforts are being made to better compensate the faculty members. while reviewing the previous literature it was revealed that there is sparingly any work done on reviewing the performance measuring processes/ criteria for faculty members. wiley et al. (2016) have also stated the same, that there are very few researches available that evaluate the existing promotion criteria of faculty members. similarly, there is scarcity of research on recommending suggestions for the change in these criteria. however, the available research revealed that the overall performance of any faculty member can be primarily determined through three factors; teaching, service and research. this study was primarily focused on faculty members of business schools; but the factors studied in this research are generalized and can be considered for evaluating the performance of faculty members of any discipline with minor amendments. to get a deeper understanding of the phenomenon, the available literature shoaib irshad, sadia irshad and sadaf kashif 90 on these three important factors of measuring performance of faculty members is being reviewed and analyzed. 2.1 teaching: the research conducted by crawford, burns, and mcnamara (2012) stated that there are very few percentages of deans of university, to be precise only 6.2 percent, that consider teaching as the most important part of a professor's job. this study reveals that the faculty members, while reaching the post of professorship, have opined that teaching is not the most integral part of their job. while analyzing this in pakistani scenario, khuram et al. (2017) have reviewed the current hec’s policies and their impact on motivation of faculty members and concluded that the existing hec’s faculty appointment criteria has increased the motivation amongst the faculty members to perform better. the performance measurement mechanism in current criteria is based solely on the research outcomes that is also only on the number of research publications (rafi, ahmad, naeem, khan, & jianming, 2020). hence, this study did not include any other factor to be associated with performance of faculty members. thus, this study is limited in scope as it ignored the distressed faculty waiting to be recruited for the next post for years owing to problems in achieving set criteria of research publications. on the contrary, malachowski (2010), in his study revealed that there is a dire need for the faculty members to bring balance between the increased demand of research productivity and teaching. for this, the teaching load needs to be rationalized and emerging teaching methodologies and approaches need to be inculcated. this study provides a solution to the debate of how much working hours of faculty members need to be dedicated towards research and how much of them should be for teaching. the balance between these two needs to be observed, but unfortunately current criteria do not address this key issue either. while focusing on teaching, the measurement of teaching performance is a most debated topic and a number of researches have been conducted to devise a strategy to measure teaching performance of faculty members. in this regard fairweather (2002); paulsen (2002) have studied the assessment mechanism of faculty’s performance with regard to teaching and came up with the conclusion that peer reviews, student assessment, and tracking student progress enable accurate assessment of faculty’s teaching performance. these studies demonstrate that if deemed appropriate by the policy makers, an accurate mechanism can be devised to measure the teaching performance of any faculty member. therefore, the study in hand aims to locate the place of teachers teaching performance in the forefront while gathering data from the faculty on hec’s faculty appointment criteria. 2.2 service: a mission of the higher education sector is to serve the communities that support it as well as those that do not. ward (2003) found that institutions of higher education are falling short of supporting this mission and one way to respond to this challenge is for faculty to be more engaged with community services. when faculty are actively engaged with the community, the campus builds a positive relationship of engagement with numerous stakeholders, including alumni, businesses, and potential journal of applied economics and business studies, volume. 5, issue 2 (2021) 87-102 https://doi.org/10.34260/jaebs.526 91 donors. in this context, mahmood (2016) states the mission of hec to “facilitate institutions of higher learning to serve as engines of growth for the socioeconomic development of pakistan”. that implies the focus of heis of pakistan should also be on provision of community services, but this pivotal part has never been gauged by any policy devised for assessing the performance of faculty members. and this is another concern for which this study is taken up. 2.3 research: research is one of the most central functions of universities throughout the world and the faculty plays the most crucial role in producing knowledge through research (tien, 2008). while reviewing the current hec’s appointment criteria, it is revealed that it is research centric. it implies that the research publication is the only parameter to assess the performance of any faculty member for his/ her appointment/ promotion. whereas other important factors such as teaching, services and industrial linkages are not considered to review overall performance. shah, akhtar, zafar, and riaz (2012) in their study revealed that job satisfaction plays an integral part in performance of faculty members at educational institutions. whereas, the current criteria is only focused on research publications. that implies the faculty members who are excellent in teaching and are serving the community through certain philanthropic tasks are not recognized in the current criteria of appointment/promotion. these factors may lead to job dissatisfaction amongst the faculty members of pakistani heis. in order to address these identified research gaps, the present study attempts to investigate the social sciences faculty of pakistani heis’ perspective on faculty appointment criteria. thereof, to fill this research gap, the following research questions have been formulated: i. what is the perspective of faculty members regarding hec’s faculty appointment criteria? ii. what problems do the faculty members face while fulfilling hec’s faculty appointment criteria? iii. why do hei’s faculty members find faculty appointment criteria a hindrance to their professional growth? 3. research methodology 3.1 research strategy for the purpose of this study the researcher has employed a qualitative research strategy with a semi-structured questionnaire for conducting interviews of social sciences faculty of pakistani universities and the gathered data was scrutinized by means of interpretive phenomenological analysis (hereafter ipa). 3.2 context and population this study is conducted in public sector higher education institutions and specifically in those institutions that are located in islamabad. due to diversity of disciplines and exemptions in hec’s faculty appointment criteria for some shoaib irshad, sadia irshad and sadaf kashif 92 disciplines, such as engineering, information technology, medical, law and arts & design, the study was limited to faculty members of social sciences discipline. the population for this study includes both male and female full time social sciences faculty members working as assistant professor or associate professor only with age group of 35 years to 50 years, serving in public sector higher education institutions of islamabad. the minimum qualification for population is ph.d. as ph.d. is the requirement for appointment/ promotion as associate professor and professor. 3.3 sampling frame: social sciences faculty members with the designations of assistant professors and associate professors who have done their ph.ds. (the minimum qualification requirement for associate professor and professor) and have attained minimum experience required for appointment/ promotion as associate professor and professor; and are serving in islamabad based public sector universities. initially a data was gathered from higher education commission to examine how many heis have adopted hec’s faculty appointment criteria, currently there are 221 hec’s recognized heis out of which the record of 159 heis regarding adoption of faculty appointment criteria was available that revealed that out of them 154 have adopted the criteria from their statutory bodies. the remaining 05 heis are newly established heis and are working to get approval of adoption of hec’s faculty appointment criteria from their respective statutory bodies. this implies that at institution level, both public and private sector heis are committed to adopt the hec’s faculty appointment criteria. 3.4 sampling technique: for the purpose of this study purposive and convenience sampling techniques were employed. purposive sampling technique was employed to select those ph.d. social sciences faculty members that were serving as assistant professors and associate professors in public sector heis of islamabad including allama iqbal open university, international islamic university islamabad and quaide-azam university; and have gained the required experience to be appointed/ promoted to associate professor and professor. they were invited to participate in this study through emails and telephonic communication. owing to the limited time allotted for this research only those faculty members were interviewed who could be conveniently approached. 1) sample size: the sample size for phenomenological studies usually ranges from 05 to 12, and particularly in case of interpretive phenomenological analysis (ipa), smith (2011) mentioned that sample size of 03 is useful for those researchers who are beginners in ipa, while 5 to 6 is considered to be a reasonable size for any ipa research. considering this in view, 07 social sciences faculty members from 03 above stated universities have been purposefully and conveniently selected as the sample for the study. 2) data collection instruments: the study employs self-devised semi structured open-ended questionnaires (see appendix) for the interviews of social sciences faculty members serving as assistant professors and associate professors to journal of applied economics and business studies, volume. 5, issue 2 (2021) 87-102 https://doi.org/10.34260/jaebs.526 93 evaluate their perceptions regarding hec’s faculty appointment criteria for appointments and promotions. a semi-structured open-ended questionnaire was devised as no previous research in pakistan supports this study. secondly, it helped direct the interview according to faculty’s responses to get insight of the actual perceptions on hec’s faculty appointment criteria. 3) data analysis technique: the qualitative data generated from the semi structured open-ended interview questionnaires on faculty perception regarding hec’s faculty appointment criteria was analyzed thematically. the study involved an interpretive phenomenological analysis (smith, 2011) which helped in understanding the perspectives of social sciences’ faculty members on hec’s faculty appointment criteria. 4. analysis of the data the interpretive phenomenological analysis provides the opportunity to analyze the subjective experiences of the social sciences faculty. for the objective presentation of these subjective experiences the interviews were analyzed thematically. this thematic interpretive phenomenological analysis helps to find the perceptions of faculty members about the hec’s faculty appointment criteria. as the questionnaire was divided into 02 parts; the first part was about background information to ensure that the respondents meet the minimum requirements to get interviewed. whereas the second part was focused on getting the perceptions of these faculty members on hec’s faculty appointment criteria. while analyzing the responses of the first part of the questionnaire, the collected data from the 07 respondents revealed that 4 faculty members were working as assistant professor and 3 as associate professor. the next question (see appendix, question 2) was asked to determine the length of their service in the current position, the primary objective of this question was to get preliminary information that the respondents meet the minimum requirement of experience for selection in the next scale. the collected data showed that all the 07 faculty members fulfill the minimum experience requirement for posting on the next scale. resultantly, it was revealed that the faculty members that have been interviewed were eligible to get promoted to the next scale in terms of minimum experience required. in answering to the question (see appendix, question 3) on how long they have been associated with this hei, 04 out of 07 informed that this was their first job in university, 02 of them informed that they have switched their university due to lack of opportunities and growth, whereas 01 of them informed that he completed his minimum tenure in his previous institution but there was no sanctioned position of associate professor in his discipline, therefore he has to switch his employer and have to join as assistant professor in this university. in answer to the question (see appendix, question 4) on whether they did phd while serving in this hei, all the 04 faculty members who have been into their first job have done their phd while serving as lecturer. it revealed that most of the respondents were loyal to their institutions. shoaib irshad, sadia irshad and sadaf kashif 94 because all of them had availed faculty development program scholarships plus paid leave for the completion of doctorate degrees, they showed integrity to their institution by resuming their services after completion of higher degree. the second part of the questionnaire was analyzed thematically and main themes were extracted while assigning the codes to core themes and sub codes to other important themes, as suggested by distinguished gurus of qualitative research (creswell & poth, 2016; silverman, 2015). whereas, attride-stirling (2001) came up with the idea of constructing a thematic network, in order to better arrange the like themes in a single network and connect these networks to each other, to analyze the qualitative data. the collected data was reviewed systematically to extract the themes, and while doing so following core themes were extracted: 4.1 requirements in faculty appointment criteria the respondents unanimously raised serious concerns on the hec’s faculty appointment criteria, specifically related to the requirements mentioned in the criteria that are related to minimum qualification, minimum experience and minimum number of publications in hec recognized journals. their opinion about the current criteria was very clear, as reflected from the statement of one of the respondents: “….how can hec say that this is an employee friendly criteria, many of employees like us are sufferers of this criteria and nothing has been done to improve this criteria in the last 14 years…..” another respondent raised questions on not defining streams in the current criteria, by stating: “why we can’t have separate streams for researchers and teachers in the faculty appointment criteria, we being teachers are spending our entire life for educating the next generation but are not eligible for selection in next grade” the responses of these deprived faculty members show that there is a dire need of reviewing the current appointment criteria, and that the involvement of stakeholders of different schools of thought need to be considered. it is evident by reviewing the current criteria that it is unidimensional and primarily focused on only one parameter of performance i.e. research publications. the respondents were also critical of no provision of promotions in the criteria i.e. each time the faculty members have to go through a selection process if they desire to achieve the next scale. 4.1.1 job experience requirements the respondents also criticized the job experience requirements, as it got considerably relaxed for those candidates who were active researchers and able to get their research published. “if one gets his or her phd at the age of 25 and he or she is a good researcher then within 05 years he or she gets to be an associate professor and in next 05 journal of applied economics and business studies, volume. 5, issue 2 (2021) 87-102 https://doi.org/10.34260/jaebs.526 95 years he or she can become a full professor, means at the age of 35 one can be a full professor, the only condition is to be a good researcher.” one of the other responded further elaborated this point by stating: “…tell me isn’t it be considered a clear discrimination and unjust to pressing the faculty to leave all other job responsibilities and just focus on research publications” it is evident from the responses that there should be a proper mechanism to devise minimum experience requirements for faculty members, by inculcating all the important parameters of performance measurement. 4.1.2 academic qualification criteria though all the respondents were phd, they were still criticizing the minimum qualification requirements set by hec in fac. one of the respondents was of the view that: “who said that only phd can be a good teacher and only he or she is eligible to become associate professor or professor” the other respondent was even more critical of this requirement as they state: “…look i am 49 now, i completed my phd in 2014 at the age of 45 to get eligible for associate professor, you can see my students’ feedback record, it’s excellent, but still i had to do ph.d.” it was revealed by analyzing responses that the faculty members were not satisfied with the qualification requirement set by hec for appointments as associate professor and professor. 4.1.3 research publications the most serious issue pointed out by all the respondents in the current criteria was related to the requirement of minimum research papers. while analyzing their responses, ethics and plagiarism were the most common issues related to publication requirements, highlighted by the respondents. some of the eye-opening comments of respondents have been coated below to throw light on the severity of the issue. “i have seen them (faculty members) forcing their supervisees to include their names as second author and even in many cases as first author; it is just because of fulfillment of that publication requirement” “it’s unethical, isn’t it? to pressurize your students, pressurizing your subordinates to include your name in the research paper; in which you have merely contributed” “…we, the teachers, the so-called builders of nation are turn out to be the most corrupt, we are at the height of academic dishonesty, just to get benefits for our self-promotion” it implies that the hec needs to reconsider the requirement of publications by incorporating the perceptions of faculty members. shoaib irshad, sadia irshad and sadaf kashif 96 4.1.4 no promotion, only selection the faculty members in general were praising this requirement that each time a faculty member has to go through a selection process, to get to the next designation. but the complexity and the delays involved in this process need to be taken care of. 4.2 lacking in criteria the respondents have also highlighted a number of shortcomings or lacking in the current appointment criteria. the first major deficiency, as raised by the respondents, was related to no proper mechanism for measuring performance of faculty members and the current criteria was termed as “unidimensional” as there is only one parameter of performance i.e., of research publications. further, the respondents also emphasized the inclusion of other important parameters such as excellence in teaching, services including administrative responsibilities, community services, and industrial linkages and so on. the respondents also put their weight on inclusion of separate streams for researchers, teachers and industrial experts so that a blend of well-equipped teachers can be recruited in the universities that would eventually enable to enhance the quality of teaching and education. 4.2.1 unidimensional in terms of performance measurement some of the responses of respondents have been placed below, which truly reflects the concerns of these faculty members on non-inclusion of other very important performance parameters in hec’s appointment criteria. “….what was the motto of hec while devising this criteria, whether research publication is the only responsibility of faculty member” “these criteria rate me as a bad teacher, because according to these criteria i am not eligible to get promoted as associate professor……” 4.2.2 no weightage to excellence in teaching the most critical and important duty of any faculty member is to teach, and his or her performance in teaching reflects how good or bad he or she is in teaching. universities, over the years, with the continuous support of hec have developed feedback mechanisms, through which each teacher is being evaluated by the student. the respondents were of the view that excellence in teaching should be the key pillar of performance measurement of any faculty, moreover it should also be considered while conducting selection boards of any faculty members. but, due to current appointment criteria, the teachers have shifted their roles from teaching to only research. as stated by one of the respondent: “go to the large public sector universities and check what is the actual credit hour load of any senior professor, you would get surprised as in majority cases they are not teaching even a single course” the other respondent highlighted the issue of ethics: journal of applied economics and business studies, volume. 5, issue 2 (2021) 87-102 https://doi.org/10.34260/jaebs.526 97 “what to say about our faculty members’ integrity, most of the times the senior professors have hired research assistants, you can call them teaching assistants and these teaching assistants are teaching their courses” one of the respondents blamed the hec’s faculty appointment criteria and hec’s ranking criteria for decline in the quality of education, he stated that: “we all are in the race of getting more and more publications, no one bothers to notice how well we are teaching, even our vice chancellors and deans are only interested in the number of research publications. this all is due to these requirements in hec’s criteria and weightage given to research in hec’s ranking criteria” it is affirmed from the respondents’ answers that excellence in teaching is the most critical missing link in the current appointment criteria. due to which most of the faculty members consider this criteria as non-friendly with respect to the sanctity of this profession. 4.2.3 no weightage to services (including administrative responsibilities, community services etc.) beside teaching and research, the teachers have to devote their services in other important tasks that include administrative responsibilities, development of industrial linkages and doing community services. it is the role of the university to act as a hub of community services in which the role of teacher is integral. further, being a teacher, he or she has to perform certain administrative duties such as dean, head of department, focal person of department for accreditation, program team member and many others. for which they have to perform several tasks related to these administrative duties, a good performance measuring mechanism should also gauge these responsibilities. 4.2.4 no privilege for linking industrial experience with academic experience the respondents were of the view that industrial experience should be linked with academic experience in order to encourage professionals to come and teach in the universities. as this would enable us to bridge the gap between industry and academia. one of the respondents summed it up by stating: “there is a dire need of linking industries with academia, without that our university education is going nowhere” 4.2.5 no relaxation for industrial gurus in qualification & publication requirements the respondents were emphasizing to include separate streams for professionals in the appointment criteria, so that they can come and join academia in reputed positions. for that the requirements of phd and research publications need to be relaxed. one of the respondents stated this in a very convincing way: “….look around the world, the gurus of their fields come to universities and teach courses even at doctoral level, though their own qualification could be of only shoaib irshad, sadia irshad and sadaf kashif 98 bachelors level. it is their rich experience that they can transfer, for which qualification doesn’t matter at all…..” 4.3 issues at the university end the most important issue raised by the teachers was related to the complexity of procedures adopted by the university for selection of faculty members. 4.3.1 sanctioned positions one of the important shortcomings at the university end, as highlighted by the respondents, is no provision of post in next grade. this is due to the pyramid designed by the university that each department would have such and such numbers of professors, associate professors and assistant professors in it. it hinders the chances of selection of other faculty members in higher ranks. as stated by one of the respondents: “….i have to leave my last job, as there was only one sanctioned position of associate professor in my discipline and one associate professor was already there. and agony is that, they were not willing to sanction another post” 4.3.2 procedural issues and procedural delays one of the respondents has explained the misery of procedural issues as: “….one of my colleagues got eligible for the next position two years back and within two months the advertisement was announced for that post, from then the shortlisting is not finalized yet….” it was revealed that in public sector universities the selection process takes at least 12 to 18 months, subject to availability of sanctioned positions. 4.4 issues at hec’s end related to promotion of research culture 4.4.1 limited local journals in social sciences on the other hand, as the respondents were from social sciences disciplines, they also highlighted the issues of scarcity of journals. one of the respondents further elaborated this issue by pointing out that: “if you look at the hec’s website for recognized journals, you will find that the number of research journals recognized in social sciences disciplines are way less than sciences discipline, most of the recognized journals are in z category, the lowest category of recognition, in which the published papers are not considered for appointments” 4.4.2 no bifurcation for social sciences and sciences continuing from scarcity of journals in social sciences, the respondents further elaborated their concerns on equal criteria of research publications for faculty recruitment for both: social sciences and sciences disciplines’ faculty of heis. according to the respondents, it brings a great deal of dissatisfaction amongst the social scientists when they find the criteria for evaluating faculty members of sciences and social sciences is the same. one of the respondents further elaborated this issue by pointing out that: journal of applied economics and business studies, volume. 5, issue 2 (2021) 87-102 https://doi.org/10.34260/jaebs.526 99 “…how come sciences and social sciences can be treated equally, when it comes to research outcomes in terms of recognized publications?” on the other hand, the other respondent pointed out the continuous neglect of hec for social sciences by stating: “since its inception in 2002, tell me what steps have been taken by hec to promote social sciences discipline, very few. so much so that the first thematic research grant for social sciences is being announced in 2016, whereas up till now the council for promotion of social sciences is still in pipeline.” 4.4.3 reputed international databases are not accepted for publications one major concern raised on hec by the respondents was non recognition of many well-reputed international databases of research journals. owing to this, they find a limited number of journals suitable for getting their research work published. 5. discussion the interpretive phenomenological analyses of the data revealed the perspective of faculty members regarding hec’s faculty appointment criteria. like previous studies (shah et al., 2012), the critical evaluation of the gathered data pointed to respondents’ serious concerns on the hec’s faculty appointment criteria, specifically related to the requirements mentioned in the criteria that are related to minimum qualification, minimum experience and minimum number of publications in hec recognized journals. the responses of the faculty members show that there is a dire need of reviewing the current appointment criteria, and for that they suggested the involvement of stakeholders of different schools of thoughts. the data analyses also highlighted the problems faced by heis’ social sciences faculty regarding the fulfillment of hec’s faculty appointment criteria. it is evident by reviewing the current criteria’s unidimensionality in-terms of its prime focus on one parameter of performance i.e. research publications. this study supports mahmood’s (2016) argument that the mission of hec should be to facilitate learning and education but this pivotal part has never been considered by any policy devised for assessing the performance of faculty members. the respondents also criticized no provision of promotions in the criteria. they emphasized that the faculty suffers owing to this as each time faculty members have to go through the selection process for the next scale. the findings of the study support khuram et al. (2017); however, this study concludes that hec’s policies affect motivation of faculty members negatively. although the faculty members in general praised no direct promotion criteria and also the requirement of going through the selection process, the delays involved in this process lead to frustration and anxiety. another problem is non-recognition of excellence in teaching as a criterion for recruitment. this most critical missing link in the current appointment criteria leads to many other problems for teachers and administration alike. an overambitious researcher may ignore his or her duties towards teaching as teaching performance is a non-existent entity in these criteria. shoaib irshad, sadia irshad and sadaf kashif 100 the study reveals that these criteria do not consider the role of university to act as a hub of community services in which the role of teacher is integral. further, it also ignores that being a teacher, he or she has to perform certain administrative duties such as dean, head of department, focal person of department for accreditation, program team member and the like. resultantly, the faculty considers such responsibilities as extra jobs and many try to avoid shouldering such services. therefore, hei’s faculty finds faculty appointment criteria a hindrance to their professional growth. 6. conclusion this study implicates that the policy makers at higher education commission may consider the perspectives of faculty members of higher education institutions on faculty appointment criteria relevant and may revisit it in decision making processes. since this study provides an insight to the problems the social sciences faculty faces in meeting with the set criteria, hec may consider it appropriate to revise faculty appointment criteria to better manage human resources at heis of the country. moreover, the study foregrounds the social sciences faculty’s concerns and worries related to faculty appointment criteria. in this way, it helps to elucidate and recognize the other important factors for gauging performances of faculty members of pakistani heis. insofar, teaching performance, faculty’s involvement in heis administrative services and other community services have not been the part of selection criteria. resultantly, this study implies an intention to review and revisit hec's faculty appointment criteria. this study is limited to public sector higher education institutions and specifically to those institutions that are located in islamabad region. due to diversity of disciplines and exemptions in hec’s faculty appointment criteria for some disciplines, such as engineering, information technology, medical, law and arts & design, the study was limited to faculty members of social sciences discipline. however, in order to better gauge problems of hec’s faculty appointment criteria a broader and detailed study may be conducted in future. the next researchers may involve private sector heis as well to better address this issue. moreover, if a future research involves a longitudinal and detailed quantitative and qualitative survey of all the heis of pakistan, it would provide with a more empirical and generalizable results of the gathered data. and such a study would better help the policy makers to take concrete steps in the light of proposed suggestions. furthermore, i recommend that the future research may involve a study of the perspective of all the stakeholders of hec’s faculty appointment criteria. so as to, come up with a profound plan, scheme and agenda for better human resource management in heis of pakistan. journal of applied economics and business studies, volume. 5, issue 2 (2021) 87-102 https://doi.org/10.34260/jaebs.526 101 references attride-stirling, jennifer. (2001). thematic networks: an analytic tool for qualitative research. qualitative research, 1(3), 385-405. centra, john a. (1977). how universities evaluate faculty performance: a survey of department heads. crawford, charles, burns, ronald, & mcnamara, robert hartmann. (2012). promotion to full professor: moving beyond tenure and associate professorship. journal of criminal justice education, 23(1), 41-64. creswell, john w, & poth, cheryl n. (2016). qualitative inquiry and research design: choosing among five approaches: sage publications. dill, david d, & van vught, frans a. (2010). national innovation policies: governments as innovation agents of higher education and research university research for innovation (pp. 107-124): economica. fairweather, james s. (2002). the ultimate faculty evaluation: promotion and tenure decisions. new directions for institutional research, 2002(114), 97-108. kang, eungoo, & lee, hyoyoung. (2021). employee compensation strategy as sustainable competitive advantage for hr education practitioners. sustainability, 13(3), 1049. khuram, waqas, bhutto, arabella, & jabeen, arifa. (2017). analyzing the impact of higher education commission policies on motivation of the faculty member of jamshoro education city. international journal of academic research in business and social sciences, 7(6), 232-257. mahmood, khalid. (2016). higher education commission pakistan. higher education commission (hec). malachowski, mitchell. (2010). unintended consequences: is cur undermining teaching at predominantly undergraduate institutions. cur quarterly, 31(1), 32-36. paulsen, michael b. (2002). evaluating teaching performance. new directions for institutional research, 2002(114), 5-18. rafi, muhammad, ahmad, khurshid, naeem, salman bin, khan, asad ullah, & jianming, zheng. (2020). knowledge-based society and emerging disciplines: a correlation of academic performance. the bottom line. shah, muhammad jamal, akhtar, gulnaz, zafar, huma, & riaz, adnan. (2012). job satisfaction and motivation of teachers of public educational institutions. international journal of business and social science, 3(8). silverman, david. (2015). interpreting qualitative data: sage. smith, jonathan a. (2011). evaluating the contribution of interpretative phenomenological analysis. health psychology review, 5(1), 9-27. tien, flora f. (2008). what kind of faculty are motivated to perform research by the desire for promotion? higher education, 55(1), 17-32. ward, kelly. (2003). faculty service roles and the scholarship of engagement. ashe-eric higher education report. jossey-bass higher and adult education series: eric. wiley, carolyn, wallingford, valerie, monllor-tormos, mireia, & konyu-fogel, gyongyi. (2016). faculty promotion in business schools: what counts and what should count? journal of higher education theory & practice, 16(4). shoaib irshad, sadia irshad and sadaf kashif 102 annexure semi structured questionnaire part-i (background information) 1. on which position are you currently working? 2. how long are you working on this position? 3. how long have you been serving in this hei? 4. did you get doctorate while serving in this hei? part-ii 5. what is your opinion about hec’s faculty appointment criteria? 6. what is your opinion about minimum qualification requirements in hec’s faculty appointment criteria for associate professor and/ or professor? do you agree or disagree with this requirement? 7. if disagreed to above question, what should be the minimum qualification for associate professor and/ or professor, in hec’s appointment criteria? (leading question) 8. what is your opinion about minimum experience requirements in hec’s faculty appointment criteria for associate professor and/ or professor? do you agree or disagree with this requirement? 9. if disagreed to above question, what should be the minimum experience for associate professor and/ or professor, in hec’s appointment criteria? (leading question) 10. what is your opinion for minimum research publication requirements in hec’s faculty appointment criteria for associate professor and/ or professor? do you agree or disagree with this requirement? 11. if disagreed to above question, whether there should be any minimum requirement of research publications for associate professor and/ or professor, in hec’s appointment criteria? (leading question) 12. why does hec consider research publications as the only performance measuring tool? what is your opinion in this regard? 13. what other factors should be included in current appointment criteria? 14. what should be the parameters for measuring faculty’s performance? 15. what are the key job responsibilities of faculty members? are they aligned with hec’s faculty appointment criteria? how? 16. how do you consider hec’s faculty appointment criteria from the perspective of its employee friendliness? 17. what is the role of student’s satisfaction in gauging faculty’s performance? 18. what is the role of community services and other services in gauging faculty’ performance? journal of applied economics and business studies, volume. 5, issue 4 (2021) 97-110 https://doi.org/10.34260/jaebs.545 97 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x dynamic institutions quality and sustainable development: evidence from developing countries muhammad azam1, zafir ullah khan2 & muhammad zubair3 1 department of economics, ghazi university, dera ghazi khan, pakistan 2 department of economics, university of science and technology bannu, pakistan 3 institute of management sciences, university of science and technology bannu, pakistan. abstract this article empirically examines the direct and indirect impact of dynamic institutions' quality on sustainable development. dynamic institution's quality is considered in the following forms; political and ethnic institutions, market legitimizing and regulating institutions and democratic institutions. for the empirical model, we use annual data set of 47 developing countries across the period 2000-2015. to address the issue of endogeneity, the dynamic panel system generalized method is employed. our findings suggest that ethnic conflicts have a negative significant impact on sustainable development while democracy shows an insignificant impact on sustainable development. market regulatory institutions are found to have a positive significant impact on sustainable development. in addition to this, the multiplicative impact of dynamic institutions through either economic growth efficiency or fdi channels indicates a positive impact on sustainable development except for the rule of laws. finally, we suggest that dynamic institutions' quality is necessary to enhance economic activities and attract foreign investment to ensure sustainable economic development in a specific set of developing countries. the policymakers should also focus on structural improvements regarding the rule of law enforcement for achieving sustainable economic development goals in selected developing countries. keywords dynamic institutions, ethnic conflicts, sustainable development, system-gmm jel classification q01, d02 muhammad azam, zafir ullah khan & muhammad zubair 98 1. introduction in recent decades, the development view has changed from simple development to sustainable development (sd). sd signifies "advancement that addresses the present age's issues without trading off the capacity of future ages to address their particular issues" (brundtland, 1987). it stimulates the adaptive capacity by way of proper allocations of resources (stafford-smith, 2017). existing policies of the united nations (un) mainly focus on sd and its implications for sustained factors productivity and natural resource conservation. accordingly, most of the empirical literature in development economics relates institutions with sd. as we move towards institutional dimensions, market institutions are in centre-stage in better implementing sd agenda and are now widely admitted dimensions towards sd(micic 2009, mitchell, wooliscroft, et al. 2010). increasing the share of political as well as the market institutional role is an important policy agenda for developing economies. market regulatory institutions are not well organized in most developing countries, to reduce consumption and increase resource efficiency. for global awareness regarding sd, the main target of most developing nations is to achieve an appropriate balance policy having the following objectives:1) to achieve sd agenda through efficiency market institutional role and, 2) to implement sd by political-institutional role. different components influence the elements of sd, including monetary, social and institutional variables ((polasky, kling et al. 2019). the new institutional schools of thought signify institutional factors in advancement to the concept of sd (veeman and politylo 2003). institutions are characterized as the "requirements that individuals force on themselves"(north 1991). a better institutional system generates a higher level of taxes that builds the financial capacity of the system and promotes sd (gambetta, azadian, et al. 2019). the existing literature focused to explain the political-institutional role and mixed findings have been reported. a few studies demonstrate a positive connection between institutions and sustainable development (berman, quinn et al. 2012, abreu, cunha et al. 2015, rosati and faria 2019). studies show that governance, rule of law, transparency and democratic quality positively influence sustainable development. however, other studies have found that the majority of institutions play a negative role in sustainable development (opschoor and van der straaten 1993). an issue of market institutional quality that has not received much attention in the economics literature is how market and political institutions matter for sd in developing countries? do improved market and political-institutional arrangements help in attaining sd in developing countries? the major focus of this study is to answer these questions specific to developing countries. given the above-mentioned background, journal of applied economics and business studies, volume. 5, issue 4 (2021) 97-110 https://doi.org/10.34260/jaebs.545 99 the objectives of this study are to examine the impact of democratic institutions, market regulations and ethnic conflicts on sustainable development of selected developing countries. the paper is organized as: section 2 discusses the existing literature relevant to dynamic institutions and sustainable development; section 3 explains econometric model and methodology; section 4 discusses the empirical results and the summary and conclusion are given in section 5. 2. literature review institutions are important to determine sustainable development. literature on sd shows that market institutions are helpful in the protection of property rights, facilitate sustainable development policies to enable total factor productivity and are used as instruments in achieving a higher level of sd. administrative quality is a major outcome for institutions which comprises several factors relevant to institutions’ policies that govern any country. moreover, administrative quality is designed to influence the incentive structure important for sustainable policies. therefore, it may influence sd and efficiency which contributes to the production and consumption aspect of the economy. following baker (2012) administrative quality encompasses five dimensions: the size of the government (government spending, taxes, and government enterprises); property rights and legal structure; effective monetary and fiscal policies; and trade policies and regulation of business (including labour and credit markets). in the case of sd, a major portion of investments is required for effective policy and the public as well as political acceptance is mandatory from regulatory bodies. the efficient level of market and political institutions prevent market failure and sustain economic development to mitigate natural resource distortions. dietz, neumayer, et al. (2007) investigated the impact of institutional quality and regular assets on adjusted genuine savings (proxy for sustainable development). the data set included from 1984 up to 2001 for gulf countries. the outcomes demonstrated that the corruption index has a positive role for adjusted genuine saving but its effect on natural resources is negative. carbonnier and wagner (2011) identified the positive impact of institutional quality on sustainable development. they utilized a data set of 108 developing countries over 24 years, starting from 1984 up to 2007. the study found a positive relationship between institutional quality and development. the study included political, power and effective checks, corruption, armed violence and conflict negative impact on sustainable development. corruption is found to have a negative effect in rich resources countries. stoever (2012) examined the effect of institutional quality on sustainable development using adjusted net savings (ans) as a muhammad azam, zafir ullah khan & muhammad zubair 100 proxy of sustainable development. for, institutional quality indicators, he used the voice of accountability, the effectiveness of government, corruption control, law and order, and political stability as proxy variables for a panel of 138 countries for the period 19702006. the general outcomes feature that institutional quality positively affects sustainability. barbier (2010) examined the long-run effect of corruption on sustainable development by applying the board information of african and asian nations for the period 1970-2003. the findings supported that corruption negatively affects the sustainable development agenda in africa, which implies that corruption control is effective for sustainable development for africa. they utilized the adjusted net saving for sustainable development. a comprehensive association has been found in the asian data set, but a different level of association for corruption and sustainable development relative to africa. abou-ali and abdelfattah (2013) examined sustainable development and intensity of natural resources for a panel of 62 countries using world development indicator data for the period 1990-2007 and found supporting evidence of environmental kuznets and resource curse hypothesis. the overall results argued that countries are not achieving sustainable development goals and the effect is negative on environmental quality. they used investment rates, inflation, and education and institution quality. the overall results indicated that countries having a well-developed rule of law is essential for institutional quality. the results showed fewer roles of institutions for environmental quality. carbonnier and wagner (2015), indicated that institutions have a weak effect on 104 developing countries' sustainability. they evaluated the negative role of institutions for sustainability. the study argued that excessive resource utilization negatively affects sustainability. the institutions are failed in developing countries, which have a negative influence on violence. the above-cited literature shows that there is limited relevant literature on the impact of democratic, ethnic and market institutional dimensions on sustainable development. the thirst of this study is to empirically examine whether democratic institutions, ethnic conflicts, and market-based institutions matter for sustainable development in developing countries or not. 3. model, data and estimation method adjusted net savings is output (ans) which is produced after considering inputs such as physical capital, investment, and institutional quality. our model analyses the journal of applied economics and business studies, volume. 5, issue 4 (2021) 97-110 https://doi.org/10.34260/jaebs.545 101 role of political and market institutions along with ethnic conflict impact on sd. using sustainable development as the dependent variable, the model is specified as: ititititit nczdians  ++++= 3210 (3.1) the above equation, shows adjusted net saving used as proxy of sustainable development using adjusted net savings as a percentage of gni(stoever,2012). on the right-hand side of the equation, the it di indicates a set of dynamic institutions as an independent variable, which consists of: (1) control over corruption(cc); (2) law and order(lo); (3) democracy by polity ii;(4) index of regulations (reg); and (5) polity relevant ethnic group share in a country (epr). it z shows a set of explanatory variables that include (gdp per capita, fdi as a percentage of gdp) and used in various studies(barbier, aidt 2010, abou-ali and abdelfattah 2013, venard 2013, carbonnier and wagner 2015). the study employed a panel of 47 countries because of the availability of data on the adjusted net savings data from the world development indicators (wdi). adjusted net savings are equivalent to net national savings which include education expenditure, subtracted net forest depletion, mineral depletion, and carbon dioxide and particular emission damages are excluded from this variable. data on institutional variables are taken from the international country risk guide (icrg) published by “the international political risk services group” (or prs). the related data set is commonly used for apprehending the stable quality of an economy in economic literature. for democracy, it calculates the governmental positivity from the general public. it is assumed that if there is low public positivity then the democratic government can perform its task more freely but will achieve less and face trouble if there is a more violent environment in the case of a non-democratic system in an economy. the assigned range started from 0, a representation of a very high level of risk and 10, allocation of a very low level of risk. democracy data is taken from the integrated network for societal conflict research (inscr). the market controlling institutions turn into a remarkable shield for declining the transaction costs which upgrades productivity that is more prominent and prompts sustainable economic outcomes. the essential reason is that efficient market institutions decrease transaction cost (anderson and swimmer 1997). ethnic disagreement is a clash in which people or groups/ parties try to identify themselves and others, including those from outside the group, in ethnic expressions and using ethnic operations. estimation methodology muhammad azam, zafir ullah khan & muhammad zubair 102 omitted variables biases are the main cause of the endogeneity issue. therefore, this study utilizes the dynamic system gmm model (gmm-sys) method for estimation. the controlled exogenous independent variables are not completely fulfilled as assumed and can be sorted out through proper utilization of instruments (islam 1995). for appropriate sorting out of time changing effects and endogeneity problem, the generalized method of moments (gmm) is the most appropriate method (caselli, esquivel, et al. 1996, bond, hoeffler, et al. 2001). arellano and bond (1991) devised a method based on two steps for an efficient way of integrating the instruments with the gmm method. the zero restrictions are supposed regarding moments' conditions in the model (arellano and bover 1995, blundell and bond 1998). in the second step, the variance-covariance matrix in an antithetical form can be presented where d_h and gmm estimators are written as: € ̂_gmm= (x^' z^* d_h z^(*') x)^(-1) x^' z^* d_h z^(*') y … … (3.3.10) the gmm estimators deliberated from the two steps arellano and bond (1991) which presents most analytical instruments for endogenous variables x_(i,t) i. empirical results the results are given in table 4.1, which are based on a panel of 47 developing countries, showing the impact of dynamic institutions on sustainable economic development. different diagnostic tests have been used to ensure the suitability of the model specifications. after using ar1 and ar2 tests, results show consistency regarding the validity requirements of the instruments in sys-gmm methodology. the estimated results show that epr is significant at the 10% level of significance. it indicates that a 10 percent increase in the epr decreases sustainable economic development by 0.9 percentage points in developing countries. ethnic conflict is a key challenge for developing economies. it negatively affects economic development through channels of income and health inequalities. gdp per capita is positively related to sustainable economic development. the gdp per capita coefficient is statistically significant and positive at a 1 per cent level in different model specifications (see tables 4.1). it suggests that for developing countries economic evolution is a preliminary condition for economic sustainability. the same kind of results has been mentioned in earlier studies (dietz, neumayer, et al. 2007). the estimated result shows that fdi inflows as a percentage of gdp are insignificant in developing countries for sustainable development. it indicates that fdi has no role in economic sustainability. the results of our study contrast with previous journal of applied economics and business studies, volume. 5, issue 4 (2021) 97-110 https://doi.org/10.34260/jaebs.545 103 studies which indicates that fdi inflows positively impacts sustainability (costantini and monni 2008, othman, jafari et al. 2014). multinational corporations (mncs) usually invest only with short-term profitability objectives and have no worries regarding economic sustainability in developing countries (abdul-gafaru 2009). law and order show a negative impact on sustainable economic development of developing countries. the estimated coefficient is -0.74 which is negative and statistically significant at 1 per cent level. the ineffectiveness of law and order on sustainable economic development is due to weak infrastructure systems in developing countries (romano 2009). as expected, the impact of control over corruption on sustainable development is positive and the findings indicate that a more efficient institutional framework can only be promoted through controlling corruption. an efficient institutional structure promotes productive economic activities through the effective utilization of economic resources that promote sustainable development in the entire economy. the index of the regulation (reg), reflecting market regulating institutions has a positive and significant effect on sustainable economic development. the study of kaldaru and parts (2005) supported this argument for sustainable economic development in developing countries. the results indicate a negative impact of democracy on sd. the reason is that democracy is not supportive of minimizing income distribution among people by way of health improvement and its lower capacity of economic sustainability programs in developing countries (de soysa, bailey, et al. 2012). also, we identify the indirect effects of dynamic institutions on sustainable development through gdp and, the fdi channel as mentioned in table 4.1 (columns 2 to7). the coefficients of gdp and dynamic institutions have a combined positive impact on sustainable development in developing countries. this confirms the hypothesis that dynamic institutions significantly improve sustainable development through economic growth in developing countries. further, our empirical results are quite interesting regarding the multiplicative term between multiple institutions and foreign direct investment in developing countries. multilevel institutions can improve sustainable development through the attraction of fdi inflows. 4. summary and conclusions the increasing importance of dynamic institutions and their association with sustainable development has remained a growing concern among policymakers and government officials from developing countries. a better way of dealing with muhammad azam, zafir ullah khan & muhammad zubair 104 sustainable development agenda is implementation issues and united nations have been actively engaged in advocating the institutional importance across countries. however, the majority of sustainable development activities in developing countries are still dependent on political factors. hence, sustainable development can only be achieved by focusing on market controlling institutions and ethnic conflict removal. based on the results of our study, market institutions and ethnic conflict promote sustainability in developing countries. given the importance of dynamic institutions, most developing countries around the globe, have started to prioritize dynamic institutional quality for sustainable economic development. above and beyond the direct importance of dynamic institutions, we have also evaluated the indirect importance of dynamic institutions via economic growth and fdi channel. after utilizing panel data set of 47 developing countries, the dynamic modelling approach i.e system gmm approach is used for estimation of cross-sectional variation. we further analyze dynamic institutions through direct and indirect channels on sustainable development at an aggregate level in developing countries. the analysis suggests that dynamic institutions matter for sustainable economic development in developing countries. therefore, a significant role of dynamic institutions will strengthen sustainable economic development. consequently, both political and market institutional quality lead to sustainable economic development in developing countries. the major implications of our study are as follows; ethnic conflicts normally deter sustainable economic development agenda in developing countries. democracy seems to be an ineffective and insignificant role in economic sustainability in developing countries. there are many structural flaws in the democratic election process, which make it ineffective for sustainable development agenda achievement in developing countries. in addition to this, the multiplicative impact of dynamic institutions through either economic growth efficiency channels or fdi is unsatisfactory from the rule of laws' sustainability point of view. the current finding of this study is also favor of prevailing literature that dynamic institutions is an important and significant factor in sustainable economic development and prosperity in developing countries (baker and mehmood 2015, avelino and wittmayer 2016). having a more effective institutional structure provide more freedom and allows individuals to think in more innovative ways and ensure efficient participation in sustainable economic development. accordingly, we argue that ethnic conflict roles should be discouraged for sustainable economic development formulation in developing countries. countries with more conflicts give the least priority to sustainable economic development. based on the result of the study, we can suggest journal of applied economics and business studies, volume. 5, issue 4 (2021) 97-110 https://doi.org/10.34260/jaebs.545 105 that, as dynamic institutional quality improve, governments and individual members in developing countries become capable of adopting sustainable economic developmentfriendly policies. to choose an alternative strategy, policymakers and regulators in developing countries should analyze ethnic conflict at the national level, in judging the impacts and risks of conflicts. thus, dynamic institutional management including ethnic conflicts is a key issue in the implementation of sustainable economic development in developing countries. the present study used a panel of homogenous countries and future studies can use the heterogeneity of dynamic institutions using both developed and developing countries. references abdul-gafaru, a. (2009). "are multinational corporations compatible with sustainable development? the experience of developing countries." ivanaj and mcintyre (eds.) multinational enterprises and the challenge of sustainable development: 50-72. abou-ali, h. and y. m. abdelfattah (2013). "integrated paradigm for sustainable development: a panel data study." economic modelling 30: 334-342. abreu, m. c. s. d., et al. (2015). "institutional dynamics and organizations affecting the adoption of sustainable development in the u nited k ingdom and b razil." business ethics: a european review 24(1): 73-90. aidt, t. s. (2010). "green taxes: refunding rules and lobbying." journal of environmental economics and management 60(1): 31-43. anderson, c. l. and e. swimmer (1997). "some empirical evidence on property rights of first peoples." journal of economic behavior & organization 33(1): 1-22. arellano, m. and s. bond (1991). "some tests of specification for panel data: monte carlo evidence and an application to employment equations." the review of economic studies 58(2): 277-297. arellano, m. and o. bover (1995). "another look at the instrumental variable estimation of error-components models." journal of econometrics 68(1): 29-51. avelino, f. and j. m. wittmayer (2016). "shifting power relations in sustainability transitions: a multi-actor perspective." journal of environmental policy & planning 18(5): 628-649. baker, s. (2012). politics of sustainable development, routledge. muhammad azam, zafir ullah khan & muhammad zubair 106 baker, s. and a. mehmood (2015). "social innovation and the governance of sustainable places." local environment 20(3): 321-334. barbier, e. the policy challenges for green economy and sustainable economic development, wiley online library. barbier, e. b. (2010). "corruption and the political economy of resource-based development: a comparison of asia and sub-saharan africa." environmental and resource economics 46(4): 511-537. berman, r., et al. (2012). "the role of institutions in the transformation of coping capacity to sustainable adaptive capacity." environmental development 2: 86-100. blundell, r. and s. bond (1998). "initial conditions and moment restrictions in dynamic panel data models." journal of econometrics 87(1): 115-143. bond, s. r., et al. (2001). "gmm estimation of empirical growth models." brundtland, g. h. (1987). report of the world commission on environment and development:" our common future.", united nations. carbonnier, g. and n. wagner (2011). "oil, gas and minerals: the impact of resourcedependence and governance on sustainable development." carbonnier, g. and n. wagner (2015). "resource dependence and armed violence: impact on sustainability in developing countries." defence and peace economics 26(1): 115-132. caselli, f., et al. (1996). "reopening the convergence debate: a new look at cross-country growth empirics." journal of economic growth 1(3): 363-389. costantini, v. and s. monni (2008). "environment, human development and economic growth." ecological economics 64(4): 867-880. de soysa, i., et al. (2012). "free to squander?: democracy and sustainable development, 1975– 2000." global environmental change and human security: 261. dietz, s., et al. (2007). "corruption, the resource curse and genuine saving." environment and development economics 12(1): 33-53. gambetta, n., et al. (2019). "the financing framework for sustainable development in emerging economies: the case of uruguay." sustainability 11(4): 1059. islam, n. (1995). "growth empirics: a panel data approach." the quarterly journal of economics: 1127-1170. journal of applied economics and business studies, volume. 5, issue 4 (2021) 97-110 https://doi.org/10.34260/jaebs.545 107 kaldaru, h. and e. parts (2005). "the effect of macro-level social capital on sustainable economic development." university of tartu faculty of economics and business administration working paper(42). micic, v. (2009). "industrial policy and development: the political economy of capabilities accumulation." econ. horiz 14: 53-55. mitchell, r. w., et al. (2010). "sustainable market orientation: a new approach to managing marketing strategy." journal of macromarketing 30(2): 160-170. north, d. c. (1991). "institutions." journal of economic perspectives 5(1): 97-112. opschoor, h. and j. van der straaten (1993). "sustainable development: an institutional approach." ecological economics 7(3): 203-222. othman, j., et al. (2014). "economic growth, foreign direct investment, macroeconomic conditions and sustainability in malaysia." applied econometrics and international development 14(1): 213-223. polasky, s., et al. (2019). "role of economics in analyzing the environment and sustainable development." proceedings of the national academy of sciences 116(12): 5233-5238. romano, m. k. (2009). the impact of migrant remittances on economic and social welfare in municipalities of el salvador, rutgers university-graduate school-newark. rosati, f. and l. g. faria (2019). "addressing the sdgs in sustainability reports: the relationship with institutional factors." journal of cleaner production 215: 1312-1326. stoever, j. (2012). "on comprehensive wealth, institutional quality and sustainable development-quantifying the effect of institutional quality on sustainability." journal of economic behavior & organization 81(3): 794-801. veeman, t. s. and j. politylo (2003). "the role of institutions and policy in enhancing sustainable development and conserving natural capital." environment, development and sustainability 5(3-4): 317-332. venard, b. (2013). "institutions, corruption and sustainable development." economics bulletin 33(4): 2545-2562.. appendix -a table 1: list of countries no. countries no. countries 1 albania 26 angola 2 argentina 27 cameroon 3 bolivia 28 china muhammad azam, zafir ullah khan & muhammad zubair 108 4 botswana 29 cuba 5 brazil 30 egypt 6 bulgaria 31 gabon 7 chile 32 ghana 8 colombia 33 guyana 9 costa rica 34 indonesia 10 india 35 iran 11 jamaica 36 jordan 12 namibia 37 malaysia 13 mexico 38 morocco 14 magnolia 39 pakistan 15 panama 40 senegal 16 paraguay 41 sudan 17 peru 42 syria 18 philippines 43 tunisia 19 romania 44 zambia 20 south africa 45 el salvador 21 sri lanka 46 ecuador 22 thailand 47 honduras 23 24 turkey uruguay 25 algeria table 2: data definitions and sources variable definition source adjusted net savings per capita the adjusted net savings per capita are derived from the division of adjusted net savings over the total population. worldwide governance indicators (wgi) control of corruption by aggregating several data sources, this index measures perceptions of the extent to which public power is exercised for private gain. higher values represent better (perceived) control of corruption. the published index ranges from −2.5 to +2.5; it has been rescaled here from 0 (for very poor control) to 10 (very high control). worldwide governance indicators (wgi) law and order this is variable is used to indicate the strength as well as neutrality of legal system in economy and further more adherence of public law. the data range is between 0 (which indicate a very high level of risk) up to 6 (which indicate a very low level of risk). journal of applied economics and business studies, volume. 5, issue 4 (2021) 97-110 https://doi.org/10.34260/jaebs.545 109 democracy (polityii) it measures the government responsiveness from the general public. the allocated range started from 0, an indication of a very high level of risk and 10, a sign of very low level of risk. market regulatory institutions the market regulating institutions turn into a notable hotspot for declining the transaction costs which improves productivity that is more prominent and prompts sustainable economic outcomes. political relevant ethnic group the ethnic conflict is a clash in whichever level where the people or groups/ parties involved try to identify themselves and others, including those from outside the group, in ethnic terms and using ethnic process. gdp per capita gdp per capita based on purchasing power parity (ppp). gdp is gross domestic product converted to international dollars using purchasing power parity rates. an international dollar has the same purchasing power over the gdp as the u.s. dollar has in the united states. foreign direct investment fdi net inflows are the values of inward investment made by nonresident investors in the reporting economy .inwards enterprises covers between resident in reporting economy, includes all liabilities and assets transfer to local residents and their direct investors. muhammad azam, zafir ullah khan & muhammad zubair 110 table 4.1the impact of institutions on sustainable development in developing economies: dependent variable (adjusted net savings as a percentage of gni) (1) (2) (3) (4) (5) (6) (7) (8) (9) variables ans ans ans ans ans ans ans ans ans epr -0.0889* -0.0459 -0.0436 -0.0234 -0.0455 0.0353 -0.0774** -0.165*** -0.203*** lgdp1 0.0426*** -0.0329*** 0.00872*** 0.0591*** 0.124*** fdi -0.000491 -0.000967 -0.00554*** 0.00125 -0.00260 lo -0.740*** -0.441*** -0.0421 -0.488*** -0.178* -0.0343 -0.0654 cc 0.0094*** 0.0142*** 0.0139*** 0.0143*** 0.0107*** 0.00728*** 0.00578*** reg 0.0856*** 0.0727*** 0.0958*** 0.0860*** 0.105*** 0.0681*** -0.0642*** polityii -0.0357*** 0.00331 -0.0157*** 0.0213*** -0.00395 0.0164*** 0.00880 lo_lgdp1 -0.0437*** cc_lgdp1 0.00164*** reg_lgdp1 0.0046*** polityii_lgdp1 -0.000396 lo_fdi -0.0455*** cc_fdi 0.000341*** reg_fdi 0.000860*** polityii_fdi 0.000989*** observations 676 676 676 682 676 676 676 682 682 numberof country1 ar(1) ar(2) 47 0.07 0.30 47 0.00 0.87 47 0.00 0.84 47 0.00 0.63 47 0.00 0.81 47 0.00 0.66 47 0.04 0.32 47 0.00 0.40 47 0.00 0.25 note: data period range from 2000-2015. robust standard errors are in parentheses. *, ** and *** denote significance at the 10, 5 and 1% level. journal of applied economics and business studies, volume. 5, issue 3 (2021) 43-58 https://doi.org/10.34260/jaebs.533 43 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x product line diversification and its impact on firm performance zia ur rehman phd1, asad khan 2, rafique ahmed khuhro3 and abdul ghafoor khan4 1assistant professor department of management sciences, university of haripur, pakistan 2lecturer, department of management sciences, university of haripur, pakistan 3assistant professor, department of management sciences, university of haripur, pakistan 4lecturer, department of management sciences, comsats university, abbottabad campus abstract the objective of the study is to measure product diversification’s impact on insurance firm’s financial performance in pakistan. analysis are carried out to examine how ownership structure, capitalization, group membership, firm size, diversification across business lines, industry concentration affects firm’s financial performance. data from 2009-2019 is collected to measure the impact of diversification (entropy) on the risk adjusted returns. findings of the study reveal that business line diversification has strong positive effect on firm performance (for both roa and roe) which means that diversified firms perform better than non-diversified firms. for managers these findings are useful as they propose the need for diversification, capitalization, increase in size and group affiliation to enhance firm profitability. keywords product line diversification, firm performance, return on assets, return on equity jel classification g32 1. introduction globalization, de-regulation of markets, advancement in technology has opened up borders for businesses across the globe. theoretically, product diversification is considered to be an important strategy for maximizing firm value and minimizing risk, but it also has certain costs associated with it. for example, the total cost for producing each unit of a product will be higher as compared to system where the production process is shared and associated with concentration (teece, 1980). the benefits of diversification can be realized if the internal governance mechanisms are efficient (williamson, 1985). in literature, product diversification has been analyzed with respect to property-liability or life-health insurance companies or companies that diversify across both types (meador, ryan & schellhorn, 2000; hardwick and adams, zia ur rehman, asad khan, rafique ahmad khuhro and ghafoor khan 44 2002; cummins, weiss, xie & zi, 2010; shim, 2011; che, liebenberg, liebenberg & powell; ai, bajtelsmit & wang 2018). despite several studies on the insurance industry, still there is limited evidence with respect to the performance effect of product diversification. several studies have focused on the analyzing the relationship between business line diversification and nonlife-insurers (elango, ma & pope, 2008; pavic & pervan, 2010; lee & kang, 2015; ai et al., 2018) to the best of our knowledge, we found no empirical evidence with respect to the product diversification’s impact on insurance firm’s performance in pakistan. since deregulation in the pakistani markets, many insurance companies attempted to diversify across non-life and life insurance products. the demand for both these insurance products is increasing as customers are realizing the need and importance of insurance, life or non-life in order to protect their future against uncertainties. this is the main motivation behind doing research on business line diversification’s impact on insurance firm’s performance as the empirical results till now are still unclear. hence, the study aims to analyze business line diversification’s impact on insurance firm’s performance in pakistan. following previous empirical studies and theory, this study attempted to test two theories: “strategic focus theory and conglomeration theory”. the strategic focus theory assumes that diversification has a negative effect on firm performance due to its associated costs whereas the conglomeration theory focuses on the benefits of diversification leading to positive effect on firm performance. the study contributes in the following ways. first, it adds to the existing literature by providing further evidence from a developing country as the empirical results pertaining to diversification’s influence on firm performance are mixed. second, for managers, the findings of the study are equally important in formulating and implementing business strategies, diversification plans in order to position them better in the market. third, this study helps in providing important information that will help policy makers at government and institutional level in forming polices aimed at incentivizing or disincentivizing product diversification, grouping, firm size, non-life diversification, competitive policies and capitalization that will greatly help in improving the financial performance of insurance firms. 2. literature review 2.1 the theory—firm value and diversification the views of business managers, stockholders and creditors greatly differ on the benefits of diversification. managers, for example, would like their respective firms to diversify internally or externally in order to minimize the overall firm-specific risk. likewise, creditors while protecting their own interests may want the firm to diversify its investments in such a manner that will increase the cash flow for the business thus reducing the risk delayed payments and loan default. additionally, shareholders journal of applied economics and business studies, volume. 5, issue 3 (2021) 43-58 https://doi.org/10.34260/jaebs.533 45 owning a diversified portfolio would not like the firm to diversify further if they are able to do this inexpensively in their own respective investment portfolios. 2.2 why diversify according to montgomery (1994) the three important reasons based on theory due to which firms prefer to diversify. these theoretical perspectives are the market power, resource-based view and agency theory. looking through the resource-based perspective, firms diversify into those areas where there is excess capacity of capabilities and resources that are easily and efficiently transferable across industries. here the focus is on scope economies whereby a diversified firm organizes economic activities efficiently (penrose & penrose, 2009). for instance, same marketing and distribution channels can be used to firm to market its products. similarly, firms may be able to utilize the services and skills of corporate legal and financial staff members to support business operations in various industries. matsusaka (2001) argued that organizations use diversification strategy to explore good matches between their organizational competences and external opportunities thus resulting in entering new business lines and leaving old ones that are not profitable. from the agency theory perspective, diversification strategy can be viewed as the pursuance of manager’s self-interest at the cost of shareholders. the theory further explains the possible motivation behind manager’s decision to diversify is that it is likely to (i) increase manager’s allowances and pay (jensen & murphy, 1990), increase in their stature and power (jensen, 1986); (ii) make their respective job positions more secure within organization through investment in projects that the specific skills that they possess thus making the firm dependent on them (shleifer & vishny, 1989); or (iii) reduce their own investment portfolio risk by lowering firm risk as managers are unable to lower their own portfolio risk through diversification (amihud & lev, 1981). lastly, the motivation for the firm to diversify is to seek market power. in a study, (villalonga, 2004) presented three non-competitive reasons for the firm to diversify. firstly, profits earned in one industry or sector can be used for predatory pricing in another by the firm. secondly, diversification will allow firms to conspire with other firms that are simultaneously in competition with the firm in different markets. lastly, firms may use diversification strategy to drive out small competitors through reciprocal buying with other large businesses. empirically, numerous studies have been undertaken to measure diversified firm’s performance as compared to specialized firms in the last three decades or so. hoyt and trieschmann (1991) were most probably the first researchers to study the impact of strategic focus and diversification on insurance firms. the findings of their study revealed that specialized insurance provider performed better over other insurers during the sample period. meador et al. (2000) analyzed link between output choices that the zia ur rehman, asad khan, rafique ahmad khuhro and ghafoor khan 46 firm makes and x-efficiency and concluded that diversification across various insurance and investment products lines by insurance providers led to greater xefficiency as compared to insurance providers that used a more focused production strategy. ai et al., (2018) found that enterprise risk management (erm) is an important determinant for property and casualty insurers and in insurance firms having quality erm programs, the effect of product diversification on firm performance was significant and positive. while providing proof on the validity of strategic focus theory versus the conglomeration theory berger, cummins, weiss and zi (2000) concluded that in some insurance providers conglomeration diversification hypothesis dominates whereas in others it is strategic focus hypothesis that dominates. moreover, liebenberg and sommer (2008) analyzed the effect of business line related diversification on firm performance and concluded that undiversified firms performed better than diversified firms. additionally, they also highlighted that size and capitalization positively affects performance i.e. unaffiliated insurers perform better than insurance groups and stock insurers performance is comparatively better than mutual insurers. cummins et al. (2010) analyzed the us insurance firms with particular focus on economies of scope. the focus of their analysis was to find out insurers who provide diversified are better off of insurers that provided specialized products. they concluded that although cost scope economies are realized by property-liability insurers, but they are quickly equalized by the resulting revenue scope diseconomies. in contrast, they find both revenue and cost diseconomies are experiences by firms that provide life-health insurance and concluded strategic focus is given more significance over diversification by insurance firms. krivokapić, njegomir, and stojić (2017) found that diversification impact on insurer’s performance is positive. ai et al. (2018) found significant positive effect on firm performance. che et al. (2017) also found that in terms of return on investment the performance of diversified firms is better as compared to nondiversified firms, but diversified firms underperform than non-diversified firms with respect to underwriting profitability. on the other hand, cummins and nini (2002), concluded that diversification has inverse effect on firm profitability. pavić and pervan (2010) and lee and kang (2015) also found that diversification has a significant negative influence on profitability. lee (2017) also analyzed product diversification’s impact on firm performance among insurance firms and found significant negative effect on insurer’s performance. shim (2011) focused on the merger and acquisitions related diversification’s impact on performance. the findings of the study revealed the acquirer firms experienced volatility in their earnings probably due to increased agency problems and frictions. li and greenwood (2004) concluded that mutual forbearance and market structuration is advantageous but not diversification. elango et al. (2008) found nonlinear and complex relationship between firm performance and product diversification. journal of applied economics and business studies, volume. 5, issue 3 (2021) 43-58 https://doi.org/10.34260/jaebs.533 47 they further added that the benefits of diversification for the firm are tied to the degree of insurer’s geographic diversification. berry‐stölzle, liebenberg, ruhland, and sommer (2012) examined business line diversification impact on firm performance and concluded that companies operating in volatile markets diversify less. with respect to managerial discretion hypothesis, they found support for diversification prediction when the measure of unrelated business line diversification was used and concluded that stock insurers should do more diversification than mutual insurers. the reason being, although the level of total diversification is very high for mutual insurers yet compared to stock insurers they engage is very less unrelated diversification. although most of the empirical results favour the strategic focus hypothesis over diversification hypothesis, still there is lack of consistency internationally. this, in addition to lack of similar studies generally in developing countries and particularly in pakistani insurance industry, worked as a motivator for conducting analytical study on this topic. based on the earlier studies and particularly on the study of liebenberg and sommer (2008), we attempt to analyze the strategic focus and conglomeration hypothesis h1 diversification positively influences firm performance (conglomeration) h2 diversification negatively influences firm performance (strategic focus) diversification and its relationship with firm performance relationship can be expressed in the following form: “performance = f (diversification; firm and industry characteristics)” (adapoted from liebenberg & sommer, 2008). hence, firm performance is also influenced by other industry and firm characteristics apart from diversification. the variables used in this study include firm performance (dependent variable) whereas the explanatory variables include product diversification (entropy) and firm specific and industry variables like firm size, ownership structure, capitalization, industry concentration. 3. methodology 3.1. sample and data since the focus of our study was on non-life insurance firms, therefore only nonlife insurance firms were considered. the final sample comprised of 28 firms and data for the selected firms was collected from 2009-2019. the reason for focusing on insurance firms is that in pakistan since deregulation of financial markets numerous opportunities arose which led to rise in new entrants into the insurance industry and many firms implemented diversification strategies to generate cost savings by cross selling products, entering into new business lines and markets, create hybrid products according to individual customer needs. moreover, some researchers are of the view since diversification acts as a natural hedging mechanism therefore, it potentially holds zia ur rehman, asad khan, rafique ahmad khuhro and ghafoor khan 48 a special value for insurance firms to reduce their reliance on formal hedging instruments for risk mitigation (cummins, philips & smith, (2001). for this study, those companies were selected that remained listed throughout the study period. data was collected from state bank of pakistan database which contains the financial data for all listed firms. data related to ownership structure, group affiliation was collected from company’s annual reports and websites. since data contained characteristics of both time series and cross-sectional data therefore panel regression was used. panel data is useful because “it reduces the collinearity between the independent variable, improves the efficiency of statistical estimates and also provide a large number data points” (saleh et al., 2008). the dependent variable representing firm performance is roa and roe whereas entropy (product diversification) is the main independent variable. in addition to this the model also contains other explanatory variables; firm size, specific line market concentration, share of life premiums, capitalization, and two dummy variables: gp and dmst. 3.2. dependent variable since the aim of the study is measuring product diversification’s impact on firm performance and firm performance being a dependent variable, therefore two common measures i.e. return on equity (roe) and return on assets (roa) are used to measure the performance of insurance firms. these measures are commonly used in literature for measuring insurer’s financial performance (amit & livnat, 1988; lai & limpaphayom, 2003; pottier & sommer, 1999; wang, jeng, & peng, 2007). estimations on both performance measures was carried which is consistent with methodology of liebenberg and sommer (2008). risk adjusted return of roa (sdroa) and risk adjusted roe (sdroe) was used to incorporate the impact of risk on returns. for this purpose, roa and roe was divided by standard deviation (sd) of observed roa and roe over the last five years. 3.3. explanatory variables product diversification which is an important variable in the context of our study is measured through the measure used by liebenberg and sommer (2008) i.e. “entropy” which differentiate firms operating in one line of business from firms operating many lines of businesses. furthermore, we used the “entropy measure (e) following elango et al. (2008) and pavić and pervan (2010). this entropy measures not only considers the number of products being offered by the company but also the weighted distribution of a company’s share in each product line. the entropy measure is calculated as e = ∑ pi ln il i=1 1 pi ………………………… (1) journal of applied economics and business studies, volume. 5, issue 3 (2021) 43-58 https://doi.org/10.34260/jaebs.533 49 where pi is the percentage of an insurer’s premiums written on product line i and il is number of insurance lines. if an insurance company operates in only one insurance line (i.e., exclusively focused company), the coefficient of entropy will take the value zero (e = 0). the maximum value that the entropy measure can take in a situation where a company has the same shares in all lines, is calculated as the natural logarithm of the number of insurance lines in the industry, ln(il)” (krivokapić et al., 2017). firm size also used in earlier studies positively affects firm performance (cummins & nini, 2002; liebenberg & sommer, 2008). natural logarithm of total assets is used for measuring firm size. based on empirical results from earlier studies we also expect positive influence of firm size on insurer’s performance. those insurance firms that are well capitalized are regarded as safe companies with respect to bankruptcy by rating agencies. for this reason, potential as well current insurance buyers are willing to pay additional premiums to the insurance providers as they are less likely to go bankrupt (sommer, 1996). capital to asset ratio is used as a measure for firm capitalization (cap). we expect a positive effect of firm capitalization on insurer’s performance. the structure–conduct–performance paradigm implies that market performance is determined by market conduct, which in turn directly and indirectly depends on market structure. according to s-c-p hypothesis, higher profitability in any market is associated with anticompetitive behavior induced by higher market concentration (njegomir & stojić, 2010). the support for the s–c–p hypothesis is abundant (see e.g.bajtelsmit and bouzouita (1998); chidambaran, pugel, and saunders (1997); elango et al. (2008); njegomir and stojić (2011)). to capture the impact of the competitiveness of firms’ markets on performance, we first calculate a herfindahlhirschman concentration index for each line of business (j = 1 to il) across all firms (i = 1 to n) in each year t hhijt = ∑ ( n i=1 dpwijt dpwjt ) ………………………….. (2) the larger the value of hhijt, the more concentrated is that line of business and ‘the greater is the potential for super-normal profits’ (liebenberg & sommer, 2008). we then calculate each firm’s (i = 1 to n) participation in each line of business (j = 1 to il) for each year t wijt = dpwijt dpwjt ………………… (3) using wijt as weights we then calculate the weighted sum of a firm’s exposure to industry concentration across all of the lines in which it operates zia ur rehman, asad khan, rafique ahmad khuhro and ghafoor khan 50 wconcit = ∑ wijt 28 j=1 hhijt ……………….. (4) firms with small values for wconc are exposed to competitive business lines”. we expect wconc to positively affect firm performance. in the modern-day financial services marketplace, group affiliation is an important firm characteristic that is very common now days (elango et al., 2008). there is a presence of both affiliated and unaffiliated insurance companies in the pakistani insurance market. consistent with the outcomes of earlier studies by cummins and sommer (1996), liebenberg and sommer (2008), we also expect group affiliation to negatively affect firm performance. group affiliation is measure through a dummy variable (gp) having a value equal to 1 if it a group affiliated and 0 otherwise. insurance companies in our sample have a diverse ownership structure i.e. public/private and majority owned by foreign or domestic owners. since we are considering only listed firms that are publicly owned therefore, we need to control ownership structure with respect to majority of ownership i.e. domestic or foreign. for controlling ownership structure in this respect, we used a dummy variable dmst which indicates that the firm is domestic majority owned. we assume companies that are majority foreign owned shall outperform companies that are majority domestic owned due to greater managerial expertise, know-how and financial strength (dorfman, 1998; puri, 2007). 3.4. descriptive statistics according to table 1, showing variable’s descriptive statistics the mean average of roa is 4.129% whereas as the mean average of roe 7.029%. the mean average of firm capitalization is 47.28%, the standard deviation values indicate that roa, roe and cap are more volatile as compared to other variables. table 1. descriptive statistics. mean median max min sd roa 4.129 4.820 145.360 -87.000 17.563 roe 7.029 10.360 162.910 -204.510 27.255 sdroa 0.234 0.273 8.222 -4.921 0.993 sdroe 0.258 0.381 5.989 -7.519 1.002 entropy 1.413 0.966 2.811 0.000 0.258 size 5276425 1421863 42287139 6306 8361969 cap 47.2872 44.86 161.64 -5.86 20.79188 journal of applied economics and business studies, volume. 5, issue 3 (2021) 43-58 https://doi.org/10.34260/jaebs.533 51 wconc 0.560 0.059 4.942 0.000 1.008 gp 0.597 0.988 1.000 0.000 0.474 dmst 0.890 1.000 1.000 0.000 0.413 note: return on assets (roa), return on equity (roe), risk adjusted return on assets (sdroa), risk adjusted return on equity (sdroe), firm size (size), firm capitalization (cap), exposure to industry competitiveness/concentration (wconc), group affiliation (gp), domestic majority owned (dmst) 3.5. regression models in this study a series of time series, pooled, time and cross-sectional ols regressions were performed. initially our focus was on finding out whether any type of diversification positively/negatively affects the performance of the firm. following a methodology similar to berger and ofek (1995) entropy is used as an indicator variable to identify whether “an insurance firm operates in one line (entropy = 0) or multiple lines (entropy = 1) in any particular year”. the basic equation to measure product diversification impact on firm performance is given below: sdroait = α0 + β1entropyit + β2capit + β3wconcit + β4sizeit + β5dmstit + β6gpit + μit …………… (5) sdroeit = α0 + β1entropyit + β2capit + β3wconcit + β4sizeit + β5dmstit + β6gpit + μit …………… (6) both above mentioned equations are estimated twice, initially with year dummies (ols1) followed by year and line dummies (ols2). in addition to ols other techniques are also used to test robustness of the model and results. however, recent empirical evidence suggests that diversification discounts in earlier studies has credited the discount observed to endogeneity bias. if entropy measure is not correlated with the error term, then the estimates of entropy measuring its impact on roa will be inconsistent and biased. the problem of endogeneity arises due to a number of factors like simultaneity bias, measurement errors, omitted variables etc. (wooldridge, 2016). for testing exogeneity of the entropy measure, hausman test is conducted, the results of which rejects the null hypothesis that entropy measure is exogenous. in literature we have found that researchers focusing on diversification discounts have used different methods for controlling endogeneity bias (campa & kedia, 2002; laeven & levine, 2007). one of these techniques is the fixed effects estimation that is commonly used by researchers in this regard. one of the main advantages of this technique is that it allows to control the omitted or unobservable firm specific effects which may be related with other independent variables used in the model. however, the fixed effects model is disadvantageous in situations due to its limited applicability where the key independent variables show sufficient within firm variations. “if the explanatory variables remain the same over time, they are swept away during the time-demeaning process which leads to the elimination of time-invariant unobserved effects or zia ur rehman, asad khan, rafique ahmad khuhro and ghafoor khan 52 explanatory variables that do not change substantially over time, the fixed effects technique can lead to estimates that are not precise and accurate” (wooldridge, 2016). because entropy being almost time invariant, therefore, fixed effect estimation is not used here. we have 28 firms in our sample. of these 28 firms 5 firms are those that only operate in one line at least in one year. off these 5 firms, 4 firms are those that operate only in single line during the period under investigation. thus variation occurs in only 1 firm or 20 percent of all firms. the second technique that is mostly used for controlling endogeneity bias is estimating equation 1 and 2 using a 2-stage least square technique and heckman technique (campa & kedia, 2002; laeven & levine, 2007). in 2sls technique, the entropy measure first is regressed on other explanatory variables in equation 1 and 2 and a set of instruments that do not apparently appear in equation 1 and 2. the second stage of 2sls comprises of estimating equation 1 and 2 using predicted values of the entropy measure derived from the first stage of 2sls. as far as the heckman approach is concern it follows procedure similar to 2sls but it also includes the selection of parameter on self-basis which is calculated from the information extracted from the first stage of 2sls. however, both methods require instrument selection of the entropy measure. campa and kedia (2002) suggested an “instrument set comprising lagged, current and historically averaged measures of industry growth, firm characteristics and general economic growth”. we followed same methodology as adopted by liebenberg and sommer (2008) for the constructing the instrumental variable set where instrument variable set “consists of lagged values of firm characteristics included in equation (1), 5-year historical averages of firm characteristics included in equation (1), 1-year growth in direct premiums written for the p/l industry, 1-year growth in pakistan gross domestic product, firm age, firm reinsurance use, and an index that captures the attractiveness of a firm’s markets to single-line insurers”. to be included in the instrumental variable set a candidate must need two important conditions. first, the candidate instrument must be relevant i.e. the candidate instrument must have higher partial correlation with entropy. the second condition is concerned with validity of the instrument, which means that the candidate instrument is unrelated with the error term in equation 1 and 2. wald test was used for the joint importance of excluded instruments to measure the relevance of instruments. as per the null hypothesis of wald test, the hypothesis was that instruments are jointly not significant. since different candidates were successful in relevance test for the instrument, therefore we manage to test the validity of the instruments for over identifying restrictions using hansen’s j-test. null hypothesis of hansen’s j-test is there is correlation between the instruments and the error term i.e. there are exogenous. the attractiveness of insurer’s market to single line insurance firm’s index, age and journal of applied economics and business studies, volume. 5, issue 3 (2021) 43-58 https://doi.org/10.34260/jaebs.533 53 reinsurance use are the only candidates that met the conditions of validity and relevance. 4. empirical results in this section, empirical results and discussion on the results is presented. initial results favour the conglomeration hypothesis, that is both the sdroa and sdroe have a positive relationship with entropy in all estimations, therefore we can conclude that product diversification effect on firm performance is positive. empirical studies from hoyt and trieschmann (1991), meador et al. (2000) and krivokapic et al., (2017) also found product diversification’s positive impact on insurer’s performance. secondly, firm size also has a significant positive influence on both measures of firm performance in all estimations. firm performance is positively affected by growth in total assets. these findings are consistent with respect to large firms as they exploit economies of scales and have comparatively lower default risk. empirically, liebenberg and sommer (2008) also concluded firm size positively impacting firm financial performance. table: 2 regression results (dependent variable sdroa) ols1 ols2 2sls heckman constant −0.893∗∗∗ −0.712∗∗∗ −0.599∗∗∗ −0.469∗∗∗ (0.024) (0.023) (0.025) (0.009) entropy 3.721∗∗∗ 2.989∗∗∗ 2.661∗∗∗ 2.311∗∗∗ (0.001) (0.003) (0.013) (0.007) size 0.137∗∗∗ 0.121∗∗∗ 0.108∗∗∗ 0.104∗∗∗ (0.000) (0.000) (0.002) (0.001) cap 0.020∗∗∗ 0.017∗∗∗ 0.013∗∗∗ 0.015∗∗∗ (0.004) (0.002) (0.012) (0.003) wconc 0.612∗∗ 0.501∗∗ 0.487 0.461∗∗ (0.049) (0.021) (0.005) (0.008) dmst -0.012 -0.01 -0.008 -0.004 (0.002) (0.001) (0.000) (0.001) gp −0.019∗∗∗ −0.018∗∗∗ −0.013∗∗ −0.011∗∗∗ (0.001) (0.001) (0.002) (0.001) hansen j-statistic 1.216 wald test statistic 378.1∗∗∗ self-selection parameter 0.032∗∗∗ (0.003) adjusted 𝑅2 0.23 0.24 0.27 0.25 significance level at 1%, 5%, and 10% is denoted by ∗∗∗, ∗∗, and ∗, respectively, p-value in parenthesis firm exposure to industry concentration measured through wconc has a strong positive effect on firm performance across all models. these findings are significant in zia ur rehman, asad khan, rafique ahmad khuhro and ghafoor khan 54 supporting our hypothesis that businesses operating in concentrated business lines earn higher profits by charging higher prices than businesses operating in less concentrated business lines. liebenberg and sommer (2008) and krivokapic et al., (2017) also found positive impact of firm exposure to industry concentration on firm performance. cap’s coefficient is positive and significant across all models for both measures of firm performance. this positive relationship supports the hypothesis that risk-averse policyholders paying higher prices to safe insurers will lead to a higher risk-adjusted performance. empirically, sommer (1996) also found positive effect of firm capitalization on firm performance. additionally, dummy variable dmst used to test whether domestic companies perform differently as compared to companies that are in foreign ownership is found insignificant. the coefficient of group affiliation (gp) is negative but significant. empirical studies from cummins and sommer (1996) and liebenberg and sommer (2008) also found negative relationship between group affiliation and firm performance. the negative value of gp can be attributed to lower prices made by the option to let a member fail, managerial discretion costs and other conglomeration related costs. table:3 regression results (dependent variable sdroe) ols1 ols2 2sls heckman constant −1.110∗∗∗ −0.994∗∗∗ −0.859∗∗∗ −0.773∗∗∗ (0.011) (0.009) (0.007) (0.009) entropy 2.931∗∗∗ 2.016∗∗∗ 2.009∗∗∗ 1.891∗∗∗ (0.000) (0.021) (0.006) (0.006) size 0.209∗∗∗ 0.197∗∗∗ 0.179∗∗∗ 0.176∗∗∗ (0.000) (0.000) (0.000) (0.003) cap 0.013∗∗∗ 0.012∗∗∗ 0.010∗∗∗ 0.008∗∗∗ (0.004) (0.001) (0.007) (0.002) wconc 0.591∗∗ 0.471∗∗ 0.458 0.398∗∗ (0.042) (0.026) (0.019) (0.005) dmst -0.010 -0.009 -0.007 -0.006 (0.080) (0.130) (0.112) (0.201) gp −0.015∗∗∗ −0.014∗∗∗ −0.010∗∗ −0.013∗∗∗ (0.000) (0.001) (0.001) (0.002) hansen j-statistic 1.049 wald test statistic 299.1∗∗∗ self-selection parameter 0.029∗∗∗ (0.004) adjusted 𝑅2 0.24 0.26 0.30 0.27 significance level at 1%, 5%, and 10% is denoted by ∗∗∗, ∗∗, and ∗, respectively, p-value in parenthesis journal of applied economics and business studies, volume. 5, issue 3 (2021) 43-58 https://doi.org/10.34260/jaebs.533 55 5. conclusion the objective of the study is to measure product diversification’s impact on insurance firm’s financial performance in pakistan. analysis were carried out to examine how ownership structure, capitalization, group membership, firm size, diversification across business lines, industry concentration affects firms financial performance. data from 2009-2019 is collected to measure the impact of diversification (entropy) on the risk-adjusted returns. based on our findings, we conclude that our results are similar to similar kind of studies conducted on developing countries. dummy variables apart from gp used in this study has an insignificant impact which emphasize on the need of urgent action to implement respective measures taken by the government to develop the insurance industry in particular and other financial services in general. for policy makers in pakistan, the findings are significant in the sense that currently the economy of pakistan desperately needs financial resources that would help in boosting economic growth considering the fact the economy is already overburdened by huge foreign and domestic debt and limited capital stocks available. for the management of insurance firms in pakistan, the findings of the study are equally important in formulating and implementing business strategies, diversification plans in order to position them better in the market. for government also the finding provides important information that will help in forming polices aimed at incentivizing or disincentivizing product diversification, grouping, firm size, non-life diversification, competitive policies and capitalization that will greatly help in improving the financial performance of insurance firms. overall performance of insurance industry is also of great interest to the government apart from the insurance firms themselves because greater good can be achieved socially and morally if the insurance industry performs well. as far as the limitation of the studies is concerned, one possible limitation of the study findings can be the insignificant relationship appearing in the models for most of the control variables. furthermore, research on the associated cost of both product as well as geographic diversification would be of great help to the managers in devising strategies that are cost effective thus leading better financial performance. future research should also focus on impact of geographic diversification on financial performance. additionally, future research can also conducted involving more countries which will help in getting more consistent cross-country estimators zia ur rehman, asad khan, rafique ahmad khuhro and ghafoor khan 56 references ai, j., bajtelsmit, v., & wang, t. (2018). the combined effect of enterprise risk management and diversification on property and casualty insurer performance. journal of risk and insurance, 85(2), 513-543. amihud, y., & lev, b. (1981). risk reduction as a managerial motive for conglomerate mergers. bell journal of economics, 12(2), 605-617. amit, r., & livnat, j. (1988). diversification and the risk-return trade-off. academy of management journal, 31(1), 154-166. bajtelsmit, v., & bouzouita, r. (1998). market structure and performance in private passenger automobile insurance. journal of risk & insurance, 65(3), 503-514. berger, a. n., cummins, j. d., weiss, m. a., & zi, h. (2000). conglomeration versus strategic focus: evidence from the insurance industry. journal of financial intermediation, 9(4), 323-362. berger, p. g., & ofek, e. (1995). diversification's effect on firm value. journal of financial economics, 37(1), 39-65. berry‐stölzle, t. r., liebenberg, a. p., ruhland, j. s., & sommer, d. w. (2012). determinants of corporate diversification: evidence from the property–liability insurance industry. journal of risk and insurance, 79(2), 381-413. campa, j. m., & kedia, s. (2002). explaining the diversification discount. the journal of finance, 57(4), 1731-1762. che, x., liebenberg, a. p., liebenberg, i. a., & powell, l. s. (2017). decomposing diversification effect: evidence from the us property-liability insurance industry. insurance markets and companies, 8(1), 16-28. chidambaran, n., pugel, t. a., & saunders, a. (1997). an investigation of the performance of the us property-liability insurance industry. journal of risk and insurance, 64(2), 371-382. cummins, j. d., & nini, g. p. (2002). optimal capital utilization by financial firms: evidence from the property-liability insurance industry. journal of financial services research, 21(1), 15-53. cummins, j. d., & sommer, d. w. (1996). capital and risk in property-liability insurance markets. journal of banking & finance, 20(6), 1069-1092. cummins, j. d., weiss, m. a., xie, x., & zi, h. (2010). economies of scope in financial sservices: a dea efficiency analysis of the us insurance industry. journal of banking & finance, 34(7), 1525-1539. dorfman, m. s. (1998). introduction to risk management and insurance: pearson education. elango, b., ma, y. l., & pope, n. (2008). an investigation into the diversification– performance relationship in the us property–liability insurance industry. journal of risk and insurance, 75(3), 567-591. journal of applied economics and business studies, volume. 5, issue 3 (2021) 43-58 https://doi.org/10.34260/jaebs.533 57 hardwick, p., & adams, m. (2002). firm size and growth in the united kingdom life insurance industry. journal of risk and insurance, 69(4), 577-593. hoyt, r. e., & trieschmann, j. s. (1991). risk/return relationships for life-health, propertyliability, and diversified insurers. journal of risk and insurance, 322-330. jensen, m. c. (1986). agency costs of free cash flow, corporate finance, and takeovers. the american economic review, 76(2), 323-329. jensen, m. c., & murphy, k. j. (1990). performance pay and top-management incentives. journal of political economy, 98(2), 225-264. krivokapić, r., njegomir, v., & stojić, d. (2017). effects of corporate diversification on firm performance: evidence from the serbian insurance industry. economic research-ekonomska istraživanja, 30(1), 1224-1236. laeven, l., & levine, r. (2007). is there a diversification discount in financial conglomerates? journal of financial economics, 85(2), 331-367. lai, g. c., & limpaphayom, p. (2003). organizational structure and performance: evidence from the nonlife insurance industry in japan. journal of risk and insurance, 70(4), 735-757. lee, c.-y. (2017). product diversification, business structure, and firm performance in taiwanese property and liability insurance sector. journal of risk finance, 18(5), 486-499. lee, s. u., & kang, j. (2015). technological diversification through corporate venture capital investments: creating various options to strengthen dynamic capabilities. industry and innovation, 22(5), 349-374. li, s. x., & greenwood, r. (2004). the effect of within‐industry diversification on firm performance: synergy creation, multi‐market contact and market structuration. strategic management journal, 25(12), 1131-1153. liebenberg, a. p., & sommer, d. w. (2008). effects of corporate diversification: evidence from the property–liability insurance industry. journal of risk and insurance, 75(4), 893-919. matsusaka, j. g. (2001). corporate diversification, value maximization, and organizational capabilities. the journal of business, 74(3), 409-431. meador, j. w., ryan jr, h. e., & schellhorn, c. d. (2000). product focus versus diversification: estimates of x-efficiency for the us life insurance industry. performance of financial institutions: efficiency, innovation, regulation, 175-199. montgomery, c. a. (1994). corporate diversificaton. journal of economic perspectives, 8(3), 163-178. njegomir, v., & stojić, d. (2010). determinants of insurance market attractiveness for foreing investments: the case of ex-yugoslavia. economic research-ekonomska istraživanja, 23(3), 96-110. zia ur rehman, asad khan, rafique ahmad khuhro and ghafoor khan 58 njegomir, v., & stojić, d. (2011). liberalisation and market concentration impact on performance of the non-life insurance industry: the evidence from eastern europe. the geneva papers on risk and insurance-issues and practice, 36(1), 94-106. pavić, i., & pervan, m. (2010). effects of corporate diversification on its performance: the case of croatian non-life insurance industry. ekonomska misao i praksa(1), 49-66. penrose, e., & penrose, e. t. (2009). the theory of the growth of the firm: oxford university press. pottier, s. w., & sommer, d. w. (1999). property-liability insurer financial strength ratings: differences across rating agencies. journal of risk and insurance, 621-642. puri, l. (2007). introduction to the trade and development aspects of insurance services and regulatory frameworks. trade and development aspects of insurance services and regulatory frameworks, 1-5. saleh, m., zulkifli, n., & muhamad, r. (2008). an empirical examination of the relationship between corporate social responsibility disclosure and financial performance in an emerging market. paper presented at the 16th annual conference on pacific basin finance economics accounting management (pbfeam). shim, j. (2011). mergers & acquisitions, diversification and performance in the us property-liability insurance industry. journal of financial services research, 39(3), 119-144. shleifer, a., & vishny, r. w. (1989). management entrenchment: the case of managerspecific investments. journal of financial economics, 25(1), 123-139. sommer, d. (1996). the impact of firm risk on property-liability insurance prices. journal of risk & insurance, 63(3), 501-514. teece, d. j. (1980). economies of scope and the scope of the enterprise. journal of economic behavior & organization, 1(3), 223-247. villalonga, b. (2004). diversification discount or premium? new evidence from the business information tracking series. the journal of finance, 59(2), 479-506. wang, j. l., jeng, v., & peng, j. l. (2007). the impact of corporate governance structure on the efficiency performance of insurance companies in taiwan. the geneva papers on risk and insurance-issues and practice, 32(2), 264-282. williamson, o. (1985). e., 1985, the economic institutions of capitalism: firms, markets, relational contracting. new york, ny: free press. wooldridge, j. m. (2016). introductory econometrics: a modern approach (5th ed.): nelson education. journal of applied economics and business studies, volume. 5, issue 2 (2021) 35-52 https://doi.org/10.34260/jaebs.523 35 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x safety performance and occupational injuries of workers (oiw): a dea efficiency analysis muhammad noman1*, ambreen fatima 2 and nooreen mujahid3 1 phd scholar, applied economics research centre (aerc), university of karachi 2 associate professor, applied economics research centre, university of karachi 3 associate professor, department of economics, university of karachi abstract the rapid pace of industrialization and sectoral transformation have not only induced rapid economic progress yet also engaged policy think tanks to consider the safety performance due to the increasing rate of injuries. these increasing workplace hazards have affected occupational efficiencies as well as worker’s performance. hence, a comprehensive analysis of occupation injuries of workers (oiw) is crucial to determine the safety performance of high and low-risk industries in pakistan. this study aims to incorporate the oiw for the estimation of the safety performance of industries employing data envelopment analysis (dea). this non-parametric technique allows calculating relative efficiencies incorporating inputs and outputs (both desirable and undesirable). the findings of the sbm-dea model and sensitivity analyses pointed out improvements in the farm sector and demanded more comprehensive analyses for the non-farm sectors. keywords safety performance, occupational injuries, dea, nonparametric technique, efficiency jel classification k32 1. introduction in the contemporary era, agricultural mechanization, the rapid pace of industrialization; and the restructuring of the services sector have not only induced economic growth yet also exerted social obligations (beriha, patanik, & mahapatra, 2011). considering developing economies, the farm sector mostly remains limited in the agriculture industry as it is the prime source of human incapacitation (bello, 2012). while the industrial sector still lags behind in terms of removing hazards of * noman.atiq19@gamil.com muhammad noman, ambreen fatima and nooreen mujahid 36 environmental pollution, labour exploitation, and occupation-related injuries (suri & das, 2016). consequently, the commodity-producing sector experiences huge social costs due to social pressure and finds it difficult to grab its true share in the market. on the other side, the services sector is not only doing good in terms of productivity yet also performing better in terms of safety performance (rodwell & toe, 2008) as compared to the traditional and manufacturing sectors. these contradictions at the sectoral level provide an intuition for the governments to set the target of zero accidents to ensure workplace safety for the disbursement of labor insurance benefits ensuring sectoral equity (yeh, 2017). hence, authorities and industries face pressure to keep workplace injuries at a minimal level (abdalla et al., 2017). for this purpose, evaluation of sectoral safety performance is considered vital for industries and organizations to retain efficiency and goodwill in the marketplace. these aspects exhibit multidimensional consequences in terms of competition and employees’ motivation ensuring higher profits and productivity. intuitively, a welldefined and efficient safety management setup can be implemented through safety practices, strategies, policies, procedures, functions, and roles with an integrated mechanism. this mechanism is supposed to control the hazards affecting worker’s health and safety at the workplace (labodová, 2004). meanwhile, safety measures are supposed to be assertively in compliance with the recent trends and legislation of the economy. additionally, worker’s involvement in the whole process is also a crucial requirement along with strong management support and commitment (zohar, 1980). given these industrial, organizational; and labour factors, occupational injuries of workers (oiw) can conveniently be utilized as a virtual measure to analyze the efficiency and safety performance (yeh, 2017). the most convincing argument of using the indicator of oiw points out the fact that injuries at work not only affect business activities yet also negatively affect the competitiveness of economies. hence, investigating the sectoral efficiencies through oiw is critical to improving the strategic paradigm with effective policymaking ensuring low or no injuries at the workplace. considering the economy of pakistan, the recent labour force survey (lfs) of pakistan, under an occupational injury, a worker who has gone through a work-related fatal and non-fatal injury can demand an economic return. this implies that both fatal and non-fatal accidents or injuries at the workplace exhibit different levels of severity demanding firms and governments to ensure better measures for worker’s safety. though, this fact had been realized quite early in pakistan as worker’s safety in industries has become a major concern for management ever since the enactment of the worker’s compensation act, 1923[online]. however, the execution of such interventions has been widely overlooked. the labour policy (2010) states that rationalization and consolidation of labour laws in pakistan are complex, anomalous, overlapping; and difficult to be implemented. the penalties prescribed for the noncompliance and offenses are very low and meanwhile being framed over the years. on the other side, the under-consideration pakistan occupational health and safety act journal of applied economics and business studies, volume. 5, issue 2 (2021) 35-52 https://doi.org/10.34260/jaebs.523 37 (2018) aims to develop a balanced and consistent framework for the health and safety of workers and workplaces. additionally, there exist several limitations in measuring the safety performance of occupations at the sectoral level. precisely, these include the impediments of data compilation, fund allocation, occupational health, and safety (ohs) setups, efficiency analyses and lack of performance indicators, etc. among all these the absence of performance indicators to measure sectoral efficiencies stands out. these insights demand the implementation of a holistic framework with a comprehensive analysis of safety performance at the sectoral level so that the efficient sectors can be highlighted. this may not only assess to analyze the efficiency of the sectors yet also provide plausible explanations for the better performance of the potential sectors and meanwhile to enhance the poor management setups in less efficient sectors. this paper measures the safety performance of 9 industries in pakistan through occupational injuries of workers (oiw), keeping in view the two core objectives. first, it follows an approach that incorporates oiw in the data envelopment analysis (dea) to evaluate the relative efficiencies at the industry level. second, it stresses the procedures to pace up the implementation process of safety strategies for inefficient industrial sectors in the context of pakistan. 2. literature review in literature, the safety performance of industries had been analyzed and measured through variety of research techniques indicating a broader horizon. considering the recent studies, namvar and bamdad (2021) utilized the resilience perspective as a theory of safety management to evaluate process industries. the authors integrated the theory with the dea methodology and found a strong evidence for the implementation of monitoring and assessment of risk in the sector. shin et al., (2021) examined the issues of construction industry indicating the need of research for the ohs regulations. hence, the authors measured the innovative efficiencies for korea through dea. the study indicated that companies with appropriate ohs facilities tend to be more inclined towards innovation thus stressed on appropriate measures from both practitioners and governments. zou et al. (2020) developed a comprehensive index for public safety incorporating outputs and inputs of the respective factor for the regions of china. the authors used the index to measure different efficiencies through dea procedure for the period of 2014-2018. the findings of this exercise revealed that at regional level the efficiencies were inefficient and redundant. the authors pointed that technological backwardness was the major facet in this respect and thus stressed on the improvement of public safety inputs through effective utilization of public safety resources. on the other side, yeh (2020) declared socioeconomic cost of occupational accidents as an important determinant for the safety and health development in an economy. for the estimation purpose, insurance payments had been used as proxy for the socioeconomic cost in different industries. the author explained that dea would be an appropriate methodology to measure the efficiency of industries. the findings of the muhammad noman, ambreen fatima and nooreen mujahid 38 study endorsed the significance of socioeconomic factors for the evaluation of health and safety for the economy of taiwan. however, it is worth mentioning here that all these recent studies had been examined through sophisticated data samples leading to divergent perspectives. intuitively, oiw are treated as accidents at job due to the substandard management set ups (tan et al., 2012). according to the available studies, most of the accidents were caused by poor management (yu & jiang, 2007) enhancing a sequence of such deficiencies. the other strand of literature precludes that the injuries among workers create stress and job dissatisfaction enhancing job switch rates (dawson & surpin, 2001). additionally, limited yet prominent economic literature considered workplace injuries as unwanted byproduct of the economic activity (el-mashaleh et al., 2010). hence, the negative connotation between oiw and economic indicators have started to contemplate job injuries as undesirable phenomenon for the economy. these negative outcomes of poor safety performance divulge the inefficiency of business maneuvers in an industry or a sector (egilmez & mcavoy, 2013). the discussion in this section covers the methodological diversities, sectoral divergences, and safety models to identify the research gap. turning to the quantitative assessments of occupational injuries, bulk of studies employed econometric modelling techniques to explore the related factors affecting the safety performance and efficiency of industries (feng at al., 2014; tan et al., 2012; ramli et al., 2011). whereas the other set of studies focused on the anticipation of errors in risky business set ups. however, these econometric techniques portray specific production functions and hence expected values are not necessarily supposed to provide sufficient assessments of the inefficiencies and safety loopholes. neither these analyses provide amended strategies to rectify the inefficiencies. contrary to econometric modelling, dea has been religiously considered for safety assessments as it is based on a simple presumption that fewer inputs enhance efficiency generating more desirable outputs. yeh (2017) developed dea model for 17 industrial sectors of taiwan and relative efficiency results declared mining and quarrying industry as the most inefficient sector. zhang et al., (2013) illustrated the advantages of slack-based measure (sbm) dea through graphs and modelled the slack inputs with undesirable outputs. beriha et al. (2011) benchmarked the occupational health and safety performance in industrial settings employing dea. the authors considered 3 industries and 30 organizations and found construction sector showed consistently low performance in terms of safety measurements. according to zhou et al. (2008), dea model was considered as a remarkable alternative for decision making of safety addressing the undesirability of outputs. feroz et al. (2001) employed dea to assess the economic costs of the occupational health and safety supervision in the textile industry. journal of applied economics and business studies, volume. 5, issue 2 (2021) 35-52 https://doi.org/10.34260/jaebs.523 39 the severity of workplace injuries also varies across industries or occupations as well pointing inclusion of occupational divergences in the model. yogyorn et al. (2020) estimated medical costs of occupational injury and illness (oii) for the agriculture sector of thailand. the authors declared that these were approximately 0.2 percent of the total agriculture output. hence, the study recommended government to use some of the oii funds for prevention programs in the farm sector rather focusing on paying compensations. hu et al. (2017) stated that given the desirability of high safety in industrial organizations, accident rates are still high. in this respect, the risk knowledge is not more than enough to bring about change in the unsafe behaviors (both employee and employers) due to irrationality of humans (saari, 1999). retzer et al. (2013) and witter et al. (2014) observed highest job fatality rate in the oil and gas extraction industry. whereas. tan et al. (2012) explored more deaths in the mining industry due to the perilous nature of the workplace conditions than the other observed occupations. on the other side, the studies of idrees et al. (2017) and harper and koehn (1998) larsson and field (2002) declared the construction industry as the most unsafe for the workers. thus, workplace injuries may be found in any setting, yet these are expected to be found in less efficient or higher risk industries. considering the diversity of capturing efficiency of industries, their exists various safety models and plans to reduce the occurrence of workplace accidents. arocena et al. (2008) explored occupational injuries considering occupational elements and perception practices for supple output technologies in spain. teo and linga (2006) modelled the effectiveness of safety management setups in the construction sector through employing safety index of technical, social and personal elements. these components were specifically related to the work safety technical model. dejoy et al. (2004) discussed that employee’s defiance also played an important part in determining safety issues. the authors also figured out that industrial injuries had not only affected the quality of human capital yet also generated huge financial losses. this would in turn lead to machinery damages, output disruptions, and bad firm’s reputation. further, the safety climate of an industry has been attributed to its compliance with safety measures (huang et al., 2006; dejoy et al., 2004; gershon et al., 2000). these include policies, workplace safety (perceived); and knowledge and ability to maintain safety (probst, 2004). mccauley-bell et al. (1999) developed a predictive model to detect risks of occupational injury and illness in modern workplace settings. the model was supposed to significantly help in controlling and preventing the surge in injuries and illnesses and thus minimizes the incidence and severity of these issues. khan and abbasi (1999) prepared a plan for hazard management and risk assessment in chemical industries. zurada et al. (1997) also proposed a diagnostic system in order to categorize the potential risks at jobs due to workplace designs. the author claimed that such system would be useful in injury anticipation and hazard analysis of industries. meanwhile, the system was effective in placing different jobs into high and low risk categories. muhammad noman, ambreen fatima and nooreen mujahid 40 given these facts, a brief review of literature reveals three plausible explanations. first, the quantitative or qualitative assessments of safety performance are diverse in nature depending on the objectives, industries, and occupations under consideration. second, dea method has been widely acknowledged to calculate the safety performance at sectoral or industrial level. third, to the bets of our knowledge, no such study has been performed to measure the occupational efficiencies through dea in the context of pakistan. 3. research methodology azadeh et al. (2016) claimed that efficiency or inefficiency of an industry is assessed by its decision making which is the choice of a technique that hefts alternatives and provides upgraded approaches for both workers and management. for the purpose, divergent parametric and non-parametric techniques have been employed for developing decision making modules. the dea is a non-parametric approach which considers the insight of decision-making unit (dmu). precisely, dea incorporates the idea of using less inputs for the high or same levels of output in order to measure the efficiency of any entity. however, the dea perspective can be incorporated through well-known divergent models including the charnes-cooper-rhodes (ccr), bankercharners-cooper (bcc), and slack-based measure (sbm). the ccr and bcc models are presumed to incorporate inefficient inputs or outputs and meanwhile are proportionally adjusted in the measurements. however, the non-radial nature of sbm model simultaneously integrates the inefficiencies ensuing from slack inputs and outputs (yu, 2010). further, this model also exhibits greater discerning power and the capacity to address undesirable outputs. this is the reason that the sbm model usually preferred than ccr and bcc models. in this respect, visbalcasdavid et al. (2017) explained that ccr and bcc models adopt either input or output orientation while the non-oriented model of sbm indulges the measurement of slack variables (through total input and output). besides, the primitive advantage of applying sbm is that it adequately allows to consider the slacks of inputs with both desirable and undesirable outputs in the production sets and models (hu & liu, 2017). hence, the principles of the model can be applied directly to measure the safety performance through efficiency of an industry. correspondingly, the sbm-dea model enables to segregate optimization constraints for both inputs and outputs. the other prominent advantage of using this method is that its application does not require any assumptions, price data and measurement units for any specific production functions while measuring efficiencies. therefore, the non-orientation procedure of the sbm model in this study allows to measure the safety performance through occupational injuries or slack variable analyses. the sbm model is considered suitable for this study to extract a strong complementary solution in the context of pakistan. precisely, the sbm-dea analysis specializes in measuring overall production efficiency incorporating all input and output journal of applied economics and business studies, volume. 5, issue 2 (2021) 35-52 https://doi.org/10.34260/jaebs.523 41 factors. further, as it mentioned earlier that no national literature yet utilized the sbmdea methodology to measure the efficiency of sectors in the occupation domain. given the suitability of the sbm model with the limitations of the study, the study incorporates a contemporaneous perspective of including oiw to measure the safety performance and efficiency of sectors in pakistan. hence, the study correctly and quantitatively modelled the innovative method of sbm -dea (hu & liu, 2017). 3.1 model the sbm-dea model treats occupational injuries rate as undesirable outputs. the study also presumes that there are n decision-making units to be assessed. every j decision-making unit (j = 1, . . . , n) use m inputs 𝑥𝑖𝑗 (i = 1, . . . , m) and gives rise p desirable outputs 𝑦𝑟𝑗 𝑑 (r = 1, . . . , p) as well as q undesirable outputs of occupational injuries rate 𝑦𝑘𝑗 𝑢 (k = 1, . . . , q, q = 1). so, the general efficiency (non-oriented) 𝜌 is determined through: 𝑀𝑖𝑛𝑖𝑚𝑖𝑧𝑒 𝜌 = 1 − 1 𝑚 (∑ 𝑠𝑖 − 𝑥𝑖0 𝑚 𝑖=1 ) 1 − 1 𝑝 + 𝑞 (∑ 𝑠𝑟 𝑑 𝑦𝑟0 𝑑 𝑝 𝑟=1 + ∑ 𝑠𝑘 𝑢 𝑦𝑘0 𝑢 𝑝 𝑘=1 ) (1) subject to: ∑ 𝜆𝑗 𝑛 𝑗=1 𝑥𝑖𝑗 + 𝑠𝑖 − = 𝑥𝑖0 (𝑖 = 1, . . . , 𝑚) (1.1) ∑ 𝜆𝑗 𝑛 𝑗=1 𝑦𝑟𝑗 𝑑 + 𝑠𝑟 𝑑 = 𝑦𝑟0 𝑑 (𝑟 = 1, . . . , 𝑝) (1.2) ∑ 𝜆𝑗 𝑛 𝑗=1 𝑦𝑘𝑗 𝑢 + 𝑠𝑘 𝑢 = 𝑦𝑘0 𝑑 (𝑘 = 1, . . . , 𝑞) (1.3) 𝜆 ≥ 0, 𝑠− ≥ 0, 𝑠𝑑 ≥ 0, 𝑠𝑢 ≥ 0 (1.4) where 𝑠𝑖 − is the slack input, 𝑠𝑟 𝑑 is the slack desirable output, and 𝑠𝑘 𝑢is the slack undesirable output. this model assumes, 𝜌∗ = 1 and 𝑠∗ = 0 and 0 < 𝜌∗ ≤ 1 are illustration of a specified decision-making unit with slack-based measure efficiency. by employing the optimal slacks in equation (1), the slack-based measure efficiency score 𝜌∗ can be disintegrated as mentioned below: 𝜌∗ = 1 − (∑ 𝛼𝑖 𝑚 𝑖=1 ) 1 + (∑ 𝛽𝑟 𝑑𝑝 𝑟=1 + ∑ 𝛽𝑘 𝑢𝑝 𝑘=1 ) (2) muhammad noman, ambreen fatima and nooreen mujahid 42 where: 𝛼𝑖 = 1 𝑚 𝑠𝑖 −∗ 𝑥𝑖0 (2.1) 𝛽𝑟 𝑑 = 1 𝑝 + 𝑞 𝑠𝑟 𝑑∗ 𝑦𝑟0 𝑑 (2.2) 𝛽𝑘 𝑢 = 1 𝑝 + 𝑞 𝑠𝑘 𝑢∗ 𝑦𝑘0 𝑢 (2.3) equation 2.3 is beneficial for calculating the magnitudes as well as sources of inefficient sectors concerning to the corresponding undesirable and desirable outputs as well as inputs of occupational injuries rate for a given decision-making unit. 3.2 data sources and variables intuitively, occupational injuries are assumed to be directly associated with the rapid increase in economic activity (li et al., 2011). therefore, scholars now prefer to include economic indicators to measure the safety performance of an economy at industrial level (el-mashaleh, & rababeh, 2010). given this insight, the study incorporates four variables to evaluate the efficiency of the industrial sector in pakistan. for inputs, gross fixed capital formation at sectoral level has been taken as proxy consumption of fixed capital (𝑥1)) while employed persons in a specific sector are considered to include the labour market turnover (𝑥2). for the estimation of desirable output and undesirable output, gross production value (𝑦𝑑 ) and occupational injury rate (𝑦𝑢) are considered, respectively. it is pertinent to mention here that the selection of these variables has been done by following the study of yeh (2017) for taiwan. table 1 provides a quick glimpse of the details of the variables. further, the study evaluates the safety performance of 9 main commodity producing and services sectors in pakistan for the year 2017-18. these sectors include agriculture, mining & quarrying, manufacturing, electricity generation and gas distribution, construction, wholesale and retail trade, transport and communication, finance and insurance and housing services (ownership and dwelling). table 1: description of input and output variables category id variables (unit) data source inputs 𝑥1 gross fixed capital formation (million rupees) pakistan economic survey 2017-18 𝑥2 employed persons (number in millions) labour force survey 2017-18 desirable output 𝑦𝑑 gross value of production (million rupees) pakistan economic survey 2017-18 undesirable output 𝑦𝑢 occupational injury rate (%) labour force survey 2017-18 source: compiled by authors journal of applied economics and business studies, volume. 5, issue 2 (2021) 35-52 https://doi.org/10.34260/jaebs.523 43 3.3 results this study examines the occupational injuries rate for safety performance, and the outcomes are summed up in the table 2. table 2: efficiency scores through slack-based measure model among sectors no. dmu 2017-18 ranks 1 agriculture 0.16 7 2 mining & quarrying 1.00 1 3 manufacturing 0.24 5 4 electricity generation and distribution & gas distribution 0.26 4 5 construction 0.19 6 6 wholesale & retail trade 1.00 1 7 transport & communication 0.33 2 8 finance & insurance 1.00 1 9 housing services (o.d) 0.27 3 average 0.4944 standard deviation 0.3821 source: estimated by authors table 2 depicts the ranks of the respective sectors based on the estimated efficiency scores of each sector. the results show that three out of nine sectors attain the efficiency score of 1 including mining and quarrying, wholesale and retail and finance and insurance. hence, these sectors can be used as benchmark to evaluate the efficiency or inefficiency of other sectors. based on the scores of efficient sectors it can be concluded that all other six sectors were inefficient in the given year of 2017-18. showing the consistency with the data, the sbm measurement affirms that agriculture is the most inefficient sector among all with the lowest score of 0.16. additionally, the average efficiency score of all sectors remains to be 0.4944. on the whole, these results indicate that pakistan’s sectors persist to have a room to enhance their efficiency through ensuring better safety measures and performance. turning to the sbm, it is presumed to evaluate the performance and improvement of the under-consideration indicators (zhao et al., 2018). therefore, in order to evaluate the inefficient variations of industries, the slack based analysis has been performed. table 3 portrays the projection values of inputs and outputs for inefficient sectors of industry along with potential addition in outputs given the potential reduction in inputs. column 3 and 4 reveals overall inefficiency of input (x1 and x2). column 5 and 6 explains slacks and projection targets of indicating that whether an increase in output or decrease in inputs results in efficiency. precisely, columns 5 is used to analyze that how the current level of output can be produced with less inputs. whereas column 6 evaluates the production of higher level of output with current level of inputs. additionally, column 7 and 8 decomposed the inputs and outputs, respectively. muhammad noman, ambreen fatima and nooreen mujahid 44 the potential improvements of the sbm-dea show areas of improvement in input-output activity needed to put an inefficient sector onto the efficient frontier. to move onto the efficient frontier, for the agriculture sector, the labour can be reduced by 78.27% and the injury rate can be reduced by 85.03 percent improving total farm output by 179.65 percent. likewise, in construction (which is 2nd most inefficient sector) needs to reduce its capital input by 36.61 percent and use 68.25 percent less injuries enhancing its share by 235.71 percent. the similar conclusions can conveniently be drawn for other inefficient sectors. the performance of the three efficient sectors (finance & insurance, mining & quarrying, and wholesale & retail trade) is up to the mark. this can be evident from column 3 and 4 where the inefficiencies of input and output are zero. the primary sources and magnitudes of most inefficient industrial sectors is presumed to be labour issues and injury rates. davies et al. (2009) observed and explained that injuries are higher in high labour employment sectors. indeed, higher number of workers can lead to greater risk of injuries, which can undermine employees’ safety awareness and health. this information may increase the pakistani government’s understanding of the improvements required for each industrial sector and enable it to make subsequent improvements. journal of applied economics and business studies, volume. 5, issue 2 (2021) 35-52 https://doi.org/10.34260/jaebs.523 45 table 3: sbm projections and inefficiency decomposition n o (1) dmu (2) efficienc y score (3) input (4) output (5) slack (6) projection (7) input reduction (%) (8) output reducti on (%) (9) outpu t additi on (%) inefficienc y inefficienc y x1 x2 yu yd x1 x2 yu yd x1 x2 yu yd 1 agriculture 0.16 0.54 1.796500 0.00 -20.42 35.37 4198046.4 1 100065 9.00 5.6 7 6.23 653481 7.41 0.00 78.2 7 85.03 179.65 2 construction 0.19 0.35 2.357100 -27574.32 0.00 11.81 824478.62 47745.6 8 4.5 1 5.49 117425 6.60 36.61 0.00 68.25 235.71 3 electricity generation and distribution & gas distribution 0.26 0.19 2.116000 -87651.00 -0.06 0.00 464405.00 155350. 00 0.2 0 0.60 683868 .00 36.07 21.9 5 0.00 211.61 4 finance & insurance 1.00 0.00 0.000000 0.00 0.00 0.00 0.00 64266.0 0 0.3 6 0.40 419690 .00 0.00 0.00 0.00 0.00 5 housing services (o.d) 0.27 0.25 1.802800 -159708.12 0.00 -2.11 1457090.7 5 514596. 88 0.6 5 1.99 226531 2.75 23.68 0.00 51.52 180.28 6 manufacturing 0.24 0.49 1.079900 0.00 -6.43 13.57 1814408.3 6 535114. 00 3.0 3 3.33 349456 9.36 0.00 67.9 5 80.29 107.99 7 mining & quarrying 1.00 0.00 0.000000 0.00 0.00 0.00 0.00 77675.0 0 0.1 0 0.30 341934 .00 0.00 0.00 0.00 0.00 8 transport & communication 0.33 0.50 0.489900 0.00 -2.39 -5.69 791725.02 546958. 00 0.7 0 2.11 240777 0.02 0.00 77.4 7 72.92 48.99 9 wholesale & retail trade 1.00 0.00 0.000000 0.00 0.00 0.00 0.00 95622.0 0 9.0 3 11.0 0 235172 6.00 0.00 0.00 0.00 0.00 muhammad noman, ambreen fatima and nooreen mujahid 46 table 4: outcomes of the sensitivity analysis dmus efficient dmus mining & quarrying construction wholesale & retail trade finance & insurance housing services score score score score score agriculture 0.359624 1 0.421506 0.359624 0.359624 mining & quarrying 1 1 1 1 manufacturing 0.518651 1 0.518651 0.52718 0.518651 electricity generation and distribution & gas distribution 0.51 0.344935 0.344935 0.344935 0.344935 construction 1 1 1 1 wholesale & retail trade 1 1 1 1 transport & communication 0.710095 1 0.673265 0.673265 0.673265 finance & insurance 1 1 1 1 housing services (o.d) 1 1 1 1 source: estimated by authors in addition to the sbm-dea outlook, the study has also performed sensitivity analysis in table 4. this mathematical exercise takes in to account the stability of efficiency scores by omitting an efficient sector and subsequently alters a reference set for the respective sector. the outcomes showed that efficiency scores of efficient sectors (mining and quarrying, construction, wholesale and retail trade, finance and insurance, and housing services) considerably affected the efficiency magnitudes of inefficient sectors. this implies that the omission of these efficient sectors would produce better scores for the inefficient sectors at the level. for instance, the omission of mining and quarrying sector from dmus improves the efficiency of electricity generation and gas distribution from 0.345 to 0.51. further, the transport and communication sector also experienced a slight increase from 0.673 to 0.71. similarly, the omission of other efficient sector (construction) enhances the efficiency scores of all other sector except the electricity and gas distribution (remains same). the exclusion of wholesale and retail trade improved the efficiency of the farming sector while all other sectors remained same in terms of efficiency scores. finally, the removal of finance and insurance and housing sector did not have any impact on other sectors of the economy. therefore, on the whole, the sensitivity analysis declared the construction sector as the most influential or safe occupation for workers among all other dmus. intuitively, the analysis points out two plausible explanations for the study. first, the efficient performance of efficient and less risky industries would have reasonable outcomes for the efficiency of high-risk sectors. second, grouping of divergent sectors would demand a unified, well defined, and an integrated policy mechanism for health and safety of workers in the economy of pakistan. journal of applied economics and business studies, volume. 5, issue 2 (2021) 35-52 https://doi.org/10.34260/jaebs.523 47 4. discussion the dea analysis of all nine sectors of pakistan has assessed to utilize the oiw in an innovative manner indicating rudimentary facts of the occupational domain. the divergence in the outcomes of sbm-dea and sensitivity analyses not only explained the variations at occupation level yet also pointed out the need to perform more efficiency analyses in order to evaluate the performance of the sectors at more disaggregated level. additionally, the divergence at the sectoral level points out the fact that still the agriculture and manufacturing sectors lag behind given its potential due to high inefficiencies. this finding is consistent with the study of yogyorn et al. (2020). despite the fact that the labour policy (2010) of pakistan includes core laws for industries, employment, service conditions; and occupational safety and health, the implementation of the policies have not been achieved yet in these sectors. meanwhile, the policy considers the construction industry as a backbone of all the development activities, and this could be the reason that it performs well in terms of efficiency in the dea evaluation. however, a detailed analysis of the sector has been recommended as the construction industry of pakistan is going through new challenges and problems due to rapid expansion (harper and koehn, 1998). it is a well-evident fact that the expansion of this particular industry causes health, safety, and occupational hazards (witter, 2014). this could be the reason that there has always been stressed by policy makers to ensure safe working conditions for the workers in the construction industry. in this respect, as prescribed earlier by teo and linga (2006) and probst (2004) in the literature, measures of workmen’s compensation, social security, old age pension interventions have been proposed to ensure the worker’s safety. considering the declining share of agriculture sector in the total output and increasing inefficiencies, the sector persists to move towards the rapid mechanization demanding technical expertise of farmers (devkota et al., 2020). consequently, unskilled farm workers are becoming unemployed (rao, 1972). additionally, no labour laws have been implemented in the sector so far depriving the benefits of various welfare legislations. these legislations were fully or partly been offered to the counterparts of farmers in the non-farm establishments. though, the governments over the years tend to extend the coverage of workmen’s compensation act, 1923 for the farm sector which states that in case of any injury or death to workers of mechanized farm in the rural sector. according to forastieri (1999), the better performance of the services sector can be attributed to the fact that informal economy (including domestic workers) is persistently improving health and safety, and social security arrangements. whereas the payments of minimum wages also contributed to this regard as the employer-employee relationship is much more evident here than the other sectors of the economy (baig, 2005). muhammad noman, ambreen fatima and nooreen mujahid 48 the sensitivity analysis of the study demands an advocacy and harmony in the rights of workers (dawson & surpin, 2001) given the occupational diversity. this goal can easily be attained through promoting the concept of providing decent work (blackett, 2011). the labour policy (2010) envisages a harmonious working relationship between workers and employers to enhance the performance and efficiency of the industry sector. this efficiency analysis offers several vital contributions in terms of prevailing knowledge and new insights which may lead to better theoretical perceptions and methodologies. correspondingly, the existing literature is not enough for precise assessments of efficiencies at occupational and industrial levels. predominantly, qualitative studies could not correctly measure the efficiencies and safety performance leaving a research gap for quantities investigations. the sbm-dea perspective not only contributes to the quantitative investigation yet also adds up slack indicators. this would tend to reduce resource consumption minimizing hazards together with enhancing production value. hence, with the help of the study findings, both theoretical and practical insights underpin the utilization of dea in measuring efficiency at the sectoral level (song et al., 2012). 5. conclusion the present study endeavors to evaluate 0the safety performance of industries through employing oiw. for this purpose, the sbm-dea perspective has assessed to derive relative efficiencies of all the nine available occupations for the year 2017-18. meanwhile, the sensitivity analysis adds up an insight to differentiate between high-risk and low-risk occupations. in a nutshell, the agriculture sector is found to be the inefficient sector with low scores while the construction sector was efficient among all other nine sectors. besides, the services sector has shown better safety performance in terms of efficiency and considered relatively safe for workers in pakistan. hence, an integrated safety policy is required to reduce workplace hazards with a policy design with the optimal set of measures. the outcomes of the study could also be helpful for the authorities for upgrading safety performance established on the situations in each industry. to attain these objectives, the study recommends a few doable policy options. the labour policy must oblige the employers to follow rules for working hours with an incentive of worker’s training specifically in high-risk industries. using inefficiency corrosions, the principal sources and magnitudes of highly inefficient manufacturing sectors must also be incorporated. additionally, expanding safety investment in the professional and management personnel could contribute to less severe injuries and deaths. future research may consider the effects of other economic factors on the safety performance of business operations as the incidence of workplace injuries varies with economic fluctuations. considering the socio-economic protection of labour, a strategy must be progressed and tracked in compliance with the idea of decent work in the labour market. journal of applied economics and business studies, volume. 5, issue 2 (2021) 35-52 https://doi.org/10.34260/jaebs.523 49 this study does not address occupational diversity through dea as the information of respective injury rates is not available. further, the available dataset incorporates only nine major sectors ignoring the heterogenous sectors in terms of injury rates. last, the dynamics of the oiw have not been considered due to simplicity purposes. given these data limitations, it can be recommended that future studies may incorporate the survey data analyses for capturing sectoral diversities. besides, the dea model of the study can be extended using the dynamic dea to quantify variations in efficiency scores over time. the model could also be employed to explore the effects of conventional components of a business on safety performance. reference abdalla, s., apramian, s. s., cantley, l. f., & cullen, m. r. (2017). occupation and risk for injuries. in injury prevention and environmental health. 3rd edition. the international bank for reconstruction and development/the world bank. arocena, p., núñez, i., & villanueva, m. (2008). the impact of prevention measures and organisational factors on occupational injuries. safety science, 46(9), 13691384. azadeh, a., zarrin, m., & hamid, m. (2016). a novel framework for improvement of road accidents considering decision-making styles of drivers in a large metropolitan area. accident analysis & prevention, 87, 17-33. baig, z. (2005). employer-employee relationship in pakistan. market forces, 1(1). bello, s. r. (2012). agricultural machinery hazards & safety practices. create space independent publishing platform (26 may 2012) beriha, g. s., patnaik, b., & mahapatra, s. s. (2011). safety performance evaluation of indian organizations using data envelopment analysis. benchmarking: an international journal. blackett, a. (2011). introduction: regulating decent work for domestic workers. canadian journal of women and the law, 23(1), 1-46. davies, r., jones, p., & nuñez, i. (2009). the impact of the business cycle on occupational injuries in the uk. social science & medicine, 69(2), 178-182. dawson, s., & surpin, r. (2001). direct-care health workers: the unnecessary crisis in long-term care. report for the domestic strategy group of the aspen institute, by the paraprofessional healthcare institute, washington, dc: the aspen institute. dejoy, d. m., schaffer, b. s., wilson, m. g., vandenberg, r. j., & butts, m. m. (2004). creating safer workplaces: assessing the determinants and role of safety climate. journal of safety research, 35(1), 81-90. devkota, r., pant, l. p., gartaula, h. n., patel, k., gauchan, d., hambly-odame, h., ... & raizada, m. n. (2020). responsible agricultural mechanization innovation muhammad noman, ambreen fatima and nooreen mujahid 50 for the sustainable development of nepal’s hillside farming system. sustainability, 12(1), 374. egilmez, g., & mcavoy, d. (2013). benchmarking road safety of us states: a deabased malmquist productivity index approach. accident analysis & prevention, 53, 55-64. el-mashaleh, m. s., rababeh, s. m., & hyari, k. h. (2010). utilizing data envelopment analysis to benchmark safety performance of construction contractors. international journal of project management, 28(1), 61-67. feng, y., teo, e. a. l., ling, f. y. y., & low, s. p. (2014). exploring the interactive effects of safety investments, safety culture and project hazard on safety performance: an empirical analysis. international journal of project management, 32(6), 932-943. feroz, e. h., raab, r., & haag, s. (2001). an income efficiency model approach to the economic consequences of osha cotton dust regulation. australian journal of management, 26(1), 69-89. forastieri, v. (1999). improvement of working conditions and environment in the informal sector through safety and health measures. geneva: international labour office, 1-17. gershon, r. r., karkashian, c. d., grosch, j. w., murphy, l. r., escamilla-cejudo, a., flanagan, p. a., ... & martin, l. (2000). hospital safety climate and its relationship with safe work practices and workplace exposure incidents. american journal of infection control, 28(3), 211-221. harper, r. s., & koehn, e. (1998). managing industrial construction safety in southeast texas. journal of construction engineering and management, 124(6), 452-457. hu, x., & liu, c. (2017). slacks-based data envelopment analysis for eco-efficiency assessment in the australian construction industry. construction management and economics, 35(11-12), 693-706. huang, y. h., ho, m., smith, g. s., & chen, p. y. (2006). safety climate and selfreported injury: assessing the mediating role of employee safety control. accident analysis & prevention, 38(3), 425-433. idrees, m. d., hafeez, m., & kim, j. y. (2017). workers’ age and the impact of psychological factors on the perception of safety at construction sites. sustainability, 9(5), 745. khan, f. i., & abbasi, s. a. (1999). hazdig: a new software package for assessing the risks of accidental release of toxic chemicals. journal of loss prevention in the process industries, 12(2), 167-181. labodová, a. (2004). implementing integrated management systems using a risk analysis based approach. journal of cleaner production, 12(6), 571-580. larsson, t. j., & field, b. (2002). the distribution of occupational injury risks in the victorian construction industry. safety science, 40(5), 439-456. journal of applied economics and business studies, volume. 5, issue 2 (2021) 35-52 https://doi.org/10.34260/jaebs.523 51 li, s. o. n. g., xueqiu, h. e., & li, c. (2011). longitudinal relationship between economic development and occupational accidents in china. accident analysis & prevention, 43(1), 82-86. mccauley-bell, p. r., crumpton-young, l. l., & badiru, a. b. (1999). techniques and applications of fuzzy theory in quantifying risk levels in occupational injuries and illnesses. in fuzzy theory systems (pp. 223-265). academic press. namvar, h., & bamdad, s. (2021). performance evaluation of process industries resilience: risk-based with a network approach. journal of loss prevention in the process industries, 104474. notice, a. (1923). workmen‟ s compensation act. pakistan occupational health and safety act. (2018) department of labour – government of pakistan policy, l. (2010). ministry of labour and manpower. government of pakistan. probst, t. m. (2004). safety and insecurity: exploring the moderating effect of organizational safety climate. journal of occupational health psychology, 9(1), 3. ramli, a. a., watada, j., & pedrycz, w. (2011). possibilistic regression analysis of influential factors for occupational health and safety management systems. safety science, 49(8-9), 1110-1117. retzer, k. d., hill, r. d., & pratt, s. g. (2013). motor vehicle fatalities among oil and gas extraction workers. accident analysis & prevention, 51, 168-174. saari, d. j. (1999). global corporations and sovereign nations: collision or cooperation?. praeger. shin, j., kim, y., & kim, c. (2021). the perception of occupational safety and health (osh) regulation and innovation efficiency in the construction industry: evidence from south korea. international journal of environmental research and public health, 18(5), 2334. song, m., an, q., zhang, w., wang, z., & wu, j. (2012). environmental efficiency evaluation based on data envelopment analysis: a review. renewable and sustainable energy reviews, 16(7), 4465-4469. suri, s., & das, r. (2016). occupational health profile of workers employed in the manufacturing sector of india. the national medical journal of india, 29(5), 277. tan, h., wang, h., chen, l., & ren, h. (2012). empirical analysis on contribution share of safety investment to economic growth: a case study of chinese mining industry. safety science, 50(7), 1472-1479. teo, e. a. l., & ling, f. y. y. (2006). developing a model to measure the effectiveness of safety management systems of construction sites. building and environment, 41(11), 1584-1592. muhammad noman, ambreen fatima and nooreen mujahid 52 visbal-cadavid, d., martínez-gómez, m., & guijarro, f. (2017). assessing the efficiency of public universities through dea. a case study. sustainability, 9(8), 1416. witter, r. z., tenney, l., clark, s., & newman, l. s. (2014). occupational exposures in the oil and gas extraction industry: state of the science and research recommendations. american journal of industrial medicine, 57(7), 847-856. yeh, l. t. (2017). incorporating workplace injury to measure the safety performance of industrial sectors in taiwan. sustainability, 9(12), 2241. yeh, l. t. (2020). using weighted data envelopment analysis to measure occupational safety and healthy economic performance of taiwan’s industrial sectors. mathematics, 8(9), 1635. yogyorn, d., slatin, c., siriruttanapruk, s., woskie, s., chantian, t., chaladlerd, p., & kongtip, p. (2020). estimating of the costs of nonfatal occupational injuries and illnesses in agricultural works in thailand. journal of public health policy, 1-15. yu, m. m. (2010). assessment of airport performance using the sbm-ndea model. omega, 38(6), 440-452. yu, x. b., & jiang, w. c. (2007). system analysis and improvement steps on safety management for chinese coal mining industry. coal science and technology, 1. zhang, j., ding, w., li, y., & wu, c. (2013). task complexity matters: the influence of trait mindfulness on task and safety performance of nuclear power plant operators. personality and individual differences, 55(4), 433-439. zhou, p., ang, b. w., & poh, k. l. (2008). a survey of data envelopment analysis in energy and environmental studies. european journal of operational research, 189(1), 1-18. zhou, z., liu, c., zeng, x., jiang, y., & liu, w. (2018). carbon emission performance evaluation and allocation in chinese cities. journal of cleaner production, 172, 1254-1272. zohar, d. (1980). safety climate in industrial organizations: theoretical and applied implications. journal of applied psychology, 65(1), 96. zou, y., he, y., lin, w., & fang, s. (2020). china’s regional public safety efficiency: a data envelopment analysis approach. the annals of regional science, 1-30. zurada, j., karwowski, w., & marras, w. s. (1997). a neural network-based system for classification of industrial jobs with respect to risk of low back disorders due to workplace design. applied ergonomics, 28(1), 49-58. journal of applied economics and business studies, volume. 6, issue 1 (2022) 109-126 https://doi.org/10.34260/jaebs.615 109 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x impact of intergovernmental fiscal transfer on interprovincial fiscal disparity in pakistan shabbir ahmed1*1& ambreen fatima2 1 phd scholar, applied economics research centre, university of karachi, karachi 2associate professor, applied economics research centre, university of karachi, karachi abstract the purpose of this research is to evaluate the disparity in the intergovernmental fiscal transfers to the sub-national governments, in the last four nfc awards specifically in the 7thnfc award in pakistan. this study measures the disparity by employing two different approaches. firstly coefficient of variation (c.v) approach i.e. variation in the regional own-source revenues to the variation in the total resource of regional government after inclusion of federal transfer and the own-source revenue, secondly, in addition, to estimate the impact of intergovernmental transfers on the fiscal disparity among the sub-national government this study uses panel data technique i.e. least square dummy variable (lsdv). the results of fiscal disparity reflect that in pakistan the disparity in combined regional own source revenue increases from 0.52 to 0.83 points during the last four nfc awards. however, at the sub-national level, the values of disparity remain constant at 0.16 in sindh province which reflects fiscal efforts for the collection of own-source revenues among provinces. in contrast to this, the disparity increases in kpk, and its value reaches 0.240 in the existing award. the results of the empirical analysis also reveal that when the intergovernmental transfer increases by 1 percent the disparity ratio increases by 1.19percent. while the expenditure of keywords fiscal disparity, own source revenues, fiscal resources, intergovernmental fiscal transfer jel classification e62, h27, h71 1 * shabbirbaqai@gmail.com this paper is part of the ph.d. thesis which is under process. the author acknowledges prof. dr. hafiz ahmed pasha for his advisory role in the study. mailto:shabbirbaqai@gmail.com shabbir ahmed & ambreen fatima 110 the government increases the disparity at the sub-national level as the positive sign of the model reflects. based on the findings this study suggests that pakistan still needs to devise a revenue-sharing formula that encourages the efforts of optimal resource generation from the regional governments. 1. introduction the constitution of pakistan (1973) assigned the responsibility for the collection of revenues and the allocation of the expenditure at the different tiers of government i.e central and the sub-nationals. the high regional variation in socioeconomic characteristics may always influence regional growth. this variation is also reflected in the fiscal position of the regional governments. the study conducted by nabi (2010) highlights that around 90percent of taxes either from direct sources or indirect sources are collected by the central government whereas only 10 percent of the taxes are collected by regional governments. further, around 28percent of the expenditures are covered by the sub-national governments while the 72percent spends by the federal government. these imbalances between revenues and expenditures especially for the regional government called fiscal imbalances in the literature. the main aim of the study is to assess these imbalances which lead to interprovincial disparity in pakistan. in this regard, pakistan has implemented multiple steps to improve the fiscal health of the sub-national governments. for instance, through the 1997 revenue-sharing formula, the central government announced fiscal incentives for the regional government to improve its fiscal position through the collection of own-source revenues (osr). similarly in the 7th nfc award, the devolution of general sales tax (gst) on services from the central to the sub-national governments is also a step towards fiscal decentralization in the country. in addition to the devolution of the fiscal resource, the current revenue sharing formula also enhanced the horizontal share by 10 percent as compared to the previous awards. overall the existing award allocated 57.5percent of federal taxes to the provinces. in contrast to the expenditure responsibility, in 2010 the 18th amendment devolved many functions to the regional governments which has also created the burden of additional expenditures on the regional governments. the consolidated expenditure of the provinces enhances from 6.2percent to 7.6percent of the gdp in the current award. while the overall federal transfers through the nfc award increase from 5.5percent to 6.3percent of gdp during the same award (fiscal operation, 2020) though intergovernmental fiscal transfers play an important role to improve the fiscal journal of applied economics and business studies, volume. 6, issue 1 (2022) 109-126 https://doi.org/10.34260/jaebs.615 111 health of the sub-national governments, however, these transfers are expected to influence directly the fiscal behavior of the regional government. the study conducted by pasha and ghaus (1994) highlighted that in pakistan, a one-rupee increase in federal transfers and grants may increase provincial expenditure by 0.61 rupees, whereas the remaining 0.39 rupees substitutes for provincial fiscal efforts. in other words, the fiscal behavior of the regional government reflects that the additional transfer of taxes may serve as a replacement for the fiscal efforts of the regional governments. therefore an in-depth analysis and evaluation of the existing fiscal position of the regional government is an important issue to analyze especially in the current context. further, the analysis of fiscal disparity during the various revenue-sharing transfers is an important question to explore. the analysis would be beneficial to reveal how the intergovernmental fiscal transfer may have impacted the inter-provincial fiscal disparity in the country. an important aspect of the study is that it will help us to identify the existing position of inter-provincial fiscal disparity and how the regional government responses to intergovernmental fiscal transfer to reduce fiscal disparity. in sum, the purpose of this paper is to measure the regional fiscal disparity created in previous nfc awards and especially in the 7th nfc award in pakistan. this study aims to highlight the contribution of intergovernmental through nfc awards which have affected inter-provincial fiscal disparity. to assess the above-stated objectives, this study extracted the data from federal/provincial budget documents and economic surveys of pakistan (various issues). this research hence fills the gap in the knowledge by exploring the existing situation of fiscal disparity within and among the sub-national governments, especially after the 7th nfc award. moreover, it assesses the behavior of the regional governments regarding federal transfers through the nfc award. overall, the article recommends that there is a strong need to encourage the fiscal health of the regional government for this purpose the inclusion of indicators or additional incentives for the growth of own-source revenue (osr) may encourage the regional governments to improve their fiscal efforts. this will also help to overcome the fiscal imbalances within and among the sub-national governments. in addition to this, the intergovernmental transfer may link the transfer with the collection of the regional revenues. further, the agreed formula should be announced at a regular interval of time i.e. after every 05years. it will help to address the regional fiscal issues and will also help in improving the fiscal health of the sub-national governments. this paper is organized as follows: section 2 develops the theoretical framework based shabbir ahmed & ambreen fatima 112 on the review of literature; section 3 focuses on the empirical model and the data sources; section 4 highlights an overview of intergovernmental fiscal relations in pakistan while, section 5, highlights the estimation results of the study and the last section draws conclusion and policy implications. 2. literature review the theoretical foundation of fiscal federalism highlights that the central government is responsible to manage macroeconomics stability, resource allocation, and redistribution of national income, musgrave (1959). in addition to this, in a federal system of governance, the central government has more taxing powers compared to expenditure responsibilities whereas the sub-national governments have less revenue collection power compared to high expenditure responsibilities. this variation in revenues and expenditures is considered a fiscal imbalance or the fiscal gap in the literature, shah (2006). fiscal inequality is defined as the difference between fiscal need and fiscal capacity. the fiscal need is the cost of providing service delivery while the fiscal capacity is the capacity to raise revenues (tannenwald, 2002). however, for the removal of the fiscal disparity, the literature suggests that increasing transfer from the central government to the sub-national government is one way to balance the budget and overcome the regional fiscal disparity (bird and smart 2002) the literature on the fiscal imbalance of the regional governments indicates that it is the inconsistency between revenue-raising and fiscal needs of the same tier of government. boadway and flatters (1982) reveal that in a federalism state, the revenue-sharing arrangements may remove the disparity in the state. similarly, garnaut and fitzgerald (2002) argue that the equalization grants from the central government to the sub-national government may reduce regional disparities and inequalities to manage the fiscal needs and fiscal capacity of the regional government. in contrast to this, martinez and zhang (2008) highlight that the expenditure of sub-national government decreased the disparities for a certain time. blochliger and charbit (2008) indicate that the balanced budget of regional government may support providing uniform public service with the minimum tax burden. in addition to this, mclure (1993) identifies the reasons for the imbalances that the central government collects taxes more easily at lower economic costs than sub-national governments. furthermore, bird (2003) argued that in many developing countries subnational governments have limited fiscal power which is insufficient to finance their expenditure needs for providing the services. journal of applied economics and business studies, volume. 6, issue 1 (2022) 109-126 https://doi.org/10.34260/jaebs.615 113 bird and smart (2003) also highlight that the formula-based fiscal transfer may help the sub-national government expenditure in the same direction because this transfer is relatively transparent and predictable. furthermore, boex and martinez (2007) point out the allocation may base on fair allocation, simple and transparent, and unconditional. in addition to the above discussion, the literature on the behavior of sub-national government on intergovernmental fiscal transfers highlights that the federal transfer may be one of the ways that may help to address the fiscal disparity and balance regional budgets. bahl linn, shah, bird, and smart (1992, 1994, and 2008). however, the other side of the literature also highlights the balance of resource distribution is based on the proper mechanism of intergovernmental transfers. prud and homme (1995) indicate that the centralized structure of the public sector produces a much-balanced distribution of resource management by directing the resources from rich to poor regions. the study on the chinese economy reflects that in a decentralized political economy the local resource distribution inversely depends on economic growth. however, rodríguez-pose and ezcurra (2009) highlight the relationship between decentralization and income inequalities. the literature on devolution reveals that non-pure fiscal decentralization is considered when the lower tier of government is responsible to allocate the resources. in contrast to this, pure decentralization considers when the regional governments generate osr and do not depend on the central government's transfers to manage their expenditure. in the context of pakistan ghaus, et al (2010) estimates that the 7th nfc award has made the biggest effort at ensuring equalization of regional revenues. a similar study conducted by pasha and ghaus (1994) indicates that in pakistan, a one-rupee increase in federal transfers and grants will raise provincial expenditure by 0.61 rupees, whereas the remaining 0.39 rupees substitutes for provincial fiscal efforts. further, this research also highlights the effect of regional gdp is low while the effect of provincial borrowings on provincial spending is high i.e. 0.89. in the context of the above discussion, this study examines the position of regional fiscal disparity during the period experiencing changes in revenue-sharing formulas. this study also estimates the impact of the intergovernmental fiscal transfer on the inter-provincial fiscal disparity as said earlier. this study focuses on the regional fiscal health of the regional economies using the provincial level data throughout the period 1990–2020. specifically, the goal is to observe whether the intergovernmental fiscal transfers from the central to the provincial government are anticipated to minimize disparity and inequality at the subnational level. shabbir ahmed & ambreen fatima 114 2.1 fiscal equalization – structural relationship 3. estimation techniques martinez and zhang (2008), measured the fiscal disparity as a ratio of the coefficient of variation of the regional own source revenues before and after the federal transfer. for empirical analysis, this study has used two different approaches to estimate the impact of fiscal disparity difference (rrc – fn fiscal need rrc revenue raising capacity rrc=fn fiscal equalization rrc<fn fiscal disparity >0 intergovernmental fiscal transfer aiming to reduce the fiscal disparity among provinces developed provinces-high fiscal capacity fiscal transfer payments from central government divisible pool transfers transferdisp grants in aids/subventions straights transfers under developed provinces low fiscal capacity journal of applied economics and business studies, volume. 6, issue 1 (2022) 109-126 https://doi.org/10.34260/jaebs.615 115 intergovernmental fiscal transfers on the inter-provincial fiscal disparity; firstly we estimate the disparity ratio is measured as the ratio of the coefficient of variations of per capita ownsource revenue to the per capita resources of the sub-national governments (including the osr and the nfc transfer). secondly, we used the least square dummy variable (lsdv) to highlight the effect of intergovernmental transfer on the fiscal disparity among the subnational governments. the following model has been used to estimate the regression results. fisdisp = βo+β1it+β2(inttra)it+β2(ce)it+β3(de)it+ β4(dv)it+uit fisdisp = ratio of disparity before and after federal transfers inttra = log of per capita federal transfers to the sub-national governments ce = per capita current expenditure of the regional governments de = per capita development expenditure of the provincial governments dv = provincial dummies u = error term t = period between 1990 to 2020 i = provinces for the estimation purpose, we employ the ratio of disparity before and after intergovernmental transfers as a dependent variable. the disparity ratio is measured as the ratio of the coefficient of variations of per capita own-source revenue to the per capita resources of the sub-national governments (including the osr and the nfc transfer). in other words, fiscal disparity is the ratio of the coefficient of variation of provincial osr to the coefficient of variation of the total fiscal resources available to the provincial governments. among the explanatory variables, the per capita intergovernmental fiscal transfer reveals the role of federal transfer (through nfc awards) in reducing the fiscal disparity of the regional governments. the remaining independent variables capture the role of current and development expenditures in reducing the fiscal disparity among the provinces. as far as the relationship among the variables is concerned, the dependent variable is a ratio of disparity before and after intergovernmental fiscal transfers which means that a positive sign of each coefficient in the model will show a reduction in the fiscal disparity. the same relationship is also hypothesized for regional expenditures. in contrast, the negative coefficient reflects the increase in the disparity within the sub-national government. 3.1 data source the provincial-level data related to the fiscal side is collected from the federal and provincial budgets documents, while the data on population is extracted from the economic shabbir ahmed & ambreen fatima 116 surveys of pakistan. for the overall analysis, we have used the data between 1990 2020 i.e. t = 30 for all four provinces. this study used the stata 12 software for the estimation of models. 3.2 results of estimation the result section is divided into two parts. the first part highlights the assessment of intergovernmental fiscal relations and an in-depth assessment of the fiscal position of regional governments in the country while the second part evaluates the inter-provincial fiscal disparity at the regional level through the c.v approach and the lsdv techniques for assessment of the various nfc awards and the 7th award separately. 3.3 intergovernmental fiscal relation in pakistan the intergovernmental fiscal relation in pakistan is constitutionally protected through article 160 of the constitution of pakistan which highlights the mechanism of intergovernmental fiscal relationship between the center and the regional governments through the national finance commission award. the principal objective of nfc is to distribute the taxes of the federal divisible pool, straight transfer, and provision of federal subventions and grants in aid after deducting collection costs among the provinces. the nature of divisible pool taxes and the straight transfers are unconditional and are distributed through a standard formula. in contrast to this, the nfc recommended a fixed proportion of subvention/grant in aid including special non-development grants to backward provinces or all provinces. in addition to this, the federal government has also allocated some emergency grants like revenue deficit and ad-hoc relief. these grants are also distributed under the head of non-development grants. all federal grants are funded from the federal consolidated funds. historically, pakistan has distributed nine revenue-sharing awards since independence in 1947. the first revenue-sharing i.e. raisman award was announced in 1951. the remaining awards were announced in 1961-62, 1964, and 1970 respectively. in 1971 pakistan was divided into east and west pakistan, and the nfc awards were re-designed in 1974 1990, 1996, and the existing nfc award in 2009. however in 2006, due to a lack of consensus among provinces on the formula the president of pakistan announced the amendments in the distribution of revenues and grants-in-aid order 1997. further, the taxes in the divisible pool were limited. table 01 reveals the historical comparison of divisible pool taxes in the last four nfc awards. in 1990 the excise duty on tobacco and sugar was part of the divisible pool it was excluded from the 5th nfc award. in addition, the federal excise duty on gas and custom duties were included in the domain of journal of applied economics and business studies, volume. 6, issue 1 (2022) 109-126 https://doi.org/10.34260/jaebs.615 117 divisible pool taxes. the 7th nfc devolved the sale tax on services to the provinces now provinces are collecting gst on services through their revenue collection authorities. however, the tax on income, sale, purchase of goods, and export duties on cotton is part of the divisible pool in these awards. table 01: comparison of divisible pool taxes by nfc awards s.n o taxes 7th nfc award 6th nfc award 5th nfc award 4th nfc award 1 taxes on income ✓ ✓ ✓ ✓ 2 wealth tax ✓ ✓ ✓  3 capital value tax ✓ ✓ ✓  4 taxes on sales and purchases of import/export/produced/manufactured/con sumed ✓ ✓ ✓ ✓ 5 sale tax on services (ce mode)  ✓   6 export duties on cotton ✓ ✓ ✓ ✓ 7 excise duties on tobacco and tobacco manufacture    ✓ 8 excise duty on sugar    ✓ 9 custom-duties ✓ ✓ ✓  10 federal excise duty excluding the excise duty on gas charged at the wellhead ✓ ✓ ✓  11 any other tax which may be levied by the federal government ✓ ✓ ✓  source: nfc reports (various years) 3.4 straight transfers the straight transfer was introduced in 1990. the composition of straight transfer consists of the provincial rights on its natural resources including the royalty on crude oil and gas, excise duty and surcharge on natural gas, and hydroelectricity profit of wapda. the kpk received the first allocation of straights transfer in 2000-01 at a capped level of rs.6 billion. the existing nfc has been allocating the arrears on net hydel profits and development surcharge on gas. 3.5 subvention/matching grants the fourth nfc award allocated revenue deficits grants to the provinces to finance their deficits. however, the fifth nfc linked these grant grants to smaller provinces only with an amount of rs.3.3billion and 4billion for kpk and balochistan respectively. in addition, the 6th nfc extended these grants from rs.8.7billion to rs.27.7billion to all the provinces. shabbir ahmed & ambreen fatima 118 however, the additional allocation was linked to the annual growth of the net divisible pool. the existing award abolished the discretionary grants-in-aid for all the provinces except for the sindh province. sindh was given a grant of 0.66 percent of the provincial divisible pool to partly offset losses due to the merger of one-sixth of gst in the divisible. 3.6 horizontal distribution of fiscal resources pakistan allocated the divisible pool taxes based on regional population share. in 2009 the 7th nfc award shifts from single to multiple indicators for the distribution of divisible pool taxes to the provinces. table 02 explains the horizontal sharing formula and provincial shares in the last four awards in pakistan. due to the adoption of multiple indicators, punjab is the only province that lost its share of around 5 percent while balochistan gained around an additional 4 percent share from the federal government. the sindh and kp received additional 1 percent shares in the existing revenue-sharing formula. table 02: horizontal resource sharing formula and provincial share by nfc awards indicators 7th nfc award 6th nfc award 5th nfc award 4th nfc award population 82.0% 100% 100% 100% poverty or backwardness 10.30%    revenue collection/generation 5.0%    inverse population density 2.70%    provincial shares by nfc awards punjab 51.74% 57.36% 57.88% 57.88% sindh 24.55% 23.71% 23.28% 23.28% kpk 14.62% 13.82% 13.54% 13.54% balochistan 9.09% 5.11% 5.30% 5.30% total 100% 100% 100% 100% source: nfc reports (various years) 4. overview of fiscal position of the provinces the fiscal position of subnational governments is dependent on the expenditure and the collection of revenues. table 03 examines the provincial share of expenditure financed through own-source revenues. in 1990, the punjab and sindh have covered around 15 percent of expenditures from their source revenues while the kp and balochistan financed 9 percent and 5 percent of their expenditures from provincial tax revenues. the 5th nfc award journal of applied economics and business studies, volume. 6, issue 1 (2022) 109-126 https://doi.org/10.34260/jaebs.615 119 allocated an incentive for the improvement of the collection of the osr. the kpk enhanced its revenues and covered 15 percent of the expenditure. however, due to the devolution of the gst on services the overall fiscal position of the regional government has improved. sindh is the only province that currently financed one-fourth of the expenditure from their osr revenues. while the smallest provinces financed their expenditures below 10 percent. the largest province punjab, just slightly improved its collection of revenues. table 03: provincial expenditures financed with osr by nfc awards (in %) punjab (rs in billion) sindh(rs in billion) kpk(rs in billion) balochistan(rs in billion) osr texp share osr texp share osr texp share osr texp share 4th nfc award 10 61 16% 4 28 14% 2 22 9% 0.48 9 5% 5th nfc award 25 138 18% 11 64 17% 4 38 11% 1 26 4% 6th nfc award 81 369 22% 28 210 13% 6 96 6% 3 63 5% 7th nfc award 128 679 19% 100 446 22% 19 247 8% 7 141 5% source: authors estimate from the budgets documents the two smallest provinces have been receiving the highest per capita intergovernmental transfers from the federal government. figure 01 indicates balochistan receives the highest per capita intergovernmental transfer i.e. above rs.6,000 as compared to the other provinces. however, the major jump appeared in the 7th nfc award as this award ensures a minimum of rs.83billion to balochistan from the divisible pool transfers. the award also transferred 1 percent of the divisible pool to the kpk in compensation for war and terror. however, in the past three decades, the provinces combined received transfers at rs.3000 per person. shabbir ahmed & ambreen fatima 120 figure 01: per capita nfc transfer by provinces source: federal and provincial budgets documents (various years) figure 02 highlights the per capita own-source revenue of federal and provincial governments. on average the per capita revenue enhanced from rs.3000 to 7000 during the past three decades while the combined provincial revenue reached only rs.2000 per person. at the regional level, sindh province collected the highest revenue compared to balochistan province which collected the lowest revenues. however, punjab collected more own-source revenues as compared to kpk province in the past three decades. figure 02: per capita own source revenues by governments source: federal and provincial budgets documents (various years) figure 03 exhibits the relationship between provincial own-source revenues and the intergovernmental fiscal transfer among the provinces. at the federal level, the trend highlights that when the federal government enhances the share of intergovernmental transfer through nfc awards, in parallel to the nfc share, the federal government also increases the 0 20 00 40 00 60 00 80 00 1990 2000 2010 2020 years combined-provinces punjab sindh kpk balochistan 0 20 00 40 00 60 00 80 00 1990 2000 2010 2020 years punjab sindh kpk balochistan combined-provincial federal journal of applied economics and business studies, volume. 6, issue 1 (2022) 109-126 https://doi.org/10.34260/jaebs.615 121 tax revenues, which shows the fiscal efforts of the federal government to collect more revenues. while at the sub-national level, the gap between the regional revenues and the federal transfer increases sharply in the past three decades. however, a major jump was observed in the balochistan and the kpk provinces. figure 03: relationship between osr and nfc transfers source: federal and provincial budgets documents (various years) 4.1 descriptive statistics in this section, we examined the fiscal disparity among the regional governments for the different nfc awards. the fiscal disparity is measured for each province separately and combined for all provinces as well. as discussed in the previous section, the expenditure is highly decentralized whereas the revenues remain highly centralized. the disparity ratio of the consolidated regional revenues reflects that the disparity in the own source revenues of the sub-national government has increased since 1990. the ratio ranges from 0.52 to 0.83 points, as in the 7th nfc award the fiscal disparity is high as its value is above 0.84 points. it means the consolidated provincial fiscal efforts have not increased as the rate of federal transfer has increased. however, the average regional revenues increased and reached rs.1320 per person table 04 reveals the inter-provincial disparity across provinces. the variation in osr over time increased in balochistan and kpk provinces and reaches 0.28 points, which is the highest among the four provinces. while the osr of these provinces has just reached rs 671 and rs 553 in these provinces respectively. only sindh province reveals the constant trend in the fiscal disparity across provinces. this means, that sindh put comparatively better fiscal efforts as compared to other provinces. in addition, the devolution of gst has created a strong fiscal position for sindh due to the large fiscal base that exists because of having a 0 50 00 10 00 0 0 50 00 10 00 0 1990 2000 2010 2020 1990 2000 2010 2020 1990 2000 2010 2020 federal consolidated punjab sindh kp balochistan nfc transfers own source revenue years shabbir ahmed & ambreen fatima 122 mega city karachi. while the two provinces have not improved the osr revenues. punjab is the largest economy which contributes around half of the national income according to pasha (2015) but still, its cv values of the osr enhanced from 0.13point to 0.15 in the past four nfc awards. the average per capita osr revenues of the two provinces like punjab and sindh are above the four consolidated revenues while the two smallest provinces i.e. kp and balochistan have around 50 percent below the consolidated per capita revenue. table 04: average per capita cv of own source revenue by awards mean four provinces c.vfour provinces mean punjab c.v punjab mean sindh c.v sindh meankpk c.v kpk mean balo c.vbaloc 4th nfc award 65 0.529 96 0.130 115 0.150 49 0.149 35 0.395 5th nfc award 145 0.554 198 0.130 284 0.170 98 0.113 75 0.121 6th nfc award 276 0.696 487 0.340 634 1.290 161 0.276 112 0.047 7th nfc award 1320 0.844 1537 0.150 3159 0.160 671 0.240 553 0.289 source: authors estimate from the budgets documents 5. estimation results for the empirical analysis of the study, we estimated the impact of intergovernmental fiscal transfers on the regional disparity during the period of the past four nfc awards as said earlier. the aim is to examine how these factors may have affected the fiscal disparity of the regional governments. table 05 presents the summary of the variable used in the model. the average disparity found is around 2 point. as discussed in the earlier section, the high value of the ratio indicates the level of fiscal disparity before and after transfer in the fiscal health of the own source revenues of the provincial governments. at the provincial level, the per capita current expenditures reach rs.2665 while the per capita development expenditures reach below rs.1000. however, the federal transfer stands at rs.5274 per person. journal of applied economics and business studies, volume. 6, issue 1 (2022) 109-126 https://doi.org/10.34260/jaebs.615 123 table 05: summary of the variables variable number of observation mean std. dev. min max fiscal disparity 120 2 2 0.021 15.232 per capita federal transfers 120 5274 5333 312 24252 per capita current expenditures 120 2665 1189 664 5970 per capita development expenditures 120 872 525 133 2306 per capita total expenditures 120 3535 1643 884 8179 source: authors estimate from the budgets documents table 06 reports the results of the lsdv model, estimated for the different nfc awards. here, the positive coefficient of total federal transfer reveals that the federal transfer in pakistan has increased the fiscal disparity in the country. according to the results, a 1 percent increase in the total federal transfer has increased the disparity in the osr of the regional government by 1.195percent during the last four nfc awards while it has increased the disparity by around 0.335percent without controlling the 7th nfc award. as far as the role of current expenditures is concerned, at the regional level, the positive coefficient of current expenditures indicates that if the current expenditures increased by rs. 1 then the fiscal disparity will be increased by 0.084paisa with the 7th nfc award, while the disparity will be reduced by 0.087paisa controlling for the 7th nfc award. however, excluding the role of the latest award, the development expenditures reduced the fiscal inequality by rs.0.183paisas while increasing the disparity by rs.0.184paisas during the different nfc awards. the results reflect that as the federal transfer increased the regional government did not focus on the collection of provincial revenues. the results also highlight the provincial positions of fiscal disparity during the different nfc awards; the disparity in the sindh province is statistically insignificant. however, the highest coefficient value of the disparity ratio is revealed by balochistan province i.e. around 3.24 which shows the low level of fiscal efforts for the collection of osr. the positive coefficient of disparity ratio indicates that the fiscal position has decreased with a rate of 5.119 in the 7th nfc award while the overall disparity has reduced by 6.03 without controlling for the 7th nfc award. shabbir ahmed & ambreen fatima 124 table 06: regression results of the fiscal disparity variables details lsdv-with 7th nfc award lsdvwithout 7th nfc dummy model 01 model 02 fiscal disparity (coefficient) -5.119 -6.036 (p-values) 0.162 0.028 log of per capita federal transfer 1.195 1.335 (p-values) 0.044 0.004 per capita regional current expenditures 0.084 -0.087 (p-values) 0.063 0.046 per capita regional development expenditures 0.184 -0.183 (p-values) 0.009 0.009 dummy variable -7th nfc 0.282 (p-values) 0.705 sindh 0.326 0.310 (p-values) 0.558 0.575 kp 1.041 1.011 (p-values) 0.048 0.051 balochistan 3.249 3.204 (p-values) 0.000 0.000 f-statistics 3.260 3.810 (p-values) 0.003 0.002 r-square 0.169 0.1683 adjusted r-square 0.117 0.1241 roost mse 1.926 1.914 number of observations 120 120 p values indicate a significance level 6. conclusion and policy recommendations this paper examines the existing position of fiscal disparity of the sub-national government and explores to what extent the intergovernmental fiscal transfer has reduced the fiscal disparity across provinces. for the existing fiscal disparity, the paper used the ratio of c.v for the regional osr and federal transfer. this paper adopts a panel data approach and journal of applied economics and business studies, volume. 6, issue 1 (2022) 109-126 https://doi.org/10.34260/jaebs.615 125 employed techniques i.e. lsdv model to evaluate the impact of the intergovernmental fiscal transfer on inter-provincial fiscal disparity for the last four nfc awards. this paper also highlights that sindh has improved its fiscal position compared to other provinces as reflected by the results. similar case for punjab province, the positive coefficient reflected the same in the model. the other provinces like kp and balochistan have not improved their fiscal efforts to collect the osr. these provinces finance only 8 percent and 5 percent of total expenditures from their osr. the overall conclusion is that the inter-provincial fiscal disparity is high even after the 7th nfc award. this paper also finds that regional government responds differently to intergovernmental fiscal transfers in pakistan. the main reason for the existing interprovincial fiscal disparity is the high variation in the tax base. as we highlighted that the share of tax collection is highly centralized compared to expenditure decentralization. further due to the small tax base provinces like kp and balochistan still depends on federal transfers. for the policy level, this article indicates that pakistan still needs to devise a revenuesharing formula that encourages the efforts of optimal resource generation by the regional governments. however, the nfc secretariat may link the intergovernmental fiscal transfer with the fiscal efforts of the provincial governments. as pakistan has implemented those practices in the previous award. for the horizontal distribution of resources, the inclusion of the indicators that supports the fiscal efforts may improve the fiscal position of the regional governments. in addition to this, the future nfc awards in pakistan may ensure to be announced at a regular interval of time because this may help to strengthen fiscal federalism and address the upcoming fiscal challenges on time. references bahl, r. w., & linn, j. f. (1992). urban public finance in developing countries. the world bank. bahl, r., & bird, r. (2008). subnational taxes in developing countries: the way forward. public budgeting & finance, 28(4), 1-25. bengali, k., & sadaqat, m. (2005). regional accounts of pakistan: methodology and estimates, 1973-2000. social policy and development centre. bird, r. m., & smart, m. (2002). intergovernmental fiscal transfers: international lessons for developing countries. world development, 30(6), 899-912. bird, r., dafflon, b., jeanrenaud, c., & kirchgässner, g. (2003). assignment of responsibilities and fiscal federalism. federalism in a changing world: learning from each other. blöchliger, h., & charbit, c. (2008). fiscal equalisation. oecd journal: economic studies, 2008(1), 1-22. shabbir ahmed & ambreen fatima 126 boadway, r., & flatters, f. (1982). efficiency and equalization payments in a federal system of government: a synthesis and extension of recent results. canadian journal of economics, 613-633. boex, j., & martinez-vazquez, j. (2007). designing intergovernmental equalization transfers with imperfect data: concepts, practices, and lessons. in fiscal equalization (pp. 291-343). springer, boston, ma. ezcurra, r., & rodríguez-pose, a. (2009). 17 measuring the regional divide. handbook of regional growth and development theories, 329. garnaut, r., & fitzgerald, v. (2002). issues in commonwealth–state funding. australian economic review, 35(3), 290-300. ghaus, a. a., pasha, h. a., & kemal, a. r. (1994). dynamic budgetary consequences of the 1991 nfc award [with comments]. the pakistan development review, 33(4), 627-645. ghaus-pasha, a., pasha, h. a., & zubair, a. (2010). fiscal equalisation among provinces in the nfc awards. the pakistan development review, 563-576. martinez-vazquez, j., qiao, b., & zhang, l. (2008). the role of provincial policies in fiscal equalization outcomes in china. china review, 135-167. martinez-vazquez, j., qiao, b., & zhang, l. (2008). the role of provincial policies in fiscal equalization outcomes in china. china review, 135-167. mclure jr, c. e. (1993). economic, administrative, and political factors in choosing a general consumption tax. national tax journal, 46(3), 345-358. musgrave, r. a. (1959). the theory of public finance; a study in public economy. kogakusha co.. nabi, i., & shaikh, h. (2010). fiscal federalism in pakistan: a radical departure and some new challenges. development policy research center working paper, lums, lahore. pakistan fiscal operation 202. published by ministry of finance, government of pakistan pasha, h. a. (2015). growth of the provincial economies. institute for policy reforms (ipr). http://ipr. org. pk/wp-content/uploads/2016/04/growth-ofprovincial-economics-. pdf. prud'homme, r. (1995). the dangers of decentralization. the world bank research observer, 10(2), 201-220. shah, a. (2006). fiscal decentralization and macroeconomic management. international tax and public finance, 13(4), 437-462. tannenwald, r. (2002). interstate fiscal disparity in 1997. new england economic review, 17-34 journal of applied economics and business studies, volume. 5, issue 3 (2021) 59-76 https://doi.org/10.34260/jaebs.534 59 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x digital divide, socio-economic outcomes and subjective well-being in higher education institutions of pakistan atta ur rahman*1, shandana2, laila taskeen qazi3, and adnan khan4 1 institute of management sciences, peshawar, pakistan 2 institute of management sciences, peshawar, pakistan 3 institute of management sciences, peshawar, pakistan 3 institute of management sciences, peshawar, pakistan abstract this research examines the relationship between internet usage, socioeconomic outcomes, and subjective well-being. social outcomes are measured in the form of formal and informal networks and economic outcomes are measured through education, employment, and financial status. subjective well-being is defined as assessment of one’s own life satisfaction. the cross-sectional study is conducted through a survey. the data was collected through a self-designed questionnaire from 500 university graduates in different region of pakistan. the cox regression is used to investigate the relationship among the variables. in cox regression the gender, education, major, subjective wellbeing (swb), positive affect (pa), negative affect (na) are significantly affecting the duration of internet usage. internet usage increases with age, likewise internet usage is found more in women and married individuals than men and unmarried individuals. economic outcomes are less for internet users than the non-users similarly social outcomes are lower for users than the nonusers. similarly, internet usage is found to decrease with increase in the level of education. this shows how online activities affect the economic and social activities of the individuals and in turn affect their subjective well-being. keywords digital divide, subjective wellbeing, socioeconomics, cox regression jel classification z1, z13, a14 1. introduction online communication has widely become common among the people as more attention is paid to the influences of online communication to people’s life like information sharing, social adaption, interpersonal trust and subjective well-being. subjective well-being is also gaining popularity in the recent years, and it has become a recognized subject matter in social sciences including economics. due to new * attaurrahman@imsciences.edu.pk atta ur rahman, shandana, laila taskeen qazi, and adnan khan 60 emerging technologies, online communication such as instant messaging and social networking sites have obvious impact on subjective well-being of users. the skills and particular use of internet results in positive outcomes similarly inadequate use and skills can lead to limiting success. discussion about digital divide now means socioeconomic inequalities while accessing and using information and communication technologies and the use of such technologies namely internet, results in certain outcomes. hence digital divide has left behind the discussion of having or not having access to internet and has come up with more factors like technical and social skills, motivation and awareness etc. peoples online and offline activities are interrelated as it can benefit those who are using internet however internet abuse can create problems too. (helsper, deursen & eynon 2015) different types of internet usage skills results in useful outcomes to use information and communication technologies such as personal, social, cultural and economic wellbeing that are caused by online activities. according to (helsper et al., 2015) these outcomes are useful in predicting the gratification of individuals which are the result of internet usage. similarly, technology acceptance model explains the objective related to the internet usage and following activities related to internet outcomes according to needs and ease of use which results in happiness or well-being. classification of economic and social capital is done as in terms of monetary assets and property for economic and for social in terms of relationship, social support and network. similarly institutional (public formal information and services) and political participation through internet also results in outcomes and those online and offline activities have impact on each other. the relevant theory can be found in the knowledge gap and usage gap hypotheses. decades old theory i.e. knowledge gap hypothesis is about the traditional mass media in the digital divide concept that infuse information into the social system. the use of traditional mass media is simple as compared to internet, as internet usage requires technology as well as skills to use that technology (bonfadelli, 2002). the difference in functionality is small in mass media e.g. print media, television, telephone as compared to internet use can create a usage gap. while the knowledge gap is the difference in getting information or knowledge from the mass media, the usage gap is about the society’s ways of using internet differently (van dijk, 2013). the knowledge gap is not the outcome of a poorly operating media system but due to the social structure of the society, the information flow is not same, and the gap exists (banfadelli, 2002). the usage gap as described by van dijk (2013) is the combination of societal tendencies and technological characteristics. where societal tendencies are the socio-cultural and socio-economic differences according to income, employment, information and communication facilities and access. technological characteristics involve the complex, expensive and multifunctional technologies regarding computer and internet which have different uses. journal of applied economics and business studies, volume. 5, issue 3 (2021) 59-76 https://doi.org/10.34260/jaebs.534 61 the internet usage classifications are derived from internet activities. the usage is classified as uses and gratifications approach, technology acceptance model and social cognitive theory. according to uses and gratifications approach internet is used for social and psychological purpose. it explains how users’ needs affect their choices regarding entertainment, problem solving, social connection, information and diversion etc. (cho, 2003). technology acceptance model explains the intended behavior regarding the information system with two attitudes towards using a technology i.e. perceived usefulness and perceived ease of use; where the former explains using the system that improves the performance and the latter explains that using the system will be easier and free of effort (davis, 1989). social cognitive theory explains the psychosocial factors of society in the form of cognitive, behavioral, environmental events that affect each other. it explores the new behavior patterns of basic psychosocial factors such as getting knowledge of innovative behaviors, taking up these behaviors in practice and spreading through social networks (bandura, 2001). thus, it is related to the internet usage outcomes such as monetary outcomes, social outcomes, activity outcomes, novel outcomes, self-reactive and internal outcomes (larose et al., 2001). in an information society, information functions in that society therefore people need social and cultural capital to use it in their social network. cultural capital includes family traditions, education and other resources (bucy and newhagen, 2004). the social capital comprises networks, norms, trust, sociability etc. to pursue share objectives (sabatini and sarracino, 2014). however economic capital consists of family size, occupation, income and geographic location. the relation between these fields brings the concept of techno-capital which is beyond the physical access to computer technology and information in digital divide concept (bucy and newhagen, 2004). figure: 1. digital divide digital divide knowledge gap hypothesis usage gap hypothesis uses & gratifications approach technology acceptance model social cognitive theory monetary outcomes shopping & price comparison activity outcomes playing games, entertainment social outcomes talk & support novel outcomes news & information atta ur rahman, shandana, laila taskeen qazi, and adnan khan 62 considering social capital in economics as trust and interaction is recognized for reduced transaction costs, facilitated investors and encouraged investment and innovation in human capital (sabatini and sarracino, 2014). thus, key indicator of social capital is quality of social structure, networks, norms and trust; however, the measure of well-being is the quality of individuals’ lives assessment by themselves with the most typical measure of life satisfaction (helliwell, huang and wang, 2013). 2. literature review information and communication technologies (ict) have the potential to offer immense advantages to users; it provides thorough information and opportunities for social and economic mobility. it can be the source of knowledge and expertise in education and health all across the world (nanthikesan, 2000). the importance of information technology (it) is still disputed in developing countries; it is viewed as a luxury for people living in poverty at the same time it opens the door to economic opportunity, social and political mobilization, health care and education. for opening economic and other such opportunities, access to information through it is necessary for the deprived sector (hijab 2001). while discussing the inequalities regarding internet, only considering the demographic differences in access about who is online and who is not, is not as important as more people use internet now for communication and information purpose therefore the medium of use might be considered now (hargittai 2010). subjective well-being has become a recognized topic of various social sciences including economics as it has become the agenda of political organizations, governments and international institutions. people’s subjective well-being can be measured through responses to the questions about life satisfaction and happiness and it can help planning socio-economic policy (praag 2001). people rely more on online networking for socializing and interacting with others and it turn out to be a new political and civic participation. social capital i.e. face to face interactions and social trust are strongly associated with well-being. subjective well-being is positively associated with face to face interactions however the use of social networking sites lowers the social trust which shows negative association with subjective well-being. the risk of fading social trust effects the life satisfaction i.e. subjective well-being. (sabatini and sarracino 2014). the use of information and communication technologies (icts) such as computers and mobile phones etc. has a significant impact on individual’s wellbeing, yet so far very little is known about its impact in spite of its growing importance (nie, poza , nimrod, 2017). journal of applied economics and business studies, volume. 5, issue 3 (2021) 59-76 https://doi.org/10.34260/jaebs.534 63 in general, digital divide indicates the socio-economic inequality in access and use of icts, those who use it might get benefits from it however those who do not use it might be excluded from involvement in the contemporary society. originally the digital divide was defined simply as the difference between having and not having the internet access, the difference was then accredited to difference in economic capital thus the financial resources get you connected. helsper et al., 2015 measure tangible outcomes for internet usage as well as benefits from online networking. the study identifies scale for measuring internet outcomes that includes uses, skills and outcomes measures. the result of the study shows that different outcomes of internet usage have different levels of satisfaction for users. people achieving social outcomes do not necessarily be achieving the economic outcomes as well. similarly, internet usage does not always help the user with outcome benefits. the study also reveals skills as important and different skill levels have different outcomes. selwyn (2010) argues that in contemporary higher education, digital divide is significantly gaining not losing; also the nature of existence of digital divide within the students of higher education needs to be understood more sophisticatedly. the use of information and communication technology (ict) in higher education is a multifaceted concept that includes many activities through different platforms and means of connectivity. it requires skills that results in different outcomes. the use of ict is a significant source of social inequality among university students. authorities of higher education need to bridge the gap making the technological resources available to address the problem of the digital, information and knowledge divide. hargittai (2010) analyzes data of young adults of almost same age and education level from different groups on internet uses and skills whether different uses of internet are caused by certain social factors or just a random distribution once the connectivity is achieved. the results show that variation exists in internet usage even when college students have the access/connectivity and students from lower socioeconomic status and female do not know much about web use regarding seeking information online. students with more skills are involved in more activities online than those who do not have much knowledge. relationship between pattern of internet use and gratifications gained in the context of digital divide has been explored. prominent difference in uses and gratifications are shown by the subgroups i.e. age and socioeconomic status. young people with high socioeconomic status are likely to use internet more and for learning, enjoyment and motivation etc. (cho et.al. 2003). atta ur rahman, shandana, laila taskeen qazi, and adnan khan 64 different factors of global digital divide have been taken in to account like economic, demographic, telecommunication infrastructure and others where a cross country econometric analysis has been done. the result shows that income is the most dominant factor in explaining the gap between the countries but there are other factors as well that at the same time are more effective e.g. telecommunication infrastructure. similarly, policy variables like pricing the telecommunication access do not explain the internet gap that shows that it has been swamp by other factors like economic, demographic, and institutional. the paper shows importance of regulatory quality in technology usage (chinn and fairlie 2004). maclaren and zappala (2002) analyze the influence of the usage and access of internet and computer within socioeconomic and demographic groups. the results show that parental education play important role in the use and access of computer and internet. student from lower socioeconomic status use computer and internet mostly at school which shows that digital gap is closing at school levels. robinson, dimaggio and hargittai (2003) observe that once the access is achieved, the difference in the usage can be different. people with more education use internet mostly for work, education, social and political engagements and very less for entertainment moreover the college educated is able to use new technology in a better way. hoffman and novak (1998) and rahman and uddin (2009) examine that different educational and income background of different races affect the access and use of computer and web. households with more/increasing level of income and education have access to computer and web than lower income households. zillien and hargittai (2009) investigates that socioeconomic status and context of use affect the online activities of users. users with upper socioeconomic status use internet for capital enhancing than the opposite. it suggests that experience and improved equipment can decrease the digital inequalities. internet has become an important part of academia as a tool for education, communication and socializing as well, as it is not just confined to faculty research and communication but also used by students forgetting to know the society. wang and wang (2011) analyze that people who spend more time communicating online have higher subjective well-being hence show a positive relation between online communication and subjective well-being. online communication has higher impact on boys as compared to the girls because of self-disclosure therefore it has more effect on subjective well-being of boys than for girls. journal of applied economics and business studies, volume. 5, issue 3 (2021) 59-76 https://doi.org/10.34260/jaebs.534 65 sabatini (2011) analyzes the relationship between e-shopping and happiness. eshopping is changing the social and economic life of people and bringing growth to it. ecommerce help consumer to have experiences than just to own goods and services and such experiences bring happiness and hence lead to subjective well-being. the results of the study indicate that e-commerce is positively and strongly related to well-being i.e. happiness. it also depicts that with increasing age, poor health and financial distress decrease happiness. economic outcomes are measured by indicators of education, employment, income and property that are related to illiteracy, unemployment and poverty e.g. finding a job online as access to job opportunity is wider, similarly online available information improves the performance, getting information online for educational purposes, availing discount offers online, buying and selling goods online. social fields include networks that give access to knowledge and social and emotional support. political and civic involvement through internet is also included in social field e.g. making opinions about politics, voting etc. this research focus on the social and economic benefits/outcomes among the students of different disciplines resulting from different online activities also the relationship between the internet use and subjective well-being examined. only social and economic fields have been selected as both are strongly interrelated. economic outcomes are measured by income, education, employment and financial indicators that includes online job seeking, banking, buying and selling or for learning purposes. social outcomes include the measures of different social networks (personal, political, formal) that give access to knowledge and support of others. subjective well-being i.e. people’s assessment of their own well-being as described by sabatini and sarracino (2014) is represented by the individual’s life satisfaction. the main objective of this research is to investigate that how socio-economic outcomes and subjective wellbeing affect duration of internet usage and this we test the hypothesis: h1: social outcomes and economic outcomes do not affect the duration of internet usage and that h2: subjective well-being does not have effect on duration of internet usage. 3. data collection and methods the participants of the study are 501 in numbers from different educational institutes in pakistan with major in different disciplines. the sample size consisting of 501 participant was calculated with the help openepi1 software. the three main factors 1 http://www.openepi.com/menu/oe_menu.htm atta ur rahman, shandana, laila taskeen qazi, and adnan khan 66 used to calculate the above-mentioned sample size include proportion of population using internet being 36%, confidence level of 90% and 5% error. data is collected online as well as offline i.e. questionnaires are distributed and filled by participants. duration of internet usage is measured as internet use intensity in hours per day. it also includes the usage of social networking sites, blogs, games, emails and business websites. the gender and marital status are dichotomous variables with male being 1 and female being 0 and 1 for married and 0 for unmarried. education level is divided in four categories i.e. higher secondary school, bachelors (2 years), bachelors (4 years) and masters or above. similarly current situation is measured in three categories i.e. studying, employed and unemployed. educational programs are divided in three major disciplines i.e. medical, engineering and others including social and management sciences, art and humanities and agricultural studies. subjective health is also measured by a 6 point scale 1 being very unhealthy to 6 being very healthy. satisfaction with life scale (swl scale; diener et al., 1985) is used to measure the subjective well-being of the individuals. subjective well-being is also measured by pana scale (positive affect and negative affect scale) by watson et al., 1988. a 7-point likert scale is used in swl scale to answer the questions with higher score showing more satisfaction with life. the pana scale includes two components i.e. positive affect and negative affect that shows pleasant and unpleasant experiences of individuals. pana scale includes 10 items; 5 positive and 5 negative. the usage measurement has two categories i.e. social and economic. internet use in social and economic field is measured by the subscales in outcomes measurement. internet usage in economics fields includes education, property, employment and income. similarly social fields includes personal, formal and political networks as used by helpser’s model of corresponding fields as it assumes the effect of digital usage on offline resources. subscales are selected to measure the social and economic outcomes i.e. eight subscales for each variable. variables of social and economic outcomes consists of sum scores based on individual items that is set on likert scale ranges from 1 being strongly agree to 7 being don’t know. cox regression is used for time dependent variables and results are interpreted accordingly. the predicted variable is internet usage, however internet usage is computed as time i.e. in hours. the econometric model can be written as; ℎ(𝑡) = ℎ0(𝑡)exp⁡(𝛽1𝑥1 + 𝛽2𝑥2 + ⋯ + 𝛽𝑘𝑥𝑘)… … … (1) where h(t) is the duration of internet usage at time t, h0(t) is the baseline hazard and the x1, x2 , xk are the k independent variables. journal of applied economics and business studies, volume. 5, issue 3 (2021) 59-76 https://doi.org/10.34260/jaebs.534 67 4. data analysis 4.1 descriptive analysis: of the respondents, 62% were male and 38% were female with mean age 24. out of them, 40% were currently enrolled as undergraduate students, 18% post graduate students, 23% currently working, 8% reported as working part time and 10% unemployed. 37% were with medical as major, 34% engineering, the rest 29% with other disciplines such as 7% each in social sciences, management sciences/business studies, arts and humanities, 3% agriculture and the rest 5% from other disciplines. furthermore, 85% are unmarried and 15% married. 98% have access to internet, where 89% use internet every day for different purposes. table 1: demographic profile of internet users/respondents age frequency percent 16-20 67 14.3 21-25 279 59.4 26-30 102 21.7 31-35 15 3.2 36-40 5 1.1 41-45 2 .4 gender frequency percent female 177 37.7 male 293 62.3 total 470 100.0 education frequency percent higher secondary school 81 17.2 bachelors 2 years 65 13.8 bachelors 4 years 201 42.8 masters/higher 123 26.2 current status frequency percent unemployed 51 10.9 employed 150 31.9 studying 269 57.2 major frequency percent medical 179 38.1 engineering 159 33.8 others 132 28.1 marital status frequency percent unmarried 400 85.1 married 70 14.9 total 470 100.0 atta ur rahman, shandana, laila taskeen qazi, and adnan khan 68 out of 470 respondents 201 are in their 4 years bachelors’ program, 123 are doing masters or ms/mphil/ph.d., 65 are in their 2 years bachelors’ program. 51 are unemployed, 150 are on job and 269 are getting education. 179 were studying medical, 159 engineering and 132 have other majors. 70 were married and 400 unmarried. according to the calculation mean score of internet usage is 5.87 hours, mean score of economic and social outcomes is 4.04 and 3.85 respectively, and similarly swb, pa and na scale score 4.50, 0.57 and 0.45 respectively. the mean score of economic outcomes in terms of education is 4.11 that show agreement of the people that internet usage has economic outcomes in the field of education. similarly, the fields of employment and income also show mean score of 4.27 and 4.11 which indicate the approval of the people about economic outcomes. social outcomes in the form of formal and political networks also indicate that mostly people agree to the fact that internet usage has social outcomes. figure 1 explores access to online communication technology of the respondents. most of the respondents have smart phone or tablet i.e. 403 out of 470 and 340 have laptop or desktop computers, 397 poses email accounts, 361 have accounts on social networking sites and 131 run a blog or a micro-blog. figure 2: internet usage intensity 84.5 76.8 27.9 85.7 72.3 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 an email account a social networking site (e.g. facebook, linkedin) a blog, a microblog (e.g. twitter, tumblr account) a smart phone, a tablet, ipad or ereader a laptop or desktop computer journal of applied economics and business studies, volume. 5, issue 3 (2021) 59-76 https://doi.org/10.34260/jaebs.534 69 figure 2 poses the daily activities of the respondents using internet. most of the people spend their time using social networking sites (sns) i.e. 310 numbers of respondent uses sns daily out of 470. 257 use search engines for different purposes every day, 155 check their email daily, 70 respondents use internet for playing games every day, 43 respondents check business websites for online buying and selling purposes and only 32 use internets for blogging. figure 3: daily online activities 4.2 cox regression cox regression is used with time dependent variable to investigate the relationship between time i.e. internet usage hours and predictor variables. table-3: estimation results of cox regression for internet usage time as dependent variables in the equation b se wald df sig. exp(b) age .010 .027 .143 1 .705 1.010 gender -.372 .166 5.034 1 .025 .690 education -.169 .094 3.209 1 .073 .845 current status -.070 .133 .274 1 .600 .933 major -.188 .109 2.982 1 .084 .828 marital status .273 .234 1.360 1 .244 1.314 subjective health -.017 .072 .053 1 .817 .984 eco outcomes -.061 .073 .708 1 .400 .941 social outcomes -.111 .085 1.717 1 .190 .895 swb .207 .075 7.586 1 .006 1.230 pa -.781 .282 7.666 1 .006 .458 na -.515 .255 4.087 1 .043 .598 33.0 6.8 66.0 54.7 14.9 9.1 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 email use intensity (days/week) blog social networking site search engine game website business website atta ur rahman, shandana, laila taskeen qazi, and adnan khan 70 the above table 3 shows the results of cox regression analysis. internet usage is 0.01 times more in a person who is a year older than the other. similarly, men use it 0.31 times less as compared to women. married person use it 31% more than the unmarried one. a person using internet gets .059 times less economic outcomes than the person not using internet and social outcomes for internet user is 10.5% lower than the nonuser. the subjective well-being of the internet user is .23 times more than the one not using internet. positive and negative analysis both shows 54% and 40% decrease respectively in the odds of using internet. p-value of the variables gender, education, swb, pa, na significantly affecting the duration of internet usage. 5. discussion fong (2009) results showed that there is significant relationship between information and communication technology (ict) and gross national income per capita (gni) of 91 developing countries is evaluated through regression analysis. this current study showed that there is no relationship between the two variables. sun, (2011) analyzes the effect of technology use on the performance of education. the findings of current study also suggest that education have significantly affecting the usage of internet. chou and hsiao (2004) in their research investigate the internet addiction of college students and the patterns they follow while using internet and gratification and communication pleasure by using internet. the results show self-reported pleasure being the strongest predictor of internet usage and internet “addicts” spend three times more hours on internet than the “non-addicts” specifically on web, email, games and online communication and socializing applications such as bbs which is used in taiwanese institutes. internet users who use internet for more hours are satisfied and happier than the one who use internet less. however, this negatively affects the daily routine and studies of users but positively affect the relationship with friends and family. the result here in this research suggests that search engines and social networking sites are the tools, respondents spend most of their time on while using internet. however, the social networks i.e. relationship with friends and family members through online communication show negative association with internet usage. results of cox regression show that economic outcomes are less for internet users than the people not using it, similarly social outcomes are lower for users than the nonusers. this shows how online activities affect the economic and social activities of the individuals and in turn affecting journal of applied economics and business studies, volume. 5, issue 3 (2021) 59-76 https://doi.org/10.34260/jaebs.534 71 their subjective well-being. both the models in the research express positive relationship of subjective well-being with internet usage and higher subjective well-being show higher satisfaction with life. similarly (dennis, et al., 2015) examine the relationship between social exclusion causing shopping and its effects on wellbeing and happiness. the results show that people who consider themselves socially excluded spend more time shopping online (through smart phone and computer) and offline (physically). the results indicate that however social exclusion negatively affects the wellbeing and happiness; people spent time shopping to overcome that feeling. research suggests that people use internet in order to lessen their depression. the results of the (robinson, dimaggio and hargittai, 2003) results show that college or higher educated people use internet more than students of lower education level (high school) in the form of education and later on job finding in the long-term personal usage in the short term by communicating with family relations and friends through internet. wang and wang (2011) use linear regression to find out the relationship online communication and subjective well-being of users. the results show positive association between subjective wellbeing and online communication also boys benefit more than girls in terms of gaining wellbeing from internet usage. according to usage and gratification theory people use internet for different purposes such as information gaining and socializing etc. through different applications and tools that results in pleasure and satisfaction of the users. nie, et al., (2017) analyze the relationship between internet usage and different measures of subjective wellbeing. the results of the study show negative association of internet usage with life satisfaction and happiness and close association with depression. the results are like this research where internet usage is positively related to life satisfaction and negatively associated with happiness and depression. cho et.al. (2003) explains the relationship between internet usage and pleasure gained from it and the finding indicates that young people with higher socio-economic status use internet to gain pleasure and feel satisfied through communication and learning. the cox regression shows that internet usage increase with age, similarly women and married people use internet more than men and unmarried people. economic and social outcomes show negative relationship with internet usage in other words internet usage has less economic and social outcomes. binary logistic regression shows same results for age, gender, marital status, discipline, for social and economic outcomes and atta ur rahman, shandana, laila taskeen qazi, and adnan khan 72 for subjective well-being. with increase in age, internet usage also increases, likewise internet usage is found more in women and married individuals than men and unmarried individuals. similarly employed people use it more than the unemployed one. the relationship between socio-demographic variables and internet use and types of internet usage is analyzed by deursen and djik (2013) through linear regression analysis. the results show that disabled and less educated people spend more time using internet than the people who are more educated. age, gender and education prove to be the strong predictors of internet usage. however the results of porter and donthu (2006) paper are opposite the above mentioned results by deursen and djik (2013). it shows that highly educated, wealthy and young people use internet more than the less educated, poor and older people. findings of the research are considered useful in terms of personal, social and economic fields as it helps in policy making and benefits the society. hargittai (2010) investigates the relationship between socio-economic status and internet users’ skills and types of internet usage. the results of the study show that not everyone has the internet knowledge therefore usage is different as white people (male gender) with educated parents have better internet usage skills than the opposite ones. similarly, the research proves socio-economic status as strong predictor of internet usage skills. maclaren and zappala (2002) also conclude that internet access to the children of lower socio-economic status is strongly connected to the education of parents also internet access at home is beneficial for the performance of children in education. the researcher suggests that internet access to people with lower socioeconomic status need to be preferred by policy makers. 6. conclusion this research investigates the impact of age, gender, education, employment and marital status, its relation to social and economic outcome and subjective well-being on internet usage. age, marital status and subjective well-being show positive affect on internet usage however gender, economic and social outcomes have negative affect on internet usage. different use of internet in different age is likely to be a temporary difference as the young generation will grow old and also the older people are getting more involved in online activities for socializing etc. however some part of inequality will remain in age and gender because of social and cultural preferences of the society. journal of applied economics and business studies, volume. 5, issue 3 (2021) 59-76 https://doi.org/10.34260/jaebs.534 73 different use of internet associated with the education has long-term effects as it is related to the information and knowledge of the society. the analysis of the data revealed the classification of internet use that people mostly prefer while using and spending time on, are social networking sites and search engines. uses and gratification theory is endorsed by the classification of internet usage that shows the motives of the users. as chou, condron & belland (2005) says that internet is changing the society’s way of living and although the use of internet can be similar, but the outcomes and effects are dissimilar that ultimately depends on the nature of use. internet usage improves the skills, abilities and potentiality of the users but at the same time it also results in inappropriate use hence leading to both positive and negative outcomes. findings of the research suggest that internet usage leads to the result of negative social outcomes. it suggests that users who get economic outcomes of internet do not necessarily get the same social outcomes as well. similarly benefits from online activities on internet do not necessarily mean that offline activities i.e. activities other than digital involvement, are not beneficial and important in our life. economic outcomes are more important for policy makers than social (formal and informal networks) outcomes as they are associated with education and employment. the relationship between internet use and subjective well-being is positive that shows positive relationship of internet with life satisfaction, it also shows higher level of depression and negative association with happiness through positive affect and negative affect analysis (pana). since internet users are new and amateur, it is possible that the use will change after some time as internet will become more dominant and users will become more experienced. future research can include personal and cultural fields to investigate the outcomes of internet usage. race, location and access to internet may also affect the internet usage therefore difference of use between the literate and illiterate, urban and rural differences showing geographical inequalities can be investigated. atta ur rahman, shandana, laila taskeen qazi, and adnan khan 74 7. references bandura, a. (2001). social cognitive theory: an agentic perspective. annual review of psychology, 52(1), 1-26. bonfadelli, h. (2002). the internet and knowledge gaps: a theoretical and empirical investigation. european journal of communication, 17(1), 65-84. bucy, e. p., & newhagen, j. e. (eds.). (2004). media access: social and psychological dimensions of new technology use. psychology press. chinn, m. d. & fairlie, r. w. (2004). the determinants of the global digital divide: a cross-country analysis of computer and internet penetration. center for global, international and regional studies uc santa cruz, escholarship. cho, j., de zúñiga, h. g., rojas, h. & shah, d. v. (2003). beyond access: the digital divide and internet uses and gratifications. it & society, 1 (4), 46-72. chou, c. & hsiao, m. c. (2000). internet addiction, usage, gratification, and pleasure experience: the taiwan college students' case. elsevier science ltd., computers & education, 35, 65-80. chou, c., condron, l. & belland, j. c. (2005). a review of the research on internet addiction. educational psychology review, 17 (4), 363-388. dennis, c., alamanos, e., papagiannidis, s., & bourlakis, m. (2016). does social exclusion influence multiple channel use? the interconnections with community, happiness, and well-being. journal of business research, 69(3), 1061-1070. davis, f. d. (1989). perceived usefulness, perceived ease of use, and user acceptance of information technology. mis quarterly, 319-340. diener, e. d., emmons, r. a., larsen, r. j., & griffin, s. (1985). the satisfaction with life scale. journal of personality assessment, 49(1), 71-75. fong, m. w. l. (2009). digital divide: the case of developing countries, informing science and information technology, 6, 471-478. hargittai, h. (2010). digital na(t)ives? variation in internet skills and uses among members of the ‘‘net generation”. sociological inquiry, 80 (1), 92-113. helliwell, j. f., huang, h., & wang, s. (2014). social capital and well-being in times of crisis. journal of happiness studies, 15(1), 145-162. journal of applied economics and business studies, volume. 5, issue 3 (2021) 59-76 https://doi.org/10.34260/jaebs.534 75 helsper, e. j., van deursen, a. j. a. m. & eynon, r. (2015). tangible outcomes of internet use, from digital skills to tangible outcomes project report of oxford internet institute. hijab, n. (2001). people's initiatives to use it for development. undp. hoffman, d. l., & novak, t. p. (1998). bridging the digital divide: the impact of race on computer access and internet use. larose, r., lin, c. a. & eastin, m. s. (2003). unregulated internet usage: addiction, habit, or deficient self-regulation? lawrence erlbaum associates, media psychology, 5, 225-253. maclaren, j. & zappala, g. (2002). the digital divide among financially disadvantaged families in australia. first monday, 7 (11), 1-22. nanthikesan, s. (2000). trends in digital divide. harvard center for population and development studies, cambridge, ma. nie, p., sousa-poza, a., & nimrod, g. (2017). internet use and subjective wellbeing in china. social indicators research, 132(1), 489-516. payton, f. c. (2003). rethinking the digital divide. communications of the acm, 46 (6), 89-92. porter, c. e. & donthu, n. (2006). using the technology acceptance model to explain how attitudes determine internet usage: the role of perceived access barriers and demographics. elsevier, journal of business research, 59, 999–1007. robinson, j. p., dimaggio, p. & hargittai, e. (2003). new social survey perspectives on the digital divide. it & society, 1 (5), 1-22. rahman, a. u., & uddin, s. (2009). statistical analysis of different socio economic factors affecting education of n.w.f.p (pakistan). journal of applied quantitative methods, 4(1), 88-94. sabatini, f. (2011). can a click buy a little happiness? the impact of businessto-consumer e-commerce on subjective well-being. mpra munich personal repec archive, 32393, posted 24. july 2011 15-40. sabatini, f., & sarracino, f. (2014). will facebook save or destroy social capital? an empirical investigation into the effect of online interactions on trust and networks. atta ur rahman, shandana, laila taskeen qazi, and adnan khan 76 selwyn, n. (2010). degrees of digital division: reconsidering digital inequalities and contemporary higher education. rusc. universities and knowledge society journal, 7 (1), 33-42. sun, j. c. y., & metros, s. e. (2011). the digital divide and its impact on academic performance. online submission. van dijk, t. a. (2013). news as discourse. routledge. van deursen, a. j. a. m. & van djik, j. a. g. m. (2014). the digital divide shifts to differences in usage. sage publications, 16 (3), 507-526. van praag, c. m., & cramer, j. s. (2001). the roots of entrepreneurship and labour demand: individual ability and low risk aversion. economica, 68(269), 45-62. wang, j. & wang, h. (2011). the predictive effects of online communication on well-being among chinese adolescents. scientific research, psychology, 2 (4), 359-362. watson, d., clark, l. a., & tellegen, a. (1988). positive and negative affect schedule (panas). journal of personality and social psychology, 54(6), 1063-1070. young, k. s. (2004). internet addiction, a new clinical phenomenon and its consequences. sage publications, 48 (4), 402-415. zillien, n., & hargittai, e. (2009). digital distinction: status‐specific types of internet usage. social science quarterly, 90(2), 274-291. journal of applied economics and business studies, volume. 5, issue 2 (2021) 103-120 https://doi.org/10.34260/jaebs.527 103 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x contribution of agriculture sector in economic growth of pakistan: an empirical analysis sumia bint zaman1, muhammad ishaq 2*, muhammad azam niazi 3 1 parc institute of advance studies in agriculture, national agriculture research centre, islamabad 2 parc institute of advance studies in agriculture, national agriculture research centre, islamabad 3 parc institute of advance studies in agriculture, national agriculture research centre, islamabad abstract there has been controversy in the field of development economics about the significance of the role of agriculture sector in economic growth. going through the data, it indicates that agriculture sector is significant contributor to the economy of pakistan as it contributes about 19% in national gdp. this study was designed to statistically test the contribution of agriculture sector in economic growth of pakistan through estimation of relationship between agriculture sector and pakistan’s economic growth using autoregressive distributed lag (ardl) bounds test and error correction model (ecm). time series data on selected variables was utilized from 1961-2018. study found that real agricultural value added has a significant positive impact on real gdp per capita in the long-run where one percent increase in real agricultural value added increases the real gdp/capita by 0.35%. this indicated that the promotion of agriculture sector leaves far reaching effects with respect to economic growth of the country. these results advocated for the development of agriculture sector in line with the long-term goals of economic growth and emphasized in investing in agriculture sector. coefficient of error correction term (ect) is -0.62 meaning that if there is any disequilibrium, it will restore @ 62 percent in the first period. results also proved the importance of capital formation both the physical capital and human capital. finding suggested that we keywords growth theory, economic growth, agriculture, ardl jel classification o4, o49, n5 * ishaqecon@gmail.com mailto:ishaqecon@gmail.com sumia bint zaman, muhammad ishaq, muhammad azamniazi 104 should investment in human health to enhance the economic growth as suggested by exogenous growth theory. moreover, it can also be suggested to create conducive environment and economic opportunities to reap the benefits of demographic dividends of decreased mortality in the long-run. as per analysis, maintaining stability is critically important for economic growth. moreover, literature hypothesize the positive effect of tot for economic growth, but analysis indicated that tot has not been able to put any significant impact on economic growth. further, trend analysis also pointed out that tot has been fluctuating over the time. it can be inferred from the analysis that there is need to stabilize tot and restructure the exports of the country to generate the significant positive impact. 1. introduction agriculture is considered one of the sectors which are of central importance for national economy of pakistan, but this hypothesis needs to be tested empirically. there has been controversial debate on the role of agriculture sector in economic development especially after the periods of colonial rule (lewis, 1954; fei and ranis, 1961; johnston and mellor, 1961; jorgenson, 1961; and schultz, 1964). but the factor is that most of these studies were qualitative and accentuated the possible influence of the inter-linked sectors of the economy, but direction of these relationships was not definite in sequence. numerous studies proclaimed that overall economic growth is contingent to the advancement of agricultural sector (schultz, 1964 and gollinet al, 2002). supporters of agriculture-led growth believed that investment in agriculture sector including supplementary infrastructure and institutions in linked sectors is a precondition for economic development of the country where many of them established that agriculture do effect the income of rural people and provide resources for industrial development to boost the economy(schultz, 1964;eicher and staatz, 1984;dowrick and gemmell, 1991;timmer, 1995;datt and ravallion, 1998; and thirtleet al, 2003). bhagwati and srinivasan (1975) exposed that some developing countries who tried to develop their economies through industrializing without developing their agriculture sector first, ended up in gloomy economic growth rates coupled with the uneven income distribution. johnston and mellor (1961) advocated the agricultural contribution in economic development through different inter-sectoral linkages includingi) provision of surplus labor from agriculture to industry, (ii) increased food supply and hence higher consumption, (iii) market for industrial production, (iv) savings for investment, and (v) earnings from agricultural export. additionally, timmer (1995) highlighted the significance of supplementary nonmarket connections that enhance the productivity of production factors, maintain stability of food prices, and hence lessen the poverty levels journal of applied economics and business studies, volume. 5, issue 2 (2021) 103-120 https://doi.org/10.34260/jaebs.527 105 in the economy. humphries and knowles (1998) also complemented the assertion that relocation of labor from agriculture to other sectors is correlated with economic expansion. in contradiction to this viewpoint, some others believed that agriculture sector does not have robust connections to other segments of the economy (lewis, 1954; fei and ranis, 1961; and jorgenson, 1961). based on this viewpoint, many policies in developing countries were made focusing the sectors other than agriculture such as manufacturing sector (okonkwo, 1989 and schiff and valdez, 2002). similarly, based on such kind of findings, some developing countries adopted policy of heavy taxation/duties on agriculture sector for example china has been imposing heavy taxes on agriculture before 1979 and the amount collected through these taxes was used to subsidize industrial development and urbanization (yao, 2000). according to tiffin and irz, 2006;, olsson and hibbs, 2005;kogel and prskawetz, 2001; humphries and knowles, 1998; echevarria, 1997; and many more, intensity of this controversy has amplified with the passage of time after extensive research both qualitative and quantitative on the topic. criticism on earlier empirical work on agriculture sector role in economic development is that most of these studies are based on cross-sectional analysis of panel data which have substantial confines so these may not have yielded conclusive inferences (tsakok and gardner, 2007). among others, awokuse (2007) described that simple bivariate causality analyses of such type are suspected of unauthentic results because they may overlook the other significant variables which may possibly have significant effect. recent studies that have used time series panel data and advance econometric techniques, also showed the diverse and at times contradictory proof so there is still a deficiency of harmony on the impact of agriculture on economic advancement (awokuse and xie 2015). as going through the data1, it seems that agriculture is very important for pakistan economy, the underlying association between agriculture sector and overall economy and its impact on economic growth of the country is an empirical question that demands to be studied. in pakistan, this issue has attracted very little consideration of the researchers. some of the research work done by ahmad and ahmad, 2018; raza et al, 2012 and azraet al, 2013 have addressed the issue. according to ahmad and ahmad (2018) agricultural exports of the country proved to be influential in economic growth. though pakistan’s export base is highly dependent on agriculture and trade of agricultural products accounts for 70% of the foreign exchange earnings of the country (rehman et al, 2016) but agricultural exports are only a component of the whole sector so impact of agricultural exports cannot depict the true picture of the effect of the whole agriculture sector on economic growth. a study by raza et al (2012) revealed that agriculture sub-sectors are significantly contributing in the economy except forestry which showed minimal 1agriculture contributes about 19% percent of the gross domestic product (gdp) of pakistan and employs 38.5% of the labor force (gop, 2018). sumia bint zaman, muhammad ishaq, muhammad azamniazi 106 contribution. the study employed ordinary least squares (ols) regression and have not accounted for co-integration. therefore, simple correlation coefficient tests might be having misspecification problems and results may be spurious. azra et al (2013) explored association between agriculture and economic growth of the country by estimating error correction model (ecm) and concluded that agriculture sector contributes significantly in the economy of a country. the current study was designed for further improvement and extension of the work done by azraet al (2013) through: using advance techniques in time series data analysis, adding other relevant variables in growth equation, and using extended data set (azra et al, 2013 used data from 1981 to 2005 while current study used data from 1961 to 2018). therefore, this study is conducted with the objective to estimate the relationship between growth of agriculture sector and pakistan’s economic growth using recent econometric techniques and suggest informed policy guidelines for future planning. 2. review of literature awokuse and xie (2015) investigated the direction of causal relationship between agriculture sector and economic growth using data from nine developing countries by applying ardl co-integration procedure along with directed acyclic graphs (dag) technique. study found that this direction of the relationship between agriculture sector and economic growth varies by countries. empirical analysis supported that agricultureled growth hypothesis is true for some of the countries while result for some countries supported long-run connecting stream from economic development to agriculture development. ansari and khan (2018) tried to study the significance of decreasing agriculture share in overall gdp of four countries including pakistan, india, sri lanka, and bangladesh by applying maximum likelihood technique proposed by johansen and juselius (1990), based on vector autoregression (var) model with application of granger causality test. results confirmed the hypothesis that agriculture sector has a significant positive relationship with the overall economic growth. herath et al (2013) examined the possible impact of agriculture in the development of north carolina consuming national level statistics from 2000 to 2010 through simultaneous equation analysis. results revealed that causal relationship between agriculture and economic development exist in two-way direction. tiffin and irz (2006) scrutinized underlying association between agricultural and economic development using data from panel of countries with bivariate granger causality tests and strongly inferred that agricultural growth significantly contribute in economic growth in developing countries, but no decisive results were found for developed countries. in contrast, gardner (2005) analyzed panel data of 85 countries and found that growth of agriculture is independent of per capita income growth for those who are employed in this sector. however, features triggering agricultural growth proved to impact growth of national gdp positively. similarly, matsuyama(1992) negated the assertion that agricultural efficiency is an apparatus of economic development by giving comparative advantage theory and theoretically journal of applied economics and business studies, volume. 5, issue 2 (2021) 103-120 https://doi.org/10.34260/jaebs.527 107 analyzed two sector economy growth model in closed and open economy scenario and concluded that agriculture is positively linked with economic development only in the case of closed economy model while negatively linked in case of open economy. agriculture sector is considered important not only for the overall economy but for the growth of other vital sectors of the economy mainly the industrial sector. hye (2009) investigated link between agriculture and industry for sustainable economic development in case of pakistan using annual data of agricultural and industrial output from 1971 to 2007 and applied ardl technique. results revealed that agriculture sector do have impact on industrial sector both in short and long run. gollin et al (2002) examined the role of agriculture in development through analyzing the link between industrial and agricultural productivity and relativity of the role of these two sectors in the growth in short and long run. analysis showed that agricultural productivity enhancement can affect overall income of the country through accelerating the industrial sector. in short run, improvement in agricultural productivity affect the overall nationalgrowth more as compared to the productivity improvement of non-agricultural sector. yao (2000) carried out cointegration analysis of five major sector of chinese economy including agriculture and found that these sector are co-integrated but agriculture is mainspring of economic expansion as well as development of the other four sectors but these sector does not contribute in agricultural development. this literature review concludes that there are mixed findings about the role of agriculture sector in the overall economies of different countries. there can be significant contribution of agriculture to economic growth as well as economic growth can be promoter of agriculture sector. two-way directional link between these two variables can also coexist according to some findings. it was found that this effect may vary from country to country. findings also pointed out that economic growth can also be independent of agriculture sector growth, but this cannot be generalized. 2.1. overview of agriculture sector growth in pakistan over time sustainable growth of agriculture sector has always been desired throughout the history. actually, there has been variations in agricultural growth over the time. after independence, there has been stagnation in this sector in 1950s. however, it started growing in later years of the decade. the sector experienced tremendous growth in 1960s especially the second half of this decade where average growth rate in second half of the decade recorded more than 6% mainly due to impressive growth in crops. this was called as green revolution (gr) where high growth was due to technology and high input use like seeds fertilizer etc, infrastructure development like dams building and introduction of new varieties. however, green revolution was criticized because of its unsustainability and resulting resource degradation, but agricultural growth was tremendous in that decade. early 1970s witnessed drop in agricultural growth mainly due to decreased yields of major crops mainly rice, cotton, and sugarcane and also, low growth in livestock except poultry. sumia bint zaman, muhammad ishaq, muhammad azamniazi 108 the growth recovered in second half of 70s. growth momentum continued in first half of 1980s where non-cereal crops and livestock appeared as main growth sources but drop in growth was observed in second half of 1980s and first half of 1990s when livestock and sugarcane crop witnessed relatively high growth but low performance of wheat, rice, and cotton. crops production has been fluctuation over the time and so as the agricultural growth (chaudhry and chaudhry, 1997). over the time of four decades (1960 to 2000), average growth rate has been around 4% which was higher than growth rate of population which has been round 2% per annum on average. however, growth has not been sustainable in 2000s and resulted in high food prices and more imports of wheat. this could be attributed to external factors like drought years in first half of that decade, flood in the end of that decade, and global food crises (gop, 2010). same fluctuating trends has been observed in agricultural growth in recent decade i.e. 2010s where drop down was observed in first half and it was minimum (0.154%) in 2016 due to water shortages while started recovering afterward (gop, 2012 2020). agricultural role has changed over the time from only increasing production and high growth to ensuring food security, which is a challenge. this challenge is going to be more complex in coming time w.r.t. decreasing available resources and deteriorating quality of these resources (farooq, 2015). 3. material and methods 3.1. data and analysis universe of the study consist whole pakistan. this study used annual national level data collected from world development indicators (wdi) database by world bank (wb) and database of united nations children’s fund (unicef).annual data from 1960 to 2018 was taken on the variables including gdp, agricultural value added, exports, imports, gross capital formation, mortality rate, and gdp deflator. firstly, data was transformed from nominal values to the real values and then transformed into log form. log transforming is commonly used to deal with data skewness, reduce the extent of variation, and/or to convert non-linear function like cobb-douglas production into linear functional form. since model of this study was based on cobbdouglas function, so data was transformed by taking natural log where coefficients serve as elasticities. ardl bounds testing technique along with error correction model (ecm) was applied in the analysis as used in more recent studies of such type for determining both long and short run relationships among different variables (awokuseet al,2015; udohet al,2015;alkhatlan, 2013;chaniet al,2011;hye, 2009;and pahlavaniet al,2005). ardl model is based on ols and it can be applied on both non-stationary series and on series of mix integration order (shrestha and bhatta,2018). ardl furnish the procedures of engendering data from general-to-specific frame by integrating appropriate lags and journal of applied economics and business studies, volume. 5, issue 2 (2021) 103-120 https://doi.org/10.34260/jaebs.527 109 integrates the short-run equilibrium through ecm along with the evidence of long-run relationship. it also provides the coefficient of speed of adjustment in case of disequilibrium. moreover, ardl is suitable for small data sets (nkoro and uko, 2016). though ardl bounds testing do not require the pre-testing of data for stationarity, but it is better to confirm that none of the variables is i(2) which is a pre-condition for the approach. stationarity characteristic of the data was examined using augmented dickyfuller (adf) test and phillipse perron (pp) test to validate the findings. schwartz bayesian criterion (sbc) was used for the selection of lag length. while performing the stationarity tests, trend component was tested and included where it was found to be significant. 3.2. model specification modified solow–swan growth model was used for examining the association between agriculture sector and economic growth analysis following the previous researches of such kind (ansari and khan, 2018;awokuseet al,2015; humphries and knowles, 1998;faridiet al,2012; and ahmad and ahmad, 2018). basically, the solow–swan growth model is an model for measuring long-run economic growth and this model was modified by mankiw et al (1992) by including variables of education and health in the model supposing that enhancement of education and health improvement of the labor force enhance the labor productivity and hence contribute in economic growth. augmented solow-swan growth model is a way of including the other relevant variables in the growth model. this model undertakes the assumption of an aggregate production function with constant-return-to scale written as: 𝑌𝑡 = 𝑓(𝐾𝑡𝜆 𝛽𝑡) (1) where yt represent real gdp per capita in the time period under consideration. kt represent the real gross capital formation in the period tand denoted as gcf later on; βt denotes productivity term in the period t; and 𝜆 is the output elasticity of capital. the effect of agriculture to overall economic growth is measured through its consequences on total factor productivity captured in the model by β. therefore, β in the model is supposed to be a function of agricultural production, terms of trade, and mortality: 𝛽 = 𝑓 (𝐴𝐺𝑅, 𝑇𝑂𝑇, 𝑀𝑂𝑅𝑇) writing this equation statistically 𝛽𝑡 = 𝜆1𝐴𝐺𝑅𝑡 + 𝜆2 𝑇𝑂𝑇 + 𝜆3𝑀𝑂𝑅𝑇 + µ𝑡 (2) https://en.m.wikipedia.org/wiki/economic_growth sumia bint zaman, muhammad ishaq, muhammad azamniazi 110 where agr is for agriculture, tot is for terms of trade, and mort is mortality rate (as proxy of health), and µ is error term that covers the remaining factors which can affect the productivity other than those included in the model. by incorporating equation (2) in equation (1) 𝑌𝑡 = 𝜆1𝐴𝐺𝑅𝑡 + 𝜆2𝐺𝐶𝐹𝑡 + 𝜆3𝑇𝑂𝑇 + 𝜆4 𝑀𝑂𝑅𝑇𝑡 + µ𝑡 (3) by including constant term and taking natural logs of this equation 𝑙𝑛𝑌𝑡 = 𝐶𝑜 + 𝜆1 𝑙𝑛 𝐴𝐺𝑅𝑡 + 𝜆2 𝑙𝑛𝐺𝐶𝑓𝑡𝜆3 𝑙𝑛 𝑇𝑂𝑇𝑡 + 𝜆4 𝑙𝑛 𝑀𝑂𝑅𝑇𝑡 + µ𝑡 (4) model in equation (4) was estimated in the study. ardl model is as under: 𝛥𝑙𝑛𝑌𝑡 = 𝐶𝑜 + 𝜆1𝑙𝑛𝑌𝑡 − 1 + 𝜆2 𝑙𝑛𝐴𝐺𝑅𝑡 − 1 + 𝜆3 𝑙𝑛𝐺𝐶𝐹𝑡 − 1 + 𝜆4 𝑙𝑛𝑇𝑂𝑇𝑡 − 1 + 𝜆5𝑙𝑛𝑀𝑂𝑅𝑇𝑡 − 1 + ∑ 𝛾 𝑝 𝑖=1 𝛥𝑙𝑛𝑌𝑡 − 𝑖 + ∑ 𝛾1 𝑞 𝑖=0 𝛥𝑙𝑛𝐴𝐺𝑅𝑡 − 𝑖 + ∑ 𝛾2 𝑞 𝑖=0 𝛥𝑙𝑛𝐺𝐶𝐹𝑡 − 𝑖 + ∑ 𝛾3 𝑞 𝑖=0 𝛥𝑙𝑛𝑇𝑂𝑇𝑡 − 𝑖 + ∑ 𝛾4 𝑞 𝑖=0 𝛥𝑙𝑛𝑀𝑂𝑅𝑇𝑡 − 𝑖 + µ𝑡 whereas, error correction form of the model is: 𝛥𝑙𝑛𝑌𝑡 = 𝛼 + ∑ 𝛽𝜊 𝑝 𝑖=1 𝛥𝑙𝑛𝑌𝑡 − 𝑖 + ∑ 𝛽1 𝑞 𝑖=0 𝛥𝑙𝑛𝐴𝐺𝑅𝑡 − 𝑖 + ∑ 𝛽2 𝑞 𝑖=0 𝛥𝑙𝑛𝐺𝐶𝐹𝑡 − 𝑖 + ∑ 𝛽3 𝑞 𝑖=0 𝛥𝑙𝑛𝑇𝑂𝑇𝑡 − 𝑖 + ∑ 𝛽4 𝑞 𝑖=0 𝛥𝑙𝑛𝑀𝑂𝑅𝑇𝑡 − 𝑖 + 𝛽5𝐸𝐶𝑇 𝑡 − 1 + 𝜀𝑡 where i is optimal number of lags and ect is error correction term. 3.3. theoretical justification of the variables in the model role of agricultural output in overall economic growth may possibly be demonstrated through its impact on total factor productivity or as input to the manufacturing sector (ruttan, 2000 and timmer,1995). development theories in the past took agriculture as an essential resource base for funding the expansion of manufacturing sector. agriculture is considered as an apparatus of growth so researches including ansari and khan,2018;awokuse and xie, 2015;heratheet al,2013; azraet al,2013; raza et al,2012; tiffin and irz, 2006; yao, 2000; and many more added agriculture in the growth equation to measure its impact on economic growth. this study was mainly focused to measure the impact of agricultural output on economic growth. capital is considered as one of the basic factors for economic growth in endogenous journal of applied economics and business studies, volume. 5, issue 2 (2021) 103-120 https://doi.org/10.34260/jaebs.527 111 as well as in exogenous growth theory (harrod, 1939;domar, 1946; solow,1956; swan, 1956;lucas, 1988; and romer, 1990).present study also included gross capital formation as independent variable in growth equation to see its impact on economy of pakistan following the other researches (ahmad and ahmad, 2018;ee, 2016;awokuse and xie, 2015;metha, 2011;awokuse,2007;and awokuse, 2005). other variables include tot and mortality (as proxy of health). there is lot of literature which described that there is relationship between terms of trade and economic growth (jebranet al, 2018; awokuse and xie, 2015; kalumbu and sheefeni, 2014; awokuse, 2005; blattmanet al,2004; and mendoza, 1997). so, based on evidences that terms of trade significantly affect the overall growth, tot was added in the model. moreover, health being a constituent of human capital is important as healthy labor force is expected to be more productive and efficient (cai and kalb (2006) and scultz (2005). due to the complexities in determining the measures for health proxy and data availability, commonly used indicators of health has been mortality rate and life expectancy (poças 2012).mortality rate at birth was added in the model as proxy of health in present analysis. education and health both are main components of human capital. education variable was also included in the initial estimation but this variabale was dropped due to collinearity issue. because of that, only health variable was included the model as proxy of human capital. 4. results and discussion 4.1. unit root test presence of unit root was examined using both adf test and phillipse perron test to validate the findings (table 1 & 2). results revealed that all of the variables were nonstationary at level but they were stationary at first difference so all of them were i(1) and none of them was i(2) which is pre-condition for the application of ardl. sumia bint zaman, muhammad ishaq, muhammad azamniazi 112 table 1:unit root tests using adf test variable levels first differences lny -1.734 -4.712 *** lnagr -0.972 -6.184 *** lngcf -1.808 -4.960** lntot -2.535 -4.470*** lnmort -0.189 -3.68* source: authors calculations; *=significant @ 10% level,**=significant @ 5% level and ***=significant @ 1% level table 2:unit root tests usingphillip-perron test variable levels first differences lny -1.320 -4.139*** lnagr -1.426 -4.576 *** lngcf -1.275 -4.457 *** lntot -1.171 -7.391 *** lnmort -0.464 -4.118 *** source: authors calculations; ***=significant @ 1% level 4.2. ardl bounds test the ardl bounds test was used check the co-integration which test the hypothesis that there is no level relationship which if rejected means that there is co-integration in the selected model. for bounds test, pesaranet al (2001) offered two critical values (upper and lower bounds) used to test the significance where co-integration exists if calculated value is greater than upper bound. however, since critical values offered by pesaranet al (2001) are calculated for large data sets, critical values offered by narayan (2005) which are calculated for small sample sizes (from30–80) were used in the present study. number of lags were selected using schwartz bayesian criterion (sbc) for its usefulness for small sample like in this study. results of ardl bounds test indicted that there is co-integration in selected model which is statistically significant where calculated f-statistic is safely greater than i(1) bound @5% level of significance (table 3). table 3: ardl bounds test for the model(4, 1, 0, 0, 3, 2) test statistics value k f-statistics 6.441** 5 critical value bound by narayan (2005) significance i(0) bound i(1) bound 10% 3.37 4.62 5% 4.04 5.54 1% 5.60 7.17 source: author’s calculations; **=significant @ 5% level journal of applied economics and business studies, volume. 5, issue 2 (2021) 103-120 https://doi.org/10.34260/jaebs.527 113 the estimated model delivered long-run and short-run coefficients along with the coefficient of speed adjustment (table 4). results showed that the independent variables included in the model were significantly affecting per capita real gdp in the long run except tot. coefficient of ect was -0.62 which means that if there is any disequilibrium, it will be corrected at the rate of 62 percent in the first period. results revealed that real agricultural value added has a positive relationship with per capita real gdp in the long run where one percent increase in real agricultural value added increases the real gdp/capita by 0.35%. results were significant and consistent with earlier researches (ansari and khan, 2018; khan et el, 2013; herath et al, 2013; raza et al, 2012; tiffin and irz, 2006 and yao, 2000). this indicated that the promotion of agriculture sector can have far-reaching effects on economic growth of the country. these results advocated for the development of agriculture sector in line with the long-term goals of economic growth and emphasized in investing in agriculture sector. in accordance with the growth theory, gross capital formation confirmed significant positive impact on economic growth of pakistan. analysis depicted that one percentage increase in gcf leads to 0.48% increase in real gdp/capita in the long-run. results were significant and consistent with theory and previous research that capital stimulates economic growth (ahmad and ahmad, 2018; ee, 2016; awokuse and xie, 2015; metha, 2011; and awokuse, 2005). statistically significant negative relationship was found between infant mortality rate and real gdp per capita where 1 % decrease in infant mortality improve the per capita real gdp by 1.5%. reduced mortality rate is considered as sign of good health status of the people which contribute in the economy. results of current study suggested that we should invest in human health to enhance the economic growth as suggested by endogenous growth theory. moreover, it also suggested to create conducive environment and economic opportunities to engage the people to reap the benefits of demographic dividends of decreased mortality. initially when the model was estimated, some consistency issues were found so dummy variable for structural break in 1971 (dsb71) was included to deal with the issue of inconsistency. moreover, variables of the study showed a trend over the time except tot, so years variable was added to capture the trend component and it was significant. sumia bint zaman, muhammad ishaq, muhammad azamniazi 114 table 4: ardl regression results ardl regression (4,1,0,0,3,2) 2 sample: 1964 2018number of observations = 55 r-squared=0.9489 adj r-squared= 0.9274 long-run statistics variable coefficient st. error t-statistics p-value lnagr .3546125 .1464733 2.42 0.020 lngcf .4847286 .1004735 4.82 0.000 lntot .0275572 .053874 0.51 0.612 lnmort -1.495776 .5651091 -2.65 0.012 dsb71 -.1862674 .1033004 -1.80 0.079 constant 60.82231 15.97745 3.81 0.000 short-run statistics lny ld. -.0355493 .0536052 -0.66 0.511 lny l2d -.0581824 .0478414 -1.22 0.231 lny l3d -.0433703 .0447619 -0.97 0.339 lnagr d1. .2943168 .1234393 2.38 0.022 lnmortd1. 3.440538 6.143966 0.56 0.579 lnmortld. 6.737057 5.864771 1.15 0.258 lnmortl2d. -.7564336 6.465748 -0.12 0.907 dsb71 d1 .1033921 .0546046 1.89 0.066 dsb71 ld .0498333 .0507127 0.98 0.332 years -.0324101 .0076329 -4.25 0.000 ect (-1) -0.621209 .2389056 -4.91 0.000 source: author’s calculations 4.3. diagnostic tests of the estimated model there are certain other diagnostic tests which were performed for measuring the heftiness of the estimated regression model. most common categories of diagnostics are coefficients diagnostics, lag structure, and residual diagnostics where category of residual diagnostics is the most critical. common established techniques for residual diagnostics include tests for heteroskedasticity, autocorrelation, model specification test of omitted variable, normality test, and stability test (shrestha and bhatta 2018). the tests performed for the estimated model included breusch-godfrey lm test, breusch-pagan test, autoregressive conditional heteroskedasticity (arch) lm test, ramsey reset test, and jarque-bera test. 2 values in parenthesis indicate the number of lags selected automatically by the software using the appropriate lag selection criteria. journal of applied economics and business studies, volume. 5, issue 2 (2021) 103-120 https://doi.org/10.34260/jaebs.527 115 durbin watson (d.w) test is commonly used for testing the autocorrelation but in the case of autoregressive models, breusch–godfrey lm test is a procedure alternative to d.w for testing higher order autocorrelation. since this test is based on lagrange multiplier procedure, it is also called as lm serial correlation test (dimitrios and hall 2011). breuschpagan test is to examine that whether the variance of the residuals of the estimated regression depend on independent variables or not. similarly, arch test is a statistical procedure developed for time series data where it analyzes the variance of the current error term in relation to the squares of error terms of the preceding time periods. the arch test is suitable when the variance of the error term in time series data consist the autoregressive properties. the time series data showing conditional heteroskedasticity or autocorrelation in the squares of the error terms is considered to have arch effects (engle 1982).ramsey reset test was used to examine the accuracy of the model specification while normality was examined through jarque-bera test.. results of these tests depicted that the estimated model was not suffering from any violation (table 5). table 5: diagnostics tests of the model test test statistics p-value breusch-pagan 0.56 0.4545 autoregressive conditional heteroskedasticity 0.46 0.4976 breusch-godfrey lm 0.200 0.6544 ramsey reset 0.09 0.9671 jarque-bera 2.847 0.2409 source: authors calculations in time series data analysis, the cumulative sum control chart (cusum) squared is a technique developed for sequential analysis. it is used to observe the change in the data and hence to test the stability of the estimated model. it tests the fluctuation in the mean, variance, and distribution function (bai 1994). the following graph shows that the all of the deviations around the regression line are within the range of control lines. https://en.m.wikipedia.org/wiki/time_series https://en.m.wikipedia.org/wiki/variance https://en.m.wikipedia.org/wiki/errors_and_residuals_in_statistics https://en.m.wikipedia.org/wiki/errors_and_residuals_in_statistics https://en.m.wikipedia.org/wiki/autoregressive sumia bint zaman, muhammad ishaq, muhammad azamniazi 116 figure 1: cusum.sq test of the estimated model 5. conclusion there has been controversy in the field of development economics about the significance of the role of agriculture sector in economic growth. going through the data it indicates that agriculture sector is significant contributor to the economy of pakistan as it contributes about 19% in national gdp through different sectors including crops, livestock, fisheries, and forestry. this study was designed to statistically test the role of agriculture sector in economic growth of pakistan through estimation of relationship between growth of agriculture sector and pakistan’s economic growth using recent econometric techniques. study found that real agricultural value added has a significant positive impact on per capita real gdp in the long-run. this indicated that the promotion of agriculture sector leaves far reaching effects with respect to economic growth of the country. these results advocated for the development of agriculture sector in line with the long-term goals of economic growth and emphasized in investing in agriculture sector. results also proved the importance of capital formation both the physical capital and human capital. finding suggested that we should invest in human health to enhance the economic growth as suggested by endogenous growth theory. moreover, it can also be suggested to create conducive environment and economic opportunities to reap the benefits of demographic dividends of decreased mortality in the long-run. as per analysis, structural break negatively impacts the economy which suggest that maintaining stability is critically important for economic growth. moreover, literature hypothesize the positive effect of tot for economic growth, but analysis indicated that tot has not been able to put any significant impact on economic growth. further, trend analysis also pointed out that tot has been fluctuating over the time. it can be inferred from this analysis that there is need to stabilize tot and to restructure the exports of the country so that significant positive impact cab be generated. c u s u m s qu ar ed years cusum squared 1967 2018 0 1 journal of applied economics and business studies, volume. 5, issue 2 (2021) 103-120 https://doi.org/10.34260/jaebs.527 117 references ahmad, d. and j. ahmad(2018). ‘an empirical analysis of agricultural export on economic growth: a case of pakistan’. international journal of advanced and applied sciences, 5, 2, pp. 25-32. akram, n., i. h. padda and m. khan(2008).‘the long term impact of health on economic growth in pakistan’. the pakistan development review, 47, 4,pp. 487-500. alkhathlan, k.a(2013). ‘contribution of oil in economic growth of saudi arabia’. applied economic letters, 20, 4, pp. 343-348. ansari, s. a. and w. khan(2018). ‘relevance of declining agriculture in economic development of south asian countries: an empirical analysis’. agris on-line papers in economics and informatics, 10,2,pp. 3-14. issn. 1804-1930. awokuse, t.o. and r. xie (2015). ‘does agriculture really matter for economic growth in developing countries’? canadian journal of agricultural economics, 63,pp. 77-99. awokuse, t.o(2007). ‘causality between exports, imports, and economic growth: evidence from transition economies’. economics letters, 94, pp. 389–95. awokuse, t.o (2005). ‘export-led growth and the japanese economy: evidence from var and directed acyclic graphs’. applied economics letters,12, pp. 849–58. azra, d.k., ahmed, e. and ahmed, s.s. 2013. agriculture and economic growth: empirical evidence from pakistan. sarhad university international journal of basic and applied sciences, 1 (1): 307-313.bai, j(1994). ‘weak convergence of the sequential empirical processes of residuals in arma models’. the annals of statistics, 22, pp. 2051-2061. blattman, c., j. hwang, and j. g. williamson (2003). ‘the terms of trade and economic growth in the peripher’.national bureau of economic research, pp. 1870-1938 (no. w9940). bhagwati, j. and t.n. srinivasan (1975). ‘foreign trade regimes and economic development: india’.columbia university press, new york. cai, l. and g. kalb (2006).‘health status and labour force participation: evidence from australia’. health economics, 15, 3, pp. 241-261. chani, m.i., z. pervaiz, s.a. jan, a. ali, and a.r. chaudhary(2011). ‘poverty, inflation and economic growth: empirical evidence from pakistan’. world applied sciences journal, 14, 7, pp. 1058-1063. chaudhry, m.g. and g. m. chaudhry. 1997. pakistan’s agricultural development since independence; intertemporal trends and explanations. the pakistan development review, 36 (4): 593-612. datt, g. and m. ravallion (1998). ‘farm productivity and rural poverty in india’. journal of developmental studies,34, 4, pp. 62–85. dowrick, s. and n. gemmell(1991). ‘industrial catching up and economic growth: a comparative study across the world’s capitalist economics’. economic journal,101, pp. 263–76. echevarria, c(1997). ‘changes in sectoral composition associated with economic growth’. international economic review,38, 2, pp. 431–52. ee, c. y (2016). ‘export-led growth hypothesis: empirical evidence from selected sub-saharan african countries’. procedia economics and finance, 35, pp. 232-240. eicher, c.k. and j.m. staatz(1984). ‘agricultural development in the third world’. johns hopkins university press, baltimore. sumia bint zaman, muhammad ishaq, muhammad azamniazi 118 faridi, m.z(2012). ‘contribution of agricultural exports in economic growth in pakistan’. pakistan journal of commerce and social sciences, 6, 1, pp. 133-146. farooq, u. 2015. demand-supply projections of food and fiber in pakistan: potentials and constraints. background draft chapter-2 for “vision 2050: addressing challenges to the agriculture sector of pakistan”. fei, j. and g. ranis (1961). ‘a theory of economic development’. american economic review,51, 4, pp. 533–65. gardner, b(2005). ‘causes of rural economic development’. in: d. colman and n. vink (eds.). reshaping agriculture’s contribution to society, pp. 21–41. proceedings of the 25th international conference of agricultural economists. durban, oxford: blackwell. gollin, d., s.l. parente, and r. rogerson(2002). ‘the role of agriculture in development’. american economic review, 92, 2, pp. 160–64. gop. (2012). economic survey of pakistan. ministry of finance, government of pakistan. gop. (2014). economic survey of pakistan. ministry of finance, government of pakistan. gop. (2015). economic survey of pakistan. ministry of finance, government of pakistan. gop. (2016). economic survey of pakistan. ministry of finance, government of pakistan. gop. (2018). economic survey of pakistan. ministry of finance, government of pakistan. gop(2018). ‘agricultural statistics of pakistan’. ministry of national food security and research, government of pakistan. gop. (2020). economic survey of pakistan. ministry of finance, government of pakistan. herath, j., s. bashir, and t. gebremedhin(2013). ‘an empirical analysis of agriculture ineconomic growth of north carolina’. paper presented at annual meeting of southern agricultural economics association, orlando, florida, 3-5 february 2013. humphries, h., and s. knowles(1998). ‘does agriculture contribute to economic growth? some empirical evidence’. applied economics, 30,6, pp. 775-781. hye, q. m. a (2009). ‘agriculture on the road to industrialization and sustainable economic growth: an empirical investigation for pakistan’. international journal of agricultural economics & rural development, 2, 2, pp. ijirshar, v.u(2015). ‘the empirical analysis of agricultural exports and economic growth in nigeria’. journal of development and agricultural economics, 7, 3, pp. 11-122. johansen, s. and k. juselius (1990). ‘maximum likelihood estimation and inference on cointegration with applications for the demand for money’. oxford bulletin of economics and statistics,52, pp. 169–210. johnston, b. and j. mellor(1961). ‘the role of agriculture in economic development’. american economic review, 51, 4, pp. 566–93. jorgenson, d(1961). ‘the development of a dual economy’. economic journal, 282,pp. 309–34. kalumbu, s. a. and j. p. sheefeni (2014). ‘terms of trade and economic growth in namibia’. international review of research in emerging markets and the global economy, 1, 3, pp. 90-101. kogel, t. and a. prskawetz(2001). ‘agricultural productivity growth and escape from the malthusian trap’. journal of economic growth,6, pp. 337–57. lewis, w.a(1954). ‘economic development with unlimited supplies of labour’. the manchester school, 22, 1, pp. 139–91. mankiw, n. g., d. romer, and d. n. weil (1992). ‘a contribution to the empirics of economic journal of applied economics and business studies, volume. 5, issue 2 (2021) 103-120 https://doi.org/10.34260/jaebs.527 119 growth’. quarterly journal of economics, 107, 2, pp. 407-437. matsuyama, k(1992). ‘agricultural productivity, comparative advantage, and economic growth’. journal of economic theory, 58, pp. 317-334. mendoza, e. g(1997). ‘terms-of-trade uncertainty and economic growth’. journal of development economics, 54, 2, pp. 323–356. metha, r(2011). ‘short-run and long-run relationship between capital formation and economic growth in india’. indian journal of management and technology, 19, 2, pp. 170-180. narayan, p. k(2005). ‘the saving and investment nexus for china: evidence from cointegration tests’. applied economics, 37, 17, pp. 1979-1990. nkoro, e. and a.k. uko (2016). ‘autoregressive distributive lag (ardl) cointegration techique: application and interpretation’. journal of statistical and econometric methods, 5, 4, pp. 63-91. okonkwo, i.c(1989). ‘the erosion of agricultural exports in an oil economy: the case of nigeria’. journal of agricultural economics,40, 3, pp. 375–84. olsson, o. and d.a. hibbs(2005). ‘biogeography and long-run economic development’. european economic review,49, 4,pp. 909–38. pahlavani, m., e. wilson, and a.c. owrthington(2005). ‘structural breaks and cointegration relations in iranian exports, imports and economic growth: an application incorporating autorgressive distributive lag (ardl) procedure’. americal journal of apllied science, 2, 7, pp. 1158-1165. pesaran, m.h., y.shin and r.j.smith(2001). ‘bounds testing approaches to the analysis of level relationships’. journal of applied econometrics, 16, 3,pp. 289–326. pocas, a. i. s. a(2012). ‘the interrelation between health, human capital and economic growth: empirical evidence from the oecd countries and portugal’. faculade de economia. universidade de coimbra ((doctoral dissertation, feuc). raza, s.a, y. ali, and f, mehboob (2012). ‘role of agriculture in economic growth of pakistan’. international research journal of finance and economics, 83,pp. 180-186. issn 14502887. rehman, a., l. jingdong, and m. i. hussain(2016). ‘pakistan’s major exports and imports role and importance for agriculture and economic development’. andamios: revista de investigacion social, 12, 1, pp. 174-179. ruttan, v(2000). ‘technology, growth and development: an induced innovation perspective’. oxford university press. schiff, m. and a. valdez(2002). agriculture and the macro economy, with emphasis on developing countries’. in:b.l. gardner and g.c. rausser (eds.) ‘handbook of agricultural economics’, 2, pp.1421–54. elsevier: amsterdam. shultz, t.w(2005). ‘productive effects of health’.in g. l. casasnovas, b. rivera, andl. currais (eds.), health and economic growth: findings and policy implications,pp. 287-311. the mit press: cambridge. schultz, t.w(1964). ‘transforming traditional agriculture’. yale university press, new haven. shrestha, m. b. and g. r. bhatta (2018). ‘selecting appropriate methodological framework for time series data analysis’. the journal of finance and data scient, 4, pp. 71-89. solow, r.m(1956). ‘a contribution to the theory of econmoic growth’. quarterly journal of economics, 70, 1,pp. 65-94. sumia bint zaman, muhammad ishaq, muhammad azamniazi 120 swan, t.w(1956). ‘economic growth and capital accumulation’. economic record, 32, 2, pp. 334361. thirtle, c., l.lin and j. piesse(2003). ‘the impact of research-led agricultural productivity growth on poverty reduction in africa, asia and latin america’. world development,31, 2, pp. 1959–75. tiffin, r., and x. irz (2006). ‘is agriculture the engine of growth? agricultural economics,35, pp. 79–89. timmer, c.p (1995). ‘getting agriculture moving: do markets provide the right signals’? food policy,20, 5, pp. 455–72. tsakok, i., and b. garner (2007). ‘agriculture in economic development: primary engine of growth or chicken and egg’? american journal of agricultural economics, 89, 5, pp. 1145-1151. udoh, e., u. afangideh, and e.a. udeaja(2015). fiscal decentralization, economic growth and human resouce development in nigeria: autoregressive distributive lag (ardl) approach’. cbn journal of applied statistics, 6, 1.a, pp. 69-93. yao, s(2000). ‘how imprtant is agriculture in china’s economic growth’? oxfrod department studies, 28, 1, pp. 33-49. journal of applied economics and business studies, volume. 5, issue 2 (2021) 53-68 https://doi.org/10.34260/jaebs.524 53 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x trade liberalization and labor demand elasticities: a firm-level analysis of pakistan’s manufacturing sector muhammad ramzan sheikh1*, misbah rauf2, irfan hussain3 and asad abbas4 1associate professor, school of economics, bahauddin zakariya university multan 2mphil student, school of economics, bahauddin zakariya university multan 3ph.d. student, department of economics, g.c university faisalabad 4lecturer in economics, comsats university islamabad, vehari campus abstract the study investigates the linkage of trade liberalization and labor demand elasticities in pakistan. the panel data are used by selecting 13 industries in pakistan's manufacturing sector for the years 1995-1996, 2000-2001, and 2005-2006. the pooled ols technique is applied to get the estimates at an aggregated level and disaggregated levels. overall findings support the positive relationship between trade liberalization and labor demand elasticity in production workers but in the case of non-production workers, the findings show the weak relationship between trade liberalization and labor demand elasticity. the study is also furnished with some policy recommendations. keywords trade liberalization, labor demand, production, and nonproduction workers jel classification f16, j23 1. introduction it is essential to step into international trade and earn the surplus to develop the country. to enhance international trade, a country has to devise such trade policies which may facilitate the exporters and importers. it has been observed that developed countries formulate and apply open trade policies to earn substantial profit and sustain economic growth. without trade liberalization, no country can improve the living standard of its people and if countries impose the tariff and other restrictions on international trade, they cannot stand in the queue of developed countries (edward, 1993, akhter and ali, 2007). * ramzansheikh@bzu.edu.pk muhammad ramzan sheikh, misbah rauf, irfan hussain and asad abbas 54 trade liberalization is considered an issue of less developed countries because they have to face cut-throat competition from developed countries. this competition creates failure in newly developed industries or infant industries of developing countries as they have to rely on developed countries for high standard materials and technology. on the other hand, when there is free trade environment, developing countries may lose competitiveness along with the decrease in government revenues in the form of tariffs and other duties. so, developing nations do not take the risk to liberalize their trade for protecting their domestic markets (akhter and ali, 2007). trade liberalization performs a significant role in the utilization of labor and creating the demand for labor in the south countries. as these countries are labor abundant countries, so through trade liberalization, these countries can take benefits from capital-abundant countries. according to the h.o theory, when labor-abundant countries enter the foreign trade market, they gain through labor-intensive commodities and labor demand also increases due to which wages and employment levels also go up (yasmin, 2011). the linkage of the trade liberalization and labor market has received more attention in developing countries. rodrik (1997) was the first who introduced the benefit of labor demand by increasing trade liberalization. owing to this, the employment level and wages also increase, and consequently the output or gdp rises of a nation (yasmin, 2011). rodrik also emphasized that labor demand elasticity goes up by opening trade and it also increases the output level and made the domestic market more competitive. through trade between the south and the north countries, laborers can become a substitute for capital products. the demand elasticity of the product market can be increased by free trade. this phenomenon shows a positive impact on labor demand elasticity due to increasing trade liberalization. bargaining power over rent gets higher in the firms and they can bargain for capital from labor by enjoying extra profit (akhter and ali, 2007). pakistan has adopted different trade policies to increase the size of the trade-in gdp. the government of pakistan has introduced the export bonus scheme in 1959 in which a voucher is given to the exporter that allows them to import raw material and machinery that enhance the percentage of exports but this policy was abolished in 1972. in 1962, exports credit guarantee scheme was introduced and it remained till 1980. this scheme was held by pakistan insurance corporation to ensure exporters' payments from abroad and commercial banks to repay loans that exporters borrow. maximum trade was observed from the period of 1960-1969 and the minimum trade percentage of gdp was in 1969. in 1978, exporters finance scheme was launched in which loans were given to exporters with zero interest rates. in the 1980s, pakistan introduced policies to liberalize the economy. in 1988, pakistan established the structural adjustment program (sap) with the help of imf in which subsidies and tariff rates are reduced to fall the protection level and abolished in 1994. strategic trade policy framework (stpf) was used from 2012-2015 that raised the share of trade in gdp in journal of applied economics and business studies, volume. 5, issue 2 (2021) 53-68 https://doi.org/10.34260/jaebs.524 55 2013. trade liberalization shifts the concept of capital-intensive production to laborintensive production in developing countries because exporters in ldcs depend on laborers. due to the exports of labor-intensive commodities, the wage rate has also increased which makes the labor market more competitive. the study provides the impact of trade liberalization on labor demand elasticity in pakistan. many researchers have investigated the impact of trade liberalization on different variables like growth, output and wage rate. still, only one study has estimated the linkage of trade liberalization and labor demand elasticity in pakistan except for akhter and ali (2007) and (yasmin, 2011) in which they examined the relationship of trade liberalization and labor demand elasticity by taking a few industries. but we have taken thirteen industries of the manufacturing sector of pakistan. moreover, the results of these studies are inconclusive as it is unable to find the exact impact of trade liberalization. we have taken the data of the latest years from cmi. our study has used the trade liberalization formula (x+m/gdp) but akhter and ali (2007) used a dummy variable for trade liberalization. furthermore, after 2011, no study has been conducted on the same topic. the paper is structured as: section 2 shows the literature review on the studies of trade liberalization and labor demand elasticity. section 3 specifies the model, data and methodology. section 4 exhibits the results of labor demand elasticities in aggregated and disaggregated forms. section 5 is about conclusions and policy recommendations. 2. review of literature this section presents the empirical findings of those studies that have explored the relationship between trade liberalization and labor demand. maloney and fajnzylber (1999) explained the effect of trade liberalization on labor demand elasticity in three countries. trade liberalization and own-wage elasticity revealed a negative relationship in the import competitive market but it showed a positive and significant effect in the imperfectly competitive market, export industries and non-tradable markets. the author also found the mixed effects of trade liberalization in the skilled and unskilled labor markets. akhter and ali (2000) investigated the relationship between trade liberalization and labor demand elasticity by using firm-level data in pakistan. the results were insignificant as labor demand behaved differently over time. jean (2000) focused on the effect of trade liberalization on the price elasticity of labor demand under perfect competition model. the author pointed out how trade liberalization affects the employment cost by changing the real wage of unqualified workers in france. the study found that there was a positive relation between unskilled labor and trade liberalization. krishna et al. (2000) demonstrated that trade liberalization enhanced the labor demand elasticity in turkey. slaughter (2001) examined the relationship between own-price labor demand elasticity and trade liberalization in the us. the demand for production workers increased in the manufacturing industry and muhammad ramzan sheikh, misbah rauf, irfan hussain and asad abbas 56 five out of eight industries showed a positive relationship between international trade and labor demand but the demand for non-production workers displayed a negative relationship in the manufacturing sector. driffield and kambhampati (2003) investigated the impact of trade liberalization and domestic competition on firm-level efficiency in india. five out of six sectors exhibited a positive effect of trade liberalization on firm efficiency level. middle-aged firms revealed more significant results instead of high-efficiency markets. hasan et al. (2003) pointed out the relationship between trade liberalization and labor demand elasticity in the indian manufacturing sector. the results indicated that when protection decreased, labor demand elasticity increased. bruno et.al (2004) explained the impact of trade liberalization on labor demand elasticity in european countries, the us, and japan. trade liberalization increased the labor demand elasticity in the uk. the positive effects were found in italy and less significant effects appeared in spain and france. wacziarg and wallack (2004) explored the effect of trade liberalization on the shifting of labor across the sectors. trade liberalization showed a positive impact on structural changes as a whole but it exhibited opposite effects by disaggregating the structural changes. michaels (2006) highlighted the effect of trade liberalization on the demand for skilled workers in manufacturing industries by using the heckscher-ohlin model. the result showed a positive impact on skilled workers. the author proved that where the skill of abundant factors was more, the demand showed a positive result comparatively where the skill was short or imperfect. yasmin (2011) presented the nexus between trade liberalization and labor demand elasticity in pakistan. the results pointed out insignificant effects of trade liberalization on labor demand elasticity. after using the time trend, the most significant results appeared. trade openness enhanced the productivity of labor in long run and consequently, skilled workers' demand increased. for advanced technology, skilled labors performed well but in pakistan, minor positive effects appeared regarding skilled labors. paz (2012) analyzed the relationship of trade liberalization and labor demand elasticity in brazil. the results predicted that when there was a tariff cut on domestic imports, it influenced the informal workers positively and negatively depending on labor market. mahomedy (2013) examined the relationship between international trade and labor demand elasticity in south africa and found it positive and highly significant. liyanaarachchi et al. (2016) investigated the nexus between trade liberalization, income distribution, and poverty in sri lanka. trade liberalization showed positive effects on macroeconomic outcomes in long run rather than in the short run. njikam (2016) examined the impact of trade liberalization on unskilled, skilled, and total labor demand in cameroon. the relationship between trade liberalization and unskilled labor was positive. the author pointed out that trade liberalization raised international relations and built new job opportunities for unskilled workers. erten et al. (2019) examined the effect of tariff reduction in south africa. tariff reduction affected the formal and informal employment sector and show a reduction in the same magnitude. dobbelaere and wiersma (2020) exhibited the effect journal of applied economics and business studies, volume. 5, issue 2 (2021) 53-68 https://doi.org/10.34260/jaebs.524 57 of trade liberalization on the firm’s product and labor market. the study pointed out that there was a positive relationship on the firm’s product market (power price-cost markup) due to a reduction in tariff in intermediate inputs but there is a negative relationship in the firm’s labor market power (wage markdowns). the above-mentioned studies demonstrate the relationship of trade liberalization with labor demand elasticity. these studies point out the different results in developed and developing countries. most developed countries represent positive results while developing countries exhibit mix or negative results. 3. model, data, and methodology 3.1 model specification the main objective of the firm is profit maximization. we have to determine the output function of the firm by assuming that the firm is in a monopolistic competition environment1. the inverse demand function of the representative firm is given: 1 * b mn n mn p p o − = (1) where: subscripts m and n represent the firm m in industry n pmn =own price p*n = average price in the industry η = scaling factor b= constant price elasticity of demand qmn= m th firm output in industry n. using the cobb douglas production function (cdpf): 0 1 z a mn omn o o z w = = (2) where womn denotes o th vector of inputs. we consider three inputs: capital (c), labor(l) and material (r) in cdpf: o c l r    = (3) firstly, by taking the partial derivative with respect to labor and equating it to zero, we get the following first-order condition: 1 1 * 11 b l mp w p c l r c r l b           − −  −         = = −        (4) 1 see chaudry, 1999 and akhter and ali, 2007 and yasmin,2011 for more details. muhammad ramzan sheikh, misbah rauf, irfan hussain and asad abbas 58 1 1 * 11 b l mp w p c l r c l r l b           − −  −         = = −        (5) 1 1 * 11 1 b l mp w p c l r l b        −  −        = = −       (6) correspondingly, we can write the first order conditions for capital and raw material: 1 1 * 11 1 ( ) [ ]b c mp i p c l r c b      − −  = = −    (7) 1 1 * 11 (1 )( ) [ ]b r mp r p c l r r b      − − = = − (8) we can write equation (6) in natural log form: 1 1 1 ln 1 1 1 1 ln ln ln ln 1 1 1 1 1 1 1 1 1 1 1 1 wb b b l c r p b b b b                    − − −                      = − + +                    − − − − − − − −                          (9) equation (9) can also be written as: 0 1 2 3 ln ln ln ln w l c r p       = + + +    (10) where: ( ) s f b = by plugging in first-order condition of capital and material in equation (10), we get: 0 1 2 3* * * ln ln ln ln w i r l p p p           = + + +            (11) the own price elasticity of demand for labor w.r.t industry wage 1 * 1 1 (1 )( ) ln 1 ln( ) 1 (1 )( ) l b w p b         − − +   = =    − − + +   (12) the partial derivative of the absolute value of own-price elasticity of labor demand (ɸ1) w.r.t demand elasticity of the product. 1 2 2 ( ) 0 1 1 1 ( ) b b b            =        − − + +        (13) journal of applied economics and business studies, volume. 5, issue 2 (2021) 53-68 https://doi.org/10.34260/jaebs.524 59 the labor demand function from equation (11) depends on input prices and output therefore it becomes: * * * * 0 1 2 3 4 ln( ) ln( ) ln( ) mnt mnt mnt mnt mnt l w i r o    = + + + + (14) where: * * * * * * * * , , , mnt mnt mnt mnt w i r o w i r o p p p p = = = = so, * * * * 0 1 2 3 4 5 ln( ) ln( ) ln( ) mnt mnt mnt mnt mnt mnt l w i i q trade      = + + + + + + (15) thus, our final estimating equation is * * * * 0 1 2 3 4 5 ln ln( ) ln( ) ln( ) ( ) ijt ijt ijt ijt ijt ijt l w r m q trade      = + + + + + + (16) where l is total employees, w is the price of labor, r is the price of capital, m is the price of raw material, q is total output and trade is trade liberalization. 3.2 data and methodology we have used the data of 13 industries of pakistan’s manufacturing sector in different years e.g. 1995-1996, 2000-2001, and 2005-20062. the data are taken from the census of manufacturing industries (cmi) which is a reliable source of data in pakistan. the main objective of cmi is the measurement of structural and production changes in the large-scale manufacturing industries over time. cmi collects annual-based survey data in which information of employment cost, values of inputs and outputs, contribution to gdp, stocks, fixed assets, value-added, industrial taxes, and material cost are included. cmi publishes the information only to those firms that are registered and not all the existing firms. gdp and trade liberalization data are taken from world development indicators. pooled ols technique is used in the study to estimate the results. 4. results and discussions we have done two types of analyses: aggregated analysis and disaggregated analysis. in aggregate analysis, we are taking all workers as a whole of 13 industries that show the combined results but in disaggregated analysis, we divide the workers into production workers and non-production workers. so, first, we explain the results of aggregated analysis followed by disaggregated analysis. 4.1 labor demand elasticities: an aggregated analysis now we are discussing the results of the aggregated analysis in which we take the log of total employees that are equal to production workers and non-production workers. the results of own and cross-price labor demand elasticities are given in table 1. in this table, the dependent variable is the log of total employees and independent 2 cmi provides the data for selected years so we have taken the latest data available at cmi. muhammad ramzan sheikh, misbah rauf, irfan hussain and asad abbas 60 variables are price of labor, price of capital, price of other materials, output, and trade liberalization. now we discuss the results according to one-by-one industry. textile industry textile industry exhibits the negative own-price elasticity of labor demand indicating that the price of labor or wages is negatively related to labor demand according to the classical theory of employment or output (dutta, 2014). the coefficient of interest rate shows cross-price elasticity of labor demand in terms of interest rate or price of capital is negative and significant. according to labor demand theory, when the price of capital increases, the demand for capital decreases but the demand for labor increases due to the substitution effect. so, the relationship between labor demand and prices of capital is positive. in the textile industry, the relationship between labor demand and interest rate is negative due to the complementary effect because in the textile industry, capital is more important and there is no or less substitute for capital. the parameter of cross-price elasticity of labor demand in terms of raw material is negative and significant. it means that when prices of raw material increase, the cost of production also increases so the output decrease, and the demand for labor also decreases. the coefficient of output elasticity of labor demand in the textile industry is positive and significant. when the output increases, the industry hires more laborers to increase their production, so the labor demand increases with the increase in output. trade liberalization elasticity of labor demand has appeared with a positive sign suggesting that when trade liberalization increases, industries hire more labor to increase their production. the value of r2 in the textile industry is 0.44. apparel and wearing industry the parameter of own-price elasticity of labor demand in apparel and wearing industry is negative and significant. the cross-price elasticity of labor demand corresponding to the interest rate is significant and negative due to the complementary effect. the cross-price elasticity of labor demand with respect to raw material is positive and insignificant. the reason is that when the prices of raw material increase, the production cost also increases but the demand for labor also increases due to the complementary effect in the production process. output and trade liberalization elasticities of labor demand have a significant and positive bearing on labor demand. the r2 value in this industry is 0.52. ginning industry in ginning industry, own-price elasticity of labor demand is negative and insignificant. cross-price elasticities of labor demand with respect to interest rate and raw material are negative and insignificant. output and trade liberalization elasticities of labor demand have positive association with labor demand in ginning industry and both parameters are significant. the value of r2 is 0.70 in ginning industry. journal of applied economics and business studies, volume. 5, issue 2 (2021) 53-68 https://doi.org/10.34260/jaebs.524 61 leather industry in the leather industry, own-price elasticity of labor demand shows a negative sign but it is insignificant. the relationship between labor demand and the interest rate is negative due to the complementary effect. the raw material, output, and trade liberalization elasticities of labor demand have a positive and significant effect on labor demand. the value of r2 in this industry is 0.42. food industry the parameter of own-price elasticity of labor demand in the food industry is negative and insignificant. the parameters of both cross-price elasticities of labor demand are positive and significant. the cost of capital is positive due to the substitution effect. output and trade liberalization elasticities of labor demand are positive in the food industry. the value of r2 is 0.54. electrical industry in the electrical industry, the parameters of own-price elasticity and cross-price elasticities of labor demand are negative and significant. the sign of output elasticity of labor demand is positive and insignificant. there is a positive link between trade liberalization and labor demand but it is insignificant. the r2 of the electrical industry is 0.55. iron & steel industry in the iron and steel industry, own-price elasticity and cross-price elasticities of labor demand have a negative association with labor demand. the link between output and labor demand is positive and significant. the coefficient of trade liberalization elasticity of labor demand is positive but insignificant. the value of r2 in this industry is 0.60. transport industry the parameter of own-price elasticity of labor demand is negative and insignificant in the transport industry. cross-price elasticities of labor demand with respect to interest rate and raw material are negative and significant but the sign of raw material elasticity of labor demand is negative and insignificant. the output elasticity parameter of labor demand is positive but insignificant. the relationship between trade liberalization and labor demand is positive and significant. the value of r2 in the transport industry is 0.82. paper and paper products industry in the paper and paper products industry, own-price elasticity of labor demand and both cross-price elasticities of labor demand are negative and significant. the coefficient of output elasticity of labor demand is positive and significant but trade liberalization elasticity of labor demand has a positive but insignificant effect in the paper and paper products industry. the value of r2 is 0.32. muhammad ramzan sheikh, misbah rauf, irfan hussain and asad abbas 62 tobacco industry the tobacco industry exhibits the negative own-price elasticity of labor demand. the coefficients of cross-price elasticity of labor demand with respect to interest rate and raw material are negative and highly significant. output and trade liberalization elasticities of labor demand show a positive link with labor demand. the r2 value in tobacco is 0.51. table 1: estimates of own and cross price labor demand elasticities dependent variable: ln(lijt) industries intercept ln (wijt*) ln (rijt*) ln (mijt*) ln (qjt*) ln (tradet) r2 textile 0.4332 -0.4737 -0.4153 -0.6673 0.3863 0.8472 0.44 (0.025) (0.045) (0.055) (0.058) (0.093) (0.041) apparel & wearing 0.6828 -0.0800 -0.4611 0.9011 0.8093 0.6625 0.52 (0.010) (0.015) (0.088) (0.405) (0.039) (0.074) ginning 0.8027 -0.3937 -0.1544 -0.0061 0.2388 0.9286 0.70 (0.070) (0.060) (0.201) (0.182) (0.079) (0.042) leather 0.7177 -0.6900 -0.0820 0.0024 0.2926 0.8187 0.42 (0.017) (0.128) (0.053) (0.076) (0.099) (0.092) food 0.7976 -0.6820 -0.4889 -0.6424 0.3032 0.5871 0.54 (0.987) (0.235) (0.074) (0.088) (0.249) (0.075) electrical 0.2132 -0.1196 -0.4409 -0.8208 0.8049 0.1092 0.55 (0.031) (0.056) (0.014) (0.075) (0.881) (0.104) iron & steel 0.8404 -0.9843 -0.1182 -0.7014 0.5339 0.0454 0.60 (0.300) (0.065) (0.042) (0.088) (0.009) (0.813) transport 0.3819 -0.2592 -0.2987 -0.7271 0.3044 0.0225 0.82 (0.098) (0.209) (0.050) (0.170) (0.106) (0.078) paper & paper products 0.6163 -0.9479 -0.7709 -0.1170 0.0521 0.1748 0.32 (0.590) (0.022) (0.069) (0.086) (0.036) (0.540) tobacco 0.5224 -0.9431 -0.5802 -0.3189 0.5010 0.7608 0.51 (0.012) (0.068) (0.000) (0.204) (0.000) (0.651) drugs & pharmaceutical 0.3638 -0.5564 -0.1028 -0.9587 0.2159 0.2461 0.65 (0.107) (0.054) (0.571) (0.077) (0.102) (0.105) chemical 0.4522 -0.7614 0.9879 -0.9002 0.1256 0.1947 0.70 (0.018) (0.539) (0.053) (0.083) (0.051) (0.032) beverages -0.2545 -0.3584 0.0711 -0.6009 0.1179 0.5408 0.67 (0.043) (0.037) (0.065) (0.092) (0.032) (0.069) drugs and pharmaceutical industry in the pharmaceutical industry, own price elasticity of labor demand and both cross price elasticities of labor demand are negative. output and trade liberalization elasticities of labor demand are positive but both are insignificant. the value of r2 is 0.65 in this industry. chemical industry the coefficient of own-price elasticity of labor demand and cross-price elasticity of labor demand with respect to raw material is negative while interest rate elasticity of labor demand is positive. the parameter of output and trade liberalization elasticities of labor demand is positive and significant. the r2 value of the chemical industry is 0.70. journal of applied economics and business studies, volume. 5, issue 2 (2021) 53-68 https://doi.org/10.34260/jaebs.524 63 beverages industry beverages is the last industry shown in table 1. own-price elasticity of labor demand and cross-price elasticity of labor demand with respect to raw material has a negative sign. interest rate elasticity of labor demand has a positive sign. the output and trade liberalization elasticities of labor demand show positive and significant influence. the value of r2 in the beverages industry is 0.67. in table 1, the minimum value of wages elasticity labor demand in thirteen industries is -0.98 (iron & steel industry) and the maximum value of wages elasticity labor demand is -0.08 (apparel & wearing industry). in the case of cross-price elasticity i.e., interest elasticity labor demand, the minimum value is -0.77 (paper & paper product industry) and the maximum value is 0.98 (chemical industry). another cross-price elasticity i.e., raw material elasticity labor demand, the minimum value of cross-price elasticity is -0.95 (drugs & pharmaceutical industry) and the maximum value is 0.90 (apparel and wearing industry). the minimum value of output elasticity labor demand in thirteen industries is 0.05 (paper & papers product industry) and the maximum value is 0.80 (apparel and wearing industry). the minimum value of trade liberalization elasticity labor demand is 0.02 (transport industry) and the maximum value of trade liberalization elasticity labor demand is 0.92 (ginning industry). the minimum value of r2 is 0.32 (paper & papers product industry) and the maximum value of r2 is 0.82 (transport industry). 4.2 labor demand elasticities: a disaggregated analysis now, we are turning to the disaggregated analysis in which workers are fragmented into production workers and non-production workers. in tables 2 and 3, estimates of production workers and non-production workers demand elasticities with respect to wages, interest rate, raw material, output, and trade liberalization in thirteen industries. the dependent variables in tables 2 and 3 are the log of production workers and log of non-production workers respectively and independent variables are the price of labor, price of capital, price of other materials, output, and trade liberalization. in table 2, the parameters of wage elasticity of production workers' demand and interest elasticity of production workers' demand are negative in all industries. raw material elasticity of production workers' demand shows mixed findings as the negative sign has appeared in nine industries while there is a positive sign in four industries. output and trade liberalization elasticities of production workers' demand are positive in all industries. muhammad ramzan sheikh, misbah rauf, irfan hussain and asad abbas 64 table 2: estimates of own and cross price demand elasticities of production workers dependent variable: ln (lijt) production workers industries intercept ln (wijt*) ln (rijt*) ln (mijt*) ln (qjt*) ln (tradet) r2 textile 0.7946 -0.0415 -0.9356 0.0627 0.2815 0.2948 0.52 (0.092) (0.095) (0.066) (0.161) (0.071) (0.057) apparel & wearing 0.5783 -0.4644 -0.1693 -0.5436 0.1362 0.7092 0.74 (0.051) (0.098) (0.010) (0.011) (0.058) (0.67) ginning 0.0023 -0.2508 -0.4648 -0.5210 0.0148 0.2264 0.68 (0.091) (0.054) (0.068) (0.081) (0.009) (0.073) leather 0.5925 -0.4563 -0.5638 -0.2196 0.5130 0.3108 0.77 (0.033) (0.092) (0.093) (0.002) (0.035) (0.023) food 0.1982 -0.3353 -0.2602 0.9259 0.1502 0.4834 0.85 (0.082) (0.084) (0.033) (0.009) (0.055) (0.070) electrical 0.9241 -0.9372 -0.9637 -0.4323 0.8842 0.2915 0.44 (0.009) (0.066) 0.058) (0.084) (0.072) (0.085) iron & steal 0.2878 -0.1310 -0.2945 0.4011 0.8779 0.3885 0.51 (0.091) (0.022) (0.033) (0.073) (0.074) (0.042) transport 0.8709 -0.4572 -0.7101 -0.4646 0.3175 0.0970 0.34 (0.053) (0.000) (0.051) (0.055) (0.075) (0.057) paper & paper products 0.0948 -0.6548 -0.7158 -0.5247 0.4991 0.6731 0.41 (0.027) (0.059) (0.055) (0.085) (0.030) (0.022) tobacco 0.1845 -0.0708 -0.1766 -0.8823 0.6146 0.5382 0.65 (0.029) (0.097) (0.018) (0.002) (0.092) (0.009) drugs & pharmaceutical 0.5455 -0.5595 -0.8367 0.7893 0.2978 0.6267 0.67 (0.018) (0.410) (0.200) (0.097) (0.010) (0.000) chemical 0.3408 -0.7434 -0.8333 -0.8936 0.5738 0.6729 0.23 (0.077) (0.076) (0.052) (0.094) (0.201) (0.073) beverages 0.1873 -0.2820 -0.6660 -0.4497 0.9959 0.7211 0.76 (0.065) (0.085) (0.051) (0.071) (0.805) (0.077) in table 3, wage elasticity of non-production workers' demand has mixed findings as the sign of own-price elasticity is negative in ten industries while in three industries, the sign is positive. both the cross-price elasticities of non-production workers' demand are negative in all industries. output and trade liberalization elasticities of non-production workers' demand are positive in all industries. in tables 2 and 3, the minimum value of wage elasticity is -0.93 and the maximum value is -0.04 in production workers but in non-production workers, the minimum value of wage elasticity is -0.78 and the maximum value is 0.62. in interest elasticity, the minimum value in production workers is -0.96 and the maximum value in production workers is -0.16 but the minimum value of interest elasticity in nonproduction workers is -0.86 and the maximum value is -0.02. the minimum value of the price of raw material elasticity in production workers is -0.89 and the maximum journal of applied economics and business studies, volume. 5, issue 2 (2021) 53-68 https://doi.org/10.34260/jaebs.524 65 value of elasticity is 0.92. in non-production workers, -0.90 is the minimum value and the maximum value -0.02. table 3: estimates of own and cross price demand elasticities of non-production workers dependent variable: ln (lijt) non-production workers industries intercept ln (wijt*) ln (rijt*) ln (mijt*) ln (qjt*) ln (tradet) r2 textile 0.6071 -0.7045 -0.0650 -0.0661 0.6403 0.1761 0.34 (0.074) (0.085) (0.057) (0.036) (0.084) (0.021) apparel & wearing 0.1241 -0.2819 -0.1766 -0.0206 0.7489 0.4887 0.45 (0.076) (0.088) (0.006) (0.080) (0.091) (0.098) ginning 0.5045 -0.4131 -0.1212 -0.3134 0.5489 0.2774 0.56 (0.083) (0.083) (0.014) (0.065) (0.045) (0.956) leather 0.4652 -0.2757 -0.4106 -0.5775 0.0482 0.5523 0.72 (0.072) (0.072) (0.009) (0.006) (0.075) (0.012) food 0.0108 0.5584 -0.3219 -0.8819 0.7314 0.0425 0.67 (0.097) (0.028) (0.078) (0.055) (0.004) (0.809) electrical 0.2102 -0.6759 -0.3013 -0.1108 0.7463 0.6660 0.56 (0.002) (0.003) (0.019) (0.033) (0.015) (0.035) iron & steel 0.4563 0.6001 -0.6366 -0.5273 0.0080 0.8843 0.78 (0.004) (0.023) (0.020) (0.318) (0.034) (0.914) transport 0.9577 -0.4196 -0.6723 -0.8915 0.5613 0.9195 0.42 (0.098) (0.052) (0.085) (0.000) (0.903) (0.358) paper & paper products 0.2736 -0.0727 -0.7662 -0.9061 0.1033 0.1207 0.67 (0.090) (0.064) (0.031) (0.109) (0.047) (0.001) tobacco 0.4584 0.6289 -0.3977 -0.6742 0.7002 0.0210 0.61 (0.082) (0.13) (0.037) (0.066) (0.006) (0.330) drugs & pharmaceutical 0.0050 -0.2074 -0.8580 -0.4120 0.7546 0.6426 0.71 (0.096) (0.003) (0.005) (0.078) (0.083) (0.032) chemical 0.8270 -0.6208 -0.8669 -0.4373 0.3008 0.1757 0.74 (0.071) (0.085) (0.047) (0.038) (0.064) (0.050) beverages 0.9826 -0.7801 -0.0232 -0.3776 0.1165 0.0663 0.72 (0.022) (0.058) (0.011) (0.022) (0.009) (0.017) the output elasticity of production workers has a minimum value of 0.01 and the maximum value is 0.99 but the minimum value of 0.008 in non-production workers and the maximum value is 0.75. the minimum value of trade liberalization elasticity of production workers is 0.09 and the maximum value of 0.72. in non-production workers, the minimum value of trade liberalization elasticity is 0.021and the maximum value of trade liberalization elasticity is 0.91. the minimum values of r2 are 0.23 and 0.34 in production and non-production workers respectively while the maximum value of r2 in production workers is 0.85 and 0.78 in non-production workers. 5. conclusions and policy recommendations the study probes the effects of trade liberalization on the labor demand elasticity in thirteen industries of pakistan’s manufacturing sector. we have executed two types muhammad ramzan sheikh, misbah rauf, irfan hussain and asad abbas 66 of analyses: aggregated and disaggregated. in aggregate analysis, we have taken all workers and in disaggregated analysis, we have separated the workers into production workers and non-production workers. the panel data are taken from cmi for the years 1995-1996, 2000-2001, and 2005-2006 and pooled least square method has been applied to estimate the effects of wage, interest rate, price of raw material, output, and trade liberalization on labor demand. at the aggregated and disaggregated levels, the findings show that trade liberalization has a significant positive effect on labor demand in these industries. the positive relationship between trade liberalization and labor demand is because pakistan is a labor-abundant country. when trade liberalization increases, it positively influences the labor demand due to an increase in more output in these industries. so, labor is a substitute for capital in these industries. following policies are suggested to increase the labor demand in pakistan: • the government may not intervene in the labor market by announcing minimum wages so that the unemployment rate may be decreased. in pakistan, the unemployment level is high and firms can hire more labors at a low wage rate. • the policymakers may help industries by decreasing the cost of raw material and capital. the government may provide loans at low interest rates. moreover, raw material costs can also be minimized. • the government may reduce the tariff on capital goods to establish industries and enhance the level of output that has a positive impact on labor demand. references akhter, n., & ali, a. (2007). does trade liberalization increase the labor demand elasticities? evidence from pakistan. online at https://mpra.ub.unimuenchen.de/3881/ mpra paper no. 3881, posted 06 jul 2007 utc bruno, g. s., falzoni, a. m., & helg, r. (2004). measuring the effect of globalization on labour demand elasticity: an empirical application to oecd countries. università commerciale luigi bocconi. chamarbagwala, r., & sharma, g. (2011). industrial de-licensing, trade liberalization, and skill upgrading in india. journal of development economics, 96(2), 314336. chaudry m.a. ahmad, e. burki, abid a. and khan m. a. (1999). industrial sector inputs demand responsiveness and policy responsiveness. pakistan development review, 38(4), 1083-1100. cox, d., & harris, r. (1985). trade liberalization and industrial organization: some estimates for canada. journal of political economy, 93(1), 115-145. journal of applied economics and business studies, volume. 5, issue 2 (2021) 53-68 https://doi.org/10.34260/jaebs.524 67 dobbelaere, s., & wiersma, q. (2020). the impact of trade liberalization on firms' product and labor market power. iza discussion papers, no. 12951, institute of labor economics (iza), bonn. driffield, n. l., & kambhampati, u. s. (2003). trade liberalization and the efficiency of firms in indian manufacturing. review of development economics, 7(3), 419430. dutta, p. b. (2014). skilled–unskilled wage inequality, product variety and unemployment: a static general equilibrium analysis. the journal of international trade & economic development, 23(1), 31-55. edwards, s. (1993). openness, trade liberalization, and growth in developing countries. journal of economic literature, 31(3), 1358–1393. erten, b., leight, j., & tregenna, f. (2019). trade liberalization and local labor market adjustment in south africa. journal of international economics, 118, 448-467. fajnzylber, p., & maloney, w. f. (2005). labor demand and trade reform in latin america. journal of international economics, 66(2), 423-446. greenaway, d., hine, r. c., & wright, p. (1999). an empirical assessment of the impact of trade on employment in the united kingdom. european journal of political economy, 15(3), 485-500. hasan, r. (2001). the impact of trade and labor market regulations on employment and wages: evidence from developing countries. east-west center working papers, economics series no 32. hasan, r., mitra, d., & ramaswamy, k. v. (2003). trade reforms, labor regulations and labor-demand elasticities: empirical evidence from india (no. w9879). national bureau of economic research. jean, s. (2000). the effect of international trade on labor‐demand elasticities: intersectoral matters. review of international economics, 8(3), 504-516. krishna, p., & mitra, d. (1998). trade liberalization, market discipline and productivity growth: new evidence from india. journal of development economics, 56(2), 447-462. krishna, p., mitra, d., & chinoy, s. (2001). trade liberalization and labor demand elasticities: evidence from turkey. journal of international economics, 55(2), 391-409. lang, k. (1998). the effect of trade liberalization on wages and employment: the case of new zealand. journal of labor economics, 16(4), 792-814. liyanaarachchi, t. s., naranpanawa, a., & bandara, j. s. (2016). impact of trade liberalisation on labour market and poverty in sri lanka. an integrated macromicro modelling approach. economic modelling, 59, 102-115. muhammad ramzan sheikh, misbah rauf, irfan hussain and asad abbas 68 mahomedy, a. c. (2013). international trade and labour demand elasticities: is there any empirical evidence from south africa? ersa working paper 348. maloney, w. f., & fajnzylber, p. (1999). labor demand and trade reform in latin america. the world bank. mehta, a., & hasan, r. (2012). the effects of trade and services liberalization on wage inequality in india. international review of economics & finance, 23, 75-90. michaels, g. (2008). the effect of trade on the demand for skill: evidence from the interstate highway system. the review of economics and statistics, 90(4), 683701. njikam, o. (2016). trade liberalization, labor market regulations and labor demand in cameroon. international review of economics & finance, 43, 525-541. paz, l. (2012). the impacts of trade liberalization on informal labor markets: an evaluation of the brazilian case. online at https://mpra.ub.unimuenchen.de/38858/ mpra paper no. 38858, posted 17 may 2012 14:07 utc robbins, d., & gindling, t. h. (1999). trade liberalization and the relative wages for more‐skilled workers in costa rica. review of development economics, 3(2), 140-154. rodrik, d., (1997). has globalization gone too far? washington dc: institute for international economics. revenga, a. (1997). employment and wage effects of trade liberalization: the case of mexican manufacturing. journal of labor economics, 15(s3), s20-s43. sankaran, u., abraham, v., & joseph, k. j. (2010). impact of trade liberalization on employment: the experience of india’s manufacturing industries. the indian journal of labour economics, 53(4), 587-405. slaughter, m. j. (2001). international trade and labor–demand elasticities. journal of international economics, 54(1), 27-56. wacziarg, r., & wallack, j. s. (2004). trade liberalization and intersectoral labor movements. journal of international economics, 64(2), 411-439. yasmin, b. (2011). trade openness: new evidence for labor-demand elasticity in pakistan's manufacturing sector. the lahore journal of economics, 16(2), 55. journal of applied economics and business studies, volume. 5, issue 2 (2021) 69-86 https://doi.org/10.34260/jaebs.525 69 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x impact of bank-specific, corporate governance and environmental factors on bank efficiency and profitability in pakistan hafiz muhammad athar1* and sumayya chughtai2 1ph.d. scholar, faculty of management sciences, international islamic university (iiu), islamabad, pakistan 2assistant professor, faculty of management sciences, international islamic university (iiu), islamabad, pakistan. abstract the study aims to investigate the impact of bank-specific, board structure, gender diversity, and environmental factors on bank efficiency and profitability in pakistan by taking a sample of seventeen commercial banks for the period 2013-2018. data envelopment analysis (dea) and return on assets (roa) are used as a proxy to measure bank efficiency and profitability. panel estimation techniques and generalized method of moments (gmm) are used to conceptualize the research framework and to test the hypotheses. the findings indicate a negative relationship of non-performing loans, advances, level of involvement of women into other committees, and csr index with roa; while more presence of women on board reveals a positive and significant impact on roa that is consistent with critical mass theory. however, ceo duality shed a positive impact on technical efficiency; while bank size signifies an inverse relationship with roa and technical efficiency. moreover, deposit influences roa positively; while board size finds a positive and significant relationship with roa and technical efficiency. the findings are important for various stakeholders as they can efficiently take their decision-making to better understand the factors influence bank performance. this study recommends future researchers do the same research by inculcating a larger sample size. keywords too big to fail paradigm, critical mass/tokenism theory, csr index, data envelopment analysis (dea), return on assets (roa) jel classification c23, e58, g21, g28, g34, k32 * athar23_abbas@yahoo.com mailto:athar23_abbas@yahoo.com hafiz muhammad athar and sumayya chughtai 70 1. introduction the financial sector plays a significant role in the economic progress, growth, and development of the country (chortareas, girardone & ventouri, 2013). history depicts that most of the economic crisis in the banking sector arises due to excessive borrowings, risky over-lending, heavy existence of non-performing loans, illiquidity of assets, weak regulatory and supervisory frameworks/regulations, inadequate credit analysis, bad governance, corruption, political interferences to write-off loans, fraudulent practices such as accounting manipulations, weak internal control systems, etc (bank for international settlement, 2015). thus, a strong and effective financial system is necessary to handle the issues related to poverty elimination, social injustice, unequal distribution of wealth, community’s development, unemployment, inflation, market openness, bad governance, corruption, money laundering, earning manipulations, political interference, global warming, environmental degradation, etc. broadly, there are two factors affecting bank performance i.e. internal and external. internal are those that are under the control of management such as bankspecific, governance, corporate social responsibility disclosure, reporting manipulations, etc. however, external are those that are beyond the control of management such as political, social, regulators (central bank and government), macroeconomic, technological, competition, demographics/contextual, etc. all these factors may positively or negatively influence the bank's performance, hence ultimately affects stakeholders’ interests. for example, poor asset quality, measured by the nonperforming loans to total loans (npls/tl) is an indicator of bank failures. this implies that the higher value of this ratio increases the probability of default risk that undermines performance (batir, volkman & gungor, 2017). liquidity, as measured by deposits to total assets and loans to total asset ratio, may also positively or negatively influence bank performance. they argue that more lending implies more generation of interest revenue if manage risk in a better way, but on the other side, more lending faces a higher risk of bankruptcy that deteriorates performance (goddard, liu, molyneux, & wilson, 2013). similarly, the higher level of deposits is a signal of illiquidity of assets which means bank provides lesser lending that not only hampers economic activities but also affects performance (batir et al., 2017). capital adequacy indicates the financial strength of banks or their ability to absorb captivated shocks or losses or unanticipated losses. larger banks positively influence bank performance as they enjoy the benefits of economies of scale, more market power, ease of obtaining equity, raise debt at a lower cost. etc (molyneux & thornton, 1992). conversely, larger banks may negatively influence bank performance due to the too big to fail paradigm (batir et al, 2017). similarly, the dimensions of corporate governance may also positively or negatively influence bank performance. the theories that better explain the relationship between dimensions of corporate governance and performance include information journal of applied economics and business studies, volume. 5, issue 2 (2021) 69-86 https://doi.org/10.34260/jaebs.525 71 asymmetry theory, agency theory, stewardship theory, tokenism theory, critical mass theory. agency theory criticizes the duality concept and argues that duality makes managerial monitoring ineffective (krause, semadeni, & cannella, 2014). however, stewardship theory promotes the concept of duality as it states that duality will facilitate strong and unified leadership and makes managerial monitoring effective as insiders have more information and knowledge about the organization than outsiders. gender diversity is one of the most burning issues regarding the corporate sector because males and females are traditionally, culturally, and socially different (conyon & he, 2017; varnita, niladri, & jamini, 2018). regarding gender diversity, few questions arise such as why it is necessary to enter women into the corporate board? what and how they will bring change into a corporate culture that enhances organization performance? whether her role in the corporate board as a showpiece (token) or key influencers? likewise, the environmental factors may also positively or negatively influence bank performance. environmental factors are those that are linked with corporate social responsibility (csr) disclosure. disclosure protocols are monitoring devices that reduce information and agency costs by sharing useful and quality information among stakeholders. the banking industry has both a direct and an indirect influence on the environment. direct impact belongs to the operations within banks such as the use of natural resources (energy management, water management, waste management), digitalization (paperless banking, using atms, installation of solar systems, dual utilization of papers), and reduction in greenhouse gas effect (conducting meeting through video conferencing). indirect impacts belong to the external environmental system i.e. consists of two factors such as the incorporation of environmental risk before lending/ advances and sustainable/green financing. the indirect impact seems to be more important than the direct one as it shed overall influence on society. as in the case of external environmental impacts, the banks provide loans to those projects that lessen pollution and involve the plantation of trees, safe water, renewable energy, etc. the bank should involve in green banking/financing. green financing includes 3ps i.e. planet, people, and profit. it refers to engage in such activity that helps to reduce the external carbon emission and provide loans to borrowers compliant with health, safety, and environmental rules. green banking initiatives include in-house environment management, introducing green finance, incorporation of environmental risk, creation of climate risk fund, introducing green marketing, employee training, customer awareness, and green events. we have also seen that well-established banks also play their role in philanthropic activities (education, health, disaster management, etc), the well-being of employees/citizens, and community development (szegedi, khan, & lentner, 2020; oyewumi, ogunmeru, & oboh, 2018). primarily, the study examines the impact of multiple factors on bank performance in the context of pakistan in a single model. a major contribution of this study is the inclusion of the bank-specific, board structure, gender diversity, and hafiz muhammad athar and sumayya chughtai 72 environmental factors in a single model to evaluate the bank efficiency and profitability. this study used dea (two-stage model) and return on assets (roa) as a proxy to measure bank efficiency and profitability. 2. literature review literature supports a positive relationship of non-performing loans (npls) with bank performance (iveta, 2015; syafri, 2012). contrary to the above, batir et al. (2017) find a negative impact of npls on bank performance. this is because of adverse borrower selection i.e. one of the implications of information asymmetry theory. the scholars argue that this happens due to an increase of defaulters that restrict the bank earning resultantly decreases the bank performance. bourke (1989), batir et al. (2017), and gitau, anyango, and rotich (2017) find a positive relation of liquidity with bank performance. contrary to the above, the scholar finds a negative relationship of liquidity with bank performance (athanasoglou, brissimis, & delis, 2008; havrylchyk, 2006). bourke (1989), syafri (2012), hasanul et al. (2017); and aziz and knutsen (2019) find a positive relationship of capitalization with bank performance. this is consistent with the signaling theory and the expected bankruptcy cost hypothesis. contrary to the above, the scholars find a negative relationship of capital adequacy with a bank’s performance (bitar, pukthuanthong & walker, 2018). this is consistent with the risk-return hypothesis. the supporters of finding a positive relationship of bank size (bs) with performance argue that larger banks enjoy the benefits of economies of scale (aziz and knutsen, 2019; hasanul et al., 2017; gitau et al., 2017). conversely, the supporters of finding negative relation of bs with performance posits that larger banks face too big to fail paradigm which implies that it may lessen profits as a result of diseconomies of scale (bourke, 1989, batir et al., 2017). the scholars find a relationship between bank size and performance looks like kuznets inverted u-curve theory, business cycle and product life cycle includes eichengreen & gibson (2001). the supporters of findings a positive relationship of board size (bods) with performance postulates that the larger board possesses versatile knowledge, skill to make better decisions, and difficult for ceo to dominate (riyadh et al., 2019; kiel & nicholson, 2003). on the other side, the scholar finds a negative relationship of bods with performance argues that larger boards are ineffective as it becomes difficult to coordinate, encourages free-riding and each member has their interest that may conflict with the interest of the firm. so, they are in favor of a smaller board size that makes every member more accountable (adnan, htay, rashid & meera, 2011). stewardship theory finds positive relation of ceo duality with performance and argues that duality will facilitate strong and unified leadership and makes managerial monitoring effective as insiders have more information and knowledge about the organization than outsiders (kiel & nicholson, 2003). contrary to the above, agency theory finds negative relation of ceo duality with performance and argues that duality makes managerial monitoring journal of applied economics and business studies, volume. 5, issue 2 (2021) 69-86 https://doi.org/10.34260/jaebs.525 73 ineffective (krause, semadeni & cannella, 2014). andersson and wallgren (2018); beate and gro (2010), riyadh et al. (2019) find a positive relation of women’s participation in the board with bank performance. this is consistent with critical mass theory. torchia, calabrò, and huse (2011) argue that most firms still have only one woman on the board that is still considered as a token and finds a negative relationship with bank performance. token women may face three types of fear such as visibility, polarization, and assimilation. these issues may be resolve when thirty percent of directors are female on the board of directors. berger, deyoung, genay, and udell (2000) present two hypotheses namely home field advantage and global advantage regarding ownership structure. the supporters of the home-field advantage hypothesis argue that domestic banks are more familiar with local culture, economic, social norms, government policy and regulations, institutional framework, and political factors (sufian & habibullah, 2010). conversely, the global advantage hypothesis advocates that foreign banks may have a comparative advantage of product differentiation, knowledge transfer, modern technology, better risk exposure, and reduction in the cost of capital (havrylchyk,2006). two opposite views are prevailing in the literature regarding csr index and firm performance i.e. (i) friedman (1970) argues that the manager’s main responsibility to increase firm profit and stakeholder’s wealth and doing anything else will be the misuse of the authority and brings additional expenses. hence reduces the profit of the company. fahad and busru (2021) investigate a sample of 386 companies for the period of 20072016 in india and find that csr disclosure negatively influences the return on assets. oyewumi, ogunmeru, and oboh (2018) elucidate a negative relationship between csr and performance. this is consistent with agency theory. (b) conversely, freeman (1984) posits that a firm task not only to meet the expectations of shareholders but also to protect the interests of various stakeholders as well. this is in line with the stakeholder theory. the scholars that find the positive relationship of csr with firm performance claim that csr disclosure would lead to improving its image/reputation, retention and loyalty of the customer, service delivery, attracting investors and prospective employees, employee productivity (mravlja, 2017, maqbool & zameer, 2018). however, riyadh et al. (2019) did not find any relation between csr and profitability. this study used the csrd index as a proxy to measure environmental factors. based on the literature and objective, the following hypothesis was tested in this study; h 0: there is a significant relationship of bank-specific factors (asset quality, liquidity, capitalization, bank size) with roa and technical efficiency. h 0: there is a significant relationship of governance (board size, ceo duality, gender diversity, ownership structure) with roa and technical efficiency. h 0: there is a significant relationship of environmental factors (csr index) with roa and technical efficiency hafiz muhammad athar and sumayya chughtai 74 3. data and methodology all banks that fall under the ambit of pakistan are the population of this study. however, this study concentrates only on commercial banks as the services provided by banks are reasonably homogenous and comparable across countries. convenience sampling is used and taken data of commercial banks from 2013 to 2018. 3.1. variables description and measurement the variable choice in the study is based on the literature review to align with the past researches. s. no. variable name symbols formulas literature support 1 asset quality npl non-performing loans/total loans batir et al., (2017); syafri (2012) 2 relative liability size dep deposits/total assets batir et al. (2017); gitau et al., (2017) 3 relative lending size loan loans(advances)/total assets batir et al., (2017); syafri(2012) 4 capitalization car shareholders equity/total assets hasanul et al., (2017); bitar et al., (2018); aziz and knutsen (2019) 5 bank size bs ln(total assets) gitau et al., (2017); batir et al., (2017); aziz and knutsen (2019) 6 board of director size bods ln (total # of board members) riyadh et al., 2019; adnan et al. (2011) 7 ceo duality ceod 1 if ceo and board chairperson are different persons and 0 otherwise krause et al., (2014); kiel and nicholson (2003) 8 presence of women on board wti 1 if the women are more than 30% of board and 0 otherwise andersson and wallgren (2018); riyadh et al. (2019) 9 level of involvement of women into other committees liw 1 if women involved in more than one committee and 0 otherwise andersson and wallgren (2018); riyadh et al. (2019) 10 ownership structure os %age of shares owned by foreign shareholders to the total number of shares issued sufian & habibullah (2010); herdjiono and sari (2017) 12 csr index csr csr score = sum of csr items/ total number of csr items. “1” disclose csr items and “0” otherwise fahad and busru (2021); mravlja, 2017, riyadh et al. (2019) 13. stock market development smd stock market capitalization/gdp sufian, kamarudin and nassir (2016) journal of applied economics and business studies, volume. 5, issue 2 (2021) 69-86 https://doi.org/10.34260/jaebs.525 75 14. demand density dd total deposits of banking sector/total area dietsch and lozano-vivas (2000) 14 return on assets roa net income/total assets mravlja (2017); batir et al. (2017); aziz and knutsen (2019) 15 data envelopment analysis te inputs: deposits, interest on deposits, fixed assets, share capital outputs: total loans, interest on loans, net income sealey & lindley (1977); batir et al. (2017); hasanul et al. (2017); majeed and zainab (2016) note: the table exhibits the description, measurements, and the previous scholars used these variables in their studies. corporate social responsibility disclosure index (csr) literature exhibits the methods to measure csrd index such as (1) use of reputation indices such as dow jones sustainability index, esg, asset4, eiris, etc (2) content analysis (3) surveys. this study uses content analysis for the collection of information related to the csrd index checklist from annual/sustainability reports, websites, etc. of individual banks. this study uses the dichotomous and unweighted disclosure index method as used by maqbool and zameer (2018), and riyadh et al. (2019). if banks disclose dimensions of csr in their annual reports or websites, it will be scored one and otherwise assigned zero. the formula used to calculate csr index is as under: csr index = sum of csr items /total number of csr items summary of csr disclosure items based on gri and on previous literature “1” if csr reported in annual/sustainability report and “0” otherwise crg csrreporting ccg csr-sustainable/csr committee-g cseg csr-stakeholders engagement cnpg csr-national policies-g cga csr-accreditation of an international organization cceg csr-s_ethics_code of conduct and ethics ccamlg csr-s_ethics_aml/kyc policys ccgg csr-s_ethics_grievance redressal policy csdg common forum for dialogue csig other information disclosure cews csr_employees well being ccps csr_customer privacy policy cpas csr-philanthropic activities description cdons csr_donations cnre csr-e_natural resources(n) cde csr-e_digitialization cedde csr-e_incorporation of environmental risk before lending (environmental due diligence) cgfe csr-eac_green financing csr csr index hafiz muhammad athar and sumayya chughtai 76 3.2 methodology the study for the evaluation of the banking sector's performance is significant as it can influence the interests/decision-making of various stakeholders. the scholars use various proxies to measure bank performance such as accounting or profitability measures (roa, roe, etc.), market measures (eps, market to book value ratio, tobin q, etc.), economic measures (economic value added), and efficiency measures (parametric and non-parametric approaches). this study uses return on assets (roa) and technical efficiency (te) as a measure of bank performance. roa indicates that how the firm utilizes its assets to generate income and is used by athanasoglou, delis, & staikouras (2006); zheng, rahman, begum, & ashraf (2017); mravlja (2017); riyadh, sukoharsono & alfaiza (2019). the other approach used in this study to measure bank performance is efficiency. efficiency implies that how the firm utilizes its assets to get the maximum results/output. this study employs data envelopment analysis (dea) that deals with many inputs and outputs in a single model as used by farrell (1957); charnes, cooper, and rhodes (1978); banker, charnes, and cooper (1984). the main challenge in dea is deciding the appropriate selection of input and output variables. this study uses the intermediation approach for the selection of inputs and outputs (sealey & lindley, 1977). literature reveals that the intermediation approach is better than other approaches as it inculcates interest expenses that contribute more than fifty percent of the total costs. two motives for the application of this approach i.e. (i) banks act as an intermediator to accept deposits and lend them for investments (ii) it is used to assess the efficiency of the entire bank. this study takes deposits, interest on deposits, fixed assets, and share capital as input and produces output in terms of total loans, interest on loans, and net income. previous scholars that widely used this approach includes sherman and gold (1985); sufian et al. (2016); alharthi (2016); majeed and zanib (2016); batir et al. (2017); hasanul, rubi, and eric (2017); and yonnedi and panjaitan (2019). 3.3. model specification to test the hypothesis of this study, the following model is used. 𝑅𝑂𝐴𝑖,𝑡=𝛼0 + 𝛽1𝑁𝑃𝐿𝑖,𝑡 + 𝛽2𝐷𝑒𝑝𝑜𝑠𝑖𝑡𝑖,𝑡 + 𝛽3𝐿𝑜𝑎𝑛𝑖,𝑡 +𝛽4𝐶𝐴𝑅𝑖,𝑡 + 𝛽5𝑆𝑖𝑧𝑒𝑖,𝑡 + 𝛽6𝐵𝑜𝐷𝑆𝑖,𝑡 + 𝛽7𝐶𝐸𝑂𝐷𝑖,𝑡 + 𝛽8𝑊𝑡𝑖𝑖,𝑡 + 𝛽9𝐿𝐼𝑊𝑖,𝑡 + 𝛽10𝑂𝑆𝑖,𝑡 + 𝛽11𝐶𝑆𝑅𝑖,𝑡 + 𝛽12𝑆𝑀𝐷𝑡+𝛽13𝐷𝐷𝑡 + 𝜇𝑡 ………………………………a 𝑇𝐸𝑖,𝑡=𝛼0 + 𝛽1𝑁𝑃𝐿𝑖,𝑡 + 𝛽2𝐷𝑒𝑝𝑜𝑠𝑖𝑡𝑖,𝑡 + 𝛽3𝐿𝑜𝑎𝑛𝑖,𝑡 +𝛽4𝐶𝐴𝑅𝑖,𝑡 + 𝛽5𝑆𝑖𝑧𝑒𝑖,𝑡 + 𝛽6𝐵𝑜𝐷𝑆𝑖,𝑡 + 𝛽7𝐶𝐸𝑂𝐷𝑖,𝑡 + 𝛽8𝑊𝑡𝑖𝑖,𝑡 + 𝛽9𝐿𝐼𝑊𝑖,𝑡 + 𝛽10𝑂𝑆𝑖,𝑡 + 𝛽11𝐶𝑆𝑅𝑖,𝑡 + 𝛽12𝑆𝑀𝐷𝑡+𝛽13𝐷𝐷𝑡 + 𝜇𝑡 ……………………………….b where, roai,t or tei,t = technical efficiency or return on asset at the ith bank and t time period journal of applied economics and business studies, volume. 5, issue 2 (2021) 69-86 https://doi.org/10.34260/jaebs.525 77 npli,t= non-performing loans to total loans at ith bank and t time period loani,t= loans to total asset ratio at the ith bank and t time period depositsi,t= deposits to total asset ratio at ith bank and t time period cari,t = capital adequacy at the ith bank and t time period bsi,t = bank size at the ith bank and t time period bodsi,t = board of directors size at the ith bank and t time period ceodi,t = ceo duality at ith bank and t time period ceowi,t = ceo women at ith bank and t time period wtii,t = percentage of women on the board at the ith bank and t time period osi,t = ownership structure at the ith bank and t time period csrii,t = corporate social responsibility disclosure index at the ith bank and t time period control variables ddt = demand density at t time period smdt = stock market development at the t time period μt = error term 3.4. estimation techniques this study uses panel estimation techniques (common effect, fixed effect, and random effect) to examine the impact of multiple factors on bank performance. panel data prefer over cross-sectional or time-series data as it addresses individual heterogeneity. however, the diagnostic test such as breush-pagan lagrange multiplier, hausman test, likelihood ratio executes to select which model is best among panel estimation techniques. this study further employs the system gmm estimation technique developed by arellano-bover/blundell-bond to handle the issues of normality, multi-collinearity, auto-correlation, heteroscedasticity, etc. the scholars who used the same technique in the literature include aziz and knutsen (2019), varnita, niladri, and kanta (2018), maqbool and zameer (2018), zheng et al., (2017). hafiz muhammad athar and sumayya chughtai 78 4. results and discussion 4.1. descriptive statistics it tells us about the mean, median, maximum, minimum, and standard deviations of different variables used in the study. table-1 presents descriptive statistics of all variables included in this study. table-1. descriptive statistics variables observations mean median maximum minimum std. dev. roa 102 0.775 0.870 2.718 -2.660 0.900 te 102 0.758 0.777 1.000 0.224 0.163 npl 102 11.598 10.244 39.418 1.522 7.298 dep 102 74.837 75.128 130.951 46.439 11.193 loan 102 38.587 37.703 61.098 15.331 7.974 car 102 8.723 7.960 25.327 2.240 4.066 bs 102 15.059 15.298 17.011 12.082 1.244 bods 102 2.157 2.079 2.565 1.609 0.198 ceod 102 0.882 1.000 1.000 0.000 0.324 wti 102 4.933 0.000 66.667 0.000 13.425 liw 102 0.275 0.000 1.000 0.000 0.448 os 102 12.172 2.313 78.878 0.000 20.692 csr 102 0.427 0.392 0.952 0.056 0.226 smd 102 28.077 28.080 33.000 24.400 2.975 d 102 100.746 100.023 122.366 80.144 15.664 note: table-1 shows the statistic summary (mean, median, minimum, maximum, and standard deviation) of dependent variables and explanatory variables used in this study. roa of pakistani banks is 0.775. this implies that banks earn 0.775% of total assets with a maximum value of 2.718 and a minimum value of -2.660, whereas standard deviation depicts the variation from means. furthermore, the median profit of pakistani banks is 0.870 which is higher than the return on assets. similarly, te indicates the efficiency of pakistani banks with average efficiency and median values as denoted in the table. the explanatory variables are also explained/interpreted in the same analogy. 4.2. correlation analysis correlation analysis exhibits the strength and direction of relationships among variables. its value lies between -1 to +1. this table indicates that if values are less than 0.70, it means no multicollinearity issue. the variance inflation factor (vif) also confirmed our previous opinion as all the variables’ values are less than 10. table-2 indicates correlation analysis and vif. journal of applied economics and business studies, volume. 5, issue 2 (2021) 69-86 https://doi.org/10.34260/jaebs.525 79 table-2. correlation analysis and variance inflation factors (vifs) probability vif roa te npl dep loan car bs bods ceod wti liw os csr smd dd roa 1.000 te -0.228 1.000 npl 1.444 -0.420 0.070 1.000 dep 1.634 -0.039 -0.023 0.222 1.000 loan 1.630 -0.456 0.148 0.256 0.097 1.000 car 1.629 0.195 -0.028 -0.151 -0.312 -0.184 1.000 bs 4.097 0.605 -0.325 -0.238 0.192 -0.376 -0.290 1.000 bods 1.214 0.229 -0.139 -0.088 0.041 0.069 0.014 0.178 1.000 ceod 2.316 0.283 -0.048 -0.053 -0.003 0.097 -0.202 0.438 0.252 1.000 wti 3.797 -0.338 0.191 0.027 -0.062 0.063 0.281 -0.529 -0.146 -0.637 1.000 liw 2.144 -0.084 0.031 -0.225 -0.076 -0.155 0.182 -0.033 -0.131 -0.185 0.551 1.000 os 1.707 0.430 -0.125 -0.225 0.033 -0.162 0.046 0.486 0.017 0.162 -0.096 0.307 1.000 csr 2.487 0.417 -0.192 -0.423 -0.106 -0.206 -0.062 0.510 0.266 0.079 -0.182 -0.056 0.140 1.000 smd 1.081 0.000 0.124 -0.070 -0.148 -0.089 -0.013 0.045 0.067 -0.051 0.063 0.085 0.019 0.165 1.000 dd 2.214 0.030 -0.023 -0.288 -0.375 -0.133 -0.114 0.155 0.145 0.037 0.160 0.200 -0.004 0.516 0.229 1.000 note: this table indicates the correlation analysis. if the values of correlation are less than 0.75 and vif values are less than 10, this implies no issue of multi-collinearity among variables. hafiz muhammad athar and sumayya chughtai 80 4.3 regression analysis after the regression, the next step is to test three hypotheses i.e. bp lagrange multiplier h0: pooled is better than random effect model (rem); (ii) hausman h0: random effect model (rem) is better than fixed effect model (fem); (iii) likelihood ratio h0: pooled is better than fixed-effect model (fem). the null hypothesis is accepted or rejected based on the p-value. if the p-value is less than .05, reject h0 and vice-versa. after exercising this practice, the fixed effect model is selected as in the case of roa and random effect model when the dependent variable is technical efficiency. table 3 presents panel estimation techniques such as pooled, fixed, and random effects. table-3. panel estimation techniques variable roa te cem fem rem cem fem rem npl -0.027 (0.009)* -0.046 (0.016)* -0.030 (0.009)* -0.001 (0.003) -0.006 (0.006) -0.001 (0.003) dep -0.003 (0.006) 0.006 (0.006) -0.003 (0.005) 0.001 (0.002) 0.001 (0.002) 0.001 (0.002) loan -0.028 (0.009)* -0.025 (0.013)*** -0.028 (0.009)* -0.001 (0.003) -0.005 (0.004) -0.002 (0.003) car 0.052 (0.018)* -0.003 (0.024) 0.047 (0.016)* -0.006 (0.005) 0.001 (0.008) -0.005 (0.005) bs 0.287 (0.092)* -0.636 (0.315)** 0.313 (0.094)* -0.060 (0.026)** -0.015 (0.110) -0.059 (0.029)** bods 0.427 (0.314) -0.871 (0.558) 0.355 (0.324) -0.099 (0.087) 0.065 (0.195) -0.073 (0.100) ceod 0.431 (0.265) 0.082 (0.307) 0.191 (0.230) 0.132 (0.074)*** 0.240 (0.107) 0.150 (0.076)*** wti 0.010 (0.008) 0.053 (0.016)* 0.008 (0.008) 0.002 (0.002) 0.004 (0.006) 0.003 (0.002) liw -0.527 (0.184)* -1.083 (0.225)* -0.600 (0.168)* -0.029 (0.051) -0.052 (0.079) -0.046 (0.055) os 0.008 (0.004)** 0.021 (0.010)** 0.009 (0.004)** 0.001 (0.001) 0.006 (0.004) 0.001 (0.001) csr 0.466 (0.394) -3.159 (0.730)* -0.112 (0.400) 0.019 (0.109) -0.454 (0.256) -0.024 (0.123) smd -0.009 (0.020) 0.009 (0.016) -0.008 (0.015) 0.009 (0.005) 0.009 (0.006) 0.009 (0.005)*** dd -0.009 (0.005)*** 0.027 (0.008)* -0.005 (0.005) 0.000 (0.001) 0.002 (0.003) 0.000 (0.002) c -2.683 (1.725) 11.418 (4.723)** -2.823 (1.659)*** 1.573 (0.479)* 0.460 (1.654) 1.479 (0.523)* r-squared 0.654 0.836 0.488 0.185 0.388 0.149 adj. rsquared 0.603 0.771 0.412 0.065 0.142 0.023 f-statistic 12.801 12.699 6.447 1.538 1.574 1.181 journal of applied economics and business studies, volume. 5, issue 2 (2021) 69-86 https://doi.org/10.34260/jaebs.525 81 prob (fstatistic) 0.000 0.000 0.000 0.120 0.062 0.306 durbinwatson stat 1.706 2.477 1.873 2.222 2.586 2.300 note: *,**,*** at 1%, 5%, 10% respectively. the figure in parenthesis shows standard error. return on assets (roa) and technical efficiency (te) are used as dependent variables. the models represent the impact of the non-performing loans, deposits, loans, capitalization, bank size, the board size, ceo duality, women participation on the board, level of involvement of women on board in other committees, ownership structure, corporate social responsibility index on bank performance. the control variables used in the analysis include demand density (dd) and stock market development (smd). r-squared is 84% and 15% in the case of roa and te. this implies that 84% and 15% variation in our models arises due to explanatory variables. the value of durbin-watson lies between 1.5 to 2.5 in the models implies no sign of auto-correlation among the predictors. bank-specific factors and performance this study reports a significant negative relationship of non-performing loans, advances with roa, however, insignificant positive relationships exist with technical efficiency. they argue that more lending may cause borrowers to default. this implies that banks may face the risk of bankruptcy that increases the financing cost and reduce profitability. the banks may select adverse borrowers due to imperfect information or high competition exists in the market. previous literature that supports the findings of this study includes aziz and knutsen (2019); batir et al. (2017) and sufian and habibullah (2010). the study elaborates a significant negative relationship of bank size with roa and technical efficiency. the scholars posit that banks' performance may decline due to diseconomies of scale, mismanagement, bureaucratic issues, and engagement in more risky investments. the scholars that are in support of this argument include bourke (1989); syafri (2012); sufian and habibullah (2010) and batir et al. (2017). corporate governance and performance the findings indicate that ceo duality influences technical efficiency significantly and positively. this is in line with stewardship theory arguing that duality creates managerial monitoring effectively. previous literature that supports this argument includes kiel and nicholson (2003). however, an insignificant and positive relationship exists between ceo duality and roa. furthermore, the result exhibits a significant positive relationship of the percentage of women on the board with roa. this aligns with critical mass theory implies that more than thirty percent of women on board enhance the profitability of banks. the supporters of this argument include andersson and wallgren (2018); beate and gro (2010) and riyadh et al. (2019). however, the level of involvement of women in other committees finds a significant negative relationship with roa. this implies that the more presence of a woman in other committees can adversely affect bank profitability. furthermore, a significant and positive relationship of ownership structure with roa. the supporters advocate that hafiz muhammad athar and sumayya chughtai 82 foreign banks may have a comparative advantage of product differentiation, knowledge transfer, modern technology, better risk exposure, and reduction in the cost of capital (havrylchyk, 2006). previous scholars that support this argument include jayati and subrata (2018). corporate social responsibility disclosure index and performance the findings indicate a negative and significant relationship of the csr disclosure index with roa. they argue that the manager’s main responsibility to increase firm profit and stakeholder’s wealth and doing anything else will be the misuse of the authority and brings additional expenses (friedman, 1970). this is in line with agency theory (fahad & busru, 2021; mravlja, 2017; maqbool & zameer, 2018). system gmm technique this study also employs the system gmm estimation technique developed by arellano-bover/blundell-bond as the above analysis includes the problem of endogeneity, heteroscedasticity, autocorrelation. the condition for the application of gmm is (i) when the time period is shorter than the number of groups (t<n); (ii) validity of instruments by applying sargan/hansen test; (iii) absence of autocorrelation at ar (2) by applying arellano-bond test. the insignificance value of the sargan/hansen test is an indicator of the validity of the instrument. this implies that instruments are not correlated with residuals. similarly, the insignificance value of ar (2) means the absence of autocorrelation in the residuals. table 4 depicts the impact of multiple explanatory variables on roa and technical efficiency by using the system gmm. the lagged values of dependent variables are used as instruments while conducting the analysis. table-4. system gmm variable roa te fem rem roa/te (-1) -0.109 (0.214) -0.148 (0.242) npl -0.050 (0.026)*** -0.003 (0.002) dep 0.008 (0.004)** 0.000 (0.001) loan -0.027 (0.012)** -0.001 (0.002) car 0.005 (0.030) -0.006 (0.004) bs -0.708 (0.150)* -0.044 (0.034) bods -1.136 (0.174)* -0.129 (0.060)** ceod 0.062 (0.321) 0.202 (0.063)* journal of applied economics and business studies, volume. 5, issue 2 (2021) 69-86 https://doi.org/10.34260/jaebs.525 83 wti 0.060 (0.034)*** 0.005 (0.002)* liw -1.277 (0.498)** -0.081 (0.067) os 0.019 (0.015) 0.001 (0.002) csr -3.156 (0.797)* -0.032 (0.067) smd 0.016 (0.023) 0.008 (0.003)* dd 0.030 (0.005)* -0.001 (0.001) c 12.632 (3.689)* 1.583 (0.645)* r-squared 0.825 0.223 adj. r-squared 0.728 0.068 durbin-watson stat 2.233 1.973 j-statistic 6.116 3.780 prob (j-statistic) 0.106 0.437 ar (1) 0.745 0.371 ar (2) 0.317 0.072 note: *,**,*** at 1%, 5%, 10% respectively. the figure in parenthesis reports the standard error. return on assets (roa) and technical efficiency (te) are used as dependent variables. r-squared is 82% and 22% in the case of roa and te. this implies that 82% and 22% variation in our models arises due to explanatory variables and the p-value in both cases is also less than 0.05, this implies that the model is statistically fit and significant. the value of durbin watson in both models indicates the presence of autocorrelation. as the p-value of j-statistics is greater than 0.10, this implies the validity of the instruments used in the study. similarly, the p-value of ar (2) is insignificant and greater than 0.05, this hails the absence of auto-correlation in both models. bank specific factors and performance this study finds a significant negative relationship between non-performing loans, advances, and bank size with roa. however, deposits influence roa positively and significantly. corporate governance and performance board size positively and significantly influence profitability and efficiency. this is because a larger board possesses versatile knowledge, skills to make better decisions. the scholars that are in favor of this argument include herdjiono and sari (2017); kiel and nicholson (2003); riyadh et al. (2019). ceo duality positively influences technical efficiency. this is in line with stewardship theory. the system gmm results are also matched with panel estimation techniques. this study showed a significant positive relationship of the percentage of women on the board with roa and technical efficiency, however, an inverse relationship of the level of involvement of hafiz muhammad athar and sumayya chughtai 84 women into other committees with roa. the result of system gmm matched with the findings of the panel estimation technique as well. corporate social responsibility disclosure index and performance the study finds a significant negative relationship of the csr index with roa. this is in line with agency theory and matched with the results of the panel estimation technique. 5. conclusion and future recommendations the crux is to examine the impact of bank-specific factors, governance, and environmental factors on bank performance by employing panel estimation techniques and gmm. the findings contend a negative relationship of non-performing loans, advances, level of involvement of women into other committees, csr index with roa; while the percentage of women on board find a positive and significant relationship with roa. however, the study reveals a positive relationship between ceo duality and technical efficiency. the study finds an inverse relation of bank size with roa and technical efficiency. it was further observed that deposit influences roa positively; while board size finds a positive and significant relationship with roa and technical efficiency. the findings of the study have a greater implication for various stakeholders as it helps them to understand that how these factors influence the efficiency and profitability of the banking sector. specifically, it helps the policymakers to understand that how the demographic dynamics, policies regarding asset-financing structure, capitalization levels, borrowers selection, governance, gender diversity, disclosure protocols, up to what extent the banks are involved in social investment (food, education, health, shelter), poverty reduction, sustainable and green financing, charities, etc. affecting bank performance and formulate their policies accordingly. the authors recommend future researchers do similar research in south asian countries by employing different techniques to estimate the relationship. secondly, to include more variables in governance such as audit committees, risk management, board meetings, etc. thirdly, to examine the moderating effect of institutional factors references adnan, m.a., htay, s.n., ab. rashid, h.m. and meera, a.k. (2011). a panel data analysis on the relationship between corporate governance and bank efficiency, journal of accounting, finance and economics, 1 (1), pp. 1-15. andersson, p. and wallgren, f.m. (2018). board gender diversity and firm financial performance: a study of 100 companies listed on nasdaq stockholm. athanasoglou, p., and brissimis, s.n., & delis, m.d. (2008). bank-specific, industryspecific and macroeconomic determinants of bank profitability. journal of international financial markets, institutions and money, 18(2), 121–136. journal of applied economics and business studies, volume. 5, issue 2 (2021) 69-86 https://doi.org/10.34260/jaebs.525 85 aziz, o.g., & knutsen, j. (2019). the banks profitability and economic freedom quality: empirical evidence from arab economies. journal of banking and financial economics, 1(11), 96–110. batir, t.e., volkman, d.a., & gungor, b. (2017). determinants of bank efficiency in turkey: participation banks versus conventional banks. borsa istanbul review, 17(2), 86-96. beate, e., & gro, l. (2010). women on corporate boards: key influencers or tokens? journal of managerial governance. berger, a. n., deyoung, r., genay, h., & udell, g. f. (2000). globalization of financial institutions: evidence from cross border banking performance. brookings-wharton papers on financial services, 3, 23-158. bitar, m. pukthuanthong, k., & walker, t. (2018). the effect of capital ratios on the risk, efficiency and profitability of banks: evidence from oecd countries. journal of international financial markets, institutions and money, 53, 227-262. bourke, p. (1989). concentration and other determinants of bank profitability in europe, north america and australia. journal of banking and finance, 13(1), 65–79. eichengreen, b., & gibson, h.d. (2001). greek banking at the dawn of the new millennium. cepr discussion paper. fahad, p. and busru, s.a. (2021). csr disclosure and firm performance: evidence from an emerging market. corporate governance: international journal of business in society freeman, r.e. (1984). strategic management: a stakeholder approach. pitman, marshfield. friedman, m. (1970). the social responsibility of business is to increase its profits. n. y. times mag., 32–33. goddard, j., liu, h., molyneux, p. and wilson, j.o.s. (2013). do bank profits converge? european financial management,19, pp. 345–65. gitau, p.m., anyango, w., & rotich, g. (2017). effect of micro economic factors on performance of listed commercial banks in kenya. the strategic journal of business and change management, 4 (2), 763-778. hasanul, b., rubi, a., & eric, h.y.k. (2017). determinants of commercial banks’ efficiency in bangladesh: does crisis matter? journal of asian finance, economics and business, 4 (3), 19-26. havrylchyk, o. (2006). efficiency of the polish banking industry: foreign versus domestic banks. journal of banking and finance, 30. herdjiono, i., and sari, i. m. (2017). the effect of corporate governance on the performance of a company. some empirical findings from indonesia. journal of management and business administration, 25(1), 33-523 iveta, r. (2015). banking efficiency determinants in the czech banking sector. procedia economics and finance, 23, pp. 191 – 196. jayati, s. and subrata, s. (2018). bank ownership, board characteristics and performance: evidence from commercial banks in india. international journal of financial studies. hafiz muhammad athar and sumayya chughtai 86 kiel, g.c., & nicholson, g.j. (2003). board composition and corporate performance: how the australian experience informs contrasting theories of corporate governance. corporate governance, 11 (2), 189-205. krause, r., semadeni, m., & cannella, a.a. (2014). ceo duality: a review and research agenda. journal of management, 40 (1), 256–286. majeed, m.t., & zanib, a. (2016). efficiency analysis of islamic banks in pakistan. humanomics, 32 (1), 19-32. maqbool, s., & zameer, m.n. (2018). corporate social responsibility and financial performance: an empirical analysis of indian banks. future business journal, 4, 84–93. molyneux, p. and thornton, j. (1992). determinants of european bank profitability: a note. journal of banking and finance, 16(6), pp. 1173–1178. mravlja, l. (2017). the impact of corporate social responsibility of banks on their financial performance. oyewumi, o.r., ogunmeru, o.a., & oboh, c.s. (2018). investment in corporate social responsibility, disclosure practices, and financial performance of banks in nigeria, future business journal, 4, 195-205. riyadh, h.a., sukoharsono, e.g. and alfaiza, s.a. (2019). the impact of corporate social responsibility disclosure and board characteristics on corporate performance, cogent business & management, 6(1). sealey, c. w., & lindley, j. t. (1977). inputs, outputs and theory of production cost at depository financial institutions. journal of finance, 32, 1251–1266. sufian, f., & habibullah, m.s. (2010). developments in the efficiency of the thailand banking sector: a dea approach, international journal of development issues, 9(3), 226-245. sufian, f., kamarudin, f., & nassir, a. (2016). determinants of efficiency in the malaysian banking sector: does bank origins matter? intellectual economics, 10, 38-54. syafri, m. (2012). factors affecting bank profitability in indonesia. international conference on business and management, 236-242. szegedi, k., khan, y. and lentner, c. (2020). corporate social responsibility and financial performance: evidence from pakistani listed banks, sustainability, 12. torchia, m., calabrò, a., & huse, m. (2011). women directors on corporate boards: from tokenism to critical mass. journal of business ethics, 102(2), 299-317. varnita, s., niladri, d. and jamini, k.p. (2018). women on boards in india: a need or tokenism? management decision. journal of applied economics and business studies, volume. 5, issue 3 (2021) 1-18 https://doi.org/10.34260/jaebs.531 1 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x flourishing the higher education in pakistan: an exploratory analysis of the role of higher education commission (hec) bashir khan1*,ghulam mustafa 2 and ahmad nawaz 3 1 phd scholar, school of economics, quaid-i-azam university, islamabad 2 phd scholar, quaid-i-azam university, islamabad 3 assistant professor, department of economics, comsat university, lahore (campus) abstract the paramount aim of the underlying study is to explore the efficacy of the higher education commission (hec) through evaluating the effectiveness of the key initiatives which have been taken by the hec to flourish the higher education and research environment in pakistan. using exploratory analysis, the study unleashes that the higher education and research culture has improved since 2002 owing to the hec’s effective initiatives. these include, but are not limited to, increasing the capacity of in-service teaching faculty, induction of phd faculty in public and private sector universities, award of indigenous and foreign scholarships, and provision of research grants, all of which had a positive influence on higher education through capacity building. a commendable increase in the number of universities along with the induction of highly qualified faculty has been witnessed, specifically the induction of indigenously produced phds. apart from these, the hec has played an instrumental role in helping to improve the gender parity by 45% in education sector. moreover, in order to trace out what problems the hec has been facing to implement its reforms, we conducted kiis to the hec officials who have been remained the part of policy implementation. the kiis discloses that the deteriorating budgetary allocation, less cooperation from public sector universities, and questioning the hec’s autonomy are the major challenges among others to the hec in promoting an effective and inclusive higher keywords hec, educational research and development, kiis, exploratory analysis jel classification a2, a22 * bkhan@hec.gov.pk bashir khan, ghulam mustafa and ahmad nawaz 2 education in pakistan. the kiis suggest that the autonomy of the hec should not be compromised, and the liaison between universities and the hec should be increased. 1. introduction education is considered as one of the key determinants of the quality of human capital. specifically, higher education results in boosting up the research and development activity, which ultimately sets nations on higher economic growth and development trajectory. higher level of innovations through creating a research environment, transforms countries into knowledge economies (sukhrev, 2020; marginson, 2010; brinkley, 2006). the higher education is supposed to be played significant a pivotal role in shaping the knowledge economy of any country. the existing literature highlights that knowledge-based economies, or alternatively countries with quality human capital, have experienced higher level of economic growth and development. therefore, improvement in quality of higher education always translates in better quality of human capital, thus, contributes significantly to economic growth (vidotto et al., 2017; lutz and samir, 2011; durrani and forbes, 2003; lucas, 1993). despite many challenges, higher education has witnessed tremendous growth and development since the inception of higher education commission (hec) in pakistan (qazi et al 2014). the hec ordinance was passed in 2002 by government of pakistan to promote higher education in pakistan. in addition, the commission was empowered to formulate policies, guidelines, and priorities for universities in order to advocate, and publicize the need for tertiary education in the country. the commission has been granted complete authority and power to reform pakistan's higher education system. according to hec (2015-16), hec has made significant achievements in the form of increasing enrollment, hiring well-qualified faculty, increasing the research and educational collaborations at national and international level, improving the ranking positions of universities at both national and international levels, and producing quality human capital by providing indigenous and foreign scholarships to the deserving students. hec has taken various initiatives to revive teaching, curriculum, research, and the quality of higher education. the most important initiatives taken, are introduction of tenure track system, interim placement for fresh phds (ipfp), international research support initiative program (irsip),national research program for universities (nrpu), local challenge fund (lcf), technology transfer support fund (ttsf), grand challenge fund (gcf), innovative & collaborative research grant (icrg), technology development fund (tdf), rapid research & innovation initiative (rrii), problem based applied interdisciplinary research program (pbairp), outstanding research awards, travel grants for presentation of research papers, hec digital library, the establishment of new heis, establishment of https://www.hec.gov.pk/english/services/faculty/icrg/pages/default.aspx https://www.hec.gov.pk/english/services/faculty/pbairp/pages/default.aspx https://www.hec.gov.pk/english/services/faculty/pbairp/pages/default.aspx journal of applied economics and business studies, volume. 5, issue 3 (2021) 1-18 https://doi.org/10.34260/jaebs.531 3 quality enhancement cells (qecs), establishment of the office of research innovation and commercialization (orics), establishment of business incubation centers (bics), and initiation of the indigenous and foreign scholarship schemes (hec, 2015-16 and 2019). capitalizing on the above-mentioned diversified initiatives, hec has changed the paradigm of higher education in pakistan by promoting access, quality, and relevance of quality education. those initiatives have eventually created a mammoth impact on the quality of higher education sector in respect of access to higher education, teaching, curriculum, research, and quality of higher education (jahangir, k. 2008; hec, 201516). however, hec is also facing numerous challenges such as limited budgetary allocation, obstacles related to institutional autonomy, and lack of cooperation from some big public sector universities. specifically, expansion in financial capacity is one of the most significant demands by hec in order to implement its effective policy agenda regarding flourishing higher education. the commission demands higher budgetary allocation from every incumbent government of pakistan. for instance, federal hec demands a budgetary allocation of pkr 120.00 billion, but only pkr 66.25 billion are allocated for the fiscal year of 2021-22 by federal government of pakistan. figure-1 demonstrates that on average, hec budget ranges between pkr 63 billion to pkr 66 billion during the fiscal years 2017-18 to 2021-22 (government of pakistan). these estimates unleash the gap of around pkr 53.75 billion (45%) to meet the demand of hec to implement the effective higher education agenda in pakistan. figure-1: budgetary demand and allocation to hec by federal government of pakistan source: hec 2021, government of pakistan 72.00 82.00 103.55 104.79 120.00 63.18 65.00 64.10 64.10 66.25 2017-18 2018-19 2019-20 2020-21 2021-22 12% 21% 38% 39% 45% grant demanded and recurring grant allocation recurring grant recurring grant allocated percentage recurring gap bashir khan, ghulam mustafa and ahmad nawaz 4 given these limitations, the role of hec is inevitable in promoting the higher education in the country. consequently, the underlying study aims to conduct an exploratory analysis of the achievements of hec since it came into existence in 2002. primarily, the focus of this study is to (i) identify the key reforms taken by hec to flourish higher education and research environment in pakistan, (ii) to highlight the achievement made by the commission in this regard, and iii) to explore the key challenges hec is facing currently through in-depth interviews of key informants. 2. data and methodology primarily, the underlying research is exploratory in nature. for descriptive analysis the data is collected from both primary and secondary sources. we have utilized two types of data: secondary data, and primary data. the former is extracted from various relevant reports published by hec, and government of pakistan about the range of indicators which includes information of hec reforms, higher education enrollment, establishment of higher education institutions (heis), faculty hiring, phd produced, and some other indicators which measure the achievement of hec. the data has further been presented through graphs and tables to depict trends over the years and to draw some meaningful conclusions. in addition to secondary data, primary data is gathered through conducting in-depth key informant interviews (kiis) to the hec senior officials. the objective is to ascertain the nature of problems and bottlenecks hec is facing in implementing its policy agenda and to track the performance of hec in achieving its targets. all our 13 respondents are from hec have been playing active roles in policy making and they belong to all the four provinces i.e., punjab, sindh, khyber pakhtunkhwa, and baluchistan1. in line with the objectives of the study questions have been asked and verified through a checklist after detail discussion with each of the respondent. the checklist includes three sections: i) questions related to hec goals and their accomplishment, ii) challenges hec is facing, and iii) questions regarding the effectiveness of tenure track system (tts), mode of hiring, policy. therefore, first two sections of checklist are specific to the objectives of the study. such questions are related to the major challenges, hec is facing and manner of dealing with it, the role of hec in developing linkages of universities with industry, hec role in placement of phd degree holders, the measures hec is taking to improve the quality of education and research, and issues regarding budgetary allocation to hec. last but not the least, questions regarding targets set by hec and their fulfillment have also been asked. in a nutshell, the overarching objective 1 advisors, director generals (dgs), directors, deputy directors (dds), and assistant directors of higher education commission (hec). journal of applied economics and business studies, volume. 5, issue 3 (2021) 1-18 https://doi.org/10.34260/jaebs.531 5 of the interviews is to critically investigate the role of hec in implementing an effective higher education policy agenda in pakistan to improve the quality of education. therefor keeping in view, the objectives of this study we have provided a situational analysis related to reforms by hec and its achievements over the years, later conclusions have been drawn from the interviews to hec officials. 3. reforms by hec and its achievements: situational analysis this section conducts the situational analysis regarding what types of reforms hec has implemented and their accomplishment. the situational analysis has been conducted through demonstration of graphics and tables. 3.1 key reforms by higher education commission (hec) hec has formulated multiple reforms to expand higher education and research environment in pakistan. some notable reforms are outlined as follows. 3.1.1 faculty development programs human resource development is the most important component of the hec reform process. very soon after its formulation, hec has made vital and significant progress in order to launch faculty development programs with twofold objectives of increasing institutional capacity and increasing local research activities. this also aimed to improve the academic qualification of the university teachers (hec, 2015-16; naqvi, 2010). initially, for 60 public sector universities, it was targeted that each university must have faculty members of 300 to 400 phds, while at department level at least 15 to 20 phds before it can be considered as a legitimate "university”. to develop a qualified cadre of bright young women and men, additional 15000 to 20,000 individuals were to graduate from foreign universities in the subjects of national priority over the next five years to carry out research and teaching activities in universities. the group was also supposed to provide consultancy to boost up industry in the country (hec, 2003-04). moreover, hec has adopted the reforms in hiring the teachers for public universities against the existing system of basic pay scale (bps). the new system is termed as tenure track system (tts) which provides additional incentives to the existing and newly being hired teachers. the primary objective of the tts is to promote educational and research productivity by believing in classical incentive theory. figure-2 demonstrates that 3410 phd teaching faculty members are employed as tts faculty while 8456 teachers are working under bps, mode of hiring in public universities. similarly, 3551 phd faculty are hired by private universities, while 508 faculty members are working on contractual basis (hec, 2020). bashir khan, ghulam mustafa and ahmad nawaz 6 figure-2: number of phd faculty in public universities by mode of hiring source: (hec, 2020) such faculty development programs have positive impacts on the induction of phd faculty in public universities. as table-1 demonstrates, out of 7840 faculty members in higher education institutions (heis), there were only 2281 phd in 2002-03, making it only 29.09% of total faculty members. table-1: total number of phd and non-phd faculty in higher education institutions (heis) year ph.d. faculty non-phd faculty total full time faculty % phd faculty 2018-19 16478 35016 51494 32.00 2017-18 14507 32683 47190 30.74 2016-17 13975 33854 47829 29.22 2015-16 11960 31314 43274 27.64 2014-15 10133 27295 37428 27.07 2013-14 9475 25121 34596 27.39 2007-08 3,683 13,069 16,752 22.00 2002-03 2,281 5,559 7,840 29.09 source: (hec, 2020) table-1 further demonstrates that after the establishment of hec in 2002, a significant increase in number of phds faculty has been witnessed by heis. similarly, overall increase in total full-time faculty is also experienced by heis from 2002-03 to 2018-19. in 2020, there is 32% phd faculty in universities and degree awarding institutions in the country, this shows the shortage of phds in universities. the target in mtdf-ii was to have 40% of phds in heis by the end of 2015. heis produces phds at a rate of 1600 per year (hec, 2020). kiis have informed that phd faculty has increased enormously due to various reforms opted by hec. however, the desired level is yet to achieve due to i) constrained level of funding, ii) brain drain at local and 8456 3410 3551 508 bps tts private contractual journal of applied economics and business studies, volume. 5, issue 3 (2021) 1-18 https://doi.org/10.34260/jaebs.531 7 international level, iii) the over usage of non-phds visiting faculty in private universities, and iv) non-availability of required level of phds in specific subjects. 3.1.2 local and international scholarships schemes to fulfill the need of phds in universities, hec has started different indigenous and foreign scholarship schemes. under indigenous scholarship scheme, over 5000 scholarships have been awarded with special emphasize on the quality of phds. the objective of this scheme was to increase the phd faculty in universities. this mega project was approved by executive committee of the national economic council (ecnec) on august 11, 2003. the total amount of the project was rs.6402.767 million. a total of 5000 scholarships were awarded in five years. almost every area of this endeavor, including science & technology, humanities, and life sciences, were addressed (hec, 2020). the list of local and foreign scholarships is presented in table 2. table-2: list of scholarships offered by hec local scholarship • 5000 indigenous phd fellowships • development of science & technology manpower through indigenous 300 phd scholarships • merit scholarship scheme for phd students in science and technology (200 scholarships) • scholarships for phd studies in social sciences, arts and humanities for university teachers • five information technology related hrd scholarship programs/projects foreign scholarships • overseas scholarship scheme for phd in selected fields • phd scholarships for sciences and engineering in germany • phd scholarships in science, technology and engineering in austria (phase-i) • phd scholarships in natural and basic sciences, austria (phase-ii) • phd in natural & basic sciences in france • phd scholarships for engineering sciences, china • phd scholarships for basic sciences, china • development of high-level science &technology manpower through split phd program • post-doctoral fellowship program for university teachers • ms/mphil leading to phd scholarship in engineering, natural and basic sciences/humanities/social sciences for the teachers of weaker universities • partial support for phd studies abroad • hrd program for strengthening of universities/institutes of higher learning • strengthening the existing teaching faculties of public universities and degree awarding institutes of pakistan through jointly sponsored scholarships program of hec & asian institute of technology (thailand) • short term foreign faculty hiring program / former expatriate faculty hiring program • reclamation of talented pakistanis working for promotion of teaching and research in professional universities (reverse brain drain) • visiting scholar program • transfer of knowledge through expatriate national program source: (hec, 2020) bashir khan, ghulam mustafa and ahmad nawaz 8 according to hec (2020), nineteen scholarships projects are successively implemented by higher education commission to increase the phd faculty in universities. there are nine foreign scholarships and fellowship programs and ten indigenous programs. the hec started a variety of scholarships programs both outside and within the country to address the shortage of faculty in universities. 3.1.3 in-service training programs since its establishment in 2002, hec has been making every attempt to foster academic vitality by offering educational and in-service training programs which maintains updating the skill level of university faculty. the empirical and theoretical literature is also suggestive of the importance of the training and skill development programs for employees (mcconnell et al, 2009). therefore, in order to update the skill level of university faculty, hec has launched in-service training programs. national academy of higher education (nahe) project was launched by hec to enhance standard of teaching and learning in universities of pakistan. nahe organized trainings, workshops and organized national conferences for the university faculty in developing high quality, efficient, relevant teaching, and learning methods. moreover, faculty development programs have been commenced in order to enhance capacity of faculty members. the faculty development programs contain such modules: i) professional development, ii) research methodologies, iii) testing & assessment, iii) curriculum planning & development, and iv) teaching & communication skills, v) international computer driving license, and vi) instructional resources. these programs have indicated beneficial influences on updating the skills of the faculty members of universities (hec, 2020). apart from these, hec initiated the english language teaching reforms initiative (eltr) in 2004 in order to improve the quality of teaching and learning at universities. for the english and social sciences faculty of colleges and universities, this was the first ever unique language-based program. this reform aimed at capacity building of english language teachers in colleges and universities in pakistan for sustainable development. two phases were completed under this initiative of hec. phase i catered 1398 faculty of colleges and universities. after completion of eltr phase i, hec launched phase ii in 2010. the initial target was 1400 english faculty to be trained through short-term and long-term courses. besides, many seminars and workshops have been arranged for faculty member of universities to align them with the latest development in teaching, pedagogical skills, and other contemporary issues (hec, 2020). in sum, above discussion has weaved up that hec has formulated multiple reforms to enhance the enrollment, producing phds, research environment among students and faculty members through launching in-service training programs, local and international scholarships, and faculty development programs. the most notable program regarding journal of applied economics and business studies, volume. 5, issue 3 (2021) 1-18 https://doi.org/10.34260/jaebs.531 9 faculty development is the implementation of tts mode of hiring against prevailing bps in public universities. 3.2 capacity building in higher education institutions (heis) and the hec after discussing the prominent reforms promulgated by hec, the underlying research moves ahead to discuss the impacts of such reforms on the capacity building of heis in term of their numbers, enrollment rates, and phds graduates. 3.2.1 hec and higher education institutions (heis) according to hec (2003), in 1947, pakistan had only one university, the university of the punjab, with the student enrollment of 644. there was no private university at that time. it was the fourth university to be established by the british in subcontinent. the british rulers established the first three universities at their strongholds of bombay, calcutta and madras. after war of independence in 1857, the university of the punjab came into existence as result of a long-drawn struggle of the educators. the second university was the university of sindh, which was established in 1951. in 1960, there were only 5 heis operating in the country. table-3: number of universities and degree awarding institutions (dais): pre-and-post hec year universities dais total public private public private before hec 1947 1 0 0 0 1 1960 4 0 1 0 5 1980 15 0 2 0 17 1990 19 2 3 0 24 2002 29 12 5 5 51 after hec 2010 62 41 11 19 133 2020 118 56 16 28 218 source: hec, pakistan table-3 contains the information of the number of universities and dais before the establishment of hec (before 2002), and after hec (since 2002). after the 2002s, rapid growth has been observed in the development of higher education institutions (heis) across pakistan, furthermore, there has been a significant growth in numbers from 51 in 2002 to 218 in 2020. it is undeniable that pakistan has seen the increase of new institutions since the inception of the higher education commission (hec) in 2002. likewise, figure-3 presents the bifurcation of heis into public and private sector universities in pakistan. trend analysis demonstrates that after the establishment of hec, on the whole an overwhelming increase in both public and private universities bashir khan, ghulam mustafa and ahmad nawaz 10 has been witnessed during 2002-03 to 2019-20. the public sector universities are higher than the private sector universities. however, these figures demonstrate the positive impacts of hec policies to flourish higher education in pakistan. figure-3: number of heis chartered from 1947-48 to 2019-2020 source: (hec, 2020) the underlying research identifies the construction of heis by provinces and other parts of the country in order to estimate the regional differences. figure-4 highlights that punjab province seems at top of the list wherein 73 heis have been established. the analysis for subsequent regions displays that sindh province has 62 heis while there are khyber pakhtunkhwa (41), islamabad (22), balochistan (10), azad jammu and kashmir (7) and gilgit-baltistan (2) number of universities and degree awarding institutions have been established respectively (figure 4). figure-4: public and private sector heis at sub-national level in pakistan during 2019-20 source:(hec, 2020) in conclusion, this section determines the positive impacts of hec policies on construction of heis in pakistan which are precursor to boost up the higher education in pakistan. 0 50 100 150 200 250 1 9 4 7 -4 8 1 9 4 9 -5 0 1 9 5 6 -5 7 1 9 5 8 -5 9 1 9 6 0 -6 1 1 9 6 2 -6 3 1 9 6 4 -6 5 1 9 6 6 -6 7 1 9 6 8 -6 9 1 9 7 0 -7 1 1 9 7 2 -7 3 1 9 7 4 -7 5 1 9 7 6 -7 7 1 9 7 8 -7 9 1 9 8 0 -8 1 1 9 8 2 -8 3 1 9 8 4 -8 5 1 9 8 6 -8 7 1 9 8 8 -8 9 1 9 9 0 -9 1 1 9 9 2 -9 3 1 9 9 4 -9 5 1 9 9 6 -9 7 1 9 9 9 -0 0 2 0 0 1 -0 2 2 0 0 3 -0 4 2 0 0 5 -0 6 2 0 0 7 -0 8 2 0 0 9 -1 0 2 0 1 1 -1 2 2 0 1 3 -1 4 2 0 1 5 -1 6 2 0 1 7 -1 8 2 0 1 9 -2 0 public total private total total 5 9 15 2 30 46 28 2 1 7 11 28 34 7 10 22 2 41 74 62 aj&k balochistan federal gb kp punjab sindh public private total journal of applied economics and business studies, volume. 5, issue 3 (2021) 1-18 https://doi.org/10.34260/jaebs.531 11 3.2.2 hec and higher education enrollment a country with a population of approximately 220 million people is the sixth largest in the world. pakistan has limited natural resources and its economy is struggling. it has 64 percent of its population comprised of young population i.e their age is less than 30 and 29 percent of its population are between 15 and 29 years of age, and this is the age group which is called youth. currently, pakistan is the country which has the youngest population in the world and this trend is forecasted to continue to till 2050. therefore, providing proper opportunities through education to youth to become an asset for a nation is the most desirable objective to the policy makers in pakistan (undp, 2018). hec aims to provide and increase opportunities of equitable access for gender balanced and quality higher education to a larger part of the youth 17-23 years old to make them able to participate in the development of a country. the foundation of the agenda of economic development is to invest in society to polish their cognitive skills, talents, and enhance their constructive competencies (hec, 2014). higher education sector of pakistan has shown satisfactory progress in term of access to higher education. in 2002, only 2.5% access to higher education of aged 17-23 years, while hec has steered numerical expansion of universities and degree awarding institutions from 51 heis in 2002 to 218 in 2020, increased enrollment 0.028 million in 2002 to 1.93 million in 2020, increased gender parity from 37% in 2002 to 45% in 2020 (hec, 2020). hence, table-4 presented annual increase in enrollment in public and private sector heis in pakistan during 2001-02 to 2019-20. enrollment of affiliated colleges is not included. the increase in enrollment over the years represents the demand in pakistan for higher education. despite its rapid growth, pakistan lags behind india, bangladesh, and sri lanka in terms of higher education access. to improve access to higher education in pakistan, more practical and calculated efforts are needed (hec, 2020). table-4: university level enrolment in pakistan (million) year male female total 2001-02 0.17 0.10 0.27 2002-03 0.20 0.13 0.33 2003-04 0.24 0.18 0.42 2009-10 0.52 0.43 0.95 2011-12 0.56 0.48 1.04 2012-13 0.62 0.52 1.14 2013-14 0.68 0.56 1.24 2014-15 0.72 0.58 1.30 2015-16 0.76 0.64 1.39 2016-17 0.84 0.65 1.49 2017-18 0.88 0.70 1.58 2018-19 1.03 0.83 1.86 2019-20* 1.07 0.86 1.93 source: (hec, 2020) bashir khan, ghulam mustafa and ahmad nawaz 12 table-5 presents percentage distribution of university enrollment by region, university sector, and gender wise. it demonstrates that, in public and private sector universities, female students and students from balochistan, khyber-pakhtunkhwa, and sindh are underrepresented. the estimates highlight those female students from balochistan in private and public institutions is 20% and 31% respectively. similarly, 21% and 27%, respectively, are female students from khyber-pakhtunkhwa in private and public universities, while sindh registered female students by 34% and 39% in public and private universities respectively. thus, in these regions, female students are severely underrepresented and access to higher education is a major concern. in punjab, female enrollment in private and public universities is 38% and 51%, respectively, while female students in islamabad are observed by 41% and 50% in public and private universities respectively. likewise, in azad jammu and kashmir, female enrollment is 55% and 52% respectively (table 5). two public sector universities run in gilgit baltistan, 50% male students and 50% female students are registered in these universities (hec, 2019). table-5: (%) distribution of university enrolment by region, sector, and gender during 201819 province/region public private total male female male female male female ajk 48 52 45 55 48 52 balochistan 69 31 80 20 70 30 federal 50 50 59 41 50 50 gilgit baltistan 50 50 0 0 50 50 khyber pakhtunkhwa 73 27 79 21 75 25 punjab 49 51 62 38 53 47 sindh 61 39 66 34 63 37 total 53 47 65 35 55 45 source: (hec, 2020) figure-5 presents enrollment from 1971 to 2020, the graph shows sharp increase in enrollment after inception of hec. it is evident that hec is succeeded in refurbishing enrollment in heis. despite of these reforms, positive shocks are needed to enhance enrollment to compete with the world in general and with south asian countries in particular. journal of applied economics and business studies, volume. 5, issue 3 (2021) 1-18 https://doi.org/10.34260/jaebs.531 13 figure-5: enrollment in pakistani heis by gender over the years source: (hec, 2020) 3.2.3 producing phds the hec policies have brought a significant increase in the university faculty in general and indigenously produced phd graduates in particular. moreover, the impact of different hec scholarships schemes implemented during last two decades is evidently demonstrating a sizeable increase in phd graduates. figure-6 exhibits that during 200304,271 phds have been produced which have reached 1778 during the year of 2019-20. the value for 2020-21 seems declining which has justification— to date data is being complied for analysis, hec has not received complete data. but the data for 2020-21 is expected to be increased for rest of the year (government of pakistan, 2020-21). the linear trend line helps comprehend that after 2008-09, an overwhelming increasing trend for phd producing is evidently observed which could be attributed to increase in budgetary allocation of hec as compared to the previous years. -500 0 500 1000 1500 2000 2500 1 9 7 0 -7 1 1 9 7 1 -7 2 1 9 7 3 -7 4 1 9 7 4 -7 5 1 9 7 5 -7 6 1 9 7 6 -7 7 1 9 7 7 -7 8 1 9 7 9 -8 0 1 9 8 0 -8 1 1 9 8 1 -8 2 1 9 8 6 -8 7 1 9 8 7 -8 8 1 9 8 8 -8 9 1 9 8 9 -9 0 2 0 0 1 -0 2 2 0 0 2 -0 3 2 0 0 3 -0 4 2 0 0 4 -0 5 2 0 0 5 -0 6 2 0 0 6 -0 7 2 0 0 7 -0 8 2 0 0 8 -0 9 2 0 0 9 -1 0 2 0 1 1 -1 2 2 0 1 2 -1 3 2 0 1 3 -1 4 2 0 1 4 -1 5 2 0 1 5 -1 6 2 0 1 6 -1 7 2 0 1 7 -1 8 2 0 1 8 -1 9 2 0 1 9 -2 0 s t u d e n t s ( 0 0 0 ) total male female linear (total) bashir khan, ghulam mustafa and ahmad nawaz 14 figure-6: number of phd produced in pakistan during 2000-01 to 2020-21 source: pakistan economic survey (2020-21) & hec (2020) table 6 gives detail about discipline wise phd produced in the country since 1947 to date. it demonstrates that 4866 and 4744 phds produced in physical sciences and social sciences disciplines respectively. similarly, 4468 phd produced in biological and medical sciences and pharmaceutics disciplines. whereas only 1053 phd in the discipline of engineering & technology has been produced, along with 789 phd in management sciences and business education. therefore, number of phd graduates in the field of engineering & technology and management sciences need to be increased to cater to the high market demand in the country table-6: discipline wise phds produced (p) by pakistani heis since 1947 disciplines female male grand total agriculture and veterinary sciences 363 2044 2407 arts & humanities 731 1645 2376 biological & medical sciences, pharmaceutics 2045 2423 4468 engineering &technology 133 920 1053 honorary 4 125 129 management sciences and business education 180 609 789 physical sciences 1579 3287 4866 social sciences 1444 3300 4744 grand total* 6479 14353 *20832 source: (hec, 2020) 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2 0 0 0 -0 1 2 0 0 1 -0 2 2 0 0 2 -0 3 2 0 0 3 -0 4 2 0 0 4 -0 5 2 0 0 5 -0 6 2 0 0 6 -0 7 2 0 0 7 -0 8 2 0 0 8 -0 9 2 0 0 9 -1 0 2 0 1 0 -1 1 2 0 1 1 -1 2 2 0 1 2 -1 3 2 0 1 3 -1 4 2 0 1 4 -1 5 2 0 1 5 -1 6 2 0 1 6 -1 7 2 0 1 7 -1 8 2 0 1 8 -1 9 2 0 1 9 -2 0 2 0 2 0 -2 1 phd produced journal of applied economics and business studies, volume. 5, issue 3 (2021) 1-18 https://doi.org/10.34260/jaebs.531 15 3.3. research and development hec has launched a series of research projects to promote the research culture in pakistan. the promotion of university teaching faculty is linked with research paper publications which instigate the faculty to enhance their research activities. despite this, at least one research paper publication has been made compulsory for phd scholars before award of phd degree. as a result, number of research publications have been increased since the year of 2002 to onward (hec, 2020). moreover, hec has launched some programs related to research grants. in this regard, the key initiatives are outlined as: national research program for universities (nrpu), pak-france peridot research program which for pakistani researchers and phd students visit france and carry out research in the universities in france, scientific instrumentation, business incubation centers (bics), oric, technology development fund, cpeccollaborative research grant, pak-turk researchers mobility grant program, grand challenge fund, local challenge fund, innovative & collaborative research grant (icrg), and problem based applied interdisciplinary research program (pbairp). these programs have created competitive environment in universities. thousands of research proposals received to hec, this show demands for funding and competitive research culture in universities (hec, 2020). 4. what hec thinks: discussion on interviews of hec officials previous section comprises discussion on situational analysis. so, to learn the reasons of success and failures to achieve some goals, we have conducted the key informant interviews (kiis) to the hec officials. following are the main outcomes we have obtained from kiis. 4.1 hec targets and its achievements hec take on access to higher education: the interviewed hec officials discloses that 2.329 million enrollment targets have been set till 2020 in hec vision 2025, but this target is not achieved, the actual enrollment during 2020 has been observed 1.942 million, which demonstrates the failure to achieve the targeted level. the respondents have informed multiple reasons not to achieve the desired level of enrollment: i) every incumbent government intends to support hec, but due to financial and economic constraints governments are found fail to support hec what governments have been pledging, ii) they have further unleashed that hec funds are static over the past few years, and iii)the other reason not to maintain the targeted level of enrollment is due to the failure of establishing new universities and sub campuses at required level, and iv) students’ affordability. quality of education in universities: the hec respondents are admitting the concerns related to quality enhancement. they disclosed the reasons of deteriorating quality of education: i) some public universities are not fully cooperating with hec, ii) most of the universities do not implement the hec policies regarding quality enhancement since bashir khan, ghulam mustafa and ahmad nawaz 16 hec is an advisory body not a regulatory body, and iii) governance and leadership issues in universities which are also affecting the performance of universities. faculty development: in various documents of hec, it is emphasized, on the enhanced phd faculty in universities, in this regard, in mtdf-ii (2010-2015) and hec vision 2025 that 40% phd faculty of total faculty in universities. currently 32% phd faculty are there in universities, which is evidently showing the missing of target. hec officials have informed that phd faculty has increased enormously due to various reforms opted by hec. however, the desired level is yet to achieve due to i) constrained level of funding, ii) brain drain at local and international level, iii) the over usage of non-phds visiting faculty in private universities, and iv) non-availability of required level of phds in specific subjects. 4.2 major challenges to hec the outcomes from key informant interviews (kiis) regarding what challenges hec has been facing in the pursuance of effective policy agenda to boost up the higher education in pakistan. they have revealed the key challenges such as i) insufficient funding allocation to hec, ii) lack of political will to prioritize higher education by incumbent governments, iii) lack of commitment by universities to follow the commissions’ recommendations, iv) apprehensions to the autonomy of hec from political governments which may predicament the commission to implement their effective agenda to boost up higher education in the country. 4.4 university-industry linkages and placement of phd degree holders the kiis are asked the role of the hec to link the educational research with industry, they reveal that for this purpose, the commission is making efforts such as: i) assisting the universities to establish office of research, innovation and commercialization (oric) to commercialize the academic research, ii) helping to establish business incubation centers (bics), and iii) expanding the internship related opportunities for students during their study in industries. apart from these, the kiis are designed to ask the steps taken by hec to fix placement of phd degree holders. they have reported that hec is offering interim placement for fresh phds (ipfp) to fix the fresh phd degree holders along with tenure track system (tts). the kiis also discloses that orics, in particular, has played an important role in boosting research activities in universities. 5. concluding remarks since its inception in 2002, hec has implemented its policy agenda to bring reforms in higher education to improve access, relevancy and improve quality of higher education in pakistan. this underlying paper has three specific objectives i.e., i) to identify the key reforms taken by hec to flourish higher education and research environment in pakistan, ii) to highlight the achievement of hec geared towards journal of applied economics and business studies, volume. 5, issue 3 (2021) 1-18 https://doi.org/10.34260/jaebs.531 17 increasing higher education and research activities in pakistan, and iii) to explore challenges hec is facing to implement its higher education policy agenda in pakistan. to that extent, we have employed situational analysis through secondary and primary data. the secondary data includes hec published reports and pakistan economic surveys, while primary data through conducting the key informant interview (kiis) is collected from hec officials who have been playing a pivotal role in policy making. exploratory analysis demonstrates that hec has implemented multiple reforms in order to improve the productivity and quality of university teaching faculty. for that purpose, hec has launched numerous indigenous and international scholarships, training programs, and tts mode of hiring which have triggered increase in phd faculty in public and private sector universities. since the inception of hec, a massive increase in establishment of universities and induction of phd faculty members in rising number of universities has been witnessed. tremendous increase in enrollment rates at university level has been observed from 2002 to 2020 owing to reforms implemented by hec, specifically an increase in females’ enrollment has been notably observed which may be helpful to enhance the gender parity by 45%. on average, 1000 to 1800 phds have been produced annually from 2008-09 to 2019-2020 with the courtesy of multiple indigenous and foreign scholarships by hec. despite these achievements, the commission has initiated multiple programs to foster research environment, and collaboration amongst the local and international researchers. the kiis demonstrate that the commission is failed to achieve the desired outcomes despite multiple successes: i) lack of political will, ii) squeezing budgetary allocation for higher education, iii) lack of cooperation from public universities, and iv) intimidating the autonomy of hec by incumbent political regimes. furthermore, they have recommended that a task force of educational experts, comprising the representatives from all provinces/regions be constituted to devise a comprehensive and systematic future strategy for the higher education in pakistan. references brinkley, i. (2006). defining the knowledge economy. the work foundation, london. durrani, t. s., & forbes, s. m. (2003, november). intellectual capital and technology strategy. in iemc'03 proceedings. managing technologically driven organizations: the human side of innovation and change (pp. 151-155). ieee. david e. bloom, philip g. altbach, and rosovsky, h. (2016). looking back on the lessons of higher education and developing countries: peril and promise” perspectives on china and india. pgda working paper no. 133, http://www.hsph.harvard.edu/pgda/ working/. bashir khan, ghulam mustafa and ahmad nawaz 18 gyimah-brempong, oliver paddison& workie mitiku (2006) “higher education and economic growth in africa, the journal of development studies”, 42:3, 509-529, doi: 10.1080/00220380600576490 hec vision 2025 report 2014. higher education commission unpublished reports: 2002-2020. higher education commission annual reports (2002 to 2016). jahangir, k. (2008). management of higher education reforms in pakistan: an implementation perspective (doctoral dissertation, utrecht university). khan m.m (2010) “issues of access in public and private higher education institutions in islamabad pakistan” open access dissertation, university of massachusetts amherst, majidkhan799@gmail.com. lucas, r. (1988) “on the mechanics of economic development, journal of monetary economics”, 22, pp. 3–42. lutz, w., & samir, k. c. (2011). global human capital: integrating education and population. science, 333(6042), 587-592. liao c,h. (2010) “how to improve research quality? examining the impacts of collaboration intensity and member diversity in collaboration networks” volume 86: issue 3,doi: https://doi.org/10.1007/s11192-010-0309-2 m. kazim bacchus (2008). the role of education in achieving equity, cultural diversity and national unity in multi-ethnic societies. singapore journal of education, 10:2, 110, doi: 10.1080/02188798908547656. mcconnell cr, brue sl, macpherson da. contemporary labor economics. new york: mcgrew-hill.; 2009. marginson, s. (2010). higher education in the global knowledge economy. procedia-social and behavioral sciences, 2(5), 6962-6980. naqvi h. s (2010) “overview of higher education commission support for academia in pakistan” cbm-ci international workshop, karachi, pakistan. osama, a., najam, a., kassim-lakha, s. et al. pakistan's reform experiment. nature 461, 38– 39 (2009). https://doi.org/10.1038/461038a qazi, w., raza, s. a., & jawaid, s. t. (2014). higher education and growth performance of pakistan: evidence from multivariate framework. quality & quantity, 48(3), 16511665. stone, m., horejs, j., & lomas, a. (1997). commonalities and differences in teacher leadership at the elementary, middle, and high school levels. action in teacher education, 19(3), 49–64. sukhdev, a. s. (2020). impact of irrigation on economic development a study of parner tahsil ahmednagar district. the task force: “the task force on higher education and society 2000 higher education in developing countries: peril and promise” washington, the world bank the world bank & unesco report: higher education in developing world: peril and promise, february 2000. undp pakistan report 2018 vidotto, j. d. f., ferenhof, h. a., selig, p. m., & bastos, r. c. (2017). a human capital measurement scale. journal of intellectual capital. https://akjournals.com/view/journals/11192/86/3/article-p747.xml?rskey=coourd&result=1 https://akjournals.com/view/journals/11192/86/3/article-p747.xml?rskey=coourd&result=1 https://akjournals.com/view/journals/11192/86/3/11192.86.issue-3.xml https://akjournals.com/view/journals/11192/86/3/11192.86.issue-3.xml https://doi.org/10.1080/02188798908547656 journal of applied economics and business studies, volume. 6, issue 3 (2022) 23-44 https://doi.org/10.34260/jaebs.632 23 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x female unemployment duration in pakistan – a survival analysis gulrukh zahid1*1 & ambreen fatima 1 m.phil scholar, applied economics research centre (aerc), university of karachi 2 associate professor, applied economics research centre (aerc), university of karachi abstract female unemployment in pakistan is very high in contrast to its neighboring countries. the finding of previously available literature concludes that staying out of the labour market for longer durations causes a lower possibility of getting a job again and for females; the probability is even lower compared to males. this study examines the unemployment situation and its duration, forcing females to drop out of pakistan's labour force. numerous factors are accountable for the higher inactivity rate of females in the labour market. some examples are low education level, higher age, home responsibilities such as taking care of elders and children, spouse job status, marital status, and nonavailability of job-specific knowledge. although in the case of pakistan, literature shows many causes of higher inactivity of female unemployment, it does not focus on the time estimations and calculate it in days or months to figure out how long a female take to get back on a job once she becomes unemployed. more specifically, the aspects which determine unemployment are unavailable. kaplan-meier survival analysis has been performed to assess the research by employing the data on the labour market and socio-economic variables from the labour force survey (lfs) source of 2014-15. this study can help policymakers to figure out the duration of female unemployment, prevalent in the labour market for females in pakistan. keywords female unemployment, survival analysis, kaplan-meier technique jel classification j64, j13, j16, p34 1 *gulrukhzahid@gmail.com gulrukh zahid & ambreen fatima 24 1. introduction one of the fascinating topics for academicians and policymakers is unemployment since it is one of the most distressing issues for developing and developed economies. henceforth it is the favourite topic for labour economists, causing it the most debated matter among them as the strong spur on both micro-level as an influence of impacting the wellbeing of households and on the macro level, as it creates its impact on the aggregate level. the economic term differentiates the self-unemployed and unemployed by identifying that one who quit the job to stay at home to take care of young children or who does it for higher studies is not considered unemployed as they are not actively searching for a job. numerous problems in society have the root cause of long-term unemployment. survival for individuals becomes very difficult as many candidates contest for fewer positions. social pressure and responsibilities make a person get into a severe depression when one fails to find any job, which ultimately pushes them to opt for any way possible to come out of that state. unemployment is equally stressful for both males and females as they spend their life's quality time acquiring skills and education for a prosperous future. unfortunately, they comprehend a vast gap between the opportunities offered and the people willing to get them when they enter the job market. undoubtedly, it is the state's responsibility to take policy measures that create job opportunities and ultimately stimulate growth in the long run. nevertheless, if the state fails to make employment in the economy, it provides foundations for many social problems. pakistan is considered the 6th largest economy globally by its size and its working-age population, and the number count is growing every minute. the statistics taken from the world bank data on development indicators reported that 65.3% was declared as the population under the working-age category for 2015, highlighting that most of its population consists of youth. the average unemployment rate was 5.46% from 1985 to 2015 and peaked in 2002 when unemployment reached 7.8%. journal of applied economics and business studies, volume. 6, issue 3 (2022) 23-44 https://doi.org/10.34260/jaebs.632 25 figure1: female unemployment source: data taken from world bank unemployment among females in pakistan is alarming because the female population is more in number than males in pakistan. following the role model of developed economies, it is high time to encourage the country's females to become part of the labour force. however, female labour force participation has increased over the years. still, it is not as desired compared to pakistan's neighbouring countries. being a developing country, pakistan should increase female participation in the labour force; it is being perceived that unemployment affects females differently than males, especially educated females. moreover, the persistent increase of females in the labour force is cumulating, causing the gap to grow faster, impacting the current employment opportunities. in terms of new arrivals and limited job opportunities, the constantly extending labour market creates a massive gap between jobs and the available labor force. the continuous increase in unemployment with fresh arrivals in the form of new graduates entering into the labour market creates havoc and will be problematic in the long run. 0 1 2 3 4 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 4 1 9 9 6 1 9 9 7 1 9 9 8 1 9 9 9 2 0 0 0 2 0 0 1 2 0 0 2 2 0 0 3 2 0 0 4 2 0 0 5 2 0 0 6 2 0 0 7 2 0 0 8 2 0 0 9 2 0 1 0 2 0 1 1 2 0 1 2 2 0 1 3 2 0 1 4 2 0 1 5 2 0 1 6 2 0 1 7 2 0 1 8 female unemployment gulrukh zahid & ambreen fatima 26 figure 2: female labour force participation in pakistan source: world bank it is witnessed that in pakistan, women stay at home, for which many factors are involved. the primary reason for others is that household responsibilities (child care is included in it) did not allow her to go out to work. other than this, lower-wage compared to males are also the reason for lower labour force participation. spouse job status is also among those causes which do not permit her as it is suggested not to work if husbands earnings are reasonable; these all only a few among all those issues which are responsible for limited participation of females in the labour market(tasnim khan et al. 2009 ahmed et al. 2013, shabir et al. 2015). however, according to the world bank statistics, the women residing in rural areas of pakistan have more share in the labour force than urban areas due to the higher rate of poverty. another cause can be less educated, or people with no education are more adaptable and ready to work than highly qualified or have any education as they prefer to work according to their education. different studies concluded that those with higher education chose to stay unemployed if they failed to acquire a job that matched their qualifications (kahn et al., 2009). the pioneering work of mincer (1962) and cain (1966) on the economic analysis of the labour force participation of females received significant interest. extensive female labour force participation improvement is observed in developing countries. a developing economy like pakistan needs more of its females to join the labour force for the country's rapid growth. if the issue is not appropriately addressed, it will discourage females from pursuing higher education in the future. it is essential to understand the importance of examining the duration of unemployment at the individual level especially long-term unemployment, because it is the journal of applied economics and business studies, volume. 6, issue 3 (2022) 23-44 https://doi.org/10.34260/jaebs.632 27 most significant factor which decides the life choices of individuals. females living in a developing country like pakistan let themselves deprives of many necessities due to the nonavailability of disposable income at their end. the meaningful life activities and better livelihood require better income resources tansel, a., & taşçı, h. m. (2010). 1.1. objective of the study in order to estimate the unemployment duration among females kaplan-meier technique has been used. it is a nonparametric (actuarial) method for approximating time-related events (the survivorship function). typically, a plot of the kaplan-meier estimate of the survival function is a sequence of horizontal steps of diminishing magnitude which, when a huge enough sample is taken, approaches the true survival function for that population. a significant advantage of the kaplan-meier curve is that the method can consider the "censored" data losses from the data before the final results are observed. here, we are concerned to evaluate the duration of female unemployment, which is an enormous share of the dejected labour force. we can witness from the example of developed economies that they grew faster when they created equal opportunities for their female to play their part in society's wellbeing. the previously available literature has shed light on the fact that if good health amenities and equal chances of education are provided, it will increase its long-term growth. specifically, the study will examine the role of age, education, gender, family size, children and other characteristics as determinants of unemployment duration. the paper is prearranged in the following manner: section ii shows the existing literature and highlights the critical factors related to the study—a review of the literature. the methodology part elaborates on the technique this study used for analysis purposes, the part after that debates the data sources; lastly, the study debated the study's findings by shedding light on the results and conclusion 2. literature review numerous social, economic, and psychological factors are involved in female unemployment. it is high time for females of our society to get encouraged and participate in the labour force to improve their economic stability and the country. it is being observed that females are the neglected part of our culture being stayed at home; however, if given a chance to them, an enormous change in society can be bought, as they can perform on both home and workplace if given few facilities. females quit their job after having children. it is one of the primary reasons that they stay at home because of the non-availability of help at home who can take care of their children in their absence. it was concluded in many studies that children under the age of 5 increase females' unemployment probability. this is the reason that they stay out of the labour market. moreover, it was recommended that to resolve this issue, more gulrukh zahid & ambreen fatima 28 childcare facilities will help counter this problem (michele a. et al. (2011), n lázaroet et al, (2000), shabbir et al, 2015). in another study performed by samer kharif (2015) concludes that higher wage reservations and discriminatory labour market conditions are one of the causes of female's longer duration of unemployment. in spain, the analysis was performed on the household expenditure survey data (199091). the study utilizes the socio-economic variables and personal characteristics such as age, education, and family background and concludes that there is a need to create a resolution for a better family and professional life. enormous individual-level data of romania and hungary was examined by daniela et al.(2012) and highlighted the impact of socio-economic variables such as marital status, education, health status, age, region and unemployment allowance on employment status and unemployment duration for women and indicated that for women unemployment duration and age is a significant variable. khan et al. (2009) determined that married women become an active part of the informal sector's labour force due to poverty. doing so increases productivity and working mothers' income, reducing poverty at the household level in the long run. anderson (1993) took a sample of about sixty working women and forty jobless women to examine glasgow women's unemployment experience and summarized the importance of paid employment for women's social identity. the study also concluded that unemployment is a crucial factor in the loss of women's economic and social identity and at the domestic level, there is no reward for doing household chores by losing paid work. it was examined by ahmed et al., (2013) on the data of rural and urban areas of bahawalpur, pakistan that the socio-economic variable like education, age, number of family members employed, education of father, mother, and husband, job status of the mother and technical education are the reason for lower employment rate in females. the study also summarizes that large family size, number of children, and joint family positively impact female unemployment. the study performed on india's labour force by stephan (2013) evaluated that india's economy is not utilizing its large working-age population. if the same scenario persists, the propensity to incorporate female labour force participation in coming years will reduce further, which will cause india to fall in the growth rate if they do not include their educated females in the labour force. pieters (2015) identified by incorporating the five extensive cross-sectional microdata for the period of 1989-2009 that higher education level, high growth rates and decline in fertility rate. the labour force participation of females is unchanged which is caused by market forces. the supply side's situation causes the main reason for such an immobile participation rate: household income, low-level selection of highly educated females, and the cultural feeling of dishonour for working females. nonyana, journal of applied economics and business studies, volume. 6, issue 3 (2022) 23-44 https://doi.org/10.34260/jaebs.632 29 jeanette zandile (2015) with the help of the survival technique, examine the semi-parametric and nonparametric estimates to understand the duration dependence and probabilities of exiting unemployment and the socio-demographic factors associated with it. brigitte (2018) worked on botosani county's labour market data. the information was taken from the employment agency of botosani county between the time duration of 2012to 2015 on the data set of 200 unemployment spells by applying kaplanmeier estimator to predict the probability of remaining unemployed with characteristics like age, cohort and gender. the findings revealed that variable cohort influences the probability of staying unemployed if gender and age were considered as controlled. by working on the data of the united states for the period of 2008 to 2015 sansale (2019) evaluated the importance of an individual's personality and its role in determining the unemployment duration among young adults. the methodology that was used for that purpose was competing risk model. it was estimated in the paper that individuals with more concern to their job status got a job as soon as they lost the previous one compared to those who have neuroticism. the data for the time 2000-01 was utilized by taşçi et al. (2010) to evaluate the duration of unemployment for both men and women on turkey's data. the study examines household and personal characteristics and their effect on the labour market state and concluded similarities in both developed and developing countries. it was stipulated that with higher education, such as having a university degree for females has no impact on unemployment, while for males, it positively impacts it. msigwa. (2013) employ multinomial logistic regression model (mlm) on the data of tanzania to evaluate youth unemployment. the study results suggest that skills, education, geographical location, gender, and marital status significantly explain the difference in youth employment status in tanzania. a semiparametric cox regression approach was employed by kavkler et al. (2009) to examine the duration of unemployment for five central and eastern european countries. the study utilizes the cox proportional hazard model and cox regression model with the inclusion of timedependent. 3. methodology this study focuses on unemployment duration using survival function; a nonparametric and parametric approaches. the nonparametric (actuarial) technique can handle the problem of censoring (both right and left). this is why, for duration analysis, the kaplan-meier approach, also known as product limit estimator, has been used. a breakthrough in medical science was witnessed, when edward kaplan and paul meier published their work in the journal of american statistical association in june 1958. they evaluated that while gulrukh zahid & ambreen fatima 30 performing a research trial, some patients may die while others survive at some stage. this motivates them to estimate the patient survival rate. they estimated the proportion of patients who survived at any point during the trial. kaplan-meier estimation is extensively used in recent years in other social science subjects as well. in economics, the technique is specifically used to research price and unemployment duration assessment by survival analysis. numerous studies have estimated the duration analysis of unemployment by using a nonparametric technique such as kaplan-meier. few of them are flek al (2015), vasilica, et al. (2011), danacicaet al (2010), gabriel et al. (2017), ciucaet al (2010). the nonparametric method of estimating time-related events generally evaluates the survivorship function. it is commonly used to investigate the death as an outcome in biostatistics, but lately, this has become popular in industrial sciences and other social sciences. an economist might get interested in estimating the duration of people who remained unemployed. the sequence of horizontal steps of reducing magnitude in the kaplan-meier estimation of survival function presents the population's survival function in a huge sample. it approaches its real survival function for that specific population. it is assumed that the value of the survival function in the distant sample is constant. due to its significant advantage of taking into account the "censored" data, kaplan-meier curve considers the lost data before the final results observed. the survival function "s(t)" starts with preliminary steps that captivate the probability of survival past time "t". it initiates with the steps before approximating the survival function "s(t)", which can evaluate the probability of survival time. for this study, "survival" means the period when an individual remains unemployed – i.e. before he/she gets re-employed. the time "t" is measured in days; s ̂(t)=∏_(j/j≤t) (u_j-a_j)/u_j (1) here, we have taken u_j as the spells of unemployment ongoing in numbers for j days, while a_j are those spells (in numbers), change into the employment promptly after j days. we have considered the concept of hazard rate λ(t). for this study, hazard rate λ(t) represents the short probability of leaving unemployment and moving into employment at time t, and it is entirely dependable on the condition of remaining unemployed till the time of unemployed until immediately before t. the universal classification of a continuous hazard rate is; λ ̂(t)=lim┬(∆t→0)⁡〖(pr⁡(t≤t<t+∆t\t≥t))/∆t〗 (2) in the equation above, t represents the spell duration of unemployment (in our analysis, we have taken the days till any random individual stays unemployed). the proportional hazard model assumes the continuous hazard rate for the ith spell abides the following form journal of applied economics and business studies, volume. 6, issue 3 (2022) 23-44 https://doi.org/10.34260/jaebs.632 31 λ_i (t)= λ_o (t).e^(x_i^' (t).π) (3) equation (3) represents the λ(t) as the assumed baseline hazard, and x_i(t) is the vector of covariates (descriptive variables). for example, the age of a person, gender, education level, marital status etc., and π presents the projected parameter vector. finally, the term e^(x_i^' (t).π) represents a proportional shift – the observed explanatory variables change the complete hazard rate up or down. semi-parametric analysis (cox regression) to estimate the relationship between survival rate and covariates selected. for the semiparametric analysis cox regression is employed. cox regression determines covariates' effect on the hazard model -its direction (increasing or decreasing) and importance. the common form of the cox regression is λ_(i )(t) =e^(x_i^' (β) ).λ_(o )(t) = c_i .λ_(o )(t) (4) i = 1,2 …n in the equation given above, x_(i ) ( x_(i 1) x_(i 2)….x_(i k)) is the vector of k covariates. the analysis includes age, marital status, spouse job status, education, head of household’s education, head of the household’s gender, head of household’s income, region, province and training taken if any. β (β_1 β_(2 ) 〖… β〗_k) is the vector of regression coefficients and λ_(i )(t) is the hazard function of individual i and λ_(o )(t) is the baseline hazard. the baseline hazard starts with 0. the impact of the covariates on the hazard function in the cox regression model does not depend on time due to the ratio (λ_i (t))/(λ_o (t))is equal to the constant ci. therefore, the baseline hazard determines the shape of the hazard function. the ratio of the hazard functions of individual si and j, namely (λ_i (t))/(λ_j (t))is called the hazard ratio. (λ_i (t))/(λ_j (t))= (〖e^(x_i^' (β) ) . λ〗_0 (t))/(〖e^(x_i^' (β) ) . λ〗_0 (t))=e^(〖(x〗_i^ 〖x_j)〗_^( ') (β) ) (5) the hazard ratio is the covariate ratio and is independent of time. this property is proportional, and its explanation is parallel to the odds ratio for logistic regression. a ratio higher than 1 shows increased risk while a ratio lower than 1 specifies decreased risk. (λ_i (t))/(λ_j (t))= e 〖β_p〗^ (6) the hazard ratio is said to be statistically significant at the given level when its confidence interval excludes 1. the null hypothesis that the variable is not correlated to survival can be rejected. this is the base for the explanation of the cox regression results. by using cox's partial likelihood estimator, it is possible to estimate the parameter vector without postulating and estimating the baseline hazard (see greene (2003) for details). the final model estimated is: h(t) = β_(1age )+ β_(〖age〗^2 )+ β_3edu+ β_(〖4edu〗^2 ) + β_5ms + β_(6spouse job gulrukh zahid & ambreen fatima 32 status) +β_7traning + β_8region + β_9province + β_(10 head edu) + β_(11 house size) +β_(12 head house) where h(t) represents hazard rate, while the independent variable includes age, age square, education square, marital status, spouse job status, training, region, province, head education, house size and head of the house. as far as the effect of independent variables on hazard rate is concerned, the variable age as a determinant of hazard rate – unemployment survival function is taken in this study as age has a greater impact on employment decision and choice. education too plays a significant role in finding a job. in order to evaluate the above statement, a non-linear term of education is also included in the model to see if after a certain education the job market condenses. it has been observed from the data that highly educated females, if once got unemployed their hazard of getting reemployed usually declines given that the offers did not match their qualifications, this is the reason, that the curve of education is decreasing after postgraduation. marital status is also assumed to have a negative impact on getting reemployed once a female got unemployed. the study has taken the spouse's job status for female employment as well indicating if a spouse has better job status, it will create a negative impact on a hazard of female employment. head education is taken to see if the head of the house education influences female employment reemployment rate or not. the covariate of training is hypothesized as having a significant impact in the case of getting a job for females. here training related to the job has a positive impact on job status. 4. data and variables the study has taken into account the labor force survey (lfs) data for the year 201415. the analysis was made after extracting the required information from the data. table 1: duration and unemployment variables variable expressions observations the span of the unemployment spell (duration of unemployment): the number of unemployment days. (the difference between the date of entry and the date of leaving unemployment). the minimum value is two days, and the maximum value is 3650 days (121.67 months and ten years). the mean value is 596.59 days, while the median duration is 730 days (months). journal of applied economics and business studies, volume. 6, issue 3 (2022) 23-44 https://doi.org/10.34260/jaebs.632 33 the status of the person (or the censored index) for analysis, all the diverse causes of the exiting are grouped into 2 major outgoing destinations: 1 -economically active or employed (first employment or re-employment); 0 economically inactive out of 1279 females in the labour force, there were 13 different reasons for leaving unemployment. the status variable allows for a distinction between the individuals who registered for the event (employment) and the individuals who did not register for the event. table 2: explanatory variables age we grouped the ages into 5 categories age_1 → 14-18 years age_2 → 19-28 years age_3 → 29-60 years age_4 → 61-80 years age_5 → 80 plus years age_6 → 85 plus years the average age of a female registered in the database on lfs is 13 years and the maximum is 78 years with the mean of 27.88. age square age square is taken to see the impact of education for an individual education the level of education is grouped into 5 categories: educ_1 → primary and secondary or vocational education educ_2 → high school and postsecondary education till graduation educ_3 → post-graduation till masters level educ_4 → mphil educ_5 → phd the minimum education of the group is 1 and the maximum is 18 years of education with the mean value of 12.73. education square we have taken the square of education to assess the impact of higher education on employment. marital status (0) unmarried ;(1) married 515 (40.27%) women were married in the data while 764(59.73%) were unmarried. spouse job status (1)employed;(2) unemployed; (3)not in labour force the total number of employed husbands is 139 with 10.87%, unemployed is 98 with 7.66% and not in labour force is 1,042 gulrukh zahid & ambreen fatima 34 with 81.47%. training (0) no; (1) yes females with no training are 1, 011 79.05% and with training are 268 20.95%. region (0) rural; (1) urban people in rural areas are 742 with 58.01% and in urban are 537 and 41.99% province (1) kpk (2) punjab (3) sindh (4) balochistan (5) islamabad the percentage of population according to the province in kpk is 22.67%, punjab with 53.79% sindh with 16.73% balochistan with 6.49% islamababad0.33% house hold size house hold size is taken as how many members of a family living in any houes head house (1) unemployed (2)employed head education (0) no; (1) yes about 89% of the head house is results specifically, in this study's analysis, we have studied and analyzed the time duration when any random individual leaves employment until he gets reemployed. the survival function generates a step for representing the time when an individual is exiting the state of unemployment in a graphical representation. in our sample data of the labour force survey of 2014-15, the information of 1279 females was available, among those 135 were employed, 289 were not employed, and 855 were out of the labour force (currently were not looking for job opportunities). the graph below shows the survival estimate. journal of applied economics and business studies, volume. 6, issue 3 (2022) 23-44 https://doi.org/10.34260/jaebs.632 35 table 3: kaplan-meier survival function 3650 2 2 0 0.0000 . . . 2920 5 3 0 0.0018 0.0012 0.0004 0.0061 2555 6 1 0 0.0044 0.0020 0.0017 0.0099 2190 11 5 0 0.0053 0.0021 0.0022 0.0111 1825 23 12 0 0.0097 0.0029 0.0052 0.0168 1460 47 24 0 0.0202 0.0042 0.0132 0.0297 1265 48 1 0 0.0413 0.0059 0.0308 0.0540 1095 91 43 0 0.0422 0.0060 0.0316 0.0550 1030 92 1 0 0.0800 0.0080 0.0652 0.0967 910 93 1 0 0.0809 0.0081 0.0660 0.0977 730 814 721 0 0.0818 0.0081 0.0668 0.0986 575 815 1 0 0.7158 0.0133 0.6888 0.7410 545 816 1 0 0.7167 0.0133 0.6897 0.7418 368 817 1 0 0.7176 0.0133 0.6906 0.7427 366 818 0 1 0.7185 0.0133 0.6915 0.7435 365 1027 209 0 0.7185 0.0133 0.6915 0.7435 300 1028 1 0 0.9020 0.0087 0.8835 0.9177 270 1030 2 0 0.9029 0.0087 0.8845 0.9185 240 1040 10 0 0.9047 0.0086 0.8864 0.9201 210 1042 2 0 0.9134 0.0082 0.8959 0.9282 180 1078 35 1 0.9152 0.0081 0.8978 0.9298 150 1086 8 0 0.9459 0.0065 0.9315 0.9573 120 1097 10 1 0.9529 0.0061 0.9394 0.9635 97 1098 0 1 0.9617 0.0055 0.9493 0.9711 90 1156 12 46 0.9617 0.0055 0.9493 0.9711 60 1188 12 20 0.9718 0.0047 0.9609 0.9797 40 1189 0 1 0.9817 0.0038 0.9726 0.9878 32 1190 1 0 0.9817 0.0038 0.9726 0.9878 30 1228 10 28 0.9825 0.0037 0.9736 0.9885 22 1229 0 1 0.9906 0.0027 0.9835 0.9946 20 1237 1 7 0.9906 0.0027 0.9835 0.9946 15 1263 2 24 0.9914 0.0026 0.9845 0.9952 10 1269 1 5 0.9930 0.0023 0.9865 0.9963 8 1270 1 0 0.9937 0.0022 0.9875 0.9969 6 1273 3 0 0.9945 0.0021 0.9885 0.9974 5 1275 0 2 0.9969 0.0016 0.9917 0.9988 3 1276 1 0 0.9969 0.0016 0.9917 0.9988 2 1279 3 0 0.9977 0.0014 0.9927 0.9992 time total fail lost function error [95% conf. int.] beg. net survivor std. gulrukh zahid & ambreen fatima 36 figure 3: kaplan-meier survival estimate author's estimation using lfs (2014-15) as per the outcomes of our study, a female's chance to get reemployed significantly reduces if she stays jobless for longer durations. the time projected in our analysis is ten years, i.e. 3650 days and in terms of months, it is 121.67 months. in graph 1 above, we can see that the starting steps are very close and comparatively small, predicting a higher probability of re-employment. in 2 days, the possibility of finding a job is 99.77%, and it continuously reduces, as the survival function increases. they are depicting a negative relation between re-employment and survival function. according to our findings, females' probability of finding a job reduces if they stay out of the labour market for more extended time durations, causing it as one of the main factors for females to quit job search voluntarily and come out of the labour force due to unavailability of the job permanently. the findings in figure 4 indicate that females with higher education such as a master’s and higher than that have even fewer chances of entering the job market after becoming unemployed. this situation will make females feel overqualified and make them demoralized in persuading higher education. it is being witnessed already in pakistan's case that females face numerous issues with respect to gender discrimination in the labour market. fewer wages for the same work than male is already a problem; females have to face while working. many studies show that females with children below five years of age and elders at home are the reason to generate impediments for them to proceed. journal of applied economics and business studies, volume. 6, issue 3 (2022) 23-44 https://doi.org/10.34260/jaebs.632 37 figure 4: kaplan-meier survival estimate by education author's estimation using lfs (2014-15 pakistan, being a developing country, belong to those societies, where females are associated with homemaking. it becomes unenviable to continue a career after marriage as uncountable household obligations do not allow them to do so. employers are least concerned with hiring females with children as the unavailability of childcare pushes them to bequeath their profession. it will be beneficial to get back this discouraged labour force into the job market if females are provided with daycare facilities and similar to them in economical fees. figure 5 is presenting the distribution of females according to their age groups. according to this graph, females in the age group of 29 to 60 have scarcer likelihoods to get employed compared to the age group of 19 to 28. in figure 6, females are represented by marital status. the graph shows that married females face more problems than unmarried females in employment. it is high time for females to anticipate into the labour force, as almost half of pakistan's population consists of females. at the macro level, this increase in the number of labour forces will spike up the opportunities of the higher growth rate in productivity and economic growth. similarly, it will boost the affluence and additional income into the house on a micro level, which helps them invest more in their offspring's education and health. gulrukh zahid & ambreen fatima 38 figure 5: kaplan-meier survival estimate by age groups author's estimation using lfs (2014-15) figure 6: kaplan-meier survival estimate by marital status author's estimation using lfs (2014-15) spouse job status is another cause of intentional unemployment in females. the social and cultural norms of society force females to stay at home as husbands are financial providers of their wives. is causes females due to family pressure or by their own free will leave their job after marriage and stay at home. the relationship of spouse’s job status and female employment is presented. in figure 7, the graph 1 indicates that spouse is employed, 2 as the spouse unemployed, and 3 as a spouse not in the labour force. our findings depict females with spouse’s unemployed are putting more effort to get employed once being unemployed. journal of applied economics and business studies, volume. 6, issue 3 (2022) 23-44 https://doi.org/10.34260/jaebs.632 39 figure 7: kaplan-meier survival estimate by spouse's employment status author's estimation using lfs (2014-15) cox regression the results presented in the table using the cox regression show hthe azard rate of being unemployed. marital status (being married) and age with the negatsignssign are the indicators showing that the hazard increases as age increases or if the female is married. on the other hand, years of education show a positive impact on females’ unemployment duration. province-wise break up shows that other than sindh and balochistan, the sign for punjab and kpk are positive. job-related training is also showing a significant impact on unemployment duration. head education and migration status also have a negative sign othe n hazard rate of being unemployed for females. table 4: _t coef. z p>z marital status -.0809233 -1.11 0.266 age -.0387243 -2.12 0.034 square of age .0004067 1.71 0.086 years of education .0193234 1.84 0.065 percapita income 5.92e-07 1.13 0.259 province punjab .0194688 0.26 0.798 sindh -.1493358 -1.54 0.123 balochistan -.1414082 -1.07 0.285 region .0483527 0.75 0.455 head education -.0119239 -1.32 0.186 gulrukh zahid & ambreen fatima 40 traning .1253401 1.66 0.097 migration -.0632756 -0.41 0.681 table 5: covariates coefficients exp(coefficients) p value age (-3.87e-02) 0.962 0.034 square of age (4.07e-04) 0.9999 0.086 years of education (1.93e-02) 0.019 0.065 training (1.25e-01) 1.133 0.097 marital status (-8.09e-02) 0.922 0.0777 5. conclusion and policy recommendations the study related to the unemployment duration of females in pakistan was evaluated by employing kaplan maier and cox regression to examine survival analysis for the extended unemployment of females in pakistan, which results in the discouraging labour force in the long run due to the unavailability of job opportunities. for that purpose, the data is taken from the labor force survey (lfs) 2014-15. the research concluded that a longer duration of female unemployment causes them hard to find a job again. the longer the duration, the fewer will be the chances to get employed again. the study concluded that the total time duration of re-entering into employment once being unemployed is 3650 days which is equal to 10 years, and this time is 121.67 months. it is concluded that the chances of remain unemployment duration increase, as the survival function decrease. the study determines that providing childcare will positively impact female labour force participation, as it is one reason females quit their jobs. this increased female labour force participation will help them to contribute to society by utilizing their talent. therefore, it is highly recommended that for the better economic development of any country, it is imperative to engage females' participation in the labour force. for doing so, the following are some of the policy recommendations. the most important measure that the government should take on an immediate basis is increasing female child education investment. by doing so, labour force participation will increase as it has a positive impact on employment. in this regard, both print and electronic media should play their part. more incentives should be offered to bring out children into school, especially female children, as studies suggest that an educated mother always puts more effort to send her child to school. this research also suggests that females with a master's journal of applied economics and business studies, volume. 6, issue 3 (2022) 23-44 https://doi.org/10.34260/jaebs.632 41 degree or higher education faces a longer unemployment rate. to address this issue, the government should encourage private-sector employers to employ females in decisionmaking posts, and for the government-sector job, they should increase the female quota in jobs. it is also observed that females are the least priority for employers due to their respective household responsibilities and childcare. the government should provide economic childcare centres and rules to provide this facility to increase the number of females to work after having children. in our society, the socio-economic culture has created an environment that discourages females from working after marriage. we, as a society, work on this norm to be changed. the government's responsibility is to generate such incentives and a safe and secure environment that facilitates females to pursue their careers and dreams without fear. we have the example of our armed forces. they employ highly talented and educated females into forces and facilitate them so that they can put their hearts and soul and work passionately for the better future of the country. references anderson, pauline. "women and unemployment: a case study of women's experiences of unemployment in glasgow." phd diss., university of warwick, 1993. arslan, muhammad, and rashid zaman. "unemployment and its determinants: a study of pakistan economy (1999-2010)." journal of economics and sustainable development, issn (2014): 2222-1700. bell, d. n., & blanchflower, d. g. (2010). youth unemployment: déjà vu? chapman, b., weatherburn, d., kapuscinski, c. a., chilvers, m., & roussel, s. (2002). unemployment duration, schooling and property crime. demidova, olga, and marcello signorelli. "determinants of youth unemployment in russian regions." post-communist economies 24, no. 2 (2012): 191-217. eberwein, c., ham, j. c., & lalonde, r. j. (2002). alternative methods of estimating program effects in event history models. labor economics, 9(2), 249-278. flek, v., hála, m., & mysíková, m. (2015). duration dependence and exits from youth unemployment in spain and the czech republic. economic research-ekonomska istraživanja, 28(1), 1063-1078. kavkler, alenka, daniela-emanuela danacica, ana gabriela babucea, ivo bicanic, bernhard bohm, dragan tevdovski, katerina tosevska, and darja borsic. "cox regression models for unemployment duration in romania, austria, slovenia, croatia and macedonia." romanian journal of economic forecasting 10, no. 2 (2009): 81-104. gulrukh zahid & ambreen fatima 42 kettunen, juha. "education and unemployment duration." economics of education review 16, no. 2 (1997): 163-170. khan, t., & yousaf, f. (2013). unemployment duration of first time job seekers: a case study of bahawalpur. asian journal of economic modelling, 1(1), 8-19. kipesha, erasmus fabian, and robert msigwa. "determinants of youth unemployment in developing countries: evidences from tanzania." (2013). kisto, m. (2014). discouraged youth in mauritius. international journal of recent research in interdisciplinary sciences, 1(3), 1-8. klasen, stephan, and janneke pieters. what explains the stagnation of female labor force participation in urban india? the world bank, 2015. meghir, costas, yannis ioannides, and christopher pissarides. "female participation and male unemployment duration in greece." european economic review 33, no. 2-3 (1989): 395-406. msigwa, r., & kipesha, e. f. (2013). determinants of youth unemployment in developing countries: evidences from tanzania. moffitt, r., & nicholson, w. (1982). the effect of unemployment insurance on unemployment: the case of federal supplemental benefits. the review of economics and statistics, 1-11. mortensen, d. t. (1986). job search and labor market analysis. handbook of labor economics, 2, 849-919. novella, m. l., & duvivier, m. (2015). the relationship between unemployment duration and education. nonyana, j. z. (2015). statistical modeling of unemployment duration in south africa (doctoral dissertation). o'higgins, niall. "the challenge of youth unemployment." international social security review 50, no. 4 (1997): 63-93. pekkala, sari, and hannu tervo. "unemployment and migration: does moving help?" scandinavian journal of economics 104, no. 4 (2002): 621-63 scarpetta, s., sonnet, a., & manfredi, t. (2010). rising youth unemployment during the crisis. sansale, r., deloach, s. b., & kurt, m. (2019). unemployment duration and the personalities of young adults' workers. journal of behavioral and experimental journal of applied economics and business studies, volume. 6, issue 3 (2022) 23-44 https://doi.org/10.34260/jaebs.632 43 economics, 79, 1-11. tansel, a., & taşçı, h. m. (2010). hazard analysis of unemployment duration by gender in a developing country: the case of turkey. labor, 24(4), 501-530. tansel, aysit, and h. mehmet tasci. "determinants of unemployment duration for men and women in turkey." (2004). verho, j. (2008). determinants of unemployment duration over the business cycle in finland. hecer discussion paper 226. verho, jouko. "determinants of unemployment duration over the business cycle in finland." (2008) gulrukh zahid & ambreen fatima 44 appendix figure: author's estimation using lfs (2014-15) figure: kaplan-meier failure function figure: survival functions showing the education related information journal of applied economics and business studies, volume. 5, issue 1 (2021) 191-214 https://doi.org/10.34260/jaebs.510 191 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x asset prices and velocity decline: an empirical investigation hina shafiq1*, wasim shahid malik2 1. assistant professor, sardar bahadur khan women’s university, quetta. 2. professor sbp chair, university of peshawar, peshawar. abstract the quantity theory of money lost much of its significance in 1980s due to the phenomenon of velocity decline and consequent instability of the money demand function; missing financial transactions and asset prices were believed to be responsible (borio, kennedy and prowse, 1994; and werner, 2012). this study, therefore, uses asset prices for pakistan to explain the velocity decline phenomenon in a regression model as well as in vector error correction model (vecm) using quarterly data for the time period 1981q1-2018q2. the study finds significant role of asset price index in explaining income velocity of money with a negative effect. moreover, the sub-sample regressions show that asset prices are helpful in explaining velocity decline phenomenon for the time period 1981-1998 and 2008-2018 but not for 19982008. moreover, there are brief periods in the sample when velocity actually increased despite an overall declining trend. to explain those short term reversals in velocity trend, the study uses indicator function. results show that the increase in velocity for brief periods is also explained by asset prices. keywords velocity decline, asset prices, money demand function, monetary policy jel classification e41, e52 1. introduction the quantity theory of money lost much of its significance in 1980s due to the phenomenon of velocity decline and at that time money demand function was also found unstable (see, for instance, hendry, 1985; belongia and chalfant, 1990; boughton, * hinasbk@hotmail.com hina shafiq, wasim shahid malik 192 1991). notwithstanding the assumption of equation of exchange, velocity of money was believed to be declining over time and the money demand was found to be an unstable function of its determinants. various sources of instability were offered that were linked to decline in the velocity of money. for instance, arguments of money demand function, like interest rate and income, had an effect on velocity, which then became unstable because of variability in the arguments; financial innovations and deregulations that changed the interest elasticity of demand for monetary aggregates ultimately led to unanticipated changes in the velocity of money; the function may also be unstable if there was continuous gap between money supplied in the system and what economic agents wanted to hold (anderson, 1985). the empirical research in the field stressed that money was highly unpredictable and the literature was unable to explain the phenomenon that when the money supply was increasing the velocity of money kept on declining. therefore, the need was felt to replace money with the interest rate as monetary policy instrument. historically, the monetary policy of the state bank of pakistan majorly focused on price stability and output growth using monetary aggregates targeting and keeping base money as its operational target. however, from 2009 onwards, it switched to having overnight money market repo rate – the short term interest rate – its operating target. the available literature in pakistan mainly deals with substantiating, or otherwise, the quantity theory of money and estimating money demand function. limited literature that is available on velocity of money concludes that the relationship between income velocity of money (derived from any of the monetary aggregates) and its determinants is stable (omer, 2010), while on the other hand, some studies show that the relationship between income velocity of money and real per capita income seems to have changed and is unstable due to financial and economic developments in post 1971 war period (omer and saqib, 2009; bilqees and rauf, 1994). the income velocity of money carries important information for the design of monetary policy in pakistan and evidence on its stability will be helpful for the debate on the adoption of monetary aggregates or inflation targeting regimes. moreover, the importance of money and hence velocity of money is still relevant for a cash based economy of pakistan where money plays importabt role in determining inflation rate (qayyum, 2006; khan and schimmelpfening, 2006; kemal, 2006; chaudhry et al., 2015). however, the existing research does not relate income velocity of money with asset prices; the relationship needs to be explored. this study adds to the existing empirical literature with regards to pakistan economy by explaining the velocity decline phenomenon using asset price index. therefore, the objective of this paper is to explain the issue of the velocity decline phenomenon for the pakistan economy and this has been done with the help of asset price indices constructed by the authors. by distinguishing between transactions that are part of gdp and those transactions that are excluded from gdp and then taking both journal of applied economics and business studies, volume. 5, issue 1 (2021) 191-214 https://doi.org/10.34260/jaebs.510 193 of these transactions in the quantitity theory equation solves the mystery of missing money due to velocity decline (werner, 2012; borio, kennedy and prowse, 1994). moreover, the study also contributes by analyzing the velocity of money in three different sub-samples divided on the basis of breaking trend and then investigating how asset prices can explain the movements in the velocity of money in different regimes. the study also attempts to explain increases in the velocity of money observed over brief periods of time using indicator function. our methodology, to accomplish these objectives, is however different from that of richard werner. instead of taking data on all real and financial transactions and then disaggregating equation of exchange for these two sectors, we have found income velocity of money based on transactions that are included in gdp and ignored assets or financial transactions. this velocity is then regressed on the asset price index to find whether or not the former can be explained by the latter. for this, regression model and vecm have been estimated using the time series data that span 1981-q1 to 2018-q2. for the sub-sample estimations, we have divided the whole sample into three subsamples; 1981q1 to 1998q2, 1998q1 to 2008q2, and 2008q3 to 2018q2. finally, for investigation of whether asset prices explain both the velocity increase and decline phenomena we have used the indicator function. as hypothesized, we find the evidence that the asset prices are significant determinant of the income velocity of money; the effect of former on the latter is negative. same result is found in first and third sub-samples while second sub-sample shows contrary evidence. furthermore, results of the indicator function conform to our hypothesis that the asset prices not only explain velocity decline but also the increase in velocity of money. remainder of the study proceeds as follows: section 2 reviews the existing literature related to the issue; section 3 sheds light on the empirical methodology; results are discussed in detail in section 4; and section 5 gives concluding remarks.. 2. review of literature the quantity theory of money remained a dominant theme in the macroeconomic models and analysis and an important instrument in the conduct of monetary policy in the late 1970s. nonetheless, from 1980s onwards the theory faced severe criticism due to the failure of some of its basic premises which led to its apparent failure in that era. the world witnessed various phenomena overtime, which the empirical literature observes as the decline in the income velocity of money, the case of the missing money, unstable money demand function, and inability to define money that were thought to be the weaknesses of the quantity theory of money (see, for instance, goldfeld, 1976; goodhart, 1989; hendry, 1985; kaldor, 1985; belongia and chalfant, 1990; boughton, 1991; pollin, 1991; cottrell, 1992). the researchers and policy makers sidelined money by formulating moneyless models and focusing on interest rate to understand the workings of the economy in the form of real business cycle (rbc) models, which hina shafiq, wasim shahid malik 194 eveolved to dsge models, later on. money was considered a weak instrument and was replaced by the interest rate as an instrument of monetary policy by almost all the central banks around the world on the assumption that the interest rate is more predictable than money. however, the relationship between interest rate and the nominal gdp growth was never well established and the correlation between the two was never more likely to be negative rather it was found positive. especially for the developing economies with low financial access indicators where most of the transactions are cash based, the monetary models, mostly perform well with money growth rules instead of taylor type interest rate rules (ahmed, pasha & rehman, 2016) . also mostly the causation runs from the economic growth to the interest rate. moreover, the failure of the decrease in interest rate to revive the japanese economy from deflation in 1990s, after the price bubble in both real estate and equity, also raised questions about the validity of the interest rate instrument and the models using it, in explaining asset prices, bank lending or international capital flows. due to this failure, the bank of japan returned back to targeting bank reserves using open market operations in march 2001 later naming it as quantitative easing, as role of money becomes the only remaining influence of monetary policy when the interest rate hits the zero nominal bound (werner, 2012; meyer, 2001). similarly, the bank of england also introduced quantitative easing after the global financial crisis in 2009 and researchers and central banks are now focusing on the quantity of money than on the price of money (werner, 2012). consequently, the economists instead of finding answers to questions posed by issues of velocity decline, inability to define money and unstable money demand function, deemed all these phenomena a failure of the quantity theory of money and found an escape route in building moneyless models and focusing on the short term nominal interest rate and supply side economics. considering the recurrent financial crisis over different time periods, failure of the interest rates in explaining relationship between financial markets and the economy and stimulating a stagnant economy, the failure of the researcher to model banks in macroeconomic theories, and inability of the supply side economics to explain real world phenomena, it is now important to analyze the factors that were responsible for the breakdown of the quantity theory of money and to find answers to questions posed by the events that followed. in order to accurately estimate the equation of exchange and reviving its place in economics, it is imperative to identify the factors that led to severe criticism and relinquishment of the quantity theory of money. as also pointed out by werner (2012), two major flaws in the equation of exchange led to its downfall. first, the original equation of exchange given by irving fisher (1911) is: mv = pt (1) which implies that the total volume of transactions equals the amount of money paid for these transactions in a typical time period. however, due to unavailability of data on total transactions, the cambridge economist replaced ‘pt’ with ‘py’ in the journal of applied economics and business studies, volume. 5, issue 1 (2021) 191-214 https://doi.org/10.34260/jaebs.510 195 equation of exchange which is the total purchases of the final goods and services only i.e. gdp of an economy, on the assumption that pt=py. the gdp statistics do not include real estate transactions and financial transactions which are a crucial element in the original equation of exchange. therefore, a distinction should be made between transactions that are included in gdp and those that are excluded because gdp is only a subset of the total transactions and it does not represent all the transactions in the economy (werner, 2012; howells & biefang-frisancho mariscal, 1992). ). if the value of transactions, not included in gdp, rises then the equation of exchange shows a fall in the velocity, which explains all the phenomena which are considered responsible for the breakdown of the quantity theory of money. moreover, economic activity in an economy cannot be defined by goods transactions only rather the asset transactions1 are also an important component of the economic activity. and the narrow focus of the monetary policy on goods transactions has serious implications for asset price bubbles and hence for financial crisis. consequently, it is imperative to consider the effects of monetary policy changes on the asset transactions because it may so happen that the asset transactions are leading monetary policy and goods transactions as well and may very well be an important source of the financial crisis. this issue was first taken up by richard werner in his papers (1992; 1997; 2012) and both sides of equation (1) were decomposed into two components – one that is related to transactions that are part of gdp and the other related to financial transactions (excluded from gdp). 𝑀𝑉 = 𝑀𝑅 𝑉𝑅 + 𝑀𝐹 𝑉𝐹 (2) 𝑃𝑇 = 𝑃𝑅 𝑇𝑅 + 𝑃𝐹 𝑇𝐹 (3) mrvr and prtr are the gdp based transactions and the value of these transactions respectively and mfvf and pftf are non gdp or financial transactions and their value respectively. therefore, the following equations should hold as well: 𝑀𝑅 𝑉𝑅 = 𝑃𝑅 𝑇𝑅 (4) 𝑀𝐹 𝑉𝐹 = 𝑃𝐹 𝑇𝐹 (5) since prtr represents the value of gdp transactions, it should be equal to nominal gdp i.e pry. 𝑀𝑅 𝑉𝑅 = 𝑃𝑅 𝑌 (6) where vr = (pry)/mr = constant eq (6) implies that with stable velocity the amount of money used for gdp transactions must be equal to nominal gdp over any time period and eq (5) implies that the quantity of money used in financial transactions must be equal to their value. for the economic growth of an economy the value of economic transactions during one time period should be greater than the previous time period. therefore, considering the net changes in the variables over time we obtain: 1 asset transactions include all those transactions that are excluded from gdp such as financial assets and second hand sale of real estate. hina shafiq, wasim shahid malik 196 ∆(𝑀𝑅 𝑉𝑅 ) = ∆(𝑃𝑅 𝑌) (7) ∆(𝑀𝐹𝑉𝐹) = ∆(𝑃𝐹𝑇𝐹 ) (8) the increase in the amount of money used for gdp transactions also increases the nominal gdp and vice versa. similarly, an increase/decrease in the amount of money used for non-gdp/financial transactions increases/decreases the value of non-gdp transactions which implies that if more money is created and directed towards fulfilling the demands for financial transactions, it may lead to the creation of asset price bubbles. second, the complexities in defining money, perhaps due to financial innovations, cast doubt on the validity of the quantity theory of money. many economists defined money as the deposit aggregates such as m0, m1,m2,m3,m4 to be used in equation of exchange but these deposit aggregates represent money that is out of circulation, whereas the m in the equation of exchange is defined as the purchasing power of money in transactions i.e. money in circulation (werner, 2012). hence, a proper definition of money used for total transactions which represents the purchasing power of money and implies money in circulation is imperative to accurately estimate the equation of exchange. the history of money suggests that the quantity theory of money had persisted through major changes in the currency systems overtime and therefore has the ability to cope with current innovations to prove its validity with empirical evidence (graff, 2008). therefore, a more proper definition of money needs to be established for which different measures of money can be constructed. one of such measures could be defining m as total credit in the economy and disaggregating it as credit for gdp/real transactions and credit for financial transactions as is suggested by werner (1997 and 2012). we did not use the credit variable in this analysis because most of the transactions in pakistan are cash based especially in asset purchases. the monetary policy of the state bank of pakistan mainly focused on maintaining price stability and output growth with initially using monetary aggregates targeting and keeping m0 as its operational target. nevertheless, from 2009 onwards the operational target was replaced by the overnight money market repo rate owing to weakening relationship between monetary aggregates and inflation rate, which can be attributed tothe structural changes in the economy, in general, and the financial sector, in particular (hanif, 2014). the existing literature on the quantity theory of money in case of pakistan has mainly focused on verifying the positive relationship between growth of money stock and the inflation and to testify the monetarist stance that inflation is always and everywhere a monetary phenomenon. such as farooq, hassan and shahid (2015), qayyum (2006), and kemal (2006) showed a positive relationship between money supply and inflation in pakistan. however, omer & saqib (2008) shows that the quantity theory of money does not hold, in its original form, in case of pakistan, owing to the absence of proportional relationship between money supply and prices, instability of the journal of applied economics and business studies, volume. 5, issue 1 (2021) 191-214 https://doi.org/10.34260/jaebs.510 197 income velocity of money, and the endogenoeity of money stock in the economy;, therefore, pakistan should rethink about monetary targeting strategy. other than above mentioned two directions, ahmed, pasha and rehman (2016) evaluates the role of money in creating business cycles in closed economy dsge model and concludes that pakistan being the country with lowest financial access indicators has a high level of currency and the most of the transactions are cash based. money has significant role in the models of the pakistan economy because monetary aggregates are not only strongly procyclical but may also be considered as leading indicators of business activity. moreover, performance of money growth rule in the form of matching to actual data is superior to the taylor type interest rate rule. moreover, hanif and hayat (2016) emphasize that the inflation in pakistan is significantly being explained by the broad money throughout all regimes including monetary targeting regime, transition phase and interest rate targeting regime. this means that the control of the money supply still has the potential to curb inflation rate but more focus on interest rate has undermined the importance of monetary targeting. these two studies highlight the importance of money in explaining the economic activities in pakistan. moreover, the literature related specifically to velocity of money in pakistan is also limited. omer (2010) shows that the velocity of base money and that of broad monetary aggregate do not dependent on the changes in interest rate but the relationship between income velocity of monetary aggregate and its determinants is stable. therefore, the study suggests to use monetary aggregates as the nominal anchor in pakistan. bilqees and rauf (1994) conclude that the relationship between income velocity of money and real per capita income has changed overtime due to the financial and economic developments in post 1971 war period. however, asset prices and asset transactions and their effect on the economic activity and velocity decline have always been ignored as an area of research in the field of monetary policy in pakistan. 3. econometric methodology 3.1 asset prices and the velocity of money from theoretical discussion on the topic, given in literature review, we hypothesize that the decline in velocity of money is associated with rising asset prices over time. therefore, velocity of money is negatively associated with asset prices. to test this hypothesis, we have estimated different models with different indices of asset prices. in the first step of the analysis we run a simple bivariate regression of income velocity of money on real aggregate asset price index. we, then estimate the multivariate models through two different approaches: regression based approach which is in conformity with engle-granger method and vecm using johansen’s approach. bivariate regression model is given as: hina shafiq, wasim shahid malik 198 (9) where py/m is the velocity of money and ap/p is the asset price index deflated by gdp deflator that has been constructed using three types of assets, namely the house prices, equity prices and the exchange rate using four different measures based on weights assigned to each asset. we run the bivariate regression using four different measures of real asset price indices constructed with each of the procedure of assigning weights along with the three more specifications that include individual asset prices namely the house price index, equity prices and the exchange rate. the coefficient on real asset prices (β) will indicate the direction and magnitude of the effect of asset prices in explaining income velocity of money. below we discuss some important steps that we follow while estimating all the specifications of bivariate regressions: 1. we have checked the stationarity of variables. if the variables are non-stationary then we may have spurious results. in that case, we have tested the presence of unit root in the residual series, which is actually the second step in engle-granger cointegration test. if residual series is found to be stationary then it signifies a long run equilibrium relationship between asset prices and the income velocity of money. 2. there are structural breaks in our data which signify important changes in the economy over time such as financial sector reforms, shift towards the floating exchange rate system, 9/11 incident, world oil price hike of 2008 and the global financial crisis. events like these change the structure of an economy and hence the behavior of the parameters. we, therefore, have used dummy variables in our analysis to incorporate those structural breaks in the data. 3. the errors are mostly autocorrelated in time series data; therefore, we have included lagged dependent variable accordingly to remove autocorrelation of errors. for selection of the number of lags (from a maximum of 12 lags in quarterly data), akaike information criterion (aic) has been used. 4. because of the inclusion of lags the model becomes a short run model. we, therefore, have adjusted short run estimates to compute long run coefficients. the bivariate analysis is subjected to the risk of specification bias and the coefficient of real asset prices may be biased due to omitted variables. moreover, it is suspected that the relationship between asset prices and velocity of money is caused by a third variable which is missing in the analysis. to prove that the above found relationship is robust we postulate a function that relates velocity of money to the asset prices and a vector of control variables including real permanent income, transitory income, and short run and long run interest rates [for detail on determinants of velocity of money in pakistan see omer (2010)]. log log log i t t t t i py ap py m p m     −       = + + +             journal of applied economics and business studies, volume. 5, issue 1 (2021) 191-214 https://doi.org/10.34260/jaebs.510 199 (10) where py/m is the velocity of money, log ap/p is log value of the ratio of asset price index to gdp deflator and xi contains other control variables such as rs (short term interest rates), rl (long term interest rates), yp (permanent income) and yt (transitory income). all these variables have some theoretical relationship with to the income velocity of money; thus, the multivariate analysis has theoretical justification. the velocity is expected to positively relate to permanent income in the economy as an increase in the latter will expand aggregate demand thereby increasing transactions which, in turn, will increase the income velocity of money. a coefficient of transitory income is expected to be positive and less than 1 which indicates that the velocity will move pro-cyclically as is consistent with the friedman’s permanent income hypothesis. over the business cycle, an increase in transitory income would increase the demand for money balances which serve as a buffer stock, thereby leading to increase in velocity. short term and long term interest rates are hypothesized to positively affect velocity as an increase in the interest rate reduces the demand for money, thereby leading to an increase in the velocity at a given level of income. asset price index is expected to have a negative sign as an increase in the prices of assets will lead to an increase in the amount of money that goes into the financial transactions whereas the gdp based transactions remain the same. this leads to a decline in the income velocity of money. as we are performing multivariate analysis, therefore, we also estimate it as a system in the form of vector error correction model that includes lagged error correction term and lags of first differenced variables. akaike information criterion has been used to select lag length. cointegration among variables has been tested using trace statistics and maximum eigen value statistics. the vecm can be written as: 1 log log log jt i t i ji ji jt i j jt i t py ap py x y m p m      − − − −        = +   +   +  +              ...(11) where i= 1, …., p denotes lag length while j represents variable. py/m is the income velocity of money, ap/p is the real asset prices, y is a vector containing dependent and independent variables, and x vector contains control variables. it becomes evident from the estimations of section 4.2 that the asset prices are significant determinant of velocity decline, however, if we look at the figure 1 of velocity of money and asset prices (see section 4.1) we observe that there are certain time periods in which this relationship is reversed or asset price do not seem to explain the velocity phenomenon. therefore, we divide our whole sample into three sublog log log log it i t t t t i py ap py x m p m      −       = + + + +             hina shafiq, wasim shahid malik 200 samples to analyze that how asset prices explain the movements in the velocity of money in different time periods. the sub-samples are 1981q1 to 1998q2, 1998q3 to 2008q2, and 2008q3 to 2018q2. we use the same regression equation 4.9 for all the subsamples. if we look at the figure of velocity of money and asset prices (see section 4.1) it can be seen that on average velocity is declining with the increase in asset prices. however, if we observe minutely, we can see that velocity is also rising in few time periods. so we have made an indicator function for both rising and declining velocity. we then estimate the regression with this indicator function to check if the inclusion of asset prices gives only one way explanation of velocity decline or asset prices are capable for a two way explanation for both rising and declining velocity. the regression equation takes the following form: 0 1 1 2 log (1 ) * log (1 ) * log log t i t t t t t i py ap ap py i i i i m p p m       −         = + − + + − + +                 ………(12) where i, is the indicator function which takes on a value of 1 if (py/m)t > (py/m)t1 and 0 otherwise. if asset prices explain the movements (rise and fall) in velocity then both the coefficients should have same value. therefore, in order to check the symmetry of coefficients of β1 and β2, we use wald statistics to test the hypothesis that β1= β2. 3.2 construction of variables we use hanif et al, (2013) and arby (2008) as data source for nominal and real gdp. arby (2008) quarterized annual gdp in pakistan for the time period 1972-2005 and hanif et al,(2013) provided quarterly estimates of the production side of annual gdp in pakistan for the time period 1999-2000 to 2009-2010 based on constant prices of 1999-2000 and also for current prices. we extend data for the time period 2011-2018 using quarterly shares found from quarterly data taken from above two sources. we find small variations in quarterly shares over time. we therefore assume that the quarterly shares, calculated as average, remain constant for the next six years for which data are to constructed. we construct quarterly data of gdp for the time period 2011-18 by multiplying the average quarterly shares with the annual gdp. broad money comprises currency in circulation, other deposits with the state bank of pakistan (sbp) excluding imf a/c nos. 1 & 2 saf loan accounts, counterpart funds, deposits of foreign central banks, foreign governments, international organizations and deposit money banks. moreover, scheduled banks’ demand deposits excluding inter-bank deposits, deposits of central and provincial governments and foreign constituents are also included. finally, scheduled banks’ time deposits are part of broad money. gdp deflator is constructed from the ratio of nominal gdp to real gdp and multiplying the result by 100. call money rate is the interest rate at which short term loans are lent and borrowed in the money market. it is the monthly average of daily minimum and maximum call money rate. long term interest rate is 5 years and above weighted average deposit rates, journal of applied economics and business studies, volume. 5, issue 1 (2021) 191-214 https://doi.org/10.34260/jaebs.510 201 which is used as a proxy for long term interest rates. short term interest rate is 6 months and below weighted average deposit rates, which proxies short term interest rate. according to ifs, beginning in september 1997 the share price index for pakistan refers to the mid-day quotes for the last day of the month for 100 common shares on the karachi stock exchange with november 1991 as base period. exchange rate is taken as price of us dollar in terms of pakistani rupees. private sector credit includes credit provided to the private sector of pakistan from conventional bank branches, islamic banks, and islamic banking branches of conventional banks. quarterly data on private sector credit was available only from 1998 q3. therefore, we have constructed quarterly data of private sector credit from 1981 q1 to 1998q2 on the assumption of fixed quarterly weights calculated from available quarterly data and the given data on stock of annual private sector credit. housing services consists of three sub-sectors, namely ownership of dwellings, real estate and cooperative housing societies. this variable is used as percent of nominal gdp to assess the size of housing services relative to the overall size of the economy. size of stock market is measured by market capitalization, which is computed as the number of outstanding shares traded on the karachi stock exchange of the listed domestic companies times the price of these shares. we have used this variable as percent of nominal gdp to calculate the size of the stock market relative to the size of the economy. foreign exchange reserves consists of reserve assets held by the state bank of pakistan and scheduled banks in foreign currency which includes foreign currencies, gold, and special drawing rights of the imf. this variable is also expressed as percent of nominal gdp to evaluate the size of foreign exchange market relative to the size of the economy. house rent index is extracted from the monthly cpi series published by pakistan bureau of statistics. we have calculated year on year inflation rate as the fourth difference of log gdp deflator multiplied by 100, which is the percentage growth rate of a particular quarter over the same quarter of previous year. 3.3 data sources quarterly data of the above mentioned variables have been used for a period spanning 1981 q1 to 2018 q2. the data source for call money rate, exchange rate and share prices is the international financial statistics (ifs). data on m2, 5 years and above deposit rates, 6 months and below deposit rates, 6 and 3 months t-bill rate, private sector credit, house rent index and market capitalization are extracted from the handbook of statistics by the state bank of pakistan. finally, economic survey of pakistan has been used as data source for housing services and foreign exchange reserves. 4. empirical findings 4.1 velocity decline and asset prices: a visual inspection hina shafiq, wasim shahid malik 202 for preliminary analysis we plot asset price index along with income velocity of money. the objective is to establish long run average relationship between these two variables. at the same time, the figure 1 is useful for identifying sub-samples, which are to be used in regression analysis. finally, the figure is helpful in constructing indicator function for positive and negative change in velocity. we are using inverse of the asset prices for this figure which means that a downward portion of the curve of the asset prices indicates a rise in the asset prices and vice versa. as can be seen from the figure, asset prices are explaining the phenomenon of velocity decline in pakistan; on average, with an increase in the asset prices, the income velocity of money is declining. for the first decade in the sample, i.e. from 1982q1 to 1991q3 the velocity of money is more or less constant and during that period asset prices have u-shape. however, from 1991q3 to 1998-q3 the asset prices are increasing and the velocity is declining. a jump in velocity can be observed from the figure in the year 1999 due to the events that took place in 1998. during 1998 pakistan conducted nuclear bomb tests which led to certain economic sanctions on the economy. this action reduced the net foreign assets of the country and the foreign exchange reserves fell to an extremely low levels of 1.5 billion dollars, thereby leading to a fall in the money supply in the economy. given the level of income, this fall in the money supply produced a jump in the velocity. from 1999 q1 to 2003 q4 asset prices are falling and so is the velocity of money. this time period shows that asset prices are not helpful in explaining the velocity decline. however, asset prices started to increase from 2004 q1, but the velocity of money continued to fall till 2007 q3. after 2007 q3 even with the increase in asset prices till 2009q1, the velocity of money is also increasing during this time period and even after that till 2011 q1. this disturbance in the relationship can be explained by variables other than asset prices. asset transaction during this period can explain the 0.5 0.6 0.7 0.8 0.9 0.15 0.2 0.25 0.3 0.35 0.4 1 9 8 2 q 3 1 9 8 3 q 4 1 9 8 5 q 1 1 9 8 6 q 2 1 9 8 7 q 3 1 9 8 8 q 4 1 9 9 0 q 1 1 9 9 1 q 2 1 9 9 2 q 3 1 9 9 3 q 4 1 9 9 5 q 1 1 9 9 6 q 2 1 9 9 7 q 3 1 9 9 8 q 4 2 0 0 0 q 1 2 0 0 1 q 2 2 0 0 2 q 3 2 0 0 3 q 4 2 0 0 5 q 1 2 0 0 6 q 2 2 0 0 7 q 3 2 0 0 8 q 4 2 0 1 0 q 1 2 0 1 1 q 2 2 0 1 2 q 3 2 0 1 3 q 4 2 0 1 5 q 1 2 0 1 6 q 2 figure 1: velocity decline and inverse asset prices velocitym2sa invlrealassetprice journal of applied economics and business studies, volume. 5, issue 1 (2021) 191-214 https://doi.org/10.34260/jaebs.510 203 increase in the velocity of money. during this time period oil prices soared to their record high levels and also the exchange rate and the consumer prices jumped up at the same time. due to increased oil prices, our foreign exchange reserves decreased to a large extent which led to decrease in the net foreign assets of the economy. therefore, money supply in the economy decreased, thereby leading to a rise in velocity of money at a given income level. moreover, it can be inferred from the figure that the relationship between the velocity of money and asset prices was disturbed over the period 2007q3 to 2011q1. asset prices do not seem to explain the velocity decline phenomenon during this time period. however, prior to 2007 and from 2011 to 2016 the increasing asset prices explain decline in velocity of money. 4.2 testing stationarity of the variables testing unit root in the variables is necessary first step in the time series analysis. analysis very much depends on the order of integration of the variables included in the model. we, therefore test unit root in each of the variable. as power of unit root tests is quite week so results from only one test cannot be relied. therefore, two unit root tests – augmented dickey-fuller test and phillips-perron test – have been used. in unit root test specifications variables are taken in logarithmic form, except interest rate that is taken in level form. all variables are assumed to have deterministic trend in log-level form but only intercept in first differenced form. akaike information criterion has been used to selecet lag length in unit root test equation. . results in table 1 show that all variables are non-stationary at level but stationary at first difference. in this case, regression results are spurious unless the variables in the regression are cointegrated. table 1: results of unit root test adf stat pp stat level fd level fd money supply -1.82 -10.14*** -2.02 -10.31*** gdp deflator -2.12 -18.02*** -2.41 -17.27*** real gdp -2.29 -6.59*** -1.04 -33.26*** asset price index -2.46 -3.18*** -1.87 -3.59** short term interest rate -2.33 -3.75*** -2.35 -7.99*** long term interest rate -1.83 -3.73*** -0.96 -7.97*** note: all variables except interest rates are in logarithmic form. **(***) show statistical significance at 5% (1%) significance level. adf referes to augmented dickey-fuller, and pp referes to phillipsperron. fd stands for first difference. 4.3 asset prices and the velocity of money: bivariate analysis hina shafiq, wasim shahid malik 204 in the first step we start our formal analysis by using bi-variate regression model. the results of this simple model of velocity and asset price indices are given in the following table (2). table 2: effect of asset prices on velocity of money: bivariate analysis velocity velocity velocity velocity constant 0.04 (0.07) -0.02 (0.02) 0.64 (0.00) 0.23 (0.00) asset price (sr) -0.02 (0.01) -0.05 (0.00) -0.24 (0.00) -0.12 (0.00) asset price (lr) -0.21 -0.39 -0.69 -0.97 adjusted rsquare 0.91 0.91 0.93 0.91 q-stat at lag 1 0.05 (0.82) 0.32 (0.57) 1.60 (0.21) 0.63 (0.43) q-stat at lag 4 3.00 (0.56) 4.35 (0.36) 6.04 (0.196) 4.52 (0.30) eg tau stats -3.09 (0.1) -2.16 (0.15) -3.95 (0.03) -2.63 (0.23) note: parentheses contain probability values for accepting null hypothesis (that the coefficient is zero). akaike information criterion is used for lag length selection. all variables are taken after seasonal adjustment. all variables except interest rate are taken in log form. eg refers to engle and granger (test of cointegration). we have performed this regression with four different measures of asset price index constructed with each of the procedure of assigning weights. in all specifications, the short run coefficient of asset price index is negative and statistically significant at less than 1% level of significance except for specification one (which is significant at 1% level of significance). the coefficient of asset price index ranges from -0.02 to -0.24. this signifies that asset prices have an inverse relationship with the income velocity of money in pakistan; thereby an increase in asset prices makes the velocity decline. our findings are in conformity with borio, kennedy and prowse (1994). also the long run coefficient of asset price in all regressions is negative and ranges between -0.21 to 0.97; long run coefficient shows that the relationship is quite strong in magnitude. all of these results show that the apparent decline in the income velocity of money over time is associated with asset prices. money is used in transactions of goods as well as asset. but in calculating income velocity of money, only goods market transactions are included, while asset market transactions are ignored. this ignorance makes income velocity of money decline over time as asset prices gain momentum. this has been captured by the negative coefficient of asset prices in the regression of income velocity of money. we have used 2 lags for model 1, 2 and 4 whereas model 3 uses 5 lags of the dependent variable. it can be seen from results that inclusion of asset prices in the journal of applied economics and business studies, volume. 5, issue 1 (2021) 191-214 https://doi.org/10.34260/jaebs.510 205 regression increases explanatory power as shown by increase in adjusted r-square. ljung box q stats for lag 1 and 4 have been reported due to quarterly frequency of our data. the results show that there is no autocorrelation in our errors. eg (engle-granger) tau stats shows that the variables are cointegrated in models 1 and 3 at 10% and 5% level of significance respectively. it is noteworthy that asset price indices are dominated by one type of asset price i.e. house prices, therefore, we have also used individual asset prices in place of asset price index in the same regression to analyze their effects on the velocity of money. the results are shown in table 3. we still have consistent results with different types of asset prices. the coefficients of house prices, exchange rate and the share prices have the negative sign and all are statistically significant at less than 1% level of significance. these results indicate that an increase in all types of asset prices leads to a fall in the velocity of money in pakistan. the long run coefficients are high in magnitude and indicates a long run negative relationship between asset prices and the income velocity of money. as all of the asset prices have increased in the past and all are ignored while calculating income velocity of money, therefore, each asset price is negatively associated with the velocity. this also shows that the velocity decline phenomenon cannot be associated with just one asset; rather, it can be generalized with respect to all asset prices. money used in each asset’s transactions is responsible for velocity decline, if that is not considered in finding the income velocity of money. table 3: effect of asset prices on velocity of money short run coefficient p-value long run coefficient intercept 0.75 (0.00) 2.44 house price index -0.08 (0.00) -0.25 exchange rate -0.21 (0.00) -0.69 share price index -0.06 (0.00) -0.20 adjusted r2 0.92 q-stat at lag 1 0.96 (0.33) q-stat at lag 4 5.04 (0.28) eg tau stats -4.88 (0.01) note: parentheses contain probability values for accepting null hypothesis (that the coefficient is zero). akaike information criterion is used for lag length selection. all variables are taken after seasonal adjustment. all variables except interest rate are taken in log form. eg refers to engle and granger (test of cointegration). we have used 3 lags for this specification. the adjusted r square is high, the q-stat is showing no autocorrelation of errors and the variables are found cointegrated. these results are in conformity with our hypothesis that asset prices play a vital role in explaining the velocity decline phenomenon in pakistan. as mentioned in section 2, two types of transactions take place in an economy, one is gdp transactions and the hina shafiq, wasim shahid malik 206 other is non-gdp transactions. if the value of non-gdp transaction rises, then even with constant gdp transactions, the equation of exchange shows a fall in the velocity, which explains all the phenomena which are considered as the breakdown of the quantity theory of money. therefore, it is important to include asset prices in the estimation of the quantity theory of money because movements in asset prices can explain the phenomenon of velocity decline and missing money appropriately. 4.4 asset prices and the velocity of money: multivariate analysis for the sake of conformity of the previous section’s results we add control variables, one by one, to our bivariate regression analysis and perform multivariate regression and vecm as well. the results are summarized in table 42. table 4: effect of asset prices on velocity of money: multivariate analysis regression analysis vecm specification 1 specification 2 specification 3 specification 4 constant 0.019 (0.24) 0.02 (0.27) 0.03 (0.24) -0.11 (0.00) 0.04 [0.49] asset prices -0.14 (0.01) -0.18 (0.05) -0.20 (0.05) -0.10 (0.00) -0.20 [-12.17] permanent income ---0.00 (0.52) 0.00 (0.45) 0.00 (0.16) 0.00 [4.27] transitory income ------0.00 (0.00) 0.00 (0.00) 0.00 [15.63] rs ---------0.04 (0.00) 0.02 [3.52] rl ---------0.00 (0.42) 0.01 [3.94] adjusted r2 0.91 0.91 0.91 0.92 0.19 q-stat at lag 1 0.36 (0.55) 0.37 (0.54) 0.93 (0.34) 1.36 (0.22) --- q-stat at lag 4 5.25 (0.26) 4.07 (0.40) 3.92 (0.42) 3.59 (0.47) --- lm stat (1) ------------431.59 (0.00) lm stat (4) ------------266.53 (0.00) trace stat ------------330.91 (0.00) max eigen stat ------------209.15 (0.00) note: parentheses contain probability values for accepting null hypothesis (that the coefficient is zero) while brackets contain t-values. akaike information criterion is used for lag length selection. 2 results of cusums test for the first equation of vecm are given in the appendix, which show that coefficients are stable. journal of applied economics and business studies, volume. 5, issue 1 (2021) 191-214 https://doi.org/10.34260/jaebs.510 207 for lag selection in vecm, different test statistics, lr, fpe, aic, sic, and hq are used. all variables are taken after seasonal adjustment. all variables except interest rate are taken in log form. eg refers to engle and granger (test of cointegration). it is found that the asset prices are significant determinant of velocity of money and the relationship is found negative in all specifications based on different weights criteria used in the construction of asset price indices. this implies that our bivariate regression results are reliable and consistent. the coefficient of permanent income is insignificant for all specifications in the regression but significant in vecm having negligible value in magnitude. however, the coefficient of transitory income is positive, trivial in magnitude but statistically significant and less than unity which means that the velocity of money is pro-cyclical. transitory income is a measure of business cycle and a positive coefficient implies that during booms velocity increases while it decreases during recessions. the regression and vecm analyses show that coefficients of short run and long run rate of interest are significant and positive. this result implies that an increase in the interest rate reduces the demand for money balances thereby leading to a rise in velocity with a given level of income. 4.5 velocity decline and asset prices: sub-samples it is evident from the results of the previous section that asset prices have an important role in explaining the velocity decline phenomenon. however, if we consider the figure 1 in section 4.1, we have seen that there are certain time periods in which asset prices do not seem to explain movements in the velocity of money. therefore, we perform regression of the previous section with three sub-samples divided according to the time periods in which the relationship between velocity of money and asset prices seemed to be strong and weak. the summary of results is given in table 5. table: 5 effect of asset prices on velocity of money: sub-samples 1981-1998 1998-2008 2008-2016 constant 0.03 (0.18) -0.15 (0.32) 0.40 (0.01) asset price (sr) -0.03 (0.00) 0.03 (0.46) -0.11 (0.00) asset price (lr) -0.11 0.35 -0.36 adjusted r2 0.80 0.81 0.78 q-stat at lag 1 0.09 (0.77) 0.05 (0.83) 0.01 (0.93) q-stat al lag 4 0.82 (0.94) 1.91 (0.75) 0.75 (0.945) note: parentheses contain probability values for accepting null hypothesis (that the coefficient is zero). akaike information criterion is used for lag length selection. all variables are taken after seasonal adjustment. all variables except interest rate are taken in log form. eg refers to engle and granger (test of cointegration). hina shafiq, wasim shahid malik 208 we have used 5 lags for the first and third sub-samples, whereas the second subsample uses 2 lags. the adjusted r-square is high in all the three samples and the q-stat is showing that there is no autocorrelation at 1st and 4th lag. the results of first and the third sub-samples are showing that asset prices explain the velocity decline phenomenon for the time period 1981-1998 and from 2008-2016. these are the two periods when continuous increase in asset prices was matched with continuous decline in velocity. however, as was observed in the figure 1, asset prices do not seem to explain the movements in velocity of money for the time period between 1998 and 2008. the coefficient of asset prices for this time period has positive sign and is also statistically insignificant which implies that no clear relationship can be explained between asset prices and the velocity of money during this time period. over this sample period, velocity was continuously declining while asset prices show a ushapped pattern initially decreasing and then increasing. that’s why we find a statistically insignificant relationship between velocity and asset prices in the sample. we can find a significant coefficient if we divide this subsample into two further subsamples based on the two sides of u (of asset prices) but instead of doing that we have done the analysis using indicator function as discussed below. 4.6 velocity decline and asset prices: indicator function in order to analyze whether the results of section 4.2 hold on average only or the asset prices explain both the phenomena of velocity rise and decline, we run the regression of asset prices and velocity of money using indicator function based on rising and declining velocity. as mentioned in section 3.1 the indicator function (i) takes the value 1 for positive change in velocity and zero otherwise. the results are given in table 6. table 6: role of asset prices in explaining fluctuations in velocity velocity of money velocity of money velocity of money velocity of money s h o rt r u n c o e ff ic ie n ts i 0.08 (0.00) 0.04 (0.00) 0.02 (0.21) 0.20 (0.00) 1-i 0.00 (0.93) -0.03 (0.00) -0.04 (0.00) 0.11 (0.03) i*asset prices -0.02 (0.00) -0.03 (0.00) 0.02 (0.00) -0.08 (0.00) (1-i)*asset prices -0.01 (0.06) -0.02 (0.03) 0.01 (0.35) -0.06 (0.01) l o n g r u n c o e ff ic ie n ts i 2.56 1.13 0.79 4.71 1-i 0.04 -0.82 -1.83 2.52 i*asset prices -0.49 -0.85 0.77 -1.86 (1-i)*asset prices -0.32 -0.58 0.25 -1.51 journal of applied economics and business studies, volume. 5, issue 1 (2021) 191-214 https://doi.org/10.34260/jaebs.510 209 adj r2 0.95 0.95 0.95 0.96 q-stat at lag 1 0.06 (0.82) 0.04 (0.84) 0.14 (0.70) 0.00 (0.98) q-stat at lag 4 4.10 (0.39) 3.88 (0.42) 5.20 (0.27) 3.78 (0.44) wald test 0.87 (0.35) 0.84 (0.36) 3.47 (0.06) 0.34 (0.56) note: parentheses contain probability values for accepting null hypothesis (that the coefficient is zero). akaike information criterion is used for lag length selection. all variables are taken after seasonal adjustment. all variables except interest rate are taken in log form. eg refers to engle and granger (test of cointegration) the coefficients of both i*asset prices and (1-i)*asset prices are negative and statistically significant in three out of four specifications which means that our results of section 4.3 are robust. this means that asset prices not only explain velocity decline but also the increase in velocity. furthermore, we have also checked the symmetry of these coefficients by wald test which accepts the null hypothesis that coefficients in both states, velocity decline and rise, are equal. the explanatory power of the models is very high and there is no autocorrelation of errors. 4 conclusion the objective of the study is to explain the increase in velocity and velocity decline phenomenon in a model that includes asset prices. we use asset prices as the candidate variable to explain the velocity decline phenomenon in pakistan for a time period spanning1981-q1 to 2018-q2. the empirical analysis shows a significant role of asset prices in explaining the velocity decline phenomenon. the asset price index has negative coefficient and ranges from -0.02 to -0.24. this indicates an inverse relationship between asset prices and the velocity whereby, an increase in the asset prices are associated with a decline in the income velocity of money. our findings are in conformity with the borio, kennedy and prowse (1994). moreover, the sub-sample regression for the weak and strong relationship between asset prices and velocity of money shows that in the first and the third sub sample asset prices were explaining the velocity decline phenomenon for the time period 1981-1998 and from 2008-2018. however, asset prices do not seem to explain the movements in velocity of money for the time period between 1998 and 2008. the results of the indicator function show that the asset prices not only explain velocity decline but also the increase in income velocity of money. the results of the study are consistent with our hypotheses and indicate that asset prices significantly explain both the velocity decline and increase phenomena in different time periods in pakistan. it is important to distinguish between transactions hina shafiq, wasim shahid malik 210 that are part of gdp and those excluded from gdp because gdp is only a subset of the total transactions and it does not represent all the transactions in the economy. the equation of exchange shows a decline in the income velocity of money when the value of financial or asset transactions rises, and that explains all the phenomena which are considered as the breakdown of the quantity theory of money. therefore, the research in the field of velocity and quantity theory of money in pakistan should focus and incorporate asset prices for an accurate estimation and explanation of these phenomena. this inclusion of asset price indices or asset prices provides useful insights into the movements of income velocity of money even if gdp based transactions remain constant. results of this paper implicate changes in policy. first, on the basis of volatility of velocity of money and therefore unstable money demand function, monetary aggregates cannot be ignored as indicators of monetary policy. velocity of money is not constant only if it is constructed using wrong proxy of economic activity or economic transactions. once, all types of economic transactions are included, velocity becomes constant and monetary aggregates become important for monetary policy. second, asset price index, being an important indicator of monetary policy, deserves appropriate weight in monetary policy decisions. the weakening of relationship between inflation and monetary aggregates, over time, is a result of specification bias through ignoring asset prices. appropriate specifications of models that include asset prices find significant relationship between asset prices and inflation. finally, more research is needed to investigate the suitability of quantity of money or its price as monetary policy instrument, especially for the case of pakistan. our study can be further extended to incorporate the effects of modern technologies in payments on velocity of money. for instance, chaudhari et al. (2020) finds statistically significant long run effect of broader measure of technology on velocity and concludes that advances in payment technologies have significant negative effect on money demand. franco (2015) finds that more use of bitcoins can lead to increase in velocity of money. durgun and timur (2015) put forward that only exceptional groups use electronic payments system, therefore the developments in payment technologies have little effect on monetary policy. holly (1999) finds insignificant effect of advances in payment technology on the co-movement of money supply, velocity, interest rate and income. in case of pakistan, mumtaz and smith (2020) finds that there is no change in velocity of money and money multiplier in pre-fintech periods as compared to those in post-fintech periods. we suggest combining the role of asset prices with advances in payment technology to see their joint effect on velocity of money. journal of applied economics and business studies, volume. 5, issue 1 (2021) 191-214 https://doi.org/10.34260/jaebs.510 211 acknowledgement: we are thankful for the valuable comments of dr. saghir pervaiz ghouri, dr. asad zaman and dr. ather maqsood. we also acknowledge the comments of sbp staff on an earlier version of the paper that the second author presented at sbp. references: ahmad, s., f. pasha, and m. rehman, (2016), the role of money in explaining business cycles for a developing economy: the case of pakistan, sbp working paper series no. 74. anderson, p. s,. (1985), the stability of money demand functions: an alternative approach. bis economic papers, no. 14. arby, m. farooq (2008), some issues in the national income accounts of pakistan (rebasing, quarterly and provincial accounts and growth accounting), phd dissertation, pakistan institute of development economics, islamabad. azim, p., ahmed, n., ullah, s., zaman, bedi and zakariya, m. (2010). demand for money in pakistan: an ardl approach. global journal of management and business research, (10), 76-80. barro, r. j. (1989). interest rate targetting, journal of monetary economics, 23, pg 330. belongia, m. t. and chalfant, j. a. (1990). alternative measures of money as indicators of inflation: a survey and some new evidence, federal reserve bank of st. louis. bilquees, f. and r. shehnaz (1994). income velocity and per capita income in pakistan: 1974-75 to 1991-92. the pakistan development review, 33:4(2): 985– 995. boughton, j. m. (1991). long-run money demand in large industrial countries, imf staff papers, vol. 38, no. 1, march, pp. 1-32 chaudhry, i. s., ismail, r., farooq, f., murtaza, g. (2015). monetary policy and its inflationary pressure in pakistan, pakistan economic and social review, vol. 53, no.02, pp 251-268. chaudhari, d. r., dhal, s., & adki, s. m. (2020). payment system innovation and currency demand in india: some applied perspective. rbi occasional papers, 40(20), 33-58. durgun, ö., & timur, m. c. (2015). the effects of electronic payments on monetary policies and central banks. procedia-social and behavioral sciences, 195, 680– 685. hina shafiq, wasim shahid malik 212 estrella, arturo and frederic. s. (1997), is there a role for monetary aggregates in the conduct of monetary policy, journal of monetary economics, 40(2), 279304. faridi, m. z. & akhtar, m. h. (2013), an estimation of money demand function in pakistan: bound testing approach to co-integration. pakistan journal of social sciences, vol. 33. no. 1, pp 11-24. farooq, omer., saeed ul hassan, qazi., and m. shahid, (2015), the applicability of quantity theory of money in case of pakistan: a time series analysis, journal of economics and sustainble development, vol.6, no.3. fisher, irving (1911), the purchasing power of money, its determination and relation to credit, interest and crises, new york: macmillan. franco, pedro. (2015). “understanding bitcoin, cryptography, engineering and economics”, wiley finance series, 1, united kingdom. goldfeld, s. m. (1976), the case of the missing money, brookings papers on economic activity, no. 3, pp. 683-730. goodhart, c. and hofmann, b., (2008). house prices, money, credit and the macroeconomy. european central bank working paper series, no. 888. gordon, r. j. (1984). the 1981-82 velocity decline: a structural shift in income or money demand? nber working paper, no. 1343. government of pakistan (2015). hand book of statistic on pakistan economy, state bank of pakistan. government of pakistan (various years). pakistan economic survey, ministry of finance, islamabad. hanif, m.n. (2014), monetary policy experience of pakistan, mpra paper, no. 60855. hanif, m.n, iqbal, j. and malik, m. j. (2013). quarterisation of national income accounts of pakitan. sbp working paper series, no. 54. hendry, d. f. (1985). monetary economic myth and econometric reality, oxford review of economic policy, vol. 1, no. 1, 72-84. holly, patrick jr. (1999) "the effect of technology growth on money supply and demand: a cointegration approach," the park place economist: vol. 7 howells, p. g. a. (1995). the demand for endogenous money. journal of post keynesian economics, 18: 89-106. kaldor, nicholas. (1985), how monetarism failed, challenge, vol. 28, no. 2 pp. 4-13. khan, m. s. and a. schimmelpfennig (2006), inflation in pakistan: money or wheat? international monetary fund working paper. sbp-research bulletin, volume 2(1), pp. 213-234. journal of applied economics and business studies, volume. 5, issue 1 (2021) 191-214 https://doi.org/10.34260/jaebs.510 213 kemal, a. (2006). is inflation in pakistan a monetary phenomenon?, pakistan development review, 45: 213–220. mumtaz m. z., and smith z. a. (2020) empirical examination of the role of fintech in monetary policy, pacific economic review, 25(5): 620-640. omer, m. and o. f. saqib (2009). monetary targeting in pakistan: a skeptical note. sbp research bulletin, 5(1): 53-81. omer, m. (2010). velocity of money functions in pakistan and lessons for monetary policy. sbp research bulletin, 6 (2), 37-55.palley, t. i. (1996). palley, t. i. (1996). accommodationism versus structuralism: time for an accommodation. journal of post keynesian economics, 18: 585–594. pollin, r. (1991). two theories of money endogeneity: some empirical evidence, journal of post keynesian economics, vol.13, no.3. qayyum, a. (2006), money, inflation, and growth in pakistan. the pakistan development review, volume 45(2), pp. 203-212. spindt, paul a. (1985), money is what money does: monetary aggregation and the equation of exchange, journal of political economy, vol. 93, pp. 175-204. taylor, j. b. (1993) discretion versus policy rules in practice. carnegie-rochester conference series on public policy 39, 195–214. werner, r. a. (1992), towards a quantity theorem of disaggregated credit and international capital flows, paper presented at the royal economic society annual conference, york, april 1993, and at the 5th annual pacap conference on pacific-asian capital markets in kuala lumpur, june 1993. werner, r. a. (1997c), towards a new monetary paradigm: a quantity theorem of disaggregated credit, with evidence from japan, kredit und kapital, vol. 30, no. 2, pp.276-309. werner, r. a. (2012), towards a new research programme on ‘banking and the economy’implications of the quantity theory of credit for the prevention and resolution of banking and debt crises, european conference on banking and the economy. werner, r. a., and lyonnet, v. (2012), lessons from the bank of england on ‘quantitative easing’ and other ‘unconventional monetary policies, international review of financial analysis. vol. 5, pp. 94-105. hina shafiq, wasim shahid malik 214 appendix figure 2: cusums test for first equation of vecm -40 -30 -20 -10 0 10 20 30 40 84 86 88 90 92 94 96 98 00 02 04 06 08 10 12 14 16 cusum 5% significance journal of applied economics and business studies, volume. 5, issue 4 (2021) 41-58 https://doi.org/10.34260/jaebs.543 41 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x corporate governance and shareholders’ confidence: a pre-post analysis of corporate governance code 2017 in pakistan sajid ullah1, muhammad zahid2, muhammad fayaz3 & muhammad saad4 1 department of management sciences, city university of science & it peshawar, khyber pakhtunkhwa, pakistan 2 department of management sciences, city university of science & it peshawar, khyber pakhtunkhwa, pakistan 3 ph.d. fellow, sarhad university of science & it, peshawar 4 institute of management sciences, university of science & technology bannu, kp, pakistan abstract this study aims to examine the impact of the recently introduced corporate governance code (cgc) of the security exchange commission of pakistan (secp) on shareholders' confidence. the four recommendations of the cgc 2017 included in the study are female directors, independent directors, independent chairman of the audit committee, and board size. the study uses a purposive sample of the top 100 companies listed security exchange commission of pakistan with a plea for better compliance. by employing ordinary least squares with panel corrected standard errors (ols-pcse) to a data collected from annual reports of the sample firms from 2015 to 2019. the secp code 2017 anticipated that cg dimensions will improve firm performance and shareholders’ confidence. the findings are fairly consistent and robust across two time periods (pre and post), whereas the results show that board diversity increases firm performance in both eras. besides, the multiple regression results indicate that firm performance is negatively and significantly linked with independent directors. nevertheless, the results show that board size and independent chairperson of the audit committee do not have a significant impact on firm performance. as these anticipations yet lack empirical support. keywords board size, female directors, independent directors, independent chairperson of the audit committee, secp 2017, pre-post analysis, shareholders confidence, pakistan jel classification g30, g39 sajid ullah, muhammad zahid, muhammad fayaz & muhammad saad 42 the findings of this study provide important insights for the regulatory bodies, policymakers, and all other key stakeholders of the public listed companies of pakistan. the study empirically investigates the recommendations of the cgc 2017 and has originality or value, particularly for the developing countries. 1. introduction over the last few years, corporate governance (cg), in general, has gained considerable public attention because of its role in corporate sustainability and development. broadly, cg aims to discipline corporations for increasing shareholders' confidence (nikhil & shil, 2017). cg is a structure that regulates and controls business entities (nodir & amonboyev, 2016). effective governance firms’ overall conduct as per the relevant regulations and regulations. it is believed that efficient use of corporate governance increases the value of a firm (nazir & afza, 2018) a good and effective corporate governance system is measured as an internal device for administration governor and monitoring. it is an active means for receiving the best financial performance of the firms and shareholders' confidence (richard & david gwilliam, 2019). therefore, a good and sound corporate governance structure will care for the right of shareholders, increase and built company transparency, and create a superior system of the closure of financial and non-financial information (bernard & woochan, 2015). thus a good corporate governance structure should guarantee that the board of directors is not only responsible for different stakeholders' communities but also encourage transparency over accurate disclosure and sharing of information among different investors communities. so if shareholders are not happy with the corporate governance internal system, level of disclosure, then they will have to compile out its management, and finally, this could compile the investors for capital flight. thus a good and sound corporate governance system has the power to impact the future position of a corporation and also plays an important role in financial market strength and stability (mugarura, 2016). corporate governance consists of internal and external mechanisms designed to minimize agency charges. the internal mechanisms of corporate governance mostly focused on the association between different contributors in an organization and their association with the board of directors. similarly, on the other side, the external mechanisms manage the association between internal and external such as shareholders, suppliers, creditors, and society for the smooth operation of an organization. (nikhil & netai, 2017). accordingly, it is expected that well-built corporation governance roles and their proper execution make journal of applied economics and business studies, volume. 5, issue 4 (2021) 41-58 https://doi.org/10.34260/jaebs.543 43 the sure proper and charming running of companies and their effect on the cash flow of a government and investors (haseeb & ibrahim, 2015). great corporate scandals and frauds such as world com, enron, and tyco, etc. in developed and industrialized countries and other cases of the east and asia economic crises raise deep blemish and questions on the efficiency and effectiveness of corporate governance mechanisms in these organizations. (bhagat & bolton, 2019). the scandals which occur due to a weak corporate governance structure, compiled regularity bodies to make rules, against protests of corruption, frauds, duplicity, and insider trading. hence these scandals and conflicts among management and owners were the key causes of the creation of codes and sound corporate governance systems all over the world (kursat, 2019). in pakistan, from the last few years' corporate governance codes repeats itself over and over to improve its structure and efficiency according to the needs, but there was still a gap as the recommendations regarding the subject to the code has not been properly implemented, as many companies are fulfilling formalities and merely making paper arrangements instead of practical implementation (fazli azim & mustapa, 2018). there is a limited literature study on pre and post contents in pakistan and around the world due to the contemporary data, so this study evaluates the pre and posts effect of the execution of the new corporate governance code 2017 on shareholders' confidence (rashid zama & ayub, 2015). the main objective of the code of corporate governance and its revision from time to time is to make the governance structure in line with updated requirements as well provide enough security to the shareholders of the company. the developing economy like pakistan also restructured the code from time to time after its inception in 2002. the current study also highlighted this issue and analyzed the situation before and after the latest code of 2017. the remainder of this paper deals with the literature review and hypotheses development followed by research methods, results, and discussion. the last section reports the conclusion and recommendations of the study. 2. literature review 2.1 theoretical framework the theoretical framework of the study discusses as follows. 2.1.1. agency theory agency and stakeholders' perceptions remain the most broadly used theories in the literature in this area, explaining the need for proper monitoring services. sound corporate sajid ullah, muhammad zahid, muhammad fayaz & muhammad saad 44 governance can decrease the possibility of earning management by managers and subsequently increase the superiority of the accounting material communicated to the stakeholders (riadh & najoua, 2020). therefore to diminish this problem and to boost the interests of owners, the company used extra costs for observing management actions (herlambang & murhadi, 2020). so from an agency theory perspective, embracing the corporate governance codes enable firms to monitor administrative actions which reduce the chances of the principal-agent problem and increase the shareholders' confidence (junaid & muzzammil, 2020). 2.1.2. shareholder primacy theory milton friedman in 1970 marked his view about shareholder dominance; he wrote that "a business administrative is a worker of the owners of the corporation. the worker is directly accountable for his/ her activities toward his owners. that responsibility includes that the management operates the business according to their mind and wishes, for the purpose to increase shareholders' confidence and business performance (lenore, 2019). thus shareholder primacy theory proposes that the main purpose of corporations is to maximize owners' profit and interests. however, managers want to increase and protect their objectives rather than investors. hence, this is the responsibility of the corporate governance to monitor the managers, but this act can be only efficiently executed if the board is independent (haseeb & ibrahim, 2015). 2.2 the context of pakistan after the scandals of wells fargo and equifax are just the most recent in the long line of scandals involving large well-known public u.s. corporations. further back in time, at the turn of the new millennium, the scandals in enron, worldcom, tyco, and qwest headed to their destroying increased the importance of corporate governance (bhagat & bolton, 2019). after each set of these scandals, policymakers raised questions about the effectiveness of corporate governance mechanisms in these companies. this led to the predictable call for more regulation and laws to oblige and regulate corporate behavior, to wit, the sarbanes oxley act of 2002 and the dodd-frank act of 2010, for diminishing conflict of interests between investors and management of various organizations (jamshaid & khalid, 2018). in pakistan, the concept of corporate governance was presented by the security exchange commission of pakistan (secp) in 2002 and later revised again and again according to the changing circumstances to increase shareholders' confidence. corporate governance systems and shareholders' confidence have always been an area of debate but journal of applied economics and business studies, volume. 5, issue 4 (2021) 41-58 https://doi.org/10.34260/jaebs.543 45 still, there is very narrow literature support in governance mechanisms (waqar & subhan, 2019). so the current study identifies the pre and post-analysis of corporate governance code 2017 and its effect on shareholders' confidence in pakistan. the development of corporate governance in pakistan took place in many phases. in pakistan before 1997, the responsible bodies were the ministry of finance, corporate law authority, and the state bank of pakistan for regulating and imposing corporate governance laws. after that during the era of 1999-2002, which was a structural formation stage, the security exchange commission of pakistan (secp) came into existence by replacing the corporate law authority (emerald, 2018). likewise to enhance confidence and create awareness among the business community about the benefits of good governance the pakistan institute of corporate governance (picg) was established in 2004 with the collaboration of the international finance corporation (arslan & ahmad, 2020). in early 2002 the security exchange commission of pakistan presented the code of corporate governance as the key step in corporate governance modifications in pakistan. the major areas of implementation exclude reforms of the board of executives to make it responsive to all investors (qaiser & harry, 2011). although the code of corporate governance 2002, is based on corporate ordinance 1984, however, the code's limited necessities on the director's independence remain intentional and deliver no direction on internal controls and board reimbursement strategies (javid & iqbal, 2010). thus the securities and exchange commission of pakistan (secp), after gaining new experiences at national and international levels reviewed and revised 2012 the corporate governance parameter of the state (zahid & haseeb ur rahman, 2019). furthermore to meet international standards, ratified the improved and updated version of corporate law in the shape of companies' code 2017, with technical and financial support provided by international organizations for the purpose to regulate and execute corporate code for getting standard passion and shareholders' confidence at a foreign and local level (atta & haider, 2020). the current study model includes the following independent variables. 2.3 board sizes and firm performance the total number of directors is recognized as the size of the board. it plays a significant role in the effectiveness of the board and also has a great impact on the quality of governance (sheikh & shah, 2018). different countries have diverse board sizes as every country has its own cultures and procedure about business (jason & michael, 2017) therefore several studies argue that a rising in board size may clue to greater coordination problems which become more difficult for board members to reach an agreement on sajid ullah, muhammad zahid, muhammad fayaz & muhammad saad 46 crucial corporate decisions which decrease the effectiveness of the firm and shareholders’ confidence. furthermore, board larger size is also less effective in observing management and, thereby, the ceo influencing increased in decisions making (aswathy & chandramohan, 2018). however, some of these studies support the opposite view that a board’s capacity to conduct monitoring increases with size. larger boards deliver an increased group of expert members with having a greater variety of skills and experience which, in reply, leads to enhanced firm performance due to upright business communication, decision making, and controlling (kamran & mehmet, 2021). as firms in belgium, france, spain, and germany tend to have a large board size (thirteen to nineteen members) therefore epstein claimed that an average of sixteen directors may be considered optimal for large companies. but on the other hand united kingdom, switzerland and holland tend to have a small board size, so lipton claimed that the board members on board should be between eight and nine while leblanc preferred eight to eleven persons on board (shafie & kamilah, 2016). however, the security exchange commission of pakistan code 2017 claimed that seven to fifteen members are an optimal board size for all listed companies. the board of directors shall act as a possessor of a business on behalf of owners to take any action and decisions regarding the best interests of the shareholders. therefore the board size plays an important role in increasing the shareholders' confidence and company efficiency (wahid & kauser, 2020). h1 (a): there will be a significant impact of board size on shareholders' confidence. h1 (b): the relationship between board size and shareholders' confidence will be improved after the promulgation of the corporate governance code 2017. 2.4 board female director and firm performance gender diversity of the board has an optimistic influence on the outcome and efficiency of the companies (al-rahahleh & ayat, 2017). the gender diversity matter has gained more importance in current years, because female directors may improve boards' negotiations by carrying various perspectives and ideas which enhance the efficiency of boards for the decisions making. female directors may also enhance board independence, by asking different questions from their male colleagues which decreases materialistic behavior, this, in turn, motivates companies to consider stakeholders' confidence and demands (shahbaz & merve, 2020). the previous studies stated that lack of diversity in a board of directors leads to follow similar ways of solving company problems that can develop group think issues, ineffective boards, poor governance, and as well as hurdles in the attainment objectives of the firms' which affect investors' confidence (srinidhi & journal of applied economics and business studies, volume. 5, issue 4 (2021) 41-58 https://doi.org/10.34260/jaebs.543 47 shiqiang, 2020). h2 (a): there will be a significant impact on female directors 'shareholder confidence. h2 (b): the relationship between female directors and shareholders' confidence will be improved after the promulgation of the corporate governance code 2017. 2.5 board independent directors and firm performance independent directors of the boards represent the ratio of the outside directors of the board who does not have a financial relationship with the business (akshita & chandan, 2016). thus it recommends that external directors have more talents and knowledge about the market, so independent directors on a board offer a brilliant opportunity to monitor the firm's performance and shareholders' confidence (junaid & muzzammil, 2020). the independent directors play a vital part in speaking out in support of strict tractability with the law and protection of minority stakeholders' benefits. moreover, independent directors show more compliance in the observance of the regulations and are more concerned about corporate performance and shareholders’’ confidence (valeria, 2019). therefore several studies claim a positive association, such as panasian (2003), which stated that if the ratio of independent directors on board increases then it will be useful for the firm performance and shareholders' confidence. while on the other hand baghat and black (2002) say that independent directors have an inverse influence on firm efficiency and shareholders' confidence, whereas postma (2002) found no relation between independent directors and the financial output of the firms (yussoff & anees, 2016). however, the security exchange commission of pakistan code 2017 makes mandatory the availability of a twothirds ratio of independent directors on the board for all listed companies. h3 (a): there will be a significant impact of independent directors on shareholders' confidence. h3 (b): the relationship between independent directors and shareholders' confidence will be improved after the promulgation of the corporate governance code 2017. 2.6 independence chairperson of audit committee and firm performance the independence of the audit committee plays a vital role in increasing shareholders' confidence by having accounting information and financial matters at hand (wahid & kauser, 2020). for the audit committee's effectiveness, the members and the chairperson of the audit committee must be free from the effect and stresses of management and the board of directors (mohammad rateb, 2018). as to increase the investors' confidence the accurate and true financial information of an organization is needed because based on this information the investment decisions are made by the investors. but, some organizations do not represent a real image of the fiscal position of a company. hence, in such a situation sajid ullah, muhammad zahid, muhammad fayaz & muhammad saad 48 a neutral body like the independent chairperson of the audit committee is needed to monitor and control such illegal events and actions of the corporation (nidhi & anil, 2016). as the conclusions of earlier studies on this relationship are questionable, an autonomous committee of audit does work more efficient as compared with a less independent committee, since the earlier is more expected to provide the best result and observing through its capacity and experience to resist tension from top management (kallamu & mohd saat, 2015). further when independent directors presided over the audit committee then there is a significant influence of the committee on monetary reporting and a lower existence of fake reporting (nekbetweenli, 2017). h4 (a): there will be a significant impact of the independent chairperson of the audit committee on shareholder confidence. h4 (b): the relationship between the independent chairperson of the audit committee and shareholders' confidence will be improved after the promulgation of the corporate governance code 2017. 2.7 control variables besides the use of explanatory variables the model also includes a set of control variables as the second group of independent variables in this study. firm size is measured as the natural log of firm total assets. there is no bond between the researchers on the connection between the business size and its financial performance as adams and ferreira (2009) and krishnan and park (2005) specify that firm size is directly related to firm performance, while carter et al. (2003) fail to do so, as the firm size changed for every firm it depends on the nature of the business (chancharat, 2019). firm age refers to the period that elapsed since the company's incorporation. so there is a vague association between business age and firm performance. mature businesses have better efficiency than newly established businesses due to experience, skill, and quality services, that they have developed over time in the market (jacob & safdar, 2020). leverage is defined as the ratio of the book value of debt to total assets. likewise, firm leverage is also opposed on account of the increasing firm performance and investors’ confidence. the higher value of leverage means that the firm financial performance is not in good condition (hexana, 2020). 3. research methodology the present study employed a positivist approach to testing the aforementioned hypotheses. positivism mechanisms always think positively and encourage exploring happenings and events. so in the positivist approach, the researcher collects the data from journal of applied economics and business studies, volume. 5, issue 4 (2021) 41-58 https://doi.org/10.34260/jaebs.543 49 various resources and then interprets that data in an impartial, proper, and objective way (andriukaitiene, 2013). 3.1 sample and sampling techniques this study examines the pre and post analyses of corporate governance code (secp, 2017) and its effect on shareholder confidence. therefore, the population of the current study is the public listed companies in pakistan. as the total number of listed companies was 559, but the study bound our sample size, so it should consist only of the top hundreds (100) listed companies of the security exchange commission of pakistan (secp). the kse-100 index companies are designed for the measurement of the performance of 100 companies which consists of the major market capitalization. to measure the kse-100 index the aggregate market value is divided by the base value and multiplied by 1000 to get the current index number. the study used the sources of the internet and downloaded the annual report of the target firms from the website of secp, from 2015 to 2019 (nazir & afza, 2018). the main reasons for the selection of these companies are that these are registered with secp, and quickly adopt any reform of the regulators (agyemang, 2017). 3.2 share price the current study adopted the share price for measuring shareholders' confidence. a share price is the price of a single share of several commercial stocks of a firm, derivative, or other financial assets. the share price is not stable but fluctuates according to market situations (joseph, 2018). 3.3 economic model following is the model developed for investigation. sc𝑖𝑡(sp) = β0+ β1indb𝑖𝑡 + β2 fdb𝑖𝑡 + β3bsiz𝑖𝑡+ β4indac𝑖𝑡+ β5fage𝑖𝑡 + β6flev𝑖𝑡+ β7fsize𝑖𝑡+ β8id𝑖𝑡 + β9yd𝑖𝑡 …………………model 1. 3.4 measurement of variables table 1 offers the operationalization of model variables along with the support of former literature. table 1: variable measurements s no parameters explanation references 1 shareholders confidence measured by earnings per share: net profit dividends paid/ shares (mustaruddin & rusnah, 2011) 2 independence directors the proportion of independent directors on (bhagat & bolton, 2019) sajid ullah, muhammad zahid, muhammad fayaz & muhammad saad 50 the board 3 board size total number of directors having board membership (bakhsh & zeeshan, 2019) 4 women directors the proportion of women directors on the board (charumathi & rahman, 2019) 5 independence chairperson of audit committee the proportion of independent directors on the board (nidhi & anil, 2016) 6 firm age present year – incorporation year (bakhsh & zeeshan, 2019) 7 firm leverage total debt ÷ total assets (zahid & haseeb ur rahman, 2019) 4. results and discussion 4.1 analysis of descriptive model the kurtosis and skewness are used in table 2 for checking normality between independent variables. the result identified that the minimum value of the variable board size is 3 and the maximum value is 17, with a standard deviation of 1.942, the mean value of the board size is 8.75. similarly, the variable of female director minimum value is 0, the maximum value is 3 with a standard deviation of 0.681 and the mean value is 0.56. moreover, the lowest valve of independent directors is 0 maximum valve is 9 and the mean value is 2.11 and possesses a 1.416 value for standard deviation. the independent chairperson of the audit committee keeps the minimum value 0, maximum value 4, the average is 1.40 and the standard deviation rate is 0.788. similarly, firm age is the control variable its lowest price is 7, the highest rate is 106 and the standard deviation value is 20.294 and the mean is 41.22. firm size possesses a minimum value of 2 million, a maximum value of 74,517 million, and a mean value of 3,437 million. likewise, the firm leverage keeps a minimum value of 67 thousand, a maximum of 57,896 million, and an average score of 2032 million. table 2: descriptive statistics min max mean s.d skewness kurtosis statistic statistic statistic statistic statistic s.e statistic s.e share price -2.807 2.674 -0.005 0.973 -0.026 0.122 -0.036 0.243 b size 3 17 8.753 1.942 1.124 0.122 1.365 0.243 f director 0 3 0.567 0.681 0.975 0.122 0.320 0.243 journal of applied economics and business studies, volume. 5, issue 4 (2021) 41-58 https://doi.org/10.34260/jaebs.543 51 ind. director 0 9 2.115 1.416 1.427 0.122 2.674 0.243 ind. chairperson 0 4 1.406 0.788 1.422 0.122 2.351 0.243 firm age 7 106 41.227 20.294 0.541 0.122 -0.390 0.243 firm size 2 m 74,517 m 3,437 m 10,251 m 0.000 0.122 -0.159 0.243 leverage 67 k 57,896 m 2032 m 6,902 m 0.000 0.122 -0.159 0.243 4.2 pearson’s correlation matrix the study employed pearson's correlation matrix to check multicollinearity among the variables in the model. it is stated that correlation between variables exists but if the association between two elements is greater than 0.900, it signs that a multicollinearity problem exists between two factors. the statistics reported in table 2 stated that the maximum correlations among various elements are equal to or lower than 0.900, thus no signs of multicollinearity problem are found (rahman & zahid, 2019). table 3: pearson’s correlation (1) (2) (3) (4) (5) (6) (7) (8) s price (1) 1 b size (2) -0.161 ** 1 f director (3) 0.068 0.006 1 ind. director (4) -0.211 ** 0.399** 0.080 1 ic audit committee (5) -0.114 * 0.220** -0.025 0.785** 1 f age (6) 0.049 0.097 0.038 0.003 0.010 1 f size (7) -0.140 ** 0.059 0.128* 0.185** 0.158** -0.013 1 f leverage (8) -0.186 ** 0.124* 0.143** 0.228** 0.173** 0.024 0.900** 1 **. correlation is significant at the 0.01 level (2-tailed). *. correlation is significant at the 0.05 level (2-tailed). pearson’s correlation matrix represented in table 3 is utilized in the study for exploring the nature of relationships among the different variables. as in the table, board size has a negative correlation with the share price, so there is a statistically negative link of shareholder confidence with board size. independent directors and independent chairpersons of the audit committee are also negatively correlated with the share price. moreover, firm size and firm leverage are control variables that possess a negative association with a share price. according to the correlation matrix, some of the variables in the table have a positive correlation with the share price. the board diversity variable like the presence of female board directors and firm age is positively correlated with shareholders' confidence. the correlation value of all the variables is less than the standard value of multicollinearity 0.900, which indicates a weak correlation. sajid ullah, muhammad zahid, muhammad fayaz & muhammad saad 52 table 4: regression analysis prepcgc post-pcgc combined ols pcse ols pcse ols pcse board size -0.004 -0.004 0.012 0.012 0.002 0.002 (0.040) (0.004) (0.037) (0.016) (0.027) (0.018) female director 0.223** 0.223*** 0.244** 0.244** 0.238*** 0.238*** (0.109) (0.049) (0.099) (0.051) (0.072) (0.058) independent director -0.086 -0.086*** -0.280*** -0.280*** -0.188*** -0.188*** (0.086) (0.032) (0.081) (0.089) (0.059) (0.067) independent audit committee -0.007 -0.007 0.363*** 0.363*** 0.195** 0.195 (0.148) (0.038) (0.132) (0.135) (0.098) (0.126) firm age 0.145** 0.145*** 0.076 0.076*** 0.109** 0.109*** (0.069) (0.049) (0.068) (0.009) (0.048) (0.019) firm size -0.000 -0.000 0.245 0.245 0.062 0.062 (0.132) (0.066) (0.211) (0.200) (0.110) (0.092) firm leverage -0.259* -0.259*** -0.548** -0.548*** -0.342*** -0.342*** (0.134) (0.044) (0.215) (0.165) (0.111) (0.102) lag of share price -0.058 -0.058 0.018 0.018 -0.025 -0.025 (0.067) (0.035) (0.067) (0.027) (0.047) (0.016) constant -1.649 -1.649* 0.634 0.634 -0.498 -0.498 (1.009) (0.938) (0.933) (0.683) (0.684) (0.734) obs. 199 199 200 200 399 399 r-squared 0.162 0.162 0.197 0.197 0.165 0.165 years dummy yes yes yes yes yes yes standard errors are in parenthesis *** p<0.01, ** p<0.05, * p<0.1 4.3 results and discussion the research failed to find any significant relation between board size and shareholders' confidence. as our results consist of pre and post scenario, so in the pre scenario board size value is negative, but in post-analysis the value of board size is positive, so the results of post-analysis are better than pre, but overall board size has no effect on shareholders’ confidence therefore both the hypotheses h1 (a), h1 (b) of the study, are rejected. this finding does not support the results of a bunch of studies e.g., adams & ferreira, 2007; kiel & nicholson, 2003; singh et al., 2018, which investigated that board size has a positive effect on shareholders’ confidence. as board size has the talent to create an extensive pool of expertise and more skills for the firm (kamran & mehmet, 2021). the current study result exposes that independent directors of the board are negatively associated with shareholders' confidence, which is not supported by hypotheses, h3(a) and h3(b) as in both pre and post scenarios the value is negative. this study results journal of applied economics and business studies, volume. 5, issue 4 (2021) 41-58 https://doi.org/10.34260/jaebs.543 53 support the earlier studies, such as reinstein and weirich (1996) argue that audit committees do not have the expertise and power to oppose management in listed firms of pakistan and also do not play a vital role in improving the value of financial reports. as they work according to the wishes and desires of management because they are hired by them. so the audit committee must have relevant experience and literacy as made mandatory in the code for all listed firms (arslan & alqatan, 2020). the statistics result in table 4 show a significant positive influence of female directors on shareholders' confidence. as in pre and post scenario, the results are significant and positive, however, the post result is better than the pre-analysis, which indicates that corporate code 2017 play important role in increasing shareholder confidence, therefore the study accepted the h2(a) and h2(b) hypotheses. these findings are also supported by the earlier literature as miller and del carmen triana (2009) discussed that existence of female directors may help in improving the skills of the firm to yield profits by using its resources and income. besides, female directors also influence the decision-making capability of the firm when it comes to improving business and investors’ confidence. simply the finding recommends that women are more conscious about shareholders' profit than male directors (subba & sujana, 2021). the result indicates that the audit committee is negatively associated with shareholders' confidence. although in the pre scenario its valve is negative, in the postanalysis the value is positive, so this is a change but statistically, they have no relation with the share price, so our results reject the (h4 (a) and h4 (b) hypotheses. previous studies also argue that the financial scandals that occurred at the start of the 2000s (enron, worldcom) confirm to influences done by some managers and showed the weakness of this governance instrument. development of audit value is therefore needed to support the audit to play its role as a governance mechanism. an upper audit dominance increases the value of financial information and stimulates good control by managers and good decision-making by shareholders (riadh & najoua, 2020). further noted that the effects of the control variables on the dependent variables as follows. the firms' age (business age), has positively associated while the size of the firm, does not affect, and the firm's level of debt hurts firm performance and shareholders 'confidence. finally, r-square values in the table present the proportion of the variation in shareholders' confidence that is due to the changes in the independents' variables. 5. conclusion and future directions this study examines the impact of corporate governance mechanisms on shareholders’ confidence, using a proxy (sp) and four corporate governance mechanisms sajid ullah, muhammad zahid, muhammad fayaz & muhammad saad 54 (board size, independent female directors, independent directors, and independent chairperson of the audit committee). a sample size of the top 100 firms listed on the security exchange commission of pakistan between 2015 and 2019 is used. from the observing phase, our results confirm the assumption that the board size has no significant impact on shareholders’ confidence. furthermore, the results of our research confirm the positive and significant effect of independent female directors on our dependent variable. there is also a negative relationship between sp and independent directors. moreover, in pre-age, there is no impact of the independent chairperson of the audit committee on shareholders' confidence but in the post-period, there exists a positive and significant relationship between them. our findings have important implications for policymakers, regulators, shareholders, companies, governments, and other countries. firstly, it contributes to the existing literature by contributing new empirical evidence on obedience and execution of the 2017 corporate governance code in pakistan. secondly, it offers suitable proof by observing whether the mentioned corporate mechanisms may affect firm shareholders’ confidence using a sample of top 100 listed firms on secp following a variation in a regulatory environment. the third contribution is that, the study split our sample period into two sub-samples pre and post to examine whether the revised cg code has any impact on the shareholders’ confidence during different sample eras. finally, the results show that the firms' confidence and corporate governance characteristics are not toughly linked and this may be due to the causes that organizations in pakistan did not follow the recommendations of the cg code in the earlier stage. journal of applied economics and business studies, volume. 5, issue 4 (2021) 41-58 https://doi.org/10.34260/jaebs.543 55 references agyemang, e. (2017). the impact of corporate board size on firm performance : evidence from ghana and nigeria. 4(2), 1–12. https://doi.org/10.5296/rbm.v4i2.11721 akshita, & chandan. (2016). corporate governance and firm performance in developing countries : evidence from india. 16(2), 420–436. https://doi.org/10.1108/cg-012016-0018 al-rahahleh, & ayat. (2017). corporate governance quality , board gender diversity and corporate dividend policy: evidence from jordan. 11(2), 86–104. https://doi.org/10.14453/aabfj.v11i2.6 andriukaitiene, r. (2013). philosophy and paradigm of scientific researchbandas, cornelia. in web of science (p. 29). https://doi.org/http://dx.doi.org/10.5772/46845 arslan, & ahmad, a. (2020). role of institutions in shaping corporate governance system: evidence from emerging economy. heliyon, 6(3), e03520. https://doi.org/10.1016/j.heliyon.2020.e03520 arslan, & alqatan. (2020). role of institutions in shaping corporate governance system: evidence from emerging economy. heliyon, 6(3), e03520. https://doi.org/10.1016/j.heliyon.2020.e03520 aswathy, & chandramohan. (2018). impact of corporate governace on firm performance: emperical evidence from india. 6(2), 209–218. atta, & haider, s. (2020). the nexus between capital structure, firm-specific factors, macroeconomic factors and financial performance in the textile sector of pakistan. heliyon, 6(8), e04741. https://doi.org/10.1016/j.heliyon.2020.e04741 bakhsh, & zeeshan. (2019). corporate governance and its impact on triple bottom line : evidence from firms listed with pakistan stock exchange abstract : 39(3), 771– 784. bernard, & woochan. (2015). how corporate governance affect firm value? evidence on a self-dealing channel from a natural experiment in korea. journal of banking and finance, 51, 131–150. https://doi.org/10.1016/j.jbankfin.2014.08.020 bhagat, & bolton. (2019). corporate governance and firm performance: the sequel. journal of corporate finance, 58(november 2018), 142–168. https://doi.org/10.1016/j.jcorpfin.2019.04.006 chancharat, s. (2019). board structure,political connection and frim performace : evidence from thailand. international journal of business and society, 20(3), 1096–1111. charumathi, b., & rahman. (2019). do women on boards influence climate change disclosures to cdp? evidence from large indian companies. australasian accounting, business and finance journal, 13(2), 5–31. https://doi.org/10.14453/aabfj.v13i2.2 emerald. (2018). corporate governance mechanisms and firm performance: evidence from the emerging market following the revised cg code. sajid ullah, muhammad zahid, muhammad fayaz & muhammad saad 56 https://doi.org/10.1108/cg-07-2018-0244/full/html fazli azim, & mustapa. (2018). arabian journal of business and management review ( kuwait chapter ) the effects of corporate governance and related party. 7(2), 9– 31. https://doi.org/10.12816/0046065 haseeb, & ibrahim. (2015). corporate governance , firm financial performance and shareholders ’ confidence : a proposed analysis of mccg 2012. 7(1), 139–147. herlambang, a., & murhadi. (2020). factors affecting the financial performance of companies based on agency theory. 115(insyma), 110–114. hexana, s. (2020). role of audit committee in the fraud pentagon and financial statement fraud. international journal of contemporary accounting, 2(1), 77–93. https://doi.org/10.25105/ijca.v2i1.7163 jacob, & safdar. (2020). the strength of ceos’influence on csr in chinese listed companies. new insights from an agency theory perspective. sustainability (switzerland), 12(6), 2–13. https://doi.org/10.3390/su12062190 jamshaid, & khalid. (2018). corporate governance in pakistan: an exploratory study of the pakistan international airlines corporation limited. 52(i), 246–260. jason, & michael. (2017). a descriptive summary of cooperative governance and performance. journal of agricultural cooperation, 32, 74–95. javid, a. y., & iqbal, r. (2010). corporate governance in pakistan: corporate valuation, ownership and financing. pide working papers, 57, 1–87. joseph, 2018. (2018). core corporate governance structure and financial performance of manufacturing companies in nigeria. 6(2), 35–49. junaid, & muzzammil. (2020). corporate governance mechanism and performance of insurers in pakistan. 2(july), 243–262. https://doi.org/10.3934/gf.2020014 kallamu, & mohd saat. (2015). audit committee attributes and firm performance: evidence from malaysian finance companies. asian review of accounting, 23(3), 8–33. https://doi.org/10.1128/aem.71.1.159-168.2005 kamran, & mehmet. (2021). workforce reductions and post-merger operating performance: the role of corporate governance. journal of business research, 122(may 2019), 109–120. https://doi.org/10.1016/j.jbusres.2020.08.042 kursat. (2019). ii . intraders uluslararas ı ticaret kongresi kongre kitab ı the second intraders international conference on international trade conference book (issue december 2018). lenore. (2019). ending shareholder primacy in corporate governance. mohammad rateb. (2018). audit committee effectiveness and company performance: evidence from jordan. 7(2), 1–48. https://doi.org/10.5430/afr.v7n2p48 mugarura, n. (2016). the juxtaposition of success and failure of corporate governance procedures the interplay between rules and practice. 23(2), 379–413. https://doi.org/10.1108/jfc-07-2013-0047 mustaruddin, & rusnah. (2011). looking for evidence of the relationship between journal of applied economics and business studies, volume. 5, issue 4 (2021) 41-58 https://doi.org/10.34260/jaebs.543 57 corporate social responsibility and corporate financial performance in an emerging market. asia-pacific journal of business administration, 3(2), 165–190. https://doi.org/10.1108/17574321111169849 nazir, m. s., & afza, t. (2018). does managerial behavior of managing earnings mitigate the relationship between corporate governance and firm value? evidence from an emerging market. future business journal, 4(1), 139–156. https://doi.org/10.1016/j.fbj.2018.03.001 nekhili, m. (2017). corporate social responsibility disclosure and market value: family versus nonfamily firms. journal of business research, 77(july 2016), 41–52. https://doi.org/10.1016/j.jbusres.2017.04.001 nidhi, & anil. (2016). audit committee , corporate governance and firm performance : empirical evidence from india. 8(3), 103–116. https://doi.org/10.5539/ijef.v8n3p103 nikhil, & netai. (2017). effect of corporate governance on firms’s performance: a review of banking sector of bangladesh. east west university of research center for research and training. nodir, & amonboyev. (2016). corporate governance and development : the case of uzbekistan. 1(6), 31–36. https://doi.org/10.18775/jibrm.1849-8558.2015.16.3005 qaiser, r., & harry. (2011). corporate governance and firm performance in pakistan : the case of karachi stock exchange ( kse ) -30. 3(august), 482–491. rahman, & zahid. (2019). boardroom gender diversity: implications for corporate sustainability disclosures in malaysia. journal of cleaner production, 118683. https://doi.org/10.1016/j.jclepro.2019.118683 rashid zama, & ayub. (2015). corporate governance and firm performance : the role of transparency & disclosure in banking sector of pakistan. 43, 152–166. https://doi.org/10.18052/www.scipress.com/ilshs.43.152 riadh, & najoua. (2020). the digital transformation of external audit and its impact on corporate governance. technological forecasting & social change, 150(december 2018), 119751. https://doi.org/10.1016/j.techfore.2019.119751 richard, & david gwilliam. (2019). audit within the corporate governance paradigm : a cornerstone built on shifting. 30, 90–105. https://doi.org/10.1111/14678551.12297 shafie, & kamilah. (2016). corporate governance practices and firm performance: evidence from top 100 public listed companies in malaysia. procedia economics and finance, 35(october 2015), 287–296. https://doi.org/10.1016/s22125671(16)00036-8 shahbaz, & merve. (2020). board attributes, csr engagement, and corporate performance: what is the nexus in the energy sector? energy policy, 143(april), 111582. https://doi.org/10.1016/j.enpol.2020.111582 sheikh, m. f., & shah, s. z. (2018). firm performance, corporate governance and executive compensation in pakistan. applied economics, 50(18), 2012–2027. sajid ullah, muhammad zahid, muhammad fayaz & muhammad saad 58 https://doi.org/10.1080/00036846.2017.1386277 srinidhi, z., & shiqiang. (2020). how do female directors improve board governance? a mechanism based on norm changes. journal of contemporary accounting and economics, 16(1), 100181. https://doi.org/10.1016/j.jcae.2019.100181 subba, & sujana. (2021). board gender diversity and corporate social responsibility: is there a case for critical mass? journal of cleaner production, 278, 123319. https://doi.org/10.1016/j.jclepro.2020.123319 valeria. (2019). corporate governance and board of directors: the effect of a board composition on firm sustainability performance. journal of cleaner production, 237, 117727. https://doi.org/10.1016/j.jclepro.2019.117727 wahid, & kauser. (2020). corporate governance and return on equity evidence from pakistan stock exchange. 6(1), 63–72. https://doi.org/10.26710/jafee.v6i1.1061 waqar, & subhan. (2019). impact of board diversity and audit on firm performance. cogent business & management, 6(01), 1–16. https://doi.org/10.1080/23311975.2019.1611719 yussoff, & anees. (2016). improving shareholder value through corporate governance mechanism in malaysian listed companies. 6, 44–49. zahid, & haseeb ur rahman. (2019). the bond between board and performance : a case of the banking industry of. 14(february 2020). . journal of applied economics and business studies, volume. 6, issue 2 (2022) 15-32 https://doi.org/10.34260/jaebs.622 15 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x welfare impact of dust pollution on human health in the district khyber ex-fata, pakistan muhammad rafiq1, luqman khan2*1 & munir khan3 1 department of economics, institute of management sciences peshawar pakistan 2department of economics, institute of management sciences peshawar pakistan 3 department of economics, institute of management sciences peshawar pakistan abstract dust pollution is among the major cause of disease and death all around the world. the lungs and respiratory system are the primary targets of ambient dust pollution. this study aims to estimate the health cost due to dust pollution from a stone crushing industry in the khyber district of ex-fata, pakistan. to examine the impact of dust pollution on respiratory illness, primary data was collected from 1278 individuals from 200 households living within a 3 km radius of the stone crushing industry. the household production function and mitigation cost function were used to determine the reduced-form equations for assessing the impact of dust pollution on human health and then estimating the monetary cost related to mitigating such diseases. the pm10 level in the crushing region was extremely high, averaging 1247g/m3, while the pm10 level in the control group was relatively low, at 45g/m3. according to our study, the regression analysis estimated that by reducing pm10 to a safe level, every individual could save pkr 4958 per year through mitigation costs. the total welfare gain for district khyber will be pkr 2453.306 million or (pkr 2.45 billion). aside from that, individual productivity loss is estimated to be .37 per day per year. the monetary value of avoiding the restricted working day is pkr 23 for one home, whereas the monetary benefit of lowering the pm10 level is pkr 161000/ for keywords cost of illness, dust pollution, health cost, workday loss, welfare analysis, respiratory illness, khyber jel classification i14, i15, i18 1* ms172606864@imsciences.edu.pk mailto:ms172606864@imsciences.edu.pk muhammad rafiq, luqman khan & munir khan 16 the entire district. the research study suggests implementing the ban on stone crush production and technology adaptation in the residential areas. 1. introduction dust pollution is made up of tiny bits of dry solid material. it starts out as airborne particles and gradually settles on flat surfaces. environmental, domestic, and industrial dust are the three main types of dust that can be produced anywhere. at home, we perform thorough cleaning without using excessive chemicals which are advised to remove the majority of the dust that is visible. the clinical literature is well versed in the devastating health impacts of dust pollution. high levels of ambient dust pollution have been linked to illnesses such as headaches, cancer, high blood pressure, asthma, heart disease, shortness of breath, etc (mao, peng, et al., 2017). industrial illnesses and injuries are associated with significant social and financial costs. this externality costs the individual a significant amount of money in terms of lost productivity and health costs. furthermore, it takes a firm stand on a public bursary in terms of greater government spending. in many areas, dust pollution exceeds the national air quality standard. the aqi is between 100 and 150. the short-term national ambient air quality standard for the protection of public health is typically equivalent to an ambient air concentration of 100 for each pollutant. in general, aqi levels of 100 or less are considered to be good. in the pakistan air quality index, the world health organization (who) listed pakistan as one of the top 10 most polluted countries on the earth (muhammad khan, 2017). the environmental hazards are caused by industrial activity including dust pollution, water pollution, and soil contamination. workers in the industrial sector; particularly in underdeveloped countries have always had to contend with filthy conditions and are regularly exposed to dust and particulate matter. the small-scale stone crushers and unorganized sectors have existed in the ex-fata residential regions. these crushers are necessary for the construction of buildings and roads. they are involved in things that require a lot of time and effort. it supplies not just raw materials for buildings, bridges, and road construction, but also a source of income for the local people. in pakistan, industrial sectors badly pollute the environment containing water contamination, soil pollution, and air pollution. the stone crushing workers have always had to contend with dirty conditions that are frequently devoid of particulate matter and dust, particularly in underdeveloped countries. furthermore, stone crushing plants are one of journal of applied economics and business studies, volume. 6, issue 2 (2022) 15-32 https://doi.org/10.34260/jaebs.622 17 pakistan's most important industrial sectors. in many hilly places, stone crushers have smallscale and unstructured sectors, yet these crushers provide the essential material for building and road construction. it not only provides a source of income for a small number of people, but it also provides raw materials for the construction of roads, buildings, and bridges, among other things. stone crushers are involved in labor-intensive operations (ilyas and rasheed 2010). air pollution such as ozone and particle matter is known to contribute to human mortality and morbidity. however, humans could inhale particulate matter (pm) with a fine size of fewer than 10 micrograms known as pm10, which subsequently travels to the respiratory system and causes major health consequences. on the other hand, household sickness and health difficulties entail social and economic implications. while, medication, travel, fees paid in treating diseases, and productivity loss are all examples of these costs. the process contains drilling, raw material transportation; screening, rock blasting, crushing, material handling, size classification, storage activities, and final product transportation are all part of the stone crushing plants. in other words, significant amounts of suspended particulate matter (spm) are emitted into the air because of mining operations. these particulates are dependent on particle weight, shape, and size since they are released from crushing machines and persist in the atmosphere for a long time, posing a risk to humans and plants alike (prajapati &tripathi, 2008). the stone crushing sector produces high levels of dust exposing men and women in the community to a variety of occupational health risks. hence, the severity of these disorders may have a significant impact on people's productivity, living standards, health, and life value. due to inhaling dust on a regular basis, rural laborers notably those working in the stone-crushing business and especially inhabitants are suffering from a variety of illnesses such as asthma, chest pain, shortness of breath, and allergies, as well as throat and skin infections. because of their illiteracy, these workers' health deteriorates, and they are not aware of the importance of using safety equipment such as a facemask, and a cloth, among other things (asante, abass, (2014). the purpose of this research was to determine the influence of dust pollution on human health in khyber that has a negative impact on the environment and those who live or work near the crushing region. however, the objective of the study is to examine the mitigation cost (mc) and workday loss (wdl) of rural households in the shakas area of district khyber who are affected by the stone crushing industry. 2. literature review the stone crushing industry is a central supplier of construction materials for the muhammad rafiq, luqman khan & munir khan 18 country's structures, canals, roads, etc. stone crushing is an important industrial sector in pakistan; where people work to produce various stones crushed and used as raw materials. in addition, stone crushing plants are the key units that could be found in the residential areas of large cities, towns, states, and even the entire country. in other words, construction operations are ongoing throughout pakistan for the benefit of society and economic progress. at the same time, it has increased the amount of dust pollution in the area which has a negative impact on occupational human health, air quality, and productivity. the results of recent medical studies on the health effects of dust pollution are astonishing. to begin with, there are no "safe" amounts of dust pollution. dust pollution affects practically every organ in the human body including the respiratory system which causes asthma and chronic obstructive pulmonary disease, as well as the heart, strokes, and even cancer. as a public health opinion, we are concerned that young children are particularly vulnerable to air pollution; new evidence suggests that air pollution exposure has a negative influence on children's cognitive development. moreover, dust pollution is still a serious public health issue around the world. according to who estimates, 92 percent of the world's population is exposed to hazardous levels of air pollution (adil zareef & muhammad rafiq, 2021). dust pollution has a negative impact on human health by generating or worsening respiratory, cardiovascular, and visual disorders, as well as influencing allergic reactions. it also has an indirect effect on mortality in the region by reducing vision and causing travel problems due to the creation of fog and smog. fog is formed when water droplets are suspended in the air at ground level. this is frequently associated with low-level air cooling and the formation of condensate, which can persist in steady conditions (e.g., low wind speeds and atmospheric inversions). these water droplets can increase the potential for fog formation in the presence of particle matter (both primary and secondary particles) by enhancing condensation nuclei and chemically interacting with fog droplets (fatima & zulifiqar, 2021). dust pollution is the leading cause of health problems all around the world but developing economies are particularly vulnerable. dust pollution is made up of a complicated mixture of gaseous and particle components, all of which have negative effects on the cardiovascular and respiratory systems. individuals in lowand middle-income countries (lmics) have distinct exposures to respiratory diseases, and thus different risk factors, than those in higher-income nations. infant lung function is harmed by increased prenatal household air pollution associated with exposure. every year, more than 4 million people die because of diseases caused by pollution (ubaid & ayat, 2021). journal of applied economics and business studies, volume. 6, issue 2 (2022) 15-32 https://doi.org/10.34260/jaebs.622 19 environmental degradation is a major issue in economics that has gotten a lot of attention from economists and environmentalists alike. due to the tremendous growth of carbon dioxide (co2) emissions, the world now faces the severe challenge of global warming. pakistan has seen strong real gdp growth in recent years as well as a significant reduction in poverty. however, the continuous availability of important ecological services and natural resources is critical to human well-being. pakistan's natural resources and ecosystem are becoming increasingly polluted and stressed. the country's most pressing environmental concerns are around the management of scarce natural resources greenhouse issues, pollution and waste management, brown issues, and possible vulnerabilities to natural disasters and climate change (khan, i., & hou, f. (2021). pakistan's ecology has suffered for many years due to poor natural resource management and very high population growth. hence, agricultural and industrial runoff have fouled water supplies, manufacturing, and car emissions have impaired air quality in metropolitan areas, both of which are caused by continuous deforestation. pakistan like other developing countries has prioritized achieving food self-sufficiency, fulfilling energy demands, and limiting its high rate of population increase over reducing pollution and other environmental hazards. as a result, "green" concerns have not been a key priority for the administration (epa, 2021). due to the life-threatening consequences of dust pollution around the world; particularly in developing countries like pakistan, groundbreaking research, massive positive input, widespread awareness, and pollution-reduction measures are all necessary. there are different environmental boosting activities from various environmental protection agencies and government or non-governmental organizations available in pakistan's larger cities. furthermore, smaller cities, towns, villages, and some divisions or districts, on the other hand, do not receive the same level of attention. however, malakand division is one of pakistan's divisions that receives insufficient attention. from a new developmental perspective, the consequences are even more dangerous. (ullah & li, 2019). pakistan faces water pollution, soil erosion, land pollution, water scarcity, global warming, air pollution, and natural disasters are only some of the climatic and environmental concerns. pakistan is on the list of countries with poor air quality, according to the latest global environment performance index (epi) ranking. furthermore, climate change and global warming are the most concerning issues, threatening the lives of millions of people around the country. carbon emissions, population growth, and deforestation are the primary causes of these environmental problems ((andam, f., & shi, 2017). the epidemiology literature has recently discovered the harmful health impacts of dust muhammad rafiq, luqman khan & munir khan 20 pollution and air pollution. although, high amounts of pollution were present which has been linked to a variety of morbidities including headaches, asthma, high blood pressure, cancer, bronchitis, and more. this externality resulted in a significant monetary loss to workers or persons in the form of decreased productivity and increased health costs (rafiq & munir 2014). the stone crushing units are one of pakistan's most important socio-economic sectors. they increase the amount of dust-related emissions and may cause a variety of health problems for the workers as well as the surrounding community. while trees developed in the area and numerous morphological characteristics were identified. the study looked into the influence of stone or marble dust on plants. the findings revealed that particulate matter (pm) of rather a significant size was created during the stone-crushing unit. plant growth is severely harmed if particle matter is deposited on vegetative stems (yashoda saini, et al, 2011). the kp environmental protection act, 2014: the kp environmental protection agency is in charge of enforcing environmental laws in the region. the environmental protection act of 2014, which was passed on december 11, 2014, outlines these regulations. it has rules and regulations in place to prevent dust, air pollution, and noise pollution. nonetheless, section 11, section 13, and section 12 all deal with dust pollution. section 17 specifies the penalties and fines that will be imposed on violators. section 11, epa act, 2014: "no individual shall omit, release, or authorize the emissions, discharge, or misuse of any sewage, misuse, dust pollution, or atmosphere pollution in quantities, attention, or levels that exceed the countrywide ecological value standard." section 13, epa act, 2014: "according to the kp protection act of 2014, no one may trade imperious substances whose chemical or physical motion is hazardous, radioactive, flammable, explosive, or creates a bad ecological effect whether used alone or in combination with other chemicals." section 14, epa act, 2014: "anywhere the provincial and federal governments are pleased that the emission and discharge of any sewage water, dust, or atmosphere pollution, or the dumping of devastating, or the treatment of hazardous materials, or any extra substance and waste are possible or happening, or has accrued in violation of the act is possible or happening." journal of applied economics and business studies, volume. 6, issue 2 (2022) 15-32 https://doi.org/10.34260/jaebs.622 21 section 17, epa act, 2014: the punishment and penalties imposed on offenders of sections 11, 13, and 14 of the protection act 2014 are discussed in this section. "such penalties under section 17 will be punishable by a minor fee of 50,000 which may be increased to 50,0000 and in the case of persistent flouting of collapse, an additional fine of 10,000 for each day that such failure and contravention continue." "the maximum penalty under section 17 is 50,0000 with a daily penalty of 100,000 if the infraction continues. this applies to any infraction under section 14 of the environmental protection act of 2014 that is related to the conduct." according to the crush production owner, they are not subject to any government taxes. owners enjoy complete flexibility for not enforcing any laws or rules, there is no governance system in place, there is no political pressure, and many more factors are imposing a significant external cost on society. as far as welfare analysis is concerned, this ongoing study is an attempt to put a value on health costs for the population of shakas in district khyber. 3. methodology 3.1 theoretical framework the health production function was used to achieve the goals of our research. the health production function is a subset of the household production function (hpf). the hypothesis leads to a necessity for the mitigation cost function (mc) and the workday loss function (wdl). demand for mitigation cost function and health production function is implicit in the utility function to assess the health benefit of reduced dust pollution. a similar theoretical lens has been employed by rafiq & munir (2014), usha (2006), and naveen (2012). we will now discuss the utility function in order to obtain the mitigation cost function and the health production function (freeman, 1993). the following is the utility function: u=u (q, h, l, x) …… (1) in this equation, q denotes dust pollution exposure, h means the number of workdays lost owing to pollution, l denotes an individual's free time, and x denotes the total consumption of commodities accessible on the market. q & h provide the most dissatisfaction, whilst l & x provide the most satisfaction. by combining mitigating efforts with a specified degree of dust pollution defined by q, a person could improve their health. they also favor socio-economic uniqueness and another good health status. the following is how the household health production function is calculated: muhammad rafiq, luqman khan & munir khan 22 h=h (z, m, q) ……. (2) z signifies a vector of health characteristics such as age, sex, and income in the given equation. where m denotes mitigation cost (mc) and q denotes the degree of dust pollution. h also counts the number of workday losses (wdl). the preceding equation implies that an individual can maintain his health status in a polluted or dusty environment by paying a mitigation cost (mc). now we will go over the individual budget constraint, which looks like this: i= w+y(h-t-l) = x + paa+ pmm…. (3) w is the wage rate, y is non-wage earnings, and (h-t-l) is the total time spent at work, according to the above equation. the t stands for total time, and pa and pm are the prices for mitigation costs and lost workdays, respectively. in the following equations, q signifies the degree of dust pollution, i means the individual's income, w denotes the wage rate, and pm reflects the price of mitigating actions and other explanatory variables. by simplifying the following problem, an individual could exploit revenue subject to (l, a, m, x) budget limitations. maxg= u (q, h, l, x)+ ƛ(w+y(h-t-l)-x-pm-pa)….(4) the ƛ lagrange function is shown in the above model. we acquire marginal willingness to pay (mwtp), health production function (hpf), workday loss (wdl) function, and further demand function for mitigation costs as a result of the foregoing models (mc). a=a (h, q, z, pm, pa, w, x)…..(5) m=m (h, q, z, pm ,pa ,w, x)…(6) mwtp= w. dh/dq +pm dm/dq+pa dm/dq+ (ƛ/uh). dq/dh ….. (7). in equation 7, the total of visible reductions in mitigation cost, further explain disutility of sickness, monetary benefit, and cost of disease is connected with marginal willingness to pay which gains health advantages by lowering dust pollution. the estimation of the (mwtp) also necessitates the estimation of the (hpf) and the mitigation cost function. the monetary gain from reducing dust pollution was captured in the first three equations above. the cost of illness (coi) is an important way to examine lower boundary estimates because avoiding expenditures is the most difficult to estimate effectively. coi= w. dq/dh+ pm dq/dm ……. (8) in equation 8, the health benefit from reduced dust pollution is represented by the mitigation cost and workday loss. journal of applied economics and business studies, volume. 6, issue 2 (2022) 15-32 https://doi.org/10.34260/jaebs.622 23 3.2 econometric models and specifications the nature of the data is always influenced by the model's estimation. we gathered primary data from rural households in december. some variables in our models are dummies. therefore, the outputs of the ols and glm models are undervalued and inaccurate. hence, in our study, we employed poisson and negative binomial regression to estimate data and then choose which model's outcome was more accurate. the models can be written as: prob yit = (yit/xit) = uit e-u it/yityit = 0, 1, 2, 3 we take log of the data to capture the linearity. lnuit=αi+ß x1 it+ ß2x2 it ……ßsxsitα+b2x2 it+ b3 x3it+ …..bs xsit the econometric model shows reduction as a form of health production function and wdl function. the estimating models, we used above are dependent on the data's nature. previous investigations on health-related disorders have concluded that the medical cost of certain observations is zero. when it comes to estimating, a linearity problem can arise, thus we utilize a tobit regression model to tackle this problem. the model is given below: μit =α + ß x it +uif the rhs >0 0 or else in the preceding equation, mit represents the ith household's probability level, which leads to positive mitigation at time t, whereas xit represents the individual's attributes. 3.3 empirical econometric models this led to the estimation of two reduced form equations, the health production function (hpf) and the workday loss (wdl) function, to assess the marginal effect of ambient dust pollution on human health and utilized the following econometric models to estimate these equations: (rafiq & munir, 2014) (usha gupta 2006). h= α1+ ß1pm10+ ß2age+ ß3age2+ ß4sex+ß5edu+ß6ocp+ß7smk+ ß8sh+ß9inc+v m= α1+ ß1pm10+ ß2age+ ß3age2+ ß4sex+ß5edu+ß6ocp+ß7smk+ ß8sh+ß9inc+v the dependent variables in the preceding equations reflect work day loss (wdl), which represents how many days labor is ill and switches over due to various diseases linked to ambient dust pollution. it also indicates the number of loss workdays per individual per year. in the second equation, it signifies mitigation cost (mc) which comprises all medical expenses spent owing to dust pollution-related illness, including medicine, transport to muhammad rafiq, luqman khan & munir khan 24 doctor clinic, hospital cost, doctor’s fee, diagnostic test, hospital stay, and so on each year per person. the independent variables explained below that affect the mitigation-related activities and health production function which is as follows; ▪ particulate matter (pm10): suspended particulate matter is the harmful matter emitted by the stone crushing industry during the production process. the spm value was measured in micrograms ug/m3 by skilled technicians from pcsir laboratories. ▪ age: age was represented by a dummy variable. if the individual was between 20 and 60 years, then ‘0’ was assigned. if an individual was less than 20 years or greater than 60 years of age, then ‘1’ was assigned. ▪ age 2: to check the non-linearity between sickness and age to a minimum. we formed a square of working individuals. ▪ sex: sex is represented by 1, if an individual was male; otherwise, it is denoted by 0. ▪ education: education is categorical. education level was measured by assigning different values ranging from 1 to 6. ▪ occupation: this pertains to one's occupation, and it indicates that those who work have a higher risk of developing various diseases than those who do not work and prefer leisure. ▪ smoking: smoking spending is the per month expenditure in rupees spent by individuals. the cost of cigarettes for smokers was calculated by taking the average price per cigarette. ▪ structure of house: the variable indicates that those who live in mud houses have a higher risk of disease, whereas those who live in cement houses have a lower risk of illness. ▪ income: income was measured in rupees earned per month by each individual. if the individual was not earning, then it was zero. 3.4 study area this research was carried out in two areas: one in the jamrud road shakas area (crushing market) in the district khyber, and the other in regi model town peshawar (without crushing region). we take a 3-kilometer radius from the study region in the crushing area (herath et al 2012). the economic cost of the disease was compared in each region to determine which region has the highest economic cost of illness. district khyber is the largest industrial town in ex-fata pakistan because the stone-crushing businesses are concentrated in the shakas khyber. khyber is known as the "city of stone." the crushing industry is the backbone of the economy and it plays a critical part in its development and building. on the other hand, khyber has the ability to utilize energy and a good transportation system both of which contribute to dust pollution. dust contamination can cause a variety of environmental and health issues. in khyber, the stone crushing industry is one of the most significant contributors to dust and air pollution. journal of applied economics and business studies, volume. 6, issue 2 (2022) 15-32 https://doi.org/10.34260/jaebs.622 25 3.5 merge area profile of district khyber district khyber is an underdeveloped section of the former federally administered tribal area (ex-fata) that is bordered by afghanistan and connected to peshawar, orakzai, and kurram districts. the district's total area is 2,576 cubic kilometers with woods accounting for around 8.22% of the total area. the jamrud, bara, landikotel, and mallagori are the four tehsils that make up the district khyber. according to a 2017 demographic survey, the total population of shakas khyber is 41216. at the study site, there are 50 crushing plants installed in a residential area. 3.6 sampling and data collection the primary data was collected for two areas: the stone crushing industry in the shakas area of district khyber and regi model town peshawar which served as the control group. with a distance of 3 km, the crushing zone is too dusty and the effect of dust pollution on local people is very strong; whereas primary data was collected within the crushing area's 3 km limit. in addition, 100 households from the crushing region and the identical household from the control group were chosen randomly. the study's goal is to determine the effect of the stone crushing industry and dust pollution on human health. for that, primary data was collected to receive selected household information (usha gupta 2006), (rafiq & munir, 2014). due to time constraints, data collection was limited to one month, i.e. november 2018. the study used the household as a key source for medical and health expenditure data. a cross-sectional household survey was used to quantify the economic repercussions of residing near this crushing region, as well as its influence on individual labor market performance due to workday loss and productivity. the pollutant-related data were acquired with the help of the pakistan council of scientific and industrial research (pcsir) peshawar. the ambient dust quality tests were conducted at two points in the target area. during the dry season, the (pcsir) measures particulate matter pm10 in surrounding communities of the crusher region for 24 hours. they scored two points in both the crushing region and the control group. after that, pm2.5 would be difficult to locate because it is made up of very fine particles, which are usually found in the pharmaceutical industry. within a week, the pcsir provided us the pm10 value, because per point and per day rates are too costly and we are facing financial constraints. the pcsir could provide an estimate for a specified time. the primary data for both workday loss (wdl) and mitigation cost (mc) were collected. the data consist of sampled household information, working individuals, health and medical costs, socio-economic traits, and other factors are included in the data. many muhammad rafiq, luqman khan & munir khan 26 researchers have suggested that the best time to collect data is in the winter (usha gupta 2006). (rafiq & munir, 2014). since crushing plants are typically operating in the day and night shifts, this period is not just suitable for primary data collection. there is a higher level of dust in the dry season, which affects the entire region, although the amount of dust is low owing to rains. due to less rainfall in the dry season, dust pollution is at an all-time high from april to september. table 01: descriptive statistics of the household survey descriptive statistics variables number of selected samples is 200, from crush and control group. observation mean std. dev. min. max. tc medical costs for one month 1278 315.6 1287.05 0 16700 workday workday loss for one month 1278 112.6 518.8 0 12 pm10 particulate matter 1278 1614.4 1145.5 45 2450 ag age of the rural respondent in years 1278 25.0 17.8 .1 80 edu education dummy variable (1 if literate) 1278 1.8 1.3 0 1 sex sex dummy variable (1 if male) 1278 1.4 7355 0 1 sm smoke dummy (1 if smokers) 1278 1.9 1438 0 1 ocp occupation of the individuals 1278 8.6 3.31 0 1 ds distance from crush to h.h in km 1278 3.3 3.65 0 1 inc income of individuals 1278 270.8 477.59 0 3000 sh structure of house dummy (1 if cement, o for mud house) 1278 1.8 5123 0 1 the descriptive statistics for the preferred variables are displayed. for the entire month, the mitigation cost (mc) ranges from 0 to 16700 pkr. the average monthly medical expense is 94. in a month, the workday loss or opportunity cost averages 12 days. this is especially true for the crush group which was severely harmed by dust pollution in the environment. the crush group had a very high pm10 concentration of 2450 g/m3, which is journal of applied economics and business studies, volume. 6, issue 2 (2022) 15-32 https://doi.org/10.34260/jaebs.622 27 exceptionally high for 24 hours. on the other hand, the average level of particle matter is 1247g/m3. females make up 48% of households, while men make up the remaining 48%. the maximum age is 80 years old, while the youngest is one year old. smokers account for 26% of the population. only 32% of the population is educated. we should point out that the literacy rate includes religious education as well as informal education and that literate people are those who can write and read, while, 65 percent of the population lives in mud and stone dwellings. the average daily wage is 270 pkr. individuals in these two places have jobs at a rate of 17 percent on average. table 02: poisson model and negative binomial regression dependent variable (wdl) poisson regression negative binomial regression independent variables coefficient std. err. coefficient std. err. const -6.76 .0357 -15.49 6.02 pm10 .0003*** 8.24 .000061*** .0009 sex -1.42 .0128 -.8435 1.02 age .064 .0007 -.1124 .109 ags2 -.0007 9.83 .0021 .001 sm .3208 .0028 -3.820 2.03 edu -.278 .0092 .285 .385 sh -.381 .005 -.4525 .497 ocp -.1776 .0010 -.224 .118 ds .0002 .0024 1.26 .3059 inc .1193 5.82 .00468 .0011 log likelihood -271589.31 -1044.11 wald chi 305127.23 58.86 total observation 1278 1278 ***, ** and * indicate significance at 1%, 5% and 10% levels table 2 represents the results of the wdl function wdl function or health production function the study’s dependent variable was workday loss (wdl), which looked at how many working days people missed owing to dust pollution and related disorders. the health production function was computed using poisson and negative binomial regression. there are 1278 observations were collected from 200 houses. muhammad rafiq, luqman khan & munir khan 28 poisson regression model the study key variable particulate matter pm10 is positive and statistically significant in the poisson model. the signals of the two variables have altered and no longer make sense. the coefficient of sex is negative. males are more afflicted than females, which indicates males are more affected than females. the coefficients of sm and ds are both positive. negative signals could be seen in edu and sh. it means that if the house structure is solid, the likelihood of disease is minimal; also as education improves, the number of workdays missed wdl reduces. with a health production function, ocp has a negative coefficient whereas inc has a positive sign. negative binomial model the major variable of the study in the negative binomial model is particulate matter pm10 which is positive and significant. so far, the sign of our other variables has been surprising. other variables positive and negative correlations have discussed above. welfare gains from reduction in the work days loss the poisson regression model could be used to calculate the number of missing days and workday loss in pkr. in the sample group, we must use poisson regression to estimate the expected value and average employment level. for thirty days, one unit rise in pm10 level can bring and save 0.00000149. according to the findings, the average daily income of working individuals is 600 pkr, while the rate of average working labor in the entire population is 17% of the sample data. the current level of employment can be extrapolated to the entire district. the following is the workday loss calculation employed by (rafiq & munir, 2014). (usha gupta 2006.) restricted days per year: α*e∑αx*ᴧpm10* 24 through the estimate of poisson regression, the preceding formula depicts the marginal savings of 0.00000149 for thirty days, resulting in a one-unit reduction in pm10. working people could save 0.37 per day on average over a year. furthermore, healthy workers may lose output because of illness, but they are still willing to go to work and earn the same wage. all of these illnesses cause employers to lose money and productivity. reduced productivity should be factored into the cost of illness coi calculation. the sample data revealed that the average wage rate of individuals in the study region is 600 pkr. according to the data, 17 percent of people are employed. the annual benefit of avoiding restricted days is predicted to be 23. according to poisson regression, the anticipated benefit to the safe level is 161000/. this is extrapolates to the entire district as a whole. journal of applied economics and business studies, volume. 6, issue 2 (2022) 15-32 https://doi.org/10.34260/jaebs.622 29 table 03: tobit & ols regressions using 1278 observations dependent variable mc tobit regression ols result independent variables coefficient std. err. coefficient std. err. const -6259.68 2874.14 -1038.30 561.41 pm10 2.26 .525 .28621 .0715 sex 345.50 283.80 39.07 52.49 age 152.94 45.74 24.22 7.82 age2 -1.54 .6242 -.199 .112 sm -2356.04 1128.82 -780.57 248.18 edu 474.69 200.98 -43.05 28.53 sh 36.24 28.27 -58.39 33.24 ocp -107.36 69.90 -11.88 13.15 ds -61.33 149.20 15.90 22.08 inc .7229*** .539 .0588*** .093 sigma (u) 4529.32 284.98 r squared 0.07 log likelihood -1899.57 -1934.05 wald chi 2 164.52 p=0.0000 total observation 1278 total observations. left censored = 1110 1278 number of observations. left censored = 1110 table 3 shows the results of mitigation cost (mc) demand function for mitigation cost in the above model, the dependent variable is mitigation cost (mc) often known as the mitigation cost function. the independent variables are the same as in the workday loss (wdl) function mentioned earlier. the demand for mitigation cost is estimated using tobit and ols regression. the number of observations in this model is 1278. tobit regression results mitigation cost is the dependent variable in the tobit regression model and the results are reported in table 3. the key variable under investigation is particulate matter pm10, which is both positive and significant. while the inc and sh structure of the house is also statistically significant. the cost of mitigation for a one-unit change in pm10 is 2.54 units. other factors' signals may be predicted. reduction of lower mitigation cost in the form of medical expenditure to lower the medical costs, we must first estimate the monetary gain and then identify the marginal effect of tobit regression as shown in table 3. we needed to check the muhammad rafiq, luqman khan & munir khan 30 probability of mitigation cost (mc) and examine the difference in pm10 from the current to safe level due to the tobit model's calculation. therefore, it devised a formula to calculate the cost of mitigation (rafiq & munir.2014) (usha gupta, 2006). reduction in mitigation cost and saving per annum: ß*prob (mc>0) *ᴧpm10* 24 the average level of particulate matter pm10 in both study locations is 1247g/m3. pm10 concentrations of 150 g/m3 are considered typical (who, pcsir). there is a total difference of 1225 in pm10 levels. we used the above formula to get particle matter pm10 down to an acceptable level. through mitigation cost, each individual could save 4958 pkr each year analyzed by tobit regression estimation. we assume that the particulate matter pm10 level in khyber is the same, and hence we find a total welfare gain of mitigation cost for the shakas district in khyber. according to a 2017 census, the total population of shakas khyber is 41216. owing to the tobit regression estimation, the entire welfare gain for shakas district khyber is pkr 2453.306 million (2.45) billion. 4. conclusion the emissions from stone crushing industries are the main reasons for respiratory, cardiovascular, skin, and eye problems. the most disastrous effect of such emissions is cancer, particularly lung cancer. the monetary cost associated with industrial emissions is evident in this study. a reduction in the spm level contributes to positive social gains. if the who's ambient air quality threshold of 50 micrograms/m3 is followed, an individual will save 4958 pkr per year in mitigation costs. the mitigation cost might be reduced to pkr 2453.306 million or (2.45) billion to reduce dust pollution to a safe level of 150 micrograms per cubic meter. the reduction in workday loss could result in a 161000/ pkr yearly average welfare gain. in addition to the direct benefits of lowering the spm level, there will be some indirect benefits in the form of reduced workday losses. however, lowering the spm level has a two-fold effect. on the one hand, it lowers the cost of mitigation; on the other, it raises social earnings. hence, the government should develop a clear policy on ambient air quality standards. the current criteria would not even meet the who's requirements. moreover, given the high health costs associated with the emission of spm into the ambient air estimated in this study, officials should alter the regulations. this is a helpful welfare analysis and we urge the official to adequately handle the situation to close the welfare gaps. otherwise, industries can implement technologies to reduce spm emissions. journal of applied economics and business studies, volume. 6, issue 2 (2022) 15-32 https://doi.org/10.34260/jaebs.622 31 references zareef, a. (2021). health impacts of air pollution and climate change. khyber medical university journal, 14(4), 254-6. jabeen, f., ali, z., & maharjan, a. (2021). assessing health impacts of winter smog in lahore for exposed occupational groups. atmosphere, 12(11), 1532. panezai, s., ali, u., zeb, a., rafiq, m., ullah, a., & saqib, s. e. (2021). quantifying the health and wealth benefits of reducing point source pollution: the case of the sugar industry in pakistan. sustainability, 13(23), 13252. khan, i., hou, f., & le, h. p. (2021). the impact of natural resources, energy consumption, and population growth on environmental quality: fresh evidence from the united states of america. science of the total environment, 754, 142222. rafiq, m., & khan, m. (2014). the health costs of the brick kilns emissions in peshawar: a policy analysis. current world environment, 9(3), 591. halwenge, j. a. (2015). dust pollution and its health risks among rock quarry workers in kajiado country, kenya (doctoral dissertation, kenyatta university). mehwish, n., & mustafa, u. impact of dust pollution on worker’s health in textile industry: a case study of faisalabad, pakistan. khan, muhammad. the health burden of dust pollution in the textile industry of faisalabad, pakistan (no. 121). saha, d. c., & padhy, p. k. (2011). effects of stone crushing industry on shorearobusta and madhucaindica foliage in lalpahari forest. atmospheric pollution research, 2(4), 463-476 adhikari, n. (2012). measuring the health benefits from reducing air pollution in kathmandu valley. sandee. shrivastaava, a. k., tikariha, a., patra, s., & sinha, m. impact of stone crushing industries on leaf anomoly architect of woody plants gupta, u. (2006). valuation of urban air pollution: a case study of kanpur city in india. environmental and resource economics, 41(3), 315-326. supe, g. n., &gawande, s. m. impact analysis of dust pollution within katraj. chougule, a. c., chougule, p. a., &kumbhoje, c. k. (2017). effect of stone crusher on ambient air quality. khan, m. m., nawaz, r., ehsan, n., ahmad, s., nawaz, m. w., & nawaz, m. h. health hazards and socioeconomic effects of stone crushing industry on its workers: a case study of sargodha, pakistan. age, 15(30), 37. ilyas, m., & rasheed, f. (2010). health and environment related issues in stone crushing in pakistan. south asia network of economic research institutes. http://saneinetwork. net/files/10_18_farook_rashee d. pdf. asante, f., abass, k., & afriyie, k. (2014). stone quarrying and livelihood transformation in peri-urban kumasi. muhammad rafiq, luqman khan & munir khan 32 mao, peng, et al. nov, 2017 "evaluation on effects of construction dust pollution on economic loss." ayesha, a., mann, a. a., &anjum, m. a. (2009). health concerns among workers in weaving industry: a case study of tehsil faisalabad, pakistan. journal of agriculture and social sciences, 5(3), 106-108 journal of applied economics and business studies, volume. 3, issue 1 (2019) 15-40 https://doi.org/10.34260/jaebs.312 15 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x do capital markets punish tunneling behaviour of business groups? agency perspective of related party transactions hamid ullah1*, attaullah shah2 & syed hamid ali shah3 1assistant professor, islamia college peshawar, pakistan 2associate professor, institute of management sciences, peshawar, 3assistant professor, quaid-e-azam college of commerce, university of peshawar abstract this study examines that do capital markets punish the tunneling behaviour of business groups in a unique institutional setting of pakistani group firms. for this purpose, data of 207 non-financial firms belonging to various business groups were taken as a sample from 2006 to 2018. the ordinary least square and two stages least square regression models are estimated. the results of various estimated models showed that both accounting performance and market valuation of the sample firms are negatively influenced by related party transactions (rpts). therefore, these transactions are assumed to be a means of conflict by the market participants and consistent with the tunneling perspective and conflict of interest hypothesis. these results support the notion that capital market pressure acts as a disciplinary device and punish the tunneling behaviour of the group firms. besides capital markets, the regulatory framework and corporate governance (cg) mechanisms, therefore, should be improved in such a way to reduce the tunneling of minority shareholder's wealth. moreover, external auditors shall devise stringent procedures to evaluate the reported rpts to minimize its negative effects. this study also contributes to agency theory and tunneling aspect of rpts in case of emerging markets. main limitation of the current study is small sample size because majority of the group firms are not listed with pakistan stock exchange and their annual reports are not publically available. keywords tunneling, rpts, business groups, accounting performance, market performance jel classification d53, l10, l25 1. introduction of the study agency theory predicts that managers or major shareholders could expropriate the wealth of minority shareholders in different ways. for example, weak corporate law and poor enforcement mechanisms can increase the fear of expropriation among minority shareholders. johnson, boone, breach and friedman (2000a) suggested that the hazard of * corresponding author: hamidullah@icp.edu.pk hamid ullah, attaullah shah & syed hamid ali shah 16 expropriation is particularly higher in firms belonging to business groups. the group firms are known to be less transparent, have more opportunities, and have complicated structures. in addition, they may have better links to the political apparatus in the country thus making it difficult to be externally interfered and monitored. in a common form of business group, a sole shareholder or family takes control of several independent firms and holds a significant ownership stake with excess cash flow rights relative to other shareholders. as such the controlling shareholder may be able to transfer or tunnel profit from the associated firms. along with the minority shareholders, tunneling also hurts development of equity market (ge et al., 2010; johnson et al., 2000b; bertrand et al., 2002; and wurgler, 2000). therefore, capital markets punish the expropriating behavior of the group firms. related party transactions (rpts) are recurring transactions of concern to the academic and practitioners due to the frequent corporate scandals. these are multifaceted complex business transactions with managers, owners, directors and associated firms etc. a party related can have variety of types. it can be a subsidiary, a joint venture, owned by a family member, or a firm owned, affiliated, or associated with some related individual(s). the inherent complexity of these transactions leads to audit risk because these transactions are difficult to be audited (johnstone & bedard, 2004). the general accounting office (2003) acknowledged that rpts are used by companies to manipulate their financial reports. the literature shows that most of the corporate collapses took place due to rpts (ge et al., 2010; and gallery et al., 2008). therefore, rpts are considered a potent threat by all stakeholders. ryngaert and thomas (2012) and ge et al., (2010) considered rpts as a mechanism to expropriate a firm’s scarce resources for a purpose other than to maximize the wealth of shareholders. so, group firms with more rpts are expected to be negatively responded by the capital markets. contrary to the above views, the “efficient transaction hypothesis” (eth) suggested that rpts are beneficial business transactions that can efficiently fulfill the economic needs of firms. thus, these transactions will advance the interests of the minority shareholders and act as an efficient contract (gordon et al., 2004). the idea that not all rpts could be the result of self-dealing are coined by kohlbeck and mayhew (2004); they opined that some more complex rpts might be in line with the eth and would thus positively support the performance of the firm. friedman et al., (2003) used the term “propping” for the first time, which refers to the relationship among the affiliated firms in which they shift resources from one firm to another through rpts. the controlling shareholders (family or government) use these resources to prop up the cash flows of the financially distressed firm through the transfer of resources from other affiliated firms. in this vein, conferring to the transaction cost hypothesis (tch) of (coase, 1937; and williamson, 1998), rpts are considered as internal dealings among the affiliated firms which provide an alternative to exchanges at the market rate and are expected to be less costly and bring ease in production (fan & goyal, journal of applied economics and business studies, volume. 3, issue 1 (2019) 15-40 https://doi.org/10.34260/jaebs.312 17 2006). thus, related party transactions could, in fact, increase performance (amzaleg & barak, 2011; and jian & wong, 2010). the above view seems to be more relevant to markets characterized by inefficient capital allocation, lack of skilled labor and product markets. as such these market imperfections could increase the probability of agency costs associated with the information asymmetry and contracting problems. however, group affiliation and internal dealings among these firms are expected to better allocate the financial, labor and material resources in a manner that would create economies of scale, bring cost-effectiveness and make easier access to the financial and other resources (ryngaert &thomas, 2007; siegel & choudary, 2012; and pizzo, 2013). few studies show that group affiliations bring technological advancement, share human, advertising, and other resource and contribute towards profitability and growth of the firms (chang and hong, 2000; siegel and choudary, 2012; lo et al. 2010; moscariello, 2007; nekhili and cherif, 2011; wahab et al., 2011). the dual nature of these transactions is one of the sources of motivation to study the economic consequences of the rpts in the perspective of pakistani business groups and the response of the capital markets towards these transactions. we add to the knowledge on the effect that rpts have on the firm accounting performance and market valuation in three ways. first, we provide new proof on the effects of rpts in the capital markets with weaker protection of minority shareholders’ rights, ownership concentration, cross-shareholdings and interlocking among the business groups and have more information asymmetry (see e.g., la porta et al., 2000) and focus on the role of the capital markets in disciplining the group firms tunneling behavior. thus, the results of this study could be directly relevant to emerging capital markets e.g., most of the asian markets having institutional setup similar to pakistan. second, this study adds to the tunneling literature and especially to the existing small number of studies that have reported a consistent decline in the financial performance of group firms and reported expropriations in the pakistani context such as ghani et al., (2005), karacaer et al., (2009), and bhutta and suleman (2016), however, they could not identify the tools that facilitate expropriations. so this study examines direct avenue i.e. rpts through which expropriation may occur and causing declines in the financial performance of group firms; and is also negatively perceived by the capital markets through adverse response to the group firms’ share prices. third, compared to the previous studies that focus on one or the other particular types of rpts separately, in this study all the rpts are considered and analyzed. moreover, apart from the conflicting results in the literature on the nature of rpts, the existing studies on the rpts are mostly based on the us, european countries and chinese firms, which have different ownership structure, investors’ protection, corporate governance (cg) systems, enforcement mechanisms than pakistan. for this purpose, the hamid ullah, attaullah shah & syed hamid ali shah 18 study investigates that either the capital markets punish the tunneling behaviour of firms belonging to business groups in pakistan. 2. literature review traditionally, research in the field of “corporate governance” has mainly focused on the conflict of interest between stockholders and management. however, particularly in the case of emerging and less developed economies, extant literature has pointed to the existence of conflict of interest between majority and minority stockholders called agency type ii (see albuquerque & schroth 2010; shleifer & vishny, 1997; barak & lauterbach, 2011; johnson et al., 2000b; and liu & magnan, 2011). la porta et al., (1998) documented that concentrated ownership due to majority shareholdings by families, institutions, or government is normal in emerging economies. these so-called insiders’ shareholders with major stakes in the firms enjoy excessive control rights and have the opportunity to expropriate resources through their operating and financial decisions (gopalan & jayaraman, 2012). according to ryngaert and thomas (2012) rpts serve as a potent mean of wealth expropriation by the majority owner(s). many studies show that several financial frauds and decline in earnings have occurred due to rpts (ge et al., 2010). statement of financial accounting standards 57 (sfas 57) and international accounting standard 24 (ias 24) defines rpts as, “the transactions between a company and its subsidiaries, affiliates, principal owners, officers or their families, directors or their families, or entities owned or controlled by its officers or their families. these transactions include sales and purchase of assets, goods and services, cash payments, loan guarantees and other types of transactions with affiliated firms.” these accounting standards require proper disclosure of rpts with the aim to reflect changes in financial position due to rpts within the financial statements for the benefits of stakeholders. the using of rpts by one party against the interests of the other is similar to the conflict of interest hypothesized earlier by berle and means (1932), followed by jensen and meckling (1976), and classens et al., (2002). johnson et al., (2000) coined the term “tunneling” for such situations where the dominant insiders transfer resources or profits from one firm to another firm. similarly, djankov et al., (2005) opined that controlling shareholders could extort cash from affiliated firms through rpts. denis and mcconnell (2003) suggested that internal mechanism (board composition, compensation system, ownership structure, and corporate transparency) and external mechanism (legal environment, protection of minority shareholders, competition for corporate control and competition within the product markets) could mitigate the incidents of rpts induced expropriation. in contrast, gordon et al., (2004) proposed the efficient transaction hypothesis and explains that rpts can fulfill the economic needs of firms. this according to friedman et journal of applied economics and business studies, volume. 3, issue 1 (2019) 15-40 https://doi.org/10.34260/jaebs.312 19 al., (2003) is propping which refers to the relation where controlling shareholders propped up the cash flows of the financially distressed firms through their private resources or group firms. siegel and choudary (2012) supported this view and explain that due to market imperfections and agency problems in the emerging markets internal dealing among group firms may better allocate the financial, labor and material resources in a manner that may create economies of scale and easier access to finance and other resources. being economical, rpts increase utilization efficiency of the firms’ resources (chien & hsu, 2010). this dual role of rpts implies that rpts could either be used to dilute wealth of shareholders (johnson et al., 2000; and cheung et al., 2006) or these could serve the economic needs of financially distress firms to enhance their financial performance (gordon et al., 2000; and jian & wong, 2004). cheung et al., (2006) categories rpts into transactions with the aim of expropriation and having a negative impact, positive impact including for the minority stockholders, and strategic importance and impact on the earnings of firms. moreover, it is worth revealing that in studies about the relationship between rpts and financial performance of firms variety of accounting and market based measures of performance are used. accounting based measures include return on asset, return on equity, return on stock price, stock price, and earnings per share, free cash flow (allgood & farrell, 2003; peng, 2004; shen & cannella, 2003; and neumann & voetmann, 2005). as these accouning-based proxies do not capture future perspectives, therefore, market-based measured are prefered in the investigation of the impact of rpts on the wealth of equity holders. however, results are inconclusive; of the two conflicting hypotheses some studies accept one and reject the other and vise versa. in line with the conflict of interest hypothesis (cih), considerable numbers of studies documented an indirect association between rpts and firms’ performance and concluded that capital markets punish the expropriating behaviour of dominant shareholders. ho et al., (2001) and gordon et al., (2004b) substantiated the cih and reported existence of the negative association of rpts with market-adjusted returns. similarly, jian and wong (2004) found that related party sales inversely influence tobin’s q and market-to-book ratio of the sample firms. yeh et al., (2012) found a negative impact of receivables to payables ratio on the firm operating performance. economic value added (eva) and market value added (mva) are also used and rpts was found to be inversely related to both these measures (chiou & huang, 2006; and cheung et al., 2009). jian and wong (2004); nekhili and cheirf, (2011) used tobin's q to report an indirect relationship with rpts. michael and shawn (2011) conducted interesting research while considering ex-ante rpts with counterparty before becoming associated firm and ex-post rpts. they documented direct relatoionship rpts and tobin's q in the ex-ante time period while the association is reported as negative in the ex-post time period. moreover, liu, qiao and joe (2007) suggested that operating hamid ullah, attaullah shah & syed hamid ali shah 20 revenue almost equal to the related party receivables in the case of excessive transactions with foreign subsidiaries of firms. they also reported relatively longer collection period of these related party receivables and thus could turn deterimental to both liquidity and performance of the firms (huang & liu, 2010). keeping in view the characteristics of pakistani economic system and following the majority, it is hypothesized that: h1: there is an inverse relationship between related party transactions (rpts) and the various accounting-based measures of financial performance (roa, roe, and earnings per share) of firms. h2: there is an inverse relationship between related party transactions (rpts) and the market-based financial performance of firms. 3. research methodology this section discusses data collection and sample framework and introduces the regression model(s) employed to analyze the data. 3.1 data collection and sample framework data of cg and ownership structure variables such as size of board, independence of board, ceo-duality and compensation, director ownership, ownership of associated firms, institutional ownership, foreign ownership, block-holdings, audit quality, number of a banking relationship, bank borrowings is acquired from the individual firm’s annual reports for the period 2006-2018. moreover, related party sales, purchases, payables, receivables, and other firm-level financial data is acquired from a “balance sheet analysis of financial statements” of sbp (state bank of pakistan) and annual reports of firms for the same sample period. we searched for firms from different sectors listed on “pakistan stock exchange” (psx) that have reported rpts for the sample period and have followed the corporate governance code 2002 and 2012 regarding evaluation and disclosure of rpts. financial and government and quasi-government firms are excluded from the sample. furthermore, data is normalized through winsorization and observation with extreme values, negative equity and losses are dropped from the data. we use cook’s d to identify and remove abnormal observations. this shall help achieve goodness of fit and generalizability of the results. finally, an unbalanced panel of 207 firms with 1098 firm-year observations is left for the estimations of the different econometric models. 3.2 statistical modeling following yezhen and wong (2015) and lin et al., (2010), the study used return on assets (roa) and returns on equity (roe) and earnings-per-share (eps) as measures of the dependent variable. the following econometric model is estimated to quantify the relationship between rpts and the dependent variable. journal of applied economics and business studies, volume. 3, issue 1 (2019) 15-40 https://doi.org/10.34260/jaebs.312 21 profi,t= α+ β1 b.si,t + β2 nedi,t + β3 b.mi,t + β4 ceo.sali,t + β5 ceo.dualityi,t + β6 insti,t + β7 m.oi,t + β8 f.oi,t + β9 assoi,t + β10 blocki,t 20% + β11 blocki,t 30% + β12 eari,t + β13 easi,t + β14 m.banksi,t + β15 l.amounti,t + β16 levi,t + β17 f.si,t + β18 f.gi,t + β19 rpti,t + ƞi + ƛt + µi,t -------------(1) profit is measured in terms of eps, roa and roe of firm i at time t and b.s, ned, b.m, ceo-sal, ceo-duality, inst, m.o, f.o, asso, block 20%, block 30%, ear, eas, m.banks, l.amount, lev, fs, f.g and rpt are explanatory variables and represent board size, ceo-duality, board meetings, ceo-compensation, board independence, institutional ownership, foreign ownership, managerial ownership, associate ownership, blockholdings20%, blockholdings30% , external audit quality, external audit fee, number of banks in relationship, loan amount from banks, leverage, firm size and related party transactions proxies respectively. moreover, α0 and µi,t are the usual constant and error terms of the regression; ƞi, and ƛt represent industry and year dummies. in the case of market-based performance, 2-stage least square (2sls) estimation technique is used because both cg and ownership structure of a firm are said to have influence on rpts where the later is used to explain the value of firm. (gordon et al., 2006; and nekhili & cherif, 2011). it is stated that both cg and ownership structure of a firm affect rpts. therefore, these studies suggested that the related party transactions that can determine the value of company are themselves affected by the company's ownership characteristics and governance mechanisms. thus, in the first equation, we check the effect of rpts, with the complete set of variables on the tobin’s q. in the second equation, we check the effect of the cg and ownership variables on the rpts. both rpts and tobin's q are endogenous variables; whereas, the cg, ownership, and the control variables are exogenous. leverage and external audit quality are used as instrumental variables in the model of rpts. nekhili and cherif (2011) used dividend as a variable for the protection of minority shareholders; we prefer leverage (debts to equity) and external audit quality (audit by big4) as monitoring devices for protection of the monitory shareholders’ wealth. 3.2.1 endogeneity test in the presence of endogeneity, ols and 2sls regressions estimates are inefficient woodridge (2001). to test that firm value and rpts are endogenously determined, durdinwu-hausman test (1978) has suggested a test that compares the ols with 2sls estimates. this test can be operationalized in two steps. in the first step, the rpts is regressed against the instrumental variables and all variables in the tobin's q regression. the instrumental variables are expected to be correlated with the rpts and uncorrelated with the error term. the residuals, predicated from this auxiliary regression, are then included in the value regression as an explanatory variable. if the residuals are found statistically significant, it is taken as endogeneity. in the case of our study, the results of the residuals are significant for hamid ullah, attaullah shah & syed hamid ali shah 22 the tobin’s q but insignificant for the roa, roe, and eps. therefore, this study has used 2sls for the tobin’s q while ols in case of roa, roe, and eps. tobinsqit= α0+β1b.sit+ β2nedit+ β3instit+ β4m.oit + β5asso.oit+ β6f.sit + β7roait + β8tangit + β9drptrpit + β10drtpspit+ ηi + λt + μi,t ---------(2) drptrpit= α0+β1b.sit+ β2nedit+ β3instit+ β4m.oit + β5asso.oit+ β6f.sit + β7 eaqit + β8levit + ηi + λt + μi,t ------------------------------------------------(3) drptspit= α0+β1b.sit+ β2nedit+ β3instit+ β4m.oit + β5asso.oit+ + β6f.sit + β7eaqit + β7levit+ ηi + λt + μi,t ----------------------------------------------(4) 3.3 operationalization and justification of variables the current study uses various variables to account for governance mechanisms, ownership structure prevailing in a country, external agents of governance such as external auditors and banks monitoring. the governance variables include corporate board size measured through number of board members of a firm in a particular year (gao & king, 2008; and ullah & shah, 2015). these members include both “executive and non-executive/independent directors”. independence of corporate board is captured through the ratio of independent directors to the board size. board meetings are used to represent the efficiency of the corporate board as required under the “code of corporate governance 2012” all rpts should be approved by the corporate board. therefore, these variables directly affect the quantity and quality of the rpts (ullah & shah, 2015). ceo-duality is a common practice in pakistani business groups and hence considered as control. dual position of the ceo may facilitate the expropriation of minority shareholders wealth through rpts (ullah & shah, 2015). the ownership structure is another commonly observed set of variables that have significant association with the rpts. in the context of pakistan most widely used variables in agency framework are managerial ownership that represents the portion of ownership held by the directors and their blood relative (gao & king, 2008). ownership of associated firms in pakistani group firms is computed as the ownership stake of other firms out of total ownership (ullah & shah, 2015). we expect that the higher level of this ownership will be positively related with the number of rpts and negatively affect performance of firms. institutional ownership and holding of the foreign investors in domestic firms assumed to play significant role in balancing interest of various stakeholders (berkman et al., 2009; and gao & kling, 2008). in emerging markets institutional investors are considered as disciplining agents due to their capacity, knowledge and access to information. the study compute institutional ownership as a portion of ownership held by financial institutions and foreign ownership as a portion of ownership held by foreign investors respectively (gao & kling, 2008). journal of applied economics and business studies, volume. 3, issue 1 (2019) 15-40 https://doi.org/10.34260/jaebs.312 23 the study use external agents such as external auditor and number of banks in relationship with firms. the study consider external auditor as an agent of governance that can closely monitor the quality, quantity and pricing mechanism used by firm in executing rpts. therefore, the external auditors can play significant role in reducing the likelihood of expropriating wealth of minority shareholders by using such transactions (gao & kling, 2008). the study also incorporate big4 as a representation for the quality of audit and has assigned a “value of 1 if the firm auditor is from big4 otherwise 0” is assigned (bennouri, nekhili & touron, 2015). the number of banks in relation with firms is used as an external governance agent for the first time in pakistan. we expect that the number of banks in relationship with a firm are positively associated with level of monitoring by these banks (ullah & shah, 2014). thus, as the number of banks increases the probability of expropriations through rpts will decrease and will positively affect firm value (ghosh, 2007). these banks are expected to control profit diluting and credit risk enhancing transactions (ullah, & shah, 2011). the control variables include firm size and leverage. earlier research studies reported positive connection of firm size with rpts and tobin’s q and negative association of leverage with rpts and tobin’s q. the study measure size as log to total assets (yeh, shu & su, 2012) and leverage as debts to equity of firm in a given year (gao & kling, 2008). 4. empirical analysis and discussion descriptive statistics, and correlation, and regression results are presented in table 1 to table 4. 4.1 descriptive statistics & pearson correlation descriptive statistics of all the variables except the rpts are reported in table 1 where these statistics for the measures of rpts are shown in table 2. as shown in table 1, the average numbers of member of corporate board are 9. on the average 35% of the members are external directors. a significant number (42%) of ceo perform dual functions. on the average the boards of the sample firms hold 6 meetings annually. the ownership of associated firms has a minimum value 0, the maximum value of 0.90 and its mean value is 0.493. we expect that this high percentage of associated firms’ ownership in pakistan might causes the firms to have a controlling decision right with the controlling shareholders. the mean value of blockholder20% is 0.7021 and the average value of the blockholder30% is 0.4584. these statistics also show that on the average managerial ownership is 0.383, institutional ownership is 0.0862, and foreign ownership is 0.0326. the average value of roa, roe, eps, and net profit margin are 0.87, 0.65, 12.46 and 0.039 respectively. to remove outliers from these variables, all corresponding rows where an abnormal value of these variables exists are dropped and as such 46 observations are dropped. the same hamid ullah, attaullah shah & syed hamid ali shah 24 procedure was used for other variables such as size & growth of firms, tangibility, and leverage. in total 212 observations were dropped. table 1 descriptive statistics table 1 shows descriptive statistics of the variables used in different regression models by using data from 2006 to 2018 for 207 firms belonging to different associated companies. board size is computed as a log of number of board members, board independence is computed as the number of non-executive directors divided by total directors, board meetings are computed as a log of the total number of board meetings. ceo-duality is equal to 1 if the chairman and ceo position is held by the same person otherwise 0, ceo compensation is computed as log of annual salary of ceo, associate ownership stands for percentage of associate ownership to total ownership, blockholder20% and blockholder30% stands for the major shareholder holding 20% and 30% shares and are coded 1 otherwise 0, “managerial ownership represents percentage of ownership of variables n mean s.d min max corporate governance board meetings 908 5.638 2.655 3.000 12.000 board independence 910 0.556 0.287 0.000 0.875 board size 909 8.209 1.614 7.000 15.000 ceo-duality 910 0.426 0.495 0.000 1.000 ceo-compensation 910 6.532 1.398 0.000 8.317 ownership structure associate ownerships 910 0.298 0.294 0.000 0.950 blockholder20 903 0.722 0.447 0.000 1.000 blockholder30 910 0.459 0.499 0.000 1.000 managerial ownership 910 0.383 0.238 0.000 0.938 institutional ownership 910 0.086 0.146 0.000 0.797 foreign ownership 910 0.033 0.116 0.000 0.085 external agent big4 910 0.431 0.59 0 1 no of external auditor 910 1.352 0.498 0.000 2 external auditor fee 910 5.991 0.637 0.000 7.444 no banks in relationship 910 4.611 2.273 0.000 15.800 bank loans 910 0.509 0.125 0.000 0.815 financial variables leverage 910 0.498 0.126 0.000 0.765 firm size 910 9.157 2.575 0.000 12.730 firm growth 910 8.935 2.847 0.000 12.670 tang 910 0.449 0.327 0.000 0.799 financial performance eps 910 0.322 0.994 -3.876 9.716 roa 910 0.002 0.019 -0.079 0.526 roe tobinsq 910 910 0.094 12.821 0.226 11.029 -0.763 0.000 0.753 24.560 journal of applied economics and business studies, volume. 3, issue 1 (2019) 15-40 https://doi.org/10.34260/jaebs.312 25 directors, their children and spouses divided by total ownership, institutional ownership represents the percentage of ownership of financial institutions divided by total ownership, foreign ownership represents the percentage of ownership of foreign investors divided by total ownership”. a number of external auditors represent the total number of audit firms which have performed their audit in a given year. the external audit fee is computed as a log of external audit fee, a number of banks in relationship represent the total number of banks that have loan contracts with the firm in a given year, and banks loan is computed as banks loan divided by the total liability of a firm in a given year. control variables include firm size is computed as log of total assets, firm growth is computed as changes in the firm fixed assets, leverage ratio is computed as debts divided by total assets, tangible is computed as fixed assets divided by total assets, eps is computed as net profit divided by outstanding shares, roa is net profit divided by total assets, roe is net profit divided by equity of a firm in a given year. in table 2 the proxy of rpts drptrp is measured as the difference of related party receivables and payables scaled by total assets. its value ranges from -0.444 to 0.3378 and has a mean of -0.030 across all the firms and the time periods. the second proxy of rpts drptsp is the difference of related party sale and purchase divided by total assets. its average value is -0.023 having a maximum (minimum) value of 0.365 (-0.686). these statistics show that on average 3% of the firm assets and 2.33% of the firms’ sales and purchases might be expropriated, however to be confirmed through the regression analysis. rploan is the third proxy of rpts and it is computed as related party loans divided by total assets. it has a maximum and minimum values of 0.68 and 0.00 respectively whereas on the average 29% of the total liabilities are loans granted through rpts and may increases up to 69 %. table 2: descriptive statistics variables n mean sd min max drptrp 910 -0.03 0.315 -0.444 0.338 drptsp 910 -0.021 0.078 -0.379 0.366 director loans 910 0.291 0.011 0 0.68 table 2 shows descriptive summary of the different types of related party transactions, drptrp stands for difference of the related party receivables and payable scaled by total assets for a firm in a given year. drptsp stands for the difference between the related party sale and purchases scaled by firm total assets. director loan represents the loan granted to different directors and is computed as a loan granted divided by total assets of a firm in a given year. n shows a number of observations, mean shows average value, sd stands for standard deviations, min stands for the minimum value of a variable and max represents the maximum value of a variable. hamid ullah, attaullah shah & syed hamid ali shah 26 the correlations between the variables are presented in table-a-1(annexued-1) which showed the magnitude of linear association between the variables. the correlation statistics showed no issue of multicollinearity in the given set of the explanatory variables. both drptrp and drptsp exhibit a positive linear association with ceo-duality, ownership of associated firms, blockholders20%, and 30%. whereas, the two measures are negatively correlated with size, independence, and meetings of board, ceo composition, institutional and foreign ownership, audit fee, an external audit from big4, number of banks in a relationship, banks loan, leverage, firm size, firm growth, and tangibility. these associations imply that firm with relatively higher managerial ownership, more ownership of associated firms, ceo-duality may have high probability of rpts. firms with large sized boards, many independent directors in the board, good compensation policy, more intuitional and foreign ownership, good audit quality and lending from more banks would have low probability of related parted transactions. these associations substantiate the theoretical predictions in the literature review with the exception that there exists a high correlation between the size of firm and growth and ownership of associated firms with blockholder20%. these simple correlations would be further analyzed through more rigorous regression models. these relationships also show that smaller firms have less tangible assets, low growth rate, lower leverage and are lesser profitable. 4.2 rpts and market financial performance table 3 presents results of the regression model with tobin’s q as the dependent variable. the table reports the estimated values of the parameters of the independent variables with their respective standard errors in the parentheses. the test statistics of durbin (score) and wu-hausman are significant at the 5% and suggests that rpts are endogenously determined by the cg and ownership structure variables (nekhili & cherif, 2011). the wald test is significant at 1% and means that the instruments used in the tobin’s q equations are not weak. the result of the sargan (score) and basmann tests of over-identification for 2sls models of tobin’s q are insignificant, which shows that the models are not overidentified. the year and industry dummy variables in these models are statistically relevant at 5%. the results of both ols and 2sls show that tobin’s q is inversely influenced by the measures of rpts. these results support the hypothesis that there is a negative relationship between rpts and the value of the firm. these results support the view that rpts are considered value destroying by the market participants. putting differently, these transactions are assumed to be a tool through which the wealth of shareholders is transferred to the controlling shareholders or associated/ affiliated firms (nekhili & cherif, 2011). these results are in line with the notion that capital markets do punish the tunneling behaviour exhibited by the group firms. the results corroborate the findings reported by journal of applied economics and business studies, volume. 3, issue 1 (2019) 15-40 https://doi.org/10.34260/jaebs.312 27 gordon (2004); gordon et al., (2006), and jian and wong (2004). size and independence of board are positive and significant in all models. the extant literature reveals that larger boards offer more benefits to firms (coles et al., 2008; and kiel & nicholson, 2003). from the resources dependency perspective, the direct association of size of board with tobin’s q is said to bring more opportunities, create linkages with other firms in industry, and access their resources such as financial and technical advice (ge et al., 2010; hillman and dalziel, 2003; kiel and nicholson, 2003; kula, 2005; liew et al., 2015; pfeffer and salancik, 1978; zahra and pearce, 1989). similarly, our results are also in line with the agency theory perspective that the larger boards are capable to better monitor firms’ managerial activities and as such may diminish the probability of expropriation by the controlling shareholders (kiel & nicholson, 2003). the positive and significant relationship of board independence and tobin’s q is due to the reason that independent directors are assumed to be neutral and could act to balance the power between insiders and outsiders by reducing the probability of expropriations (nekhili & cherif, 2011; gordon et al., 2006; liew et al., 2015). furthermore, these independent directors usually have more expertise and industry-specific knowledge that may bring effectiveness in their decisions, which may positively affect the value of a fund (dalton et al., 1998). managerial ownership and ownership of associated firms have significant but an inverse impact on the dependent variable. the relationship is significant in 2sls regressions only. in accordance to shleifer and vishny (1997) large shareholders with more controlling power may allocate firm resources in a manner that may lead to expropriation (nekhili & cherif, 2011; gordon et al., 2006). these results reveal that an increase in the ownership by managers and associated firms lead to more control over the resources distribution and increase both agency costs and chances of expropriation by them. ultimately it would deteriorate value of the fund. these findings of the inverse relationship support the entrenchment and expropriation hypotheses and fail to support the alignment of interest hypothesis proposed by jensen and meckling (1976). the negative and insignificant effect of institutional ownership suggests that either these institutional investors are passive or collude with the associated firms or managers and thus are collectively involved in the expropriation of firms’ resources or they may buy shares for speculating market for excess returns only (demsetz & villalonga, 2001). size of fund and tangibility show direct and significant association with value of firms. these findings imply that small firms and firms with fewer tangible assets face greater probability of bankruptcy and less investors’ confidence to invest in these firms (titman & wessels, 1988). moreover, these firms relative to large firms have the disadvantage to acquire economies of scale and have relatively lesser market exposure (petit & singer, 1985). https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib24 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib65 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib55 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib55 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib65 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib66 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib83 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib83 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib65 hamid ullah, attaullah shah & syed hamid ali shah 28 table 3 regression results of related party transactions and firm value. (tobinsq) (tobinsq) (tobinsq) (tobinsq) (tobinsq) (tobinsq) variables ols 2sls ols 2sls ols 2sls drptrp -0.2077* -0.2829** (0.1101) (0.1371) drptsp -0.2451* -0.1202** (0.1350) (0.6050) dirloan -0.3151*** -0.4132** (0.1251) (0.1051) board size 0.2129* 0.2243* 0.2158* 0.2219* 0.2058* 0.2118* (0.1169) (0.1312) (0.1160) (0.1172) (0.1160) (0.1071) board independence 0.7351*** 0.1364** 0.8017*** 0.7672** 0.8923*** 0.9912*** (0.2149) (0.6620) (0.1937) (0.3151) (0.1937) (0.3151) associated ownership -0.8339* -0.2874** -0.2504 -0.22928* -0.3230 -0.3195** (0.499) (0.1215) (4.370) (0.1151) (3.2101) (0.1551) managerial ownership -0.2848 -1.090* -0.4249 -0.6238** -0.2765 -0.8453** (0.8532) (0.491) (0.8883) (0.2988) (0.8953) (0.2988) institutional ownership -0.3915 -0.4700 -0.9117 -0.6036 -0.7114 -0.5926 (7.148) (10.15) (7.207) (0.7294) (7.207) (0.7252) firm size -0.2054 0.5215* -0.1373 0.9895* -0.1643*** 0.9895** (0.2336) (0.2988) (0.2406) (0.3203) (0.0406) (0.3203) tangibility -0.5555 0.4933* -0.5197 0.4591** -0.4127 0.5151*** (0.4713) (0.2575) (0.4615) (0.2516) (0.3415) (0.1011) roa 0.6680 0.1676 0.7294 0.4972 0.7294 0.4552 (0.5327) (0.5188) (5.353) (5.366) (5.369) (3.252) constant -0.1476* -0.2490* -1.617* -1.692* -0.610 -0.692 (0.0844) (148.3) (0.8271) (0.8958) (.7350) (0.6478) year dummy yes yes yes yes yes yes sector dummy yes yes yes yes yes yes observations 885 885 885 885 885 885 r-squared test of endogeneity durbin (score) wu-hausman wald test test of over identifying sargan (score) basmann 0.012 5.72(0.016) 5.68(0.017) 16.56(0.000) 2.56(0.107) 1.65(0.197) 0.011 9.01(0.006) 8.68(0.007) 14.16(0.000) 1.23(0.302) 1.01(0.301) 0.045 6.52(0.010) 7.54(0.011) 18.16(0.000) 1.06(0.507) 1.12(0.297) table 3 show results of ordinary least square (ols) and 2 stages least square (2sls). in these models, tobin’s q act as dependent whereas drptrp is the main independent variables which is computed as the difference of the related party receivables and payables scaled by total assets for a firm in a given year and drptsp which is computed as the difference of the related party sales and purchases scaled by firm total assets. refer to note to table 1 & 2 for the definition of the other variables. robust standard errors in parentheses showing different level of significance at 1% 5% and 10% represented by *** p<0.01, ** p<0.05, * p<0.1. journal of applied economics and business studies, volume. 3, issue 1 (2019) 15-40 https://doi.org/10.34260/jaebs.312 29 4.5 rpts and accounting-based measures of financial performance table 4 reports results of regression analysis of different models with the three different accounting-based measures of financial performance of the firms. the main focus of the analyses is on drptrp and drptsp as primary variables of interest. all other variables including the year and industry dummy variables are included in the estimated models. consistent with the agency theory and expropriation hypothesis, coefficients of the difference of the related party receivables and payables in column 2 through 4 are showing an inverse relationship with the dependent variables eps, roa, and roe. these results indicate that an increase in rpts dilute financial performance of the firms and rpts are not used for the benefit of the associated firms. these findings coincide with those reported by kohlbeck and mayhew (2010), gordon et al., (2006) and cheung et al., (2006). we find that the coefficients of size and independence of boards of the firms are positive however statistically significant in three of these models. larger boards are expected to have skilled members with diversified experiences and therefore could increase the financial performance of firms. moreover, larger boards would better monitor managerial activities and therefore will effectively diminish the probability of expropriation by the controlling shareholders or associated firms (kiel & nicholson, 2003). from the resource dependency perspective, large board may create more opportunities for the firms, establish corporate linkages with other firms in the industry to access their resources (hillman & dalziel, 2003; kiel & nicholson, 2003; kula, 2005; pfeffer & salancik, 1978; zahra and pearce, 1989). the direct association of board independence and accounting based financial performance measures is in accordance to the agency theory that independent directors are assumed to be neutral and thus reduce the expropriations capacity of insiders (nekhili & cherif, 2011; gordon et al., 2006). the findings are similar to the previous studies (see e.g., coles et al., 2008; and kiel & nicholson, 2003). further, results show that financial performance is an increasing function of ceo compensation. however, this relationship lacks statistical significance. ceo compensation mitigates the negative effects of rpts on the performance of the firms. ceo compensation is used as internal governance tool to motivate him/her to safeguard the interests of all investors and involve him/herself in profit creating activities (core et al., 1999; gordon et al., 2004; gao and king, 2008). the estimated coefficients of managerial ownership and ownership of associated firms reflect that an increase in managerial ownership (entrenched managers) and ownership of associated firms may provide more control over the resources distribution of the firms and as such increase chances of expropriation. the results further reveals that increase in institutional ownership and external audit fee both enhances performance and mitigate the negative impact of rpts on the performance of the firms. financial performance of the firms declines with an increase in leverage, bank loans, and number of banks. though statistically insignificant however as expected, audit from the big4 auditor increase earnings https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib65 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib55 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib55 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib65 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib66 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib83 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib83 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib24 https://www.sciencedirect.com/science/article/pii/s2314721016300238#bib65 hamid ullah, attaullah shah & syed hamid ali shah 30 per share and return on assets and equity. last, size and tangibility show positive and significant association with the accounting-based measures of financial performance of firms. this suggests that such firms have the advantage of economies of scale and secure earning more profits (petit & singer, 1985). the results of the proxy of rpts measured as the difference of the related party sales and purchases in column 5th to 7th are consistent and support predictions of the agency theory and expropriation hypothesis. as was the case in preceding paragraph, the estimated results show financial performance of firms computed through accounting-based measures is the decreasing function of rpts. in general, results of the other independent variables show similar effect as found in the earlier discussions. to sum up, this inverse relationship of the financial performance of firms and the rpts substantiate the conflict of interest and expropriation hypotheses and major shareholders are using these transactions to the disadvantage of other shareholders, however, to advance their own interest. further, this relationship in the case of market-based measure of financial performnace of the firms support the notation that capital markets do penalize group firms for tunneling behavior. table 4. regression analysis of firm performance rpts and corporate governance (eps) (roa) (roe) (eps) (roa) (roe) variables model 1 model 2 model 3 model 1 model 2 model 3 drptrp -0.319** -0.0649*** -0.0313* (0.162) (0.0270) (0.0163) drptsp -0.628*** -0.00704 -0.0683*** (0.198) (0.0566) (0.0200) board size 0.251** 0.0832** 0.0102 0.247 0.0825** 0.00987 (0.125) (0.0417) (0.0262) (0.224) (0.0417) (0.0262) board independence 0.334** 0.0169 -0.0302* 0.326** 0.0183 -0.0294* (0.161) (0.0122) (0.0162) (0.160) (0.0121) (0.0161) ceo-duality -0.0777** -0.00334 -0.00384 -0.0771** -0.00329 -0.00377 (0.0330) (0.00601) (0.00366) (0.0337) (0.00606) (0.00372) ceo-compensation 0.0293** 0.00596** 0.00351*** 0.0310** 0.00631* 0.00367*** (0.0123) (0.00287) (0.00124) (0.0121) (0.00383) (0.00123) associate ownership 0.2901*** -0.0200 0.7071*** -0.0227 -0.0213* -0.7691*** (0.0625) (0.0122) (0.0771) (0.0626) (0.0122) (0.0771) managerial ownership -0.0860** -0.0141 -0.0690* 0.0489** -0.0119 -0.00789 (0.0437) (0.0152) (0.0391) (0.0240) (0.0153) (0.00894) institutional ownership 0.268*** 0.0533*** 0.0310*** 0.264*** 0.0528*** 0.0306*** (0.0883) (0.0179) (0.0100) (0.0875) (0.0176) (0.00993) external audit fee 0.0819*** 0.0152*** 0.00778*** 0.0830*** 0.0154*** 0.00789*** (0.0265) (0.00579) (0.00270) (0.0268) (0.00585) (0.00273) external audit big4 0.00568 0.00414 0.00200 0.00195 0.00342 0.00238 (0.0485) (0.00994) (0.00508) (0.0487) (0.00999) (0.00509) no banks in relationship -0.0423 -0.0353*** -0.00745 -0.0414 -0.0351*** -0.00738 journal of applied economics and business studies, volume. 3, issue 1 (2019) 15-40 https://doi.org/10.34260/jaebs.312 31 (0.0622) (0.0124) (0.00649) (0.0621) (0.0126) (0.00648) banks loan -0.0345*** -0.00612*** -0.00376*** -0.0341*** -0.00600*** -0.00373*** (0.00855) (0.00170) (0.00101) (0.00860) (0.00174) (0.00101) firm size 0.0230*** 0.000894 0.00237*** 0.0229*** 0.000890 0.00236*** (0.00507) (0.00110) (0.000528) (0.00508) (0.00111) (0.000528) tang 0.0310 0.0247** 0.00578 0.0298 0.0251** 0.00567 (0.0554) (0.0104) (0.00625) (0.0560) (0.0104) (0.00630) leverage -0.629*** -0.184*** -0.0699*** -0.623*** -0.183*** -0.0693*** (0.0588) (0.0139) (0.00676) (0.0587) (0.0138) (0.00674) constant 0.129 -0.00509 0.0328 0.101 -0.0112 0.0302 (0.328) (0.0577) (0.0374) (0.330) (0.0576) (0.0375) industry dummy year dummy yes yes yes yes yes yes yes yes yes yes yes yes observations 948 948 948 948 948 948 f-test 9.45(0.000) 7.68(0.000) 10.45(0.000) 8.75(0.000) 9.54(0.000) 8.59(0.000) r-squared hausman test 0.301 37.98(0.00) 0.325 41.41 (0.000) 0.193 39.89(0.000) 0.299 49.78(0.000) 0.323 51.45(0.000) 0.192 48.59(0.000) table 4 show results of different regression models with dependent variables eps, roa and roe and main independent variables drptrp which is computed as the difference of the related party receivables to payables scaled by total assets for a firm in a given year and drptsp which is computed as the difference of the related party sales and purchases scaled by firm total assets. refer to note to table 1 & 2 for the definition of other variables.. robust standard errors in parentheses showing different level of significance at 1% 5% and 10% represented by *** p<0.01, ** p<0.05, * p<0.1. 5. conclusion and future scope this study is an endeavor to determine that do capital markets punish the tunneling behavior exhibited by the group firms major shareholders. the study used agency framework to analyze the influence of rpts on the financial performance in the pakistani listed firms. for this purpose data for the period, 2006-2018 of 207 firms listed in pakistan stock exchange is analyzed through ols and 2sls regression estimation techniques. using multiple/alternative measures for the variables, it is found that financial performance and rpts have a persistent negative association. rpts are considered as conflicting by investors in pakistan because these transactions have an adverse effect on the profitability of firms and market value. these results are consistent with the cih and contradictory to the eth. the results also supported that the existence of large board size and more independent members are negatively associated with the rpts, whereas, higher managerial ownership and ownership of associated firms also found to have a negative influence on the rpts. the results highlighted importance of the internal and external audit quality in controlling the expropriations through rpts. these results support the view that capital markets penalize group firms for their expropriating behavior. on this note, a stern regulatory framework and corporate governance mechanisms are required to be in placed in order to reduce the expropriation of minatory shareholders wealth through the rpts in pakistan. future studies hamid ullah, attaullah shah & syed hamid ali shah 32 may use data of firms of multiple countries. furthermore, the issue of the rpts could be investigated with respect to dividend payout, performance shock etc, to test the agency conflict and expropriation of minority shareholders by group firms. in future, other measures of performances may also be considered. the findings of this study have significance for academia, investors, auditors and policymakers. this is first study of its nature on the rpts and its economic consequences in pakistan; we expect more follow up studies and further research. the findings showed that rpts are used as a source of expropriations in pakistani group firms. so auditors should handle such transactions more carefully and closely examine the pricing mechanisms used by firms in rpts related contracts to ensure that these transactions are executed at armslength-price with no harm to minority shareholders. the study has showed a negative response of investors towards share pricing of the group firms. therefore, investors in stock market may avoid investing in such firms with higher intensity of rpts. the findings of negative response of investors towards rpts also invite the attention of corporate boards and independent directors; they shall carefully evaluate and examine rpts in order to minimize its value destroying impact. main limitation of the current study is the small sample size because data of business group’s firms in majority cases is not available. mostly group firms are not listed with pakistan stock exchange and their annual reports are not publically available. moreover, the data is collected from the annual reports of the firms so the validity of the results entirely depends on the data provided by firms in their audited reports. the study has used only tobin’s q to measure the market response towards the rpts, other measures may be considered for more robust results. references albuquerque, rui, and enrique schroth. 2010. “quantifying private benefits of control from a structural model of block trades.” journal of financial economics 96 (1): 33–55. https://doi.org/10.1016/j.jfineco.2009.12.003. allgood, sam, and kathleen a. farrell. 2003. “the match between ceo and firm*.” the journal of business 76 (2): 317–41. https://doi.org/10.1086/367752. amzaleg, yaron, and ronen barak. 2011. “ownership concentration and the value effect of related party transactions.” ssrn electronic journal. https://doi.org/10.2139/ssrn.1959557. aswadi, effiezal, and abdul wahab. 2011. “does corporate governance matter ? evidence from related party transactions in malaysia does corporate governance matter ? evidence from related party transactions in malaysia.” journal of applied economics and business studies, volume. 3, issue 1 (2019) 15-40 https://doi.org/10.34260/jaebs.312 33 ballantine, h. w., adolf a. berle, and gardiner c. means. 1932. “the modern corporation and private property.” california law review 21 (1): 78. https://doi.org/10.2307/3475545. barak, ronen, and beni lauterbach. 2011. “estimating the private benefits of control from partial control transfers: methodology and evidence.” international journal of corporate governance 2 (3/4): 183. https://doi.org/10.1504/ijcg.2011.044374. bennouri, m., nekhili, m. and touron, p. (2015) ‘does auditor reputation “discourage” related-party transactions? the french case’, auditing: a journal of practice & theory, 34(4), pp. 1–32. doi: 10.2308/ajpt-51036. berkman, henk, rebel a. cole, and lawrence j. fu. 2009. “expropriation through loan guarantees to related parties: evidence from china.” journal of banking & finance 33 (1): 141–56. https://doi.org/10.1016/j.jbankfin.2007.11.001. bertrand, m., p. mehta, and s. mullainathan. 2002. “ferreting out tunneling: an application to indian business groups.” the quarterly journal of economics 117 (1): 121–48. https://doi.org/10.1162/003355302753399463. bhutta, aamir, and muhammad tahir suleman. 2016. “performance of pakistani business groups during energy crisis.” ssrn. https://doi.org/10.2139/ssrn.2868971. bona-sánchez, carolina, carmen lorena fernández-senra, and jerónimo pérez-alemán. 2017. “related-party transactions, dominant owners and firm value.” brq business research quarterly 20 (1): 4–17. https://doi.org/10.1016/j.brq.2016.07.002. chang, s. j., and j. hong. 2000. “economic performance of group-affiliated companies in korea: intragroup-resource sharing and internal business transactions.” academy of management journal 43 (3): 429–48. https://doi.org/10.2307/1556403. cheung, y, p rau, and a stouraitis. 2006. “tunneling, propping, and expropriation: evidence from connected party transactions in hong kong☆.” journal of financial economics 82 (2): 343–86. https://doi.org/10.1016/j.jfineco.2004.08.012. cheung, yan-leung, lihua jing, tong lu, p. raghavendra rau, and aris stouraitis. 2009. “tunneling and propping up: an analysis of related party transactions by chinese listed companies.” pacific-basin finance journal 17 (3): 372–93. https://doi.org/10.1016/j.pacfin.2008.10.001. chien, chu-yang, and joseph c. s. hsu. 2010. “the role of corporate governance in related party transactions.” ssrn electronic journal. https://doi.org/10.2139/ssrn.1539808. https://doi.org/10.2307/1556403 hamid ullah, attaullah shah & syed hamid ali shah 34 claessens, stijn, simeon djankov, joseph p. h. fan, and larry h. p. lang. 2002. “disentangling the incentive and entrenchment effects of large shareholdings.” the journal of finance 57 (6): 2741–71. https://doi.org/10.1111/1540-6261.00511. coase, r. h. 1937. “the nature of the firm.” economica 4 (16): 386–405. https://doi.org/10.1111/j.1468-0335.1937.tb00002.x. dalton, dan r., catherine m. daily, alan e. ellstrand, and jonathan l. johnson. 1998. “meta-analytic reviews of board composition, leadership structure, and financial performance.” strategic management journal 19 (3): 269–90. https://doi.org/10.1002/(sici)1097-0266(199803)19:3<269::aidsmj950>3.0.co;2-k. demsetz, harold, and belén villalonga. 2001. “ownership structure and corporate performance.” journal of corporate finance 7 (3): 209–33. https://doi.org/10.1016/s0929-1199(01)00020-7. denis, d. 1999. “ownership and board structures in publicly traded corporations.” journal of financial economics 52 (2): 187–223. https://doi.org/10.1016/s0304405x(99)00008-2. djankov, simeon, rafael laporta, florencio lopez-de-silanes, and andrei shleifer. 2005. “the law and economics of self-dealing.” cambridge, ma. https://doi.org/10.3386/w11883. fan, joseph p. h., and vidhan k. goyal. 2006. “on the patterns and wealth effects of vertical mergers*.” the journal of business 79 (2): 877–902. https://doi.org/10.1086/499141. friedman, eric, simon johnson, and todd mitton. 2003. “propping and tunneling.” journal of comparative economics 31 (4): 732–50. https://doi.org/10.1016/j.jce.2003.08.004. gallery, gerry, natalie gallery, and matthew supranowicz. 2008. “cash‐based related party transactions in new economy firms.” accounting research journal 21 (2): 147–66. https://doi.org/10.1108/10309610810905935. gao, l. and kling, g. (2008) ‘corporate governance and tunneling: empirical evidence from china’, pacific-basin finance journal, 16(5), pp. 591–605. doi: 10.1016/j.pacfin.2007.09.001. gao, simon s., gordon gao, and tianxi zhang. 2008. “corporate governance reform and firm performance: evidence from china.” in, 189–209. https://doi.org/10.1016/s1479-3563(08)08007-9. journal of applied economics and business studies, volume. 3, issue 1 (2019) 15-40 https://doi.org/10.34260/jaebs.312 35 ge, wenxia, donald h. drury, steve fortin, feng liu, and desmond tsang. 2010. “value relevance of disclosed related party transactions.” advances in accounting 26 (1): 134–41. https://doi.org/10.1016/j.adiac.2010.02.004. ghani, waqar i., and ashraf junaid. 2005. “corporate governance, business group affiliation, and firm performance: descriptive evidence from pakistan,” no. 05. ghosh, s. (2007) ‘external auditing, managerial monitoring and firm valuation: an empirical analysis for india’, international journal of auditing, 11(1), pp. 1–15. doi: 10.1111/j.1099-1123.2007.00353.x. gopalan, radhakrishnan, and gregory f udell. 2010. “why do firms form new banking relationships ? ∗ why do firms form new banking relationships ?” business, no. 812. gordon, elizabeth a. 2004. “determinants of related party transactions and their impact on firm value determinants of related party transactions and their impact on firm value.” harper, d., kim, j.-o., mueller, c.w., 1980. introduction to factor analysis: what it is and how to do it. contemp. sociol., http://dx.doi.org/10.2307/2066279. hausman, j. (1978). specification tests in econometrics. econometrica 41, 1251-1271. hillman, a. j., & dalziel, t. (2003). boards of directors and firm performance: integrating agency and resource dependence perspectives. the academy of management review, 28(3), 383. doi:10.2307/30040728 ho, d.c.k., duffy, v.g., shih, h.m., 2001. total quality management: an empirical test for mediation effect. international journal of production research 39, 529–548. huang, d., & liu, z. 2010. a study of the relationship between related party transactions and firm value in high technology firms in, 4(9), 1924–1931. huang, derek-teshun, and zhien-chia liu. 1924. “a study of the relationship between related party transactions and firm value in high technology firms in taiwan and china.” african journal of business management 4 (9). http://www.academicjournals.org/ajbm. iasb. international financial reporting standard 124 related party disclosure (2005), london: iasb jensen, m. c., & meckling, w. h. 1976. theory of the firm: managerial behavior, agency costs and ownership structure. journal of financial economics, 3(4), 305-360. doi:10.1016/0304-405x(76)90026-x http://dx.doi.org/10.2307/2066279 hamid ullah, attaullah shah & syed hamid ali shah 36 jian, m., wong, t.j. 2004, earnings management and tunneling through related party transactions: evidence from chinese corporate groups. proceeding of the 13th asia pacific, 698-704. jiang, g., lee, c. m., & yue, h. 2010. tunneling through intercorporate loans: the china experience☆. journal of financial economics, 98(1), 1-20. johnson, s., boone, p., breach, a., & friedman, e., 2000a, corporate governance in the asian financial crisis, journal of financial economics, 58(1-2), 141-186. doi:10.1016/s0304-405x(00)00069-6 johnson, s., la porta, r., lopez-de-silanes, f., & shleifer, a. 2000. tunneling. american economic review, 90(2), 22-27. doi:10.1257/aer.90.2.22 johnstone, k. m., & bedard, j. c. 2004. audit firm portfolio management decisions. journal of accounting research, 42(4), 659-690. doi:10.1111/j.1475679x.2004.00153.x kaiser, h. 1960. "the application of electronic computers to factor analysis." educational and psychological measurement 20: 141-151. karacaer, s., gohar, r., aygün, m., & sayin, c. 2009. effects of personal values on auditor’s ethical decisions: a comparison of pakistani and turkish professional auditors. journal of business ethics, 88(1), 53–64. https://doi.org/10.1007/s10551009-0102-4 kiel, g. c., & nicholson, g. j. 2003. board composition and corporate performance: how the australian experience informs contrasting theories of corporate governance. corporate governance, 11(3), 189-205. doi:10.1111/1467-8683.00318 kiel, g. c., & nicholson, g. j. 2003. board composition and corporate performance: how the australian experience informs contrasting theories of corporate governance. corporate governance, 11(3), 189-205. doi:10.1111/1467-8683.00318 kohlbeck, m., & mayhew, b. w. 2010. valuation of firms that disclose related party transactions. journal of accounting and public policy, 29(2), 115-137. doi:10.1016/j.jaccpubpol.2009.10.006 kohlbeck, m., & mayhew, b. w. 2010. valuation of firms that disclose related party transactions. journal of accounting and public policy, 29(2), 115-137. doi:10.1016/j.jaccpubpol.2009.10.006 kula, v. 2005. the impact of the roles, structure and process of boards on firm performance: evidence from turkey. corporate governance, 13(2), 265-276. doi:10.1111/j.1467-8683.2005.00421.x journal of applied economics and business studies, volume. 3, issue 1 (2019) 15-40 https://doi.org/10.34260/jaebs.312 37 la porta, r., lopez-de-silanes, f., shleifer, a., & vishny, r. w. 2000. agency problems and dividend policies around the world. the journal of finance, 55(1), 1-33. doi:10.1111/0022-1082.00199 liew, c. y., & munir, s. 2011. expropriation-related variables & firm performance : evidence from malaysian family firms expropriation-related variables & firm performance : evidence from malaysian family firms, (february 2016). lin, w.-y., liu, y. a., & keng, i. 2010. related party transactions, firm performance and control mechanisms: evidence from taiwan. international research journal of finance and economics-issue (vol. 35). retrieved from http://www.eurojournals.com/finance.htm liu, qiao and joe lu, 2007, corporate governance and earnings management in the chinese listed companies: a tunneling perspective, journal of corporate finance, vol.13, issue 5, 881-906. liu, m., & magnan, m. 2011. self-dealing regulations, ownership wedge, and corporate valuation: international evidence. corporate governance: an international review, 19(2), 99-115. doi:10.1111/j.1467-8683.2010.00839.x lo, a. w. y., wong, r. m. k. and firth, m. 2010 ‘can corporate governance deter management from manipulating earnings? evidence from related-party sales transactions in china’, journal of corporate finance. elsevier b.v., 16(2), pp. 225– 235. doi: 10.1016/j.jcorpfin.2009.11.002. moscariello, n. 2007. le operazioni tra ‘‘parti correlate’’ nella comunicazione d’azienda. padova: cedam. nekhili, m., & cherif, m. 2011. related parties transactions and firm's market value: the french case. review of accounting and finance, 10(3), 291-315. doi:10.1108/14757701111155806 newman, d. p., patterson, e. r., & smith, j. r. 2005. the role of auditing in investor protection. the accounting review, 80(1), 289-313. doi:10.2308/accr.2005.80.1.289 newman, d. p., patterson, e. r., & smith, j. r. 2005. the role of auditing in investor protection. the accounting review, 80(1), 289-313. doi:10.2308/accr.2005.80.1.289 peng, m. w. 2004. outside directors and firm performance during institutional transitions. strategic management journal, 25(5), 453–471. https://doi.org/10.1002/smj.390 pettit, r. r., & singer, r. f. 1985. small business finance: a research agenda. financial management, 14(3), 47. doi:10.2307/3665059 hamid ullah, attaullah shah & syed hamid ali shah 38 pfeffer, j. and salancik, g. r. 1978, the external control of organizations: a resource dependence perspective, harper & row, new york. pizzo, m. 2013 ‘related party transactions under a contingency perspective’, journal of management & governance. springer us, 17(2), pp. 309–330. doi: 10.1007/s10997011-9178-1. qian, m., pan, h., & yeung, b. y. 2010. expropriation of minority shareholders in politically connected firms. ssrn electronic journal. doi:10.2139/ssrn.1719335 ryngaert, m., & thomas, s. 2012. not all related party transactions (rpts) are the same: ex ante versus ex post rpts. journal of accounting research, 50(3), 845882. doi:10.1111/j.1475-679x.2012.00437.x shen, w., & cannella, a. a. 2003. will succession planning increase shareholder wealth? evidence from investor reactions to relay ceo successions. strategic management journal, 24(2), 191–198. https://doi.org/10.1002/smj.280 shleifer, a., & vishny, r. w. 1997. a survey of corporate governance. the journal of finance, 52(2), 737-783. doi:10.1111/j.1540-6261.1997.tb04820.x siegel, j., & choudhury, p. 2012. a reexamination of tunneling and business groups: new data and new methods. review of financial studies, 25(6), 1763-1798. doi:10.1093/rfs/hhs008 titman, s., & wessels, r. 1988. the determinants of capital structure choice. the journal of finance, 43(1), 1-19. doi:10.1111/j.1540-6261.1988.tb02585.x ullah, h. and shah, a. (2011) ‘a literature survey of value creation through relationship banking’, 5(35), pp. 13517–13522. doi: 10.5897/ajbmx11.029. ullah, h. and shah, a. (2014) ‘the mediating role of multiple banking relationships in managerial ownership and firm value’, 32(7), pp. 1369–1380. doi: 10.5829/idosi.wasj.2014.32.07.543. ullah, h. and shah, a. (2015) ‘related party transactions and corporate governance mechanisms : evidence from firms listed on the karachi’, pp. 663–680. wan, y., & wong, l. 2015. ownership, related party transactions and performance in china. accounting research journal, 28(2), 143–159. https://doi.org/10.1108/arj-082013-0053 williamson, o. e. 1998. "transaction cost economics: how it works; where is it headed." de economist 146(april): 23-58. https://doi.org/10.1108/arj-08-2013-0053 https://doi.org/10.1108/arj-08-2013-0053 journal of applied economics and business studies, volume. 3, issue 1 (2019) 15-40 https://doi.org/10.34260/jaebs.312 39 wooldridge, j. m. 2011. econometric analysis of cross section and panel data. retrieved from https://jrvargas.files.wordpress.com/2011/01/wooldridge_j_2002_econometric_analysis_of_cross_section_and_panel_data.pdf wurgler, j. 2000. financial markets and the allocation of capital. journal of financial economics, 58(1-2), 187-214. doi:10.1016/s0304-405x(00)00070-2 yeh, y., shu, p., & su, y. 2012. related-party transactions and corporate governance: the evidence from the taiwan stock market. pacific-basin finance journal, 20(5), 755776. doi:10.1016/j.pacfin.2012.02.003 zahra, s. a., & pearce, j. a. 1989. boards of directors and corporate financial performance: a review and integrative model. journal of management, 15(2), 291334. doi:10.1177/014920638901500208 hamid ullah, attaullah shah & syed hamid ali shah 40 table a-1 matrix of correlations variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (1) drptrp 1.000 (2) drptsp -0.028 1.000 (3) board size -0.105 0.067 1.000 (4) board independence -0.024 0.018 0.006 1.000 (5) ceo-duality 0.015 0.130 0.083 0.074 1.000 (6) ceo-compensation -0.198 0.094 0.191 0.031 0.096 1.000 (7) associate ownership 0.119 0.083 0.066 0.158 0.090 0.151 1.000 (8) block20 0.054 0.081 0.087 0.157 0.127 0.186 0.792 1.000 (9) block30 0.050 0.063 0.047 0.181 0.089 0.146 0.894 0.832 1.000 (10) managerial ownership 0.011 0.116 0.129 0.082 0.115 0.267 0.364 0.347 0.332 1.000 (11) institutional ownership -0.049 0.103 0.202 0.094 0.094 0.157 0.006 0.127 0.089 0.198 1.000 12) foreign ownership -0.050 0.038 0.010 0.037 0.109 0.089 0.004 0.002 0.008 0.114 0.093 1.000 (13) no. external auditor -0.147 0.123 0.030 0.008 0.083 0.017 0.052 0.053 0.046 0.053 0.044 0.017 1.000 (14) audit fee -0.103 0.077 0.202 0.117 0.009 0.296 0.123 0.104 0.079 0.066 0.063 0.114 0.092 1.000 (15) no. banks -0.153 0.064 0.083 0.039 0.114 0.160 0.051 0.178 0.083 0.072 0.080 0.088 0.074 0.305 1.000 (16) bank loan -0.065 0.203 0.163 0.019 0.074 0.085 0.079 0.117 0.168 0.021 0.048 0.103 0.109 0.046 0.104 1.000 (17) firm size 0.040 0.101 0.090 0.289 0.086 0.048 0.044 0.020 0.035 0.045 0.069 0.070 0.022 0.185 0.000 0.014 1.000 (18) firm growth -0.091 0.132 0.142 0.236 0.140 0.156 0.077 0.067 0.064 0.066 0.085 0.090 0.025 0.235 0.009 0.047 0.827 1.000 (19) leverage -0.039 0.015 0.009 0.007 0.008 0.034 0.023 0.036 0.009 0.002 0.035 0.016 0.065 0.007 0.041 0.087 0.019 0.027 1.000 (20) tobinsq -0.032 0.040 0.148 0.052 0.083 0.080 0.170 0.118 0.129 0.080 0.029 0.081 0.169 0.206 0.043 0.112 0.034 0.122 0.025 1.000 (21) eps -0.000 0.005 0.177 0.069 0.063 0.127 0.002 0.081 0.047 0.076 0.017 0.263 0.103 0.195 0.055 0.113 0.099 0.146 0.021 0.392 1.000 (22) roe -0.002 0.004 0.185 0.063 0.046 0.134 0.039 0.084 0.064 0.088 0.028 0.269 0.099 0.198 0.047 0.113 0.100 0.149 0.018 0.453 0.983 1.000 (23) roa -0.043 0.090 0.228 0.038 0.143 0.188 0.052 0.063 0.043 0.199 0.087 0.172 0.142 0.142 0.003 0.197 0.075 0.122 0.032 0.367 0.615 0.619 1.000 refer to note to table 1 & 2 for the definition of the variables. journal of applied economics and business studies, volume. 5, issue 4 (2021) 1-12 https://doi.org/10.34260/jaebs.541 1 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x investigating the effect of contingent valuation question formats on households’ willingness to pay for pesticides free fruits in khyber pakhtunkhwa jahangir khan1,syed attaullah shah2,khurram nawaz saddozai3, aditya r. khanal4 1department of agricultural & applied economics, the university of agriculture, peshawar, kp, pakistan 2 assistant professor, department of agricultural & applied economics the university of agriculture, peshawar, kp, pakistan 3 assistant professor, department of agricultural & applied economics the university of agriculture, peshawar, kp, pakistan 4 assistant professor of agribusiness, college of agriculture, tennessee state university abstract the purpose of this study is to examine the effect of contingent valuation questions format on households ‘preferences revelation and willingness to pay (wtp) for chemicals free fruits in khyber pakhtunkhwa (kp) province of pakistan. a sample of 1000 randomly selected households, from rural and urban areas of the kp province, were confronted with three different wtp questions for chemicals free fruits-double bounded dichotomous choice question, open-ended question and payment card question. results show that households’ willingness to pay price premium for chemicals free fruits were 14 percent from open-ended question, 16 percent from payment card format and 25 percent from double bounded question. regression analyses identified head’s age, education, monthly income, awareness regarding use of pesticides in fruit production and their impact on human health and natural environment as important determinants of households’ stated wtp for chemicals free fruits. the study concludes that households are willing to pay substantial price premium for chemicals free fruits and recommends growers to start their production on commercial scale. findings of this split-sample study reveals that results of contingent valuation method are sensitive to wtp question format and recommends that proper guidelines must be followed in order to get accurate and reliable results. keywords contingent valuation method, wtp question formats, pesticidesfree fruits, split sample study, pakistan jelclassification q13, d12 jahangir khan, syed attaullah shah, khurram nawaz saddozai,& aditya r. khanal 2 1. introduction contingent valuation method (cvm) is a survey-based technique used for eliciting consumers’ perceived values on goods, services and amenities. cvm has mainly been used in develop countries to determine economic values for environmental quality improvements, and which are not possible to measure through other methods such as the revealed preference method (carson et al., 2005). its wide applications have got the interest of researchers in developing countries. beside these advantages, cvm are more sensitive to the wtp question formats and payment mechanism (carson et al., 2005). in this direction, to provide best guidelines for the application of cvm in developing countries, the research work of dale whittington (2002) has played a significant role. he found that the primary causes for the failure of most cvm research in developing nations are poor survey administration, lack of best survey design and lack of research on the robustness of the results against variations in designing research and methods adopted for survey. he suggested using split-sample experiments to test the robustness of the results against differences in cvm designs. one issue that researchers continuously reported is the selection and choice of suitable wtp question formats. the question formats mostly used by researchers in cvm studies are open ended, single dichotomous choice, payment card and double bounded dichotomous choice questions. research studies showed that information regarding households’ willingness to pay obtained through single bounded dichotomous choice wtp question is limited and also result higher estimated wtp (zhongmin et al., 2006). the single dichotomous choice wtp question also needs huge sample size (carson and hanemann, 2005). the use of an open-ended question format may give high zero wtp responses, and payment card method suffers from initial values selection bias. this study is designed to analyze the performance of different wtp question formats using a split sample case study of households’ wtp for organic/ pesticides free fruits in khyber pakhtunkhwa, pakistan. organic/pesticides free products are fresh or processed food item produced without application/use of pesticides and chemicals (khan and jan, 2018). in this modern age, food consumers are extremely conscious about their health safety and best practices for environment, so they prefer to buy and consume more healthy food such as organic products (denver & christensen, 2015). consumers are more aware of the fact that irrational use of pesticides on fruits and vegetables causes harmful effects on their health; they prefer to buy and use organic products. many studies have shown that fresh fruits journal of applied economics and business studies, volume. 5, issue 4 (2021) 1-12 https://doi.org/10.34260/jaebs.541 3 and vegetables have health advantages and can help protect from diseases including diabetes, stroke, and cancer. (stan et al., 2008). according to slavin and lioyd (2012), fruits and vegetables are universally recognized as healthy and beneficial. 2. materials and methods 2.1 study area pakistan is endowed with plenty of fresh water and fertile land. it has comparative advantage in rice and cotton production and its fruits and vegetables are famous for taste and freshness. however, the use of pesticides and chemicals on fruit and vegetable is common which is dangerous for health as well as wildlife animals and ecosystem. farmers in pakistan have the potential to produce fruits and vegetables organically but facing some hurdles. firstly, the possible reduction in pesticides free fruits production and secondly, the uncertainty about local consumers’ willingness to pay high prices for such produce. that’s, why, this study used the case of pesticides free fruits in khyber pakhtunkhwa province of pakistan and investigated consumers’ wtp for pesticides free version, using three different cv question formats. the delicious fruits such as citrus, dates, apricot, apple, peaches, mango and pesticides free dry fruits of northern areas of khyber pakhtunkhwa province of pakistan is producing. kp is also fortunate to have fertile terrain and ample water resources for agriculture. a total of 1,000 interviewees were selected at random for this research study. for sample selection, two districts namely peshawar and mardan were purposively selected because of having high population and huge demand for fruits. each district was divided into urban and rural union councils (uc) and then two from each stratum were selected randomly. thus in total 4 urban and 4 rural uc’s were selected. from each uc a sample of 125 households were randomly selected. thus in total 1000 households were selected from both districts. 2.2 contingent valuation survey designing three different cv surveys were designed. they were parallel and same in all questions except in the wtp questions. open-ended (oe), dichotomous choice’ (dc), and payment card (pc), are the three primary question formats. the open ended elicitation design collect information regarding household’s wtp for pesticides free fruit such as apple, peach and mango by simply asking how much he or she would be willing to pay for pesticides free version of them if available in local market. jahangir khan, syed attaullah shah, khurram nawaz saddozai,& aditya r. khanal 4 the double bounded dichotomous choice format presents a dollar amount ($a) to the respondents and asked for their willing to pay the presented amount for improved version. the second question is asked depending upon the response to the first question. in the second question, a higher amount is offered to the respondent if his/ her reply is in yes to the first question, or vice versa. payment card (pc) elicitation format is to some extent is a hybrid of the open ended and dichotomous choice formats. in pc format, ranges of willingness to pay values are presented to the respondents and are asked to select or encircle the value which indicates their stated wtp for the improved quality version. the values on payment card are based on open ended responses from a piolet survey or findings from past studies. 2.3 empirical models for data analysis 2.3.1 for open ended cvm for open-ended nature data, to estimate wtp and identify its determinants, a linear regresssion model was used. the mathematical form of open ended model is given below ├ wtp_j=β_0+∑_(i=0)^n▒〖βix_ij+ε〗┤ where, wtpj, represent the jth households wtp for fruits free from pesticide use in pakistani rupees. x_j, is the set of socio-economic characteristics of the jth household. β's are the coefficients and ε is the error term. 2.3.2 payment card cvm to analyze data collected on payment card format, an ordered logit technique was used. a respondent select a value from the given set of values provided on the card. this selected value and next higher value to it are considered as lower and upper bounds, respectively and this represents true wtp (y∗). for such set up, under latent variable framework, the wtp model can be employ as follows yi ∗ = xi ′β + εi where yi ∗ shows an unobserved latent variable or true wtp for household i; β is a vector of coefficients; xi ′explain socio-economic and demographic variables of the respondent that affect wtp; and εiis the error term. journal of applied economics and business studies, volume. 5, issue 4 (2021) 1-12 https://doi.org/10.34260/jaebs.541 5 based on the assumption about the distribution of the true willingness to pay in the selected interval, different models are used to estimate the parameters of the model. this study used maximum likelihood method to estimate in order logit form (see cameron and huppert, 1989 for more detail). 2.3.3 double bounded dichotomous choice cvm in this method, two bids are given to the respondent, where the second bid is contingent on the first bid's response. if the response of the consumer for the first bid is yes, than the second bid bu which is a higher amount then the first bid is asked (bu> b), and if the respondent is not agree with the first offered bid, then a slightly lesser amount than the first offered bid (bd <b) is presented. as a result, there are four alternative outcomes: a yes to the first bid then a yes to the second bid, the probability is shown by (yy); a yes and then a no, indicated by (yn); a no then a yes indicated by (ny) and a no followed by a no indicated by (nn). according to hanemann et al. (1991), the possibility that a respondent give a yes response to both the questions asked is pryy(b, b u) = pr[b ≤ wtp, bu ≤ wtp] = 1 − f(bu) = pr[b ≤ wtp \bu ≤ wtp] pr [bu ≤ wtp] = pr [bu ≤ wtp] = 1 – f(bu) the probability that respondent gives a yes response to the first question asked and regret the second question is: pryn(b, b u) = pr [b ≤ wtp < bu] = f(bu) – f(b) the chance that respondent reject the first bid and accept the second bid with yes prny(b, b d ) = pr [bd ≤ wtp < 𝐵] = f(b) – f(bd) the possibility that the respondent answers in no for both questions is prny (b, b d) = pr [b > 𝑊𝑇𝑃, bd > 𝑊𝑇𝑃] = f(bd) where, wtp: willingness to pay; pryy: the possibility of answering both bids with yes; pryn: the possibility of answering yes to first bid and no to second bid; prny: the possibility of answering a no to the first question and a yes to the second; prnn: the possibility of a no response to both the bids asked; b: shows the price in the first question; bu: the jahangir khan, syed attaullah shah, khurram nawaz saddozai,& aditya r. khanal 6 higher price of the second question; bd: shows the lower price of the second question and f: the cumulative distribution function (cdf). to estimate the likelihood of household responses, the maximum likelihood estimate was used. for the n sample respondents, the bids are bi, bid and biu that are used for the ith respondent, the probability function is: lnl = ∑ { n i=1 yyilnpryy (bi, bi u) + ynilnpryn (bi, bi u) + nyilnprny (bi, bi d) + nnilnprnn (bi, bi d)} where, dummy variables are ‘yy’, ‘yn’, ‘ny’ and ‘nn’. if a respondent answers yes to both questions with a yes-yes (yy), then yy = 1 and the other is zero. 2.4 willingness to pay comparison from cv question format to compare the households’ estimated wtp for pesticides free fruits from different willingness to pay elicitation format an independent sample t-test was used. it is a parametric test used to determine whether the mean values obtained are significantly different from each other. 3. results and discussion 3.1 socio-economic characteristics of the respondents summary statistics for the socioeconomic characteristics of the selected households are given in table 1. these information’s are collected from their heads. the table shows that majority of the households were headed by male and this is a prominent feature of pakistani society. the average age of the household’s head was around 41.2 years, and his education level was on average 9.2 years of schooling. household size was around 7 members and their average monthly income was rs.52935. results related to households’ response to questions on pesticides use on fruits and their negative impacts on human health and environment are also given in table 1. results of the research study highlighted more than half (60 %) of the respondents were aware of the fact that farmers are using excess pesticides spray during on fruit production. furthermore, 80 percent of the respondents perceived pesticides application on fruits and vegetables production is unsafe for human health and natural environment. these numbers indicate that respondents in the study area had not much information regarding the serious application of pesticides on fruits; however,regarding the effects of pesticides on human health and the environment, they have been correctly assessed. journal of applied economics and business studies, volume. 5, issue 4 (2021) 1-12 https://doi.org/10.34260/jaebs.541 7 table 1: household’s socio-economic characteristics and awareness for pesticides use variables sample-1* (600 households) sample-2** (200 households) sample-3*** (200 households) all (1000) head’s age (years) 42.0 40.2 40.0 41.2 head’s gender (%) 90 90 90 90 household size (no.) 7.2 8.3 6.2 7.2 head’s education (years) 8.2 10.2 11.4 9.2 households monthly income (pkr****) 53465.0 48750.0 55530.0 52935.0 heads aware of pesticides used on fruits (%) 60 70 60 60 heads concerned for pesticides effect on human health (%) 90 80 70 80 heads concerned for pesticides effect on environment (%) 80 80 80 80 * sample-1 of 600 households was used for dc, cv survey ** sample-2 of 200 households was used for pc, cv survey *** sample-3 of 200 households was used for oe, cv survey ****pakistani rupees( 1pkr=0.0063 us dollars) 3.2 estimated wtp models the estimated results of willingness of households to pay for pesticide-free fruit based on different contingent valuation survey models are given in table 2. the table shows that the estimated coefficients for the age of the respondents, level of education, and awareness regarding use of pesticides on fruits and their effects on the health of human being and also the natural environment are consistently positive and statistically significant for all three models, suggesting that they are directly related to the willingness of households to pay for fruits free from pesticide used. household monthly income is another factor having positive and statistically significant coefficient means that households’ with more disposable income gives more importance to fruit free of pesticides and like to pay more price premium compared to those having lower income. an increase in income raises purchasing power of consumers that divert their purchasing jahangir khan, syed attaullah shah, khurram nawaz saddozai,& aditya r. khanal 8 behavior. household’s size is the single factor having an inverse relationship with wtp for pesticides free fruits pay and according to our prior expectation. a household with more people does not pay attention to high quality food because it is difficult to feed more mouths. these findings are consistent with (haghjou et al., 2013) conducted a research study on factors influencing consumer willingness to pay for organic food in tabriz iran. by adopting ordered logistic regression model, results of the study showed that factors such as respondent's income, family dimension; in addition to consumer awareness, environmental and health concerns were important determinants of wtp for organic foods. in another research study on consumer perception regarding organic food in bangkok, roitner-schobesberger et al., (2008) found that more than 1/3 of respondents bought organic fruit or vegetables because consumer expected that food produced organically would be good both for human health and natural environment. based on their research findings, respondents of older age, higher income and higher education was the main purchaser. in addition, according to the findings of our study respondents with a positive attitude toward environmental change were willing to pay a greater price for better quality fruits. consumers who care more about their health are more willing to pay a premium price for pesticide-free fruits, according to the findings. these results are in line with khan et al (2018). journal of applied economics and business studies, volume. 5, issue 4 (2021) 1-12 https://doi.org/10.34260/jaebs.541 9 table 2: determinants effecting respondents willingness to pay for pesticidefree fruit variables open ended model payment card model double bounded model coefficient s t values coefficien ts t values coefficien ts t values constant -6.174109 1.42 -8.614601 -2.37 age 0.2348861 5.02 0.0698807 4.74 0.294755 4.62 gender -2.828394 0.91 0.2469171 -0.29 -0.133885 0.03 education 0.243413 3.15 0.0821316 3.68 0.4521878 2.87 households size -0.3936806 -2.24 0.0975395 -1.94 0.6690743 -2.68 urban/rural 0.587075 0.56 0.2783974 0.94 0.2295739 0.17 income 0.0001237 7.90 0.0000277 5.51 0.0001208 4.26 health concern 5.567941 3.80 2.600153 4.86 6.165207 3.28 environmen t concern 0.2999708 0.15 0.3993532 0.69 6.739389 3.28 awareness 3.567843 3.56 0.9595126 3.38 8.194937 4.84 no of observation = 200 f (9) = 28.59 pvalue=0.000 no of observation = 200 wald chi2 (9) =158.49 probability >chi2=0.000 no of respondents = 600 wald chi2 (9) =272.30 probability >chi2= 0.000 source: survey data calculation, 2016-17 3.3 comparison of elicitation questions formats table 3 illustrates the estimated mean wtp for the double dichotomous, the payment card, and the open survey formats. household willingness to pay is estimated using averages for the independent variables. result of the predictable models confirms mean values of pakistani rupees 25.76, 16.32 and 14.37 for the double-dichotomous, payment cards and linear regression models respectively, for pesticide-free fruit in the study area. these results are in accord with other research studies such as ngo et al., 2013 and rodriguez et al., 2008. the independent sample t-test results for comparison of estimated wtp for pesticides free fruit for different cvm formats are also given in jahangir khan, syed attaullah shah, khurram nawaz saddozai,& aditya r. khanal 10 the same table 3. they are found statistically different from each other and it proves that a wtp question format has a significant effect on estimated wtp. table 3: average willingness to pay based on the survey format of the cv questions cv questions format mean wtp open ended (14.37) mean wtp payment card (16.32) mean wtp payment card (16.32) m.d (1.94) t-value (2.50) p-value (0.006) ____ mean wtp double bounded (25.76) m.d (11.43) t-value (14.83) p-value (0.000) m.d (9.49) t-value (12.22) p-value (0.000) source: survey data calculation, 2016-17,1m.d= mean difference: 1us$=105 pkr in 2017 3.4 respondents’ response for zero wtp table 4 shows zero response rates to the three utilized wtp elicitation formats. the zero response rates to payment card format found minimum (13%) as compared to double bounded (30%) and open ended (40%) formats. from these results, it is suggested that the payment card format is valid, easy and contributes in an efficient in data collection. table 4: zero wtp responses by cv question format response to wtp question double bounded payment card open ended no-no or zero wtp responses 30% 13% 40% source: survey data calculation, 2016-17 4. conclusion and recommendations regarding the case study of pesticides use on local fruits produce, finding of the study reveals that around two third of the households were aware of pesticides uses on local fruits and more than three fourth of the households were worried about the negative effects of pesticide residues on human health and the natural ecosystem. consumers’ were willing to pay on average 18 % higher premium for fruits produced without the use of pesticides. based on the estimated wtp, the study recommends fruit growers journal of applied economics and business studies, volume. 5, issue 4 (2021) 1-12 https://doi.org/10.34260/jaebs.541 11 produce pesticide-free products as households in the local market are willing to pay a substantial amount for the pesticide-free version of fruit. on technical side, the estimated mean wtp for the three elicitation formats such as open ended, payment card and double bounded were 14.37, 16.32 and 25.76 percent, respectively. one survey format is not clearly better than the other when it comes to priorities. which elicitation format to use depends upon the circumstances of the study. for instance, if the survey has a small sample size, then payment card elicitation format best suits. however, with large sample size, it is preferred to adopt double bounded elicitation format. findings of this split samples study recommends that results of contingent valuation method are sensitive to wtp question format and the adoption of suitable wtp question’s format depends on the sample size of the respondents and their awareness for such studies. jahangir khan, syed attaullah shah, khurram nawaz saddozai,& aditya r. khanal 12 references carson, r. t., & hanemann, w. m. (2005). contingent valuation. handbook of environmental economics, 2, 821-936. denver, s., & christensen, t. (2015). organic food and health concerns: a dietary approach using observed data. njas – wageningen journal of life sciences. http://dx.doi. org/10.1016/j.njas.2015.05.001. haghjou, m., b. hayati, e. pishbahar, r.mohammadrezaei,. & g.dashti.(2013). factors affecting consumers’ potential willingness to pay for organic food products in iran: case study of tabriz. journal of agricultural science and technology. 15(2), 191202. hanemann, m., loomis, j., & kanninen, b. (1991). statistical efficiency of doublebounded dichotomous choice contingent valuation. american journal of agricultural economics, 73(4), 1255-1263. khan, j., khanal, a. r., lim, k. h., jan, a. u., & shah, s. a. (2018). willingness to pay for pesticide free fruits: evidence from pakistan. journal of international food & agribusiness marketing, 30(4), 392-408. khan , j. & jan, a. u. (2018). factors influencing households' willingness to pay for pesticides-free fresh apple in district peshawar, khyber pakhtunkhwa (kp) pakistan. sarhad journal of agriculture, 34(3),599-605 ngo minh, h., m, masahiro andf. susumu .2013. willingness to pay for organic vegetables in vietnam: an empirical analysis in hanoi capital. j. fac. agr., kyushu univ. 58(2): 449-458. roitner-schobesberger, b., darnhofer, i., somsook, s., & vogl, c. r. (2008). consumer perceptions of organic foods in bangkok, thailand. food policy, 33(2), 112-121. rodríguez.e., v. lacaze & b. lupín. (2008).contingent valuation of households’ willingness to pay for pesticide free food in argentina.12th congress of the european association of agricultural economists. doi: 10.22004/ag.econ.43947 slavin, j. l., & lloyd, b. (2012). health benefits of fruits and vegetables. advances in nutrition, 3(4), 506-516. stan, s. d., kar, s., stoner, g. d., & singh, s. v. (2008). bioactive food components and cancer risk reduction. journal of cellular biochemistry, 104(1), 339-356. whittington, w. (2002). improving the performance of contingent valuation studies in developing countries. environmental and resource economics.22, 323-367 zhongmin, x., loomis, j., zhiqiang, z., & hamamura, k. (2006). evaluating the performance of different willingness to pay question formats for valuing environmental restoration in rural china. environment and development economics, 11(5), 585-601 journal of applied economics and business studies, volume. 6, issue 1 (2022) 85-108 https://doi.org/10.34260/jaebs.614 85 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x crowding out effect of tobacco usage on household education and health expenditures an econometric analysis kashif haneef1, syed faizan iftikhar2, & ambreen fatima3 1 research scholar, applied economics research centre, university of karachi, karachi 2 research economist / assistant professor, applied economics research centre, university of karachi, karachi 3 senior research economist / associate professor, applied economics research centre, university of karachi, karachi abstract tobacco consumption is growing day by day everywhere in the world in countless forms such as supari, tobacco raw, chew, pan, cigarettes, bidi and naswar etc. tobacco consumption may avert the limited family resources far away from advantageous uses between low-slung income households. the relatively underprivileged households frequently allocate an ample share of the domestic budget on tobacco expenditures. these uneconomical expenditures on tobacco could be a diversion of household monetary resources from necessary things, such as foodstuff, healthcare and education. explicitly this paper discovers the crowding-out impact of tobacco expenditures on a household’s education and health payouts. the facts for this study has been engaged from hies (household integrated economic survey) conducted regularly by pbs (pakistan bureau of statistics) that offers evidence at household level and covered the whole nation. the estimation has been carried out by employing the engel curves, the quadratic terms using quaids (quadratic almost ideal demand system) were found to be empirically important in describing household budget behavior. by employing data on outlays, study evaluated a user demand structure that comprises numerous main spending classifications (tobacco, education and healthcare). ols (ordinary least square) was used to estimate the effect of tobacco outlay and other covariates on share of education and health in the total budget of a household. the results found out were just alike john (2006) that a rise in the tobacco expenditure brings a reduction in the budget quota dedicated to health and education. the keywords tobacco, household, health, education, engel curve, quaids jel classification d10, d15 kashif haneef, syed faizan iftikhar, & ambreen fatima, 86 quantitative results aim to help the policy makers to formulate and apply such policies to restrict the consumption of tobacco products and diverting those expenses for the formation of human capital by increased spending on health and education. 1. introduction tobacco consumption is a lavish custom. the expenditures on tobacco amongst lowslung income clusters can avert limited domestic means away from advantageous consumptions. poor families often allocate a substantial part of the family budget on tobacco. this will double if the indirect cost of health expenditures included. the healthcare costs of smoking include direct and indirect costs. direct costs–payments include health care facilities, medicines, transportation facilities, and extra food costs (diet, food in hospital) while indirect costs i.e. productive time lost when sick, time of caregivers. the burden of healthcare costs of smoking will bear by the government (national health insurance, care for the poor), individuals (out of pocket) and private sector (employers, insurance companies). most of the household has a fixed budget, while outlay on tobacco and related goods could possibly turn away household cost-effective resources from necessary goods and services, like foodstuff, healthcare and education that stay certainly important for human progress. hypothetically, the dimension and practice of crowding-out effect of consuming tobacco over foodstuff, healthcare and education may vary among households. the influence is expected to be minor on food related goods compared with education and healthcare assumed a fewer elastic nature of foodstuff consumption. unlike its influence upon household outlay on foodstuff and education, although, the relationship among tobacco consumption and healthcare outlays is blurry. the opportunity cost of consuming tobacco and related goods are regularly assessed on state level considerations for health overheads and loss of efficiency due to diseases caused by tobacco, another substantial but less evident micro-level tool by which the consumption of tobacco and related good could possibly have adverse significances for the fiscal comfort of tobacco users and their direct family members. this adverse influence has been pronounced as the “crowding-out effect” of tobacco related goods usage on household consumption. the consumption of tobacco and related goods is very common among individuals from low economic clusters living with a below average standard of living, low or no education at all in urban and rural areas respectively. tobacco consumption is growing day by day everywhere in the world in countless forms such as supari, tobacco raw, chew, pan, cigarettes, bidi and naswar etc. according to a report who (2013), pakistan ranks highest journal of applied economics and business studies, volume. 6, issue 1 (2022) 85-108 https://doi.org/10.34260/jaebs.614 87 among the south east asian region in consumption of tobacco and related goods. another global report (2014) found pakistan among few countries where smokeless (chew and raw tobacco) is extensively used. as per ctc (coalition for tobacco control, pakistan) there are more than 22 million smokers and 52% to 55% of the families have at least one tobacco user in any form. there are some clear indications that low standard of living (sol) or low socioeconomic position (sep) that is linked with higher probability of consume smoke and smokeless tobacco. not merely individual status level indicators (for instance qualification, earnings, and occupation) correlated with tobacco consumption, but regional level indicators of interruption are exposed to be substantial interpreters of using tobacco, for instance the people residing in deprived zones are probable smokers. the relationship between sep/sol and tobacco consumption has been glowingly recognized, but the deviations in sep/sol regarding household tobacco outlay has received inadequate response. it should be an apprehension if a household with lesser financial resources report more tobacco outlay or devote a greater fraction of his finance on tobacco and related goods. there are numerous studies observed the facts that the association of tobacco outlay is directly correlated with healthcare disbursements; the spending on healthcare is higher among families with tobacco user, as a consequence of uninterrupted health-related overheads due to tobacco related diseases. conversely, the fiscal influence of consuming tobacco and related goods on healthcare outlay could also slog in reverse path as tobacco consumption might lessen one-use income for healthcare. therefore, the route and degree of the consequential relationship between healthcare expenditures and tobacco consumption depends upon the relative degree of these two offsetting properties. 1.1 tobacco consumption – pakistan perspective in pakistan, tobacco consumption has materialized as a severe health consequence because of its standing among the first four countries of the world with a swift rise in tobacco market. a probable figure of about 40% of males and 9% of females are user of tobacco and related goods with an upsetting increasing growth rate. nearly about 100,000 deaths per year endorsed by tobacco consumption in pakistan. most of these deaths are amongst men, the main earners of households. this damage could increase the probability of tumbling beneath the poverty line for the families who have lost their bread-winner. hence, tobacco outlay deteriorates poverty in short-run and long-run respectively. in short-run, the outcome could be a possible inverse effect on disposable income while the chances of morbidity and kashif haneef, syed faizan iftikhar, & ambreen fatima, 88 mortality increases the healthcare outlay for household with reduced level of savings. this fact should be realizing by the habituated government of pakistan that the public spending of rs.100 billion on diseases caused by tobacco and related goods, is greater than the revenue of rs.75 billion generates from the tobacco industry per year. (faiza ilyas, published in dawn, may 27th, 2015) according to a report on “the economic cost of tobacco-induced diseases in pakistan” by nayyab et al (2021), pakistan institute of development economics (pide), the researchers find that the total costs of all smoking-attributable diseases and deaths in pakistan in 2019 amount to 615.07 billion (us$ 3.85 billion), with indirect costs making up around 70% of the total. the direct costs account for a significant portion of pakistan's total health expenditures: 8.3%. the total costs of cancer, cardiovascular diseases, and respiratory diseases are rs 437.76 billion (us$ 2.74 billion), or 71% of the total smoking-induced costs. 1.2 tobacco at present tobacco and smoking mixtures are more synchronized today. the companies engaged in tobacco production and sales have lost numerous litigations and are now enforced to visibly tag their merchandises as having an injurious influence on the healthiness of a person. moreover, tobacco promotion is rigorously restricted and controlled. the estimates show that there are around 1 billion tobacco users globally today and tobacco companies earning billions of dollars’ income every year, abolishing the healthiness of others. the loss triggered by this habit and its peddler’s amounts in trillions of dollars of healthiness overheads and environmental injuries and additional struggle should be made to inform society, particularly teens and adults, about the hazards of tobacco consumption. journal of applied economics and business studies, volume. 6, issue 1 (2022) 85-108 https://doi.org/10.34260/jaebs.614 89 figure 1: proportional distributions of various forms of tobacco consumed among men and women in nine south and south-east asian countries source: sreeramareddy et al (2014) 1.3 objectives this study’s prime objective is to discover the association among tobacco expenditures and its crowding out effect on household education and health expenditures in pakistan. this study is therefore motivated to find out as how the expenditures on tobacco crowd-out other purchases. this study investigates how much income a households allocate on tobacco. this is estimated by explaining the deviations in budget segments, assigned for several things of interest, due to variations in spending allotted to tobacco. the ultimate focus of this study is mainly on tobacco spending by households and its crowding-out effect on health and education expenditures. the explicit purposes of the study are: ➢ to investigate the tobacco usage pattern by province, region, occupational group, educational group and income group. ➢ to relate the health and education expenditures with tobacco overheads and explain it’s crowding out effect on education and health by using the conditional demand function. kashif haneef, syed faizan iftikhar, & ambreen fatima, 90 1.4 contributions tobacco use in low income household has an inverse relationship with household outlay on health and education, proposing potentially adverse impact on investment in social capital improvement. the outlay on tobacco smoked, non-smoke or chewable particularly cigarette portrayed health related issues. by now, no investigation was performed in pakistan to find out the crowding out impact of tobacco usage on health and education outlays of a household. the consumption of tobacco is not limited to the poor families as there are several households with a sound social and economic status consumed tobacco more than the underprivileged families. 1.5 research hypothesis the following null hypotheses have been scrutinized in this study: ➢ consumption of tobacco is independent of its price. ➢ earning intensities of household cause no variance in tobacco intake. ➢ education does not impact the consumption of tobacco. ➢ the region either rural or urban has no association with tobacco consumption. ➢ the decision of using tobacco has been influenced by several socio-economic and demographic variables. 2. literature review the origin of study about consumer’s budget allocation was started from northern europe in early 1840’s but the most remarkable work in this arena was done by stigler (1954) followed by engel (1857). engel acclaimed a rule that became the base of forthcoming exploration. as stated in the introduction thoroughly that the overheads interrelated with household tobacco intake are frequently hypothesized by the cost burden on economy, endorsed to tobacco via death, amplified healthcare outlay and mislaid efficiency in the household volinia et.al (2006), chaloupka et.al (2000), menke et.al (2003), and individuals linked to the substitution of their possessions due to tobacco usage. the focus of our paper is mainly on the crowding out influence of tobacco consumption on household health and education outlay. efroymson et.al (2001) highlighted the likely crowding out impact of tobacco outlay. he involved data on consumption of tobacco and integrated household expenditures from the bbs (bangladesh bureau of statistics). the findings from the research shown that male tobacco users disbursed more than two times on tobacco as on apparel, accommodation, education and health combined. journal of applied economics and business studies, volume. 6, issue 1 (2022) 85-108 https://doi.org/10.34260/jaebs.614 91 vermeulen (2003) targeted his research on belgian household budget survey of 19871988 which embraces the expenditures of 3235 household made during the whole year. he used the banks et.al (1997) engel curve technique of (quaids) quadratic almost ideal demand system. he incorporated household features, conditioning outlay on tobacco and probable heterogeneous preference of tobacco users and non-users. busch et.al (2004) empirically analyzed that how tobacco outlay crowd-out depletion of other commodities by approximating the configurations of changeover and reciprocity among tobacco products and other classes of household expenditures. the outcomes from demand system showed that price of cigarettes and demand for food is directly proportional while in few mockups negatively associated with demand for clothing and accommodation procured. expressive figures specified that tobacco users devote less money on accommodation. busch et.al (2004) in usa, evaluate a consumer demand system that is inclusive of numerous main spending sets (cigarettes, nutrition, alcohol, accommodation, health, transportation, apparel) and regulation for socio-economic variables along with other bases of evident diversification. expressive figures show that smokers spend less on accommodation as compared to non-smokers. wang et.al (2006) examined the influence of tobacco outlay on domestic spending patterns in rural china. the dominance of tobacco consumption is high among men and earnings are low in this country. the author evaluated the association among tobacco outlay and disbursements and found that the disbursements on tobacco can hurt other household members by decreasing overheads on elementary prerequisites like foodstuffs, durable commodities consumption and utilities. therefore, tobacco can have significant influence on intra-family distributions. john (2008) employing the data from india’s national sample survey. he employed instrumental variable techniques to explain the probable endogenous nature of tobacco consumption in a structured demand system. the outcome by john was not different as anticipated that tobacco outlay crowded out foodstuff, entertainment, healthcare, apparel, fuels, and entertainment expenditures. pu et.al (2008) cast-off statistics from survey of family income & expenditure of taiwan, 2004 to display the effects of crowding-out of tobacco inclusive of alcohol in a republic that had truncated spending could be quiet severe. he further found that tobacco crowded out apparel, health upkeep and conveyance among others by applying john’s (2008) technique using spending data from taiwan. kashif haneef, syed faizan iftikhar, & ambreen fatima, 92 an auxiliary to the works was provided through using a diverse mechanism for tobacco spending by koch and tshiswaka-kashalala (2008). the tool chosen was a combined smoking dominance measure established on predominance evaluations for south africa, computed in van walbeek (2002). the outcomes revealed that tobacco crowded out disbursements on education, energy, healthcare, apparel, and conveyance aimed at the complete sample of smoking households. block and webb (2009), requiring suitable influential variables, they employed a secondary method to recognize the underlying effect of tobacco spending on household’s spending choices. the elementary idea underlying is that if a set of mutual covariates diminishes the share to nutrition and lessens teenager tallness however at the same time rises the share to tobacco, then this is an expression of crowding out. john et.al (2011) using the national sample survey (nss) comprises comprehensive household spending records in classifications like nutrition, apparel, education, and tobacco goods, estimate (saf) i.e. smoking-attributable fraction. then he multiply (saf) by total health spending in households using tobacco to obtain tobacco-attributable health overheads. he found that tobacco and health expenditure (caused by tobacco) in tobacco consuming households rises the poverty in india. the well-being of the family can be further increase by not spending on tobacco and investing in to education instead. chelwa, g., & van walbeek (2014), their exploration displays that tobacco outlay inversely effects household spending on nutrition, education, water, and apparel. they employed spending statistics from zambia (a sub-saharan african country with low income and most of the households are underprivileged). they estimate engel curves by means of the (quaids) quadratic almost ideal demand system developed by banks et.al (1997). yao et.al (2014) outcomes indicate that certain group of smokers in china admit that the amount spend on cigarettes could be better expended on household essentials. the influence of smoking on crowding out or depriving of other household expenditures in china is predominantly significant as china stands the prevalent user of tobacco in the world. nasir et.al (2015) concludes the relationship among socioeconomic dynamics with smoking and its comparative influence on the male gender of pakistan. the data was taken from pakistan demographic and health survey data set 2010-13. the socioeconomic factors that found statistically significant are wealth index, residence by province, education, age, access on media particularly radio and newspaper. the numerical conclusions will provide guidance in control and management of smoking. san s. & chaloupka f. j. (2016), in turkey has been performed by using turkish journal of applied economics and business studies, volume. 6, issue 1 (2022) 85-108 https://doi.org/10.34260/jaebs.614 93 household budget surveys (2007 and 2011), projected the (qcec) quadratic conditional engel curve to conclude domestic expenditure configurations. the study revealed a crowding out effect for both years studied, whereby expenditures on smoking effects in reduced household spending on nutrition, accommodation, durable/non-durable goods and education. the methodology in this study is thoroughly brought into line with the utmost latest group of observed studies on the crowding out effect of tobacco consumption. the study’s foremost role is to use the usual influential variables recycled in the literature while making less rigid assumptions about the behavior of these influential variables. 3. methodology and data the paper estimates a system of consumer demand that integrates a series of considerable outlay classifications (health, education and tobacco) and regulators for demographics and socio-economic variables as well as further bases of evident diversity. 3.1 data source, limitations and sample size the data used for analysis purpose is taken from pakistan social living standards measurements (pslm) survey (2015-2016). there are certain limitations of the data such as: ➢ the tobacco outlay at distinct level is not provided, therefore, individuals’ features cannot be incorporated in the analysis. ➢ due to tobacco intake the specific illnesses cannot be estimated. ➢ the data does not furnish explicit advice regarding the prices of tobacco products the statistics provides information on household characteristics and utilization of many foodstuff and non-food stuffs with tobacco products. this dataset relates to household integrated economic survey (hies) which is conducted in 2015-16. this hies data covers 24,238 households. it provides substantial data on consumption patterns, savings, liabilities, and household income at national and provincial level with urban/rural itemization. 3.2 empirical methodology the main objective of this study is to calculate the budget share (net of tobacco expenditures) of a household for various socioeconomic classes supplemented with the demographic variables. to realize this concern, the paper needs a precise econometric model accompanied with an appropriate statistical technique. while conducting the study it has been observed that this can be achieved in a number of ways but every technique that could be chosen has both its pros and cons. therefore, the choice of finest technique among many obtainable is of utmost importance to attain a firm and vital conclusion. the paper estimates kashif haneef, syed faizan iftikhar, & ambreen fatima, 94 a system of consumer demand that integrates a series of considerable outlay classifications (health, education and tobacco) and regulators for demographics and socio-economic variables as well as further bases of evident diversity. 3.3 econometric modeling and estimation technique inspiring after the previous efforts, the study incorporates the approach of bhalotra & attfield (1998) and john rm (2008), hence, left with the lean alternatives after eliminating linear models. the analysis carries out using the stipulation of aids. as of yet, this approach is not frequently applied for investigation of the household data records, aside from iqbal & anwar (2014) and haider & zaidi (2017). with a view to use this method, subsequent variables are mandatory; income and shares of several household consumption goods in total household budget. the figures on income are not very much authentic in hies, as masses are reluctant to disclose their true earnings, hence, to deal with this dilemma houthakker (1970), suggested to employ total spending as an alternate of persistent income. the use of total spending as a substitute of persistent income may usually carry off the issue of economies of scale (savings). the total spending of households can be slice into these two i.e. income and substitution effect. the income effect pertains to a rise in household size at certain level of income which leads to a fall in per capita income of household and left them even poorer, while the substitution effect holds an expansion in bare necessities due to enlarged household size. to undertake this issue, the variable per capita expenditure is used, that is obtained by dividing total expenditures with household size. since hies data records does not gather data for prices of goods consumed. however, the spending data along with consumed quantities of the particular goods are obtainable from the hies data records, that can be employed to locate a closest alternate of commodities price. the study engages the almost ideal demand system to observe the household demand and substitution patterns between tobacco items and other spending categories. deaton and muellbauer (1980) projected a subjective first-order approximation with (aids) to any demand system which could be persistent with the assumption of inadequacy (i.e. entities compelled to make an alternative decision) and gratify the theorem of individual choice. the study using facts estimate a consumer demand scheme that comprises of several core spending classifications (medical care, education and tobacco). the budget shares of various commodities (medical care, education and tobacco) disclose the real total expenditures using this scheme of consumer demand equations. because the (pslm) survey contains information about demographics, therefore the study can be capable to act as regulator for socio-economic variables and other causes of observable assortment. the study evaluates this scheme for entire households inclusive of low-income and tobacco users. to estimate journal of applied economics and business studies, volume. 6, issue 1 (2022) 85-108 https://doi.org/10.34260/jaebs.614 95 the hypothesis of crowding out due to tobacco outlay, engel curve technique is employed as applied by bazlul h khondker; mandeep k virk-baker; muhammad jami husain (2016). the influence of covariates on tobacco outlay was estimated using ols regression by m siapush (2003). in this study the estimation is carried out using quaids to observe non-linear seemingly unrelated regression. quaids is the propagation of la-aids formulated by deaton and muellbauer (1980a, b). while the non-linear propagation of la-aids was projected by banks et.al (1997). the chief benefits from these models are the simple estimation of the micro-economic theory (neo-classical), while the macro-economic evaluation has a tendency to be discriminatory as they do not consider society’s behaviors and heterogeneousness. quaids additionally amplifies the aids of non-linear (quadratic) label of log expenditure and log income, therefore, it forms a non-linear “camel hump” shaped engel curves. the elasticity that arises due to the engel curve was needed to observe the quadratic relationships among different socio-economic variables as preceding model was restricted to linear relations while observing income-consumption association. the evaluation could be biased specifications due to lack of this prerequisite for preceding studies. application of the household theory call for a particular model. the econometric studies of demand generally comprise of single equation and a system of equations respectively. for a household or consumer buying “n” commodities, the function of demand can be comprehensively shown as: qi = qi (p1, p2, … pj, … pn, i), i = 1,2, …, n. (eq.3.1) where; quantity demanded denoted by “qi”, “p” represents price, “i” the subscript represents the merchandises; and “i” is income. there could be two restrictions necessary for utility maximization, although the budget along with utility function of an individual is understood. it can also be executed numerically, by way of lagrangian, where the system of equations determined by first derivatives, and resolved by rendering the utility function to the constraint demonstrated by the budget: max u = (x1 , x2) ; s.t. m = p1x1 + p2x2 (eq.3.2) consumers are characterized by eagerness and quantity boosts the growth of utility functions, while the only constraint to consumption is a fixed budget. the second part of the maximization problem is provided but the budget constraint. the only thing need to be adjusted is the utility that derive from the consumption within budget. suppose and individual have to select two goods, 1 and 2, then the constraint will be written as follows: m ≥ p1q1 + p2q2 (eq.3.3) kashif haneef, syed faizan iftikhar, & ambreen fatima, 96 pursuing pollak (1969), assumed a predetermined consumption of one good by a household’s. in this study, suppose a household pre-allocated a certain amount on tobacco consumption. in practical terms, it implies that the household seeking utility maximization now has to maximize utility, subject to the spending in surplus of the pre-allocated tobacco expenditure. if the nth commodity is tobacco, suppose that first n − 1 commodity is presented in the market on the prices {p1, ..., pn−1} above which the household has no power and the total spending on these commodities are denoted by m [(m = y − ptqt), where pt is the price of tobacco and qt the quantity consumed]. now the utility maximization dilemma for the household will turn out to be: u = (x1, x2… xn) ; s.t. ∑ 𝑃𝑖𝑥𝑖 = 𝑀 𝑛−1 𝑖=1 (eq.3.4) and the supplementary restriction xn = x̅n (eq.3.5) where household allotment of tobacco denoted by x̅n. the resolution for this maximization issue, resolving only for n − 1 commodities, provides conditional demand functions which can be transcribed as: xi = g i,n (p1, ….., pn-1, m, xn; a) (eq.3.6) the conditional demand function here is represented by gi,n for the ith commodity, i.e. conditional on the use of nth commodity (tobacco). therefore, by integrating household features (f), customized spending on tobacco (ptt), and probable diversified choices of tobacco consumers and tobacco non-users along with an explicit variable “d” with an objective of testing for consumer separability. the study estimated the subsequent engel curves (conditional) for two extensive commodities of concern i.e. education and health. ωi = (α1i + α2id + α3iptqt + δi 0∑f) + (β1i + β2id) ln m + (γ1i + γ2id) (ln m) 2 (eq.3.7) the above equation 3.7 is structured in a form of budget share. for a set of consumers, the extension of demand functions for individual consumer requires the incorporation of demographics, aside from income and prices in leading empirical requisitions. where; ωi = 𝑝𝑖𝑞𝑖 𝑥 represents the good “i” budget share in the leftover budget surplus of spending on tobacco, “ptqt” represents tobacco spending, while “m” denotes the total spending lacking the outlay on tobacco. “∑f” denotes the demographics and socio-economic attributes of the households that comprises household size, gender of the head, age of the journal of applied economics and business studies, volume. 6, issue 1 (2022) 85-108 https://doi.org/10.34260/jaebs.614 97 head, education of the head, number of children, number of earners, standard of living index (ref: no access to any facility e.g. occupation, electricity, gas, drinking water, drainage), durable assets index, and log of food expenditures. 4. results and discussion this chapter discusses the empirical results based on the models that were presented in chapter 4. the results of the quaids model show a negative relationship between household tobacco expense and share of education and health in the total budget of a household that can be verified from table.4.1, 4.2 and table.4.3 respectively. the null hypothesis in this study is that, the factors related along with the dichotomous variable “d” in equation.4.7 are mutually not significant i.e. ho: α2i = β2i = γ2i = 0. the joint significance of the separability test is mentioned at the end of each table. it became apparent that consumer separability is rejected in pakistan, both rural and urban region respectively. it shows the diverse preferences of tobacco users and non-users as both act differently. the findings show that tobacco expense has a significant impact in crowding out other expenditures mainly education and health. this negative relationship provides us a very useful and visible picture that if an increase occurs in the household tobacco expenditure, falls in the budget allocated to education and health will be administer. the results are quite similar with the descriptive analysis carried out in the earlier chapter of the paper. thesis do not try to categorize the commodities in to necessities and luxuries here except education and health, but based on the results, however, when we applied quadratic parameter with the share of education the results confirmed the existence of an engel curve formulation. similarly, the application of quaids model provides us a robust description of the household behavior representing crowding out. 4.1 assessing crowding out effect on education the results on the crowding out effects of tobacco on share of education in a household budget presented in table 4.1. according to the standard demographic separability hypothesis, budget share on education is not detachable from the intake of tobacco. the expenditures on education has shown both income and substitution effect in table.4.1. as our prime concern is to find out the effect (crowding out) of tobacco outlay on household budget along with those socio demographic factors that are inclined by the conditional demand. in order to achieve this concern, we have incorporated household tobacco expenses, logarithm of tobacco expenses (net of tobacco), region, household size, age and education of household head, durable assets, logarithm of food expenditures, the logarithm of total kashif haneef, syed faizan iftikhar, & ambreen fatima, 98 household expenditure squared (to consider relationships that are not linear between total expenditure and portion of budget to other things by engel curve). the selection of these variables is based on economic theory provided by studies like paraje (2017), which shows a negative (positive) coefficient associated with a lower (higher) budget portion of a household that consumes tobacco. these variables are broadly used in the articles alike as a regulator (control variable) for socioeconomic differences. the results present the outcomes on share of education (dependent variable) followed by a series of relevant independent variables effecting the share of education in the total budget of a household. the nature and results of all these independent variables is also discussed in detail. the dichotomous variable “d” (dummy of tobacco user shows if a household consuming any type of tobacco the value is “1” else “0”,) has an absolutely significant (positive) impact at 1% level of significance at country level and 5% level of significance at rural and urban region respectively, thus inclusion of this variable indicates household’s tobacco expenditure status. in table 4.1, the tobacco expense shows a negative and significant impact at 10% level of significance that shows a rise in tobacco expenditure leads to a fall in the share of education ultimately support the theoretical framework of thesis. the negative sign associated with the square of log of expenses confirms the formation of engel curve. the role of durable assets in the demand for tobacco is blurred. the finding of this paper shows that the household with durable assets and high income have more ability to purchase finest forms of tobacco but due to their financial status it does not influence the share of education. while on the flip, we have a lot of poor people in the country who forgo the necessary expenditures like education and health due to tobacco consumption. the result shows negative association of this variable at 1% level, i.e. it leads to a reduced share of education in the total budget of a household due to tobacco expenditures. table 4.1: crowding out of tobacco usage on education share independent variables all rural urban dummy tobacco user 0.049* (0.016) 0.062** (0.037) 0.061** (0.023) household tobacco expenses -1.55e-07*** (9.59e-08) -1.12e-07 (1.69e-07) -1.87e-07*** (1.16e-07) log of expenses (net of tobacco expenditures) 0.020* (0.001) 0.015* (0.004) 0.025* (0.002) journal of applied economics and business studies, volume. 6, issue 1 (2022) 85-108 https://doi.org/10.34260/jaebs.614 99 square of log of expenses (net of tobacco expenditures) -0.0007* (6.77e-05) -0.0005* (0.00019) -0.0009* (9.56e-05) dummy tobacco user * log of expenses -0.007* (0.002) -0.009 (0.005) -0.008** (0.003) dummy tobacco user*square of log of expense 0.0002* (9.63e-05) 0.0003 (0.0002) 0.0003** (0.0001) household size 0.0001* (1.78e-05) 2.33e-05 (3.44e-05) 0.0001* (2.22e-05) age of hh head -3.2e-05* (3.61e-06) -3e-05* (6.41e-06) -3.2e-05* (4.32e-06) education of hh head 8.41e-05* (1.21e-05) 0.00015* (2.58e-05) 6.86e-05* (1.35e-05) durable assets -1.97e-09* (7.42e-10) -1.07e-08* (2.09e-09) -1.27e-09** (6.27e-10) log of food expenditures 0.001* (0.0002) 0.001* (0.0003) 0.002* (0.0003) constant -0.156* (0.011) -0.126* (0.029) -0.196* (0.017) number of observations 24238 8083 16155 f-test 28.05 8.86 20.36 joint significance-consumer separability test 0.000 0.000 0.000 4.2 assessing crowding out effect on health similarly, table 4.2 and 4.3 discussed the impact of household tobacco expenditure on health share along with other control variables, using engel curve and without engel curve respectively. the expenditures on health has shown both income and substitution effect. the other factors that served as control variables are logarithm of expenses (net of tobacco), region, gender, standard of living, age of household head, number of children and number of earners in a household are used. the nature and results of each variable is discussed in detail as under. the categorical variable “d” (dummy of tobacco user) as discussed earlier in this chapter, as expected, negatively associated with share of health. this shows that if the spending on tobacco increases it will lead a decline in share of health in total budget of a household. the kashif haneef, syed faizan iftikhar, & ambreen fatima, 100 results mentioned below present the outcomes of share of health (dependent variable) followed by a series of relevant independent variables effecting the share of health in the total budget of a household. as our primary concern is to find out the effect of those socio demographic factors that are inclined by the conditional demand. in order to achieve this concern, we have incorporated age of household head, region, gender, standard of living, number of children and number of earners. the insertion of these variables provide us with negative (positive) coefficient and associated with a lower (higher) budget slice of a household consumes tobacco. these variables are broadly used and discussed in the literature as a regulator for socioeconomic differences. journal of applied economics and business studies, volume. 6, issue 1 (2022) 85-108 https://doi.org/10.34260/jaebs.614 101 table 4.2: crowding out of tobacco usage on health share using non-linear model independent variables all rural urban dummy tobacco user -0.032 (0.058) 0.169 (0.133) -0.074 (0.073) household tobacco expenses -3.74e-07**** (2.86e-07) -2.27e-07 (5.83e-07) -4.96e-07*** (3.11e-07) log of expenses (net of tobacco expenditures) 0.007 (0.005) 0.0186 (0.014) 0.011** (0.006) square of log of expenses (net of tobacco expenditures) -0.0003** (0.0002) -0.0007 (0.0005) -0.0005** (0.0002) dummy tobacco user * log of expenses 0.003 (0.008) -0.028 (0.020) 0.009 (0.010) dummy tobacco user*square of log of expense -4.6e-05 (0.0003) 0.001 (0.0008) -0.0003 (0.0004) gender 0.003* (0.0008) 0.004* (0.001) 0.001 (0.001) age of hh head 0.000089* (1.08e-05) 3.14e-05 (0.000021) 9.88e-05* (1.21e-05) standard of living (o) -0.002** (0.001) -0.002 (0.001) -0.004** (0.001) standard of living (oe) -0.006* (0.001) -0.005* (0.001) -0.007* (0.001) standard of living (oeg) -0.011* (0.001) -0.008* (0.001) -0.010* (0.001) standard of living (oegd) -0.015* (0.001) -0.016* (0.002) -0.013* (0.001) standard of living (oegdd) -0.015* (0.001) -0.022* (0.005) -0.013* (0.002) number of children (hh) -0.0001* (7.25e-05) -0.0006* (0.0001) -0.0001** (8.13e-05) number of earners (hh) 0.0003* (0.0001) 0.0005** (0.0002) 0.0005* (0.0001) constant -0.004* (0.037) -0.084 (0.091435) -0.03976 (0.044493) number of observations 24238 8083 16155 f-test 7.66 3.34 3.60 joint significance-consumer separability test 0.000 0.000 0.000 kashif haneef, syed faizan iftikhar, & ambreen fatima, 102 table 4.3: crowding out of tobacco usage on health share using linear model independent variables all rural urban dummy tobacco user -0.02* (0.005) -0.03* (0.01) -0.01* (0.006) household tobacco expenses -3.77e-07**** (2.85e-07) -1.81e-07 (5.83e-07) -5.13e-07** (3.11e-07) log of expenses (net of tobacco expenditures) 0.007** (0.004) 0.004 (0.01) 0.014* (0.005) square of log of expenses (net of tobacco expenditures) -0.0003** (0.0001) -0.0001 (0.0004) -0.0006* (0.0001) dummy tobacco user * log of expenses 0.001* (0.0004) 0.002* (0.0008) 0.001* (0.0004) gender 0.003* (0.0008) 0.004* (0.001) 0.001 (0.001) age of hh head 0.00008* (1.08e-05) 3.06e-05 (0.000021) 9.87e-05* (1.21e-05) standard of living (o) -0.002** (0.001) -0.002 (0.001) -0.004** (0.001) standard of living (oe) -0.006* (0.001) -0.005* (0.001) -0.007* (0.001) standard of living (oeg) -0.011* (0.001) -0.008* (0.001) -0.01* (0.001) standard of living (oegd) -0.01* (0.001) -0.01* (0.002) -0.01* (0.001) standard of living (oegdd) -0.01* (0.001) -0.02* 0.005) -0.01* (0.002) number of children (hh) -0.0001* (7.25e-05) -0.0006* (0.0001) -0.0001** (8.13e-05) number of earners (hh) 0.0003** (0.0001) 0.0005** (0.0002) 0.0005* (0.0001) constant -0.007 (0.029) 0.009 (0.067) -0.06** (0.03) number of observations 24238 8083 16155 f-test 11.49 3.87 5.09 joint significance-consumer separability test 0.000 0.0209 0.0062 journal of applied economics and business studies, volume. 6, issue 1 (2022) 85-108 https://doi.org/10.34260/jaebs.614 103 the results from this paper show that the tobacco expense has an absolute and significant (negative) impact at 20% level on share of health at pakistan level while its impact is at 10% level in urban region. it is accompanied with a dummy variable “d” which holds the value “0” if non-user and “1” if user of tobacco in any form. while few authors claims that coefficient (as projected here), computes crowding out of outlays and a differential budget share allocation [see john (2006) and (2012)]. this paper also incorporated the logarithm of total household expenditures (exclusive of tobacco expenditures) to measure the impact of total household expenses on the share of health and found a negative significant impact at 5% level on the health share in the total budget. while constructing the variable standard of living, this study incorporated the access of household on five major necessary facilities (occupation, electricity, gas, drinking water, drainage) that may have some influence on tobacco consumption. the statistical results show a negative significant impact ranging from -0.30% to -1.30% in the above mentioned facilities at 5% level of significance collectively. many authors like (john, 2011, chelwa, 2014, do yk et.al, 2015) discussed the same in diverse protocols like low-income or living conditions etc. the independent variable number of children in a household has shown a negative significant impact at 1% level on the health share of a household. based on findings, the higher the number of children the higher the tobacco expenses of a household. as discussed earlier that a larger part of the whole population living in rural areas with low income and a high illiteracy left them with a mindset that children empowered them financially. the results found out was just alike john (2006) that a rise in the tobacco expenditure bring a reduction in the budget quota dedicated to health and education. it has been preassumed that the tobacco using household must have different preferences over non-tobacco users. this fact cannot be denied that there is an opportunity cost of tobacco expense in the form of letting far away many essential expenditures on other goods. 5. conclusion and policy recommendations an integral objective of this paper is to examine the distribution of budgetary resources among tobacco consuming households as compared to non-users of tobacco and related goods. rather more consideration was paid to the socio-demographic factors responsible to consume or not to consume tobacco by households. there are several influential factors like age, gender and education of a household head, standard of living, number of children and kashif haneef, syed faizan iftikhar, & ambreen fatima, 104 earners, availability of durable assets and food expenditures. while estimating the influence on share of education and health, all these factors were found with a significant impact (positive or negative) for a household consuming tobacco and related goods. 5.1 conclusions the estimates achieved by employing share of education and health in total budget as dependent variable while tobacco expenditures act as an independent variable along with a set of some demographic characteristics of a household as control variables to get appropriate results. we have used ols estimation to evaluate our results. the estimation will be certainly helpful in the long run forecasting of the influence on household demand of education and health particularly due to consumption of tobacco and related goods, but also in evaluation of prosperity intensities across diverse households. assuming a practical form for demand models is a common practice. the assessments obtained from the quaids model shows a negative relationship between tobacco expenditures and share of education and health in a household budget. while evaluating the results it has also been found that the household with a positive tobacco expenditures tend to divert the valuable monetary resource from most of the essential expenditures of a family that could be beneficial for them in some other way. around 61% of the total population of pakistan is living in rural areas where poverty, unemployment, and illiteracy are the common factor for all masses. they are living below averages standard of living with least concerns for education and health. illiteracy is the root cause of unawareness about education and health consciousness. people living in rural areas of pakistan traditionally consume tobacco in different forms like huqqah is frequently use in punjab province; naswar in kpk, guthka in balochistan; pan, supari and other chewable forms of tobacco in sindh respectively as observed by bile et.al (2010). the most common and regular form of tobacco, used in all provinces is cigarette which has been became a practice among the respondents age 35 and above. although, household outlay on tobacco constitutes a low percentage impact while assessing its crowding-out effect on share of education and health in total budget it certainly reduces the amount of disposable family income. the habitual tobacco user has a preallocated slot of outlay for tobacco and use to spend the left amount on other necessary expenditures. consequently, tobacco becomes not only costly but an addiction too, sendoff the household using tobacco with lesser earnings for further necessary things. in this perspective, we try to illuminate the essence of crowding out of education and health mainly that arises in household outlay arrangements as an outcome of tobacco disbursements. journal of applied economics and business studies, volume. 6, issue 1 (2022) 85-108 https://doi.org/10.34260/jaebs.614 105 pakistan is an under-developed country facing a huge burden of internal and external debts. being an agrarian country with low level of industrialization brings about a poor standard of living among the masses due to unemployment and inflationary pressure. the disposable incomes for a large part of population is low and spending on tobacco becomes inflated. while considering the opportunity cost of consuming tobacco the detriment of outlay on children’s education, consuming unhealthy foods, and sacrifice on health expenses is obvious. same pattern has found out by becker (1965) where he observed a household as an alliance of more than one individuals performing as a sole entity of decision maker for consumption and the hypotheses that these entities perform for utility maximization facing constraint by scarce means is debatable. the descriptive analysis of the pslm 2015-2016 (pakistan social living standards measurements survey), discovered numerous remarkable annotations. our investigation proposes that poor and illiterate households using tobacco disbursed more of their finances on tobacco. we have included numerous socio-demographic variables like age, education, gender, region, standard of living, availability of durable assets, number of children and earners in a household to control the spending effects and prove our hypotheses. we have incorporated a method of engel curves (conditional) to estimate the influence of tobacco expenditure on education and health among necessary goods and services. 5.2 policy recommendations based on the above research conducted following are some recommendations that should be adopted to lowering down the expenditures on tobacco: i. there must be a pronounced concern on stringent pricing policy on cigarette (tobacco) consumption which results in lower consumption of cigarettes (tobacco) for both higher and lower income groups respectively. ii. the tobacco control strategies should at least be fully implemented by pre-created laws and regulations at public areas. iii. penalties should be imposed on usage and spitting of tobacco in different forms (cigarettes, bidi, tobacco raw, chew, pan, supari and naswar etc.) on public places. iv. imposition of higher taxes on tobacco merchandising and extensive restrictions on smoking in civic places should be applied as demand reduction policies. both the suggestive measures help in decreasing tobacco consumption and its subsequent destructions to societal health and economic development. v. there are a series of undertakings and rules that can restrain the consumption of tobacco in pakistan. some of the impressive precedents have been enforced in kashif haneef, syed faizan iftikhar, & ambreen fatima, 106 pakistan through the proclamation of prohibition of smoking ordinance 2002 are sales restrictions to minors, clean air policies, and tobacco taxation and policy like (tc) tobacco cess, (ced) central excise duty, and (gst) general sales tax, adding about 78% of the trade amount. if appropriate enquiry on the practice of precedents and execution proceedings can be carried out, these laws can be operative. references ahmad, k., jafary, f., jehan, i., hatcher, j., khan, a. q., chaturvedi, n., and jafar, t. h. (2005). ‘prevalence and predictors of smoking in pakistan: results of the national health survey of pakistan’. european journal of cardiovascular prevention and rehabilitation, 12(3), pp.203-208. alachkar, a. (2008). ‘household expenditure on cigarettes and tobacco in syria’ (no. 08, 18). school of economics discussion papers. alam, a. y., iqbal, a., mohamud, k. b., laporte, r. e., ahmed, a., and nishtar, s. (2008). ‘investigating socio-economic-demographic determinants of tobacco use in rawalpindi, pakistan’. bmc public health, 8(1), pp.50. ali, s., sathiakumar, n., and delzell, e. (2006). ‘prevalence and socio-demographic factors associated with tobacco smoking among adult males in rural sindh, pakistan’. southeast asian journal of tropical medicine and public health, 37(5), pp.1054. banks, j., blundell, r., and lewbel, a. (1997). ‘quadratic engel curves and consumer demand’. review of economics and statistics, 79(4), pp.527-539. bile k.m., j.a. shaikh, h.u.r. afridi and y. khan, ‘smokeless tobacco use in pakistan and its association with oropharyngeal cancer’, emhj, vol.16, 2010. block, s., and webb, p. (2009). ‘up in smoke: tobacco use, expenditure on food, and child malnutrition in developing countries’. economic development and cultural change, 58(1), pp.1-23. bobak, m., jha, p., nguyen, s., jarvis, m., and mundial, b. (2000). ‘poverty and smoking’. in tobacco control in developing countries. oxford university press. pp.41-61 bolin, k., and lindgren, b. (2007). ‘smoking, healthcare cost, and loss of productivity in sweden 2001’. scandinavian journal of public health, 35(2), pp.187-196. bonu, s., rani, m., peters, d. h., jha, p., and nguyen, s. n. (2005). ‘does use of tobacco or alcohol contribute to impoverishment from hospitalization costs in india?’ health policy and planning, 20(1), pp.41-49. busch, s. h., jofre-bonet, m., falba, t. a., and sindhelar, j. l. (2004). ‘burning a hole in the budget’. applied health economics and health policy, 3(4), pp.263-272. chaloupka, f. j., and warner, k. e. (2000). ‘the economics of smoking’. handbook of health economics, 1, pp.1539-1627. chelwa, g., and van walbeek, c. (2014). ‘assessing the causal impact of tobacco expenditure on household spending patterns in zambia’. econ res southern afr, 453, pp.1-40. cutler, d. m., and lleras-muney, a. (2006). ‘education and health: evaluating theories and evidence’ (no. w12352). national bureau of economic research. deaton, a., and kozel, v. (2005). ‘the great indian poverty debate’. macmillan. deaton, a., and muellbauer, j. (1980). ‘an almost ideal demand system’. the american economic journal of applied economics and business studies, volume. 6, issue 1 (2022) 85-108 https://doi.org/10.34260/jaebs.614 107 review, 70(3), pp.312-326. djikanovic, b., marinkovic, j., jankovic, j., vujanac, v., and simic, s. (2010). ‘gender differences in smoking experience and cessation: do wealth and education matter equally for women and men in serbia?’ journal of public health, 33(1), pp.31-38. efroymson, d. (ed.). (2003). ‘tobacco and poverty: observations from india and bangladesh’. path canada. efroymson, d., ahmed, s., townsend, j., alam, s. m., dey, a. r., saha, r., and rahman, o. (2001). ‘hungry for tobacco: an analysis of the economic impact of tobacco consumption on the poor in bangladesh’. tobacco control, 10(3), pp.212-217. efroymson, d., pham, h. a., jones, l., and fitzgerald, s. (2011). ‘tobacco and poverty: evidence from vietnam’. tobacco control, 20(4), ‘296-301. government of pakistan 2001. ministry of finance: economic survey of pakistan islamabad. guindon, g., and boisclair, d. (2003). ‘past, current and future trends in tobacco use’. ucsf: center for tobacco control research and education. haider, a., and zaidi, m. (2017). ‘food consumption patterns and nutrition disparity in pakistan’. forthcoming in: pakistan development review. hu, t. w., mao, z., liu, y., de beyer, j., and ong, m. (2005). ‘smoking, standard of living, and poverty in china’. tobacco control, 14(4), pp.247-250. johansen, l. (1974). ‘establishing preference functions for macroeconomic decision models: some observations on ragnar frisch's contributions’. european economic review, 5(1), pp.41-66. john, r. m. (2008). ‘crowding out effect of tobacco expenditure and its implications on household resource allocation in india’. social science and medicine, 66(6), pp.1356-1367. john, r. m., ross, h., and blecher, e. (2012). ‘tobacco expenditure and its implications for household resource allocation in cambodia’. tobacco control, 21(3), pp.341-346. john, r. m., sung, h. y., max, w. b., and ross, h. (2011). ‘counting 15 million more poor in india, thanks to tobacco’. tobacco control, tc-2010. koch, s. f., and tshiswaka-kashalala, g. (2008). ‘tobacco substitution and the poor’ (no. 200832). koning, p., webbink, d., and martin, n. g. (2015). ‘the effect of education on smoking behavior: new evidence from smoking durations of a sample of twins’. empirical economics, 48(4), pp.14791497. liu, y., rao, k., hu, t. w., sun, q., and mao, z. (2006). ‘cigarette smoking and poverty in china’. social science and medicine, 63(11), pp.2784-2790. majra, j. p., and gur, a. (2009). ‘poverty, tobacco, and health: an indian scenario’. journal of health, population, and nutrition, 27(3), pp.316. menke, f. l., kang, h. g., chen, z., park, j. m., kumar, d., and klessig, d. f. (2005). ‘tobacco transcription factor wrky1 is phosphorylated by the map kinase sipk and mediates hr-like cell death in tobacco’. molecular plant-microbe interactions, 18(10), pp.1027-1034. nasir, j. a., imran, m., chohan, a. a., and zaidi, s. a. a. (2015). ‘impact of sociodemographic factors on cigarette smoking in pakistani males’. khyber medical university journal, 7(3). nayab, d., nasir, m., memon, j.a. and siddique, o., 2021. the economic cost of tobacco-induced diseases in pakistan. pakistan institute of development economics. nizami, s., sobani, z. a., raza, e., baloch, n. u. a., and khan, j. (2011). ‘causes of smoking in pakistan: an analysis of social factors’. journal of the pakistan medical association, 61(2), pp.198. paraje, g., and araya, d. (2018). ‘relationship between smoking and health and education spending in chile’. tobacco control, 27(5), pp.560-567. pu, c. y., lan, v., chou, y. j., and lan, c. f. (2008). ‘the crowding-out effects of tobacco and alcohol kashif haneef, syed faizan iftikhar, & ambreen fatima, 108 where expenditure shares are low: analyzing expenditure data for taiwan’. social science and medicine, 66(9), pp.1979-1989. resolution, u. g. a. (2000). 55/2, united nations millennium declaration, un doc. a/res/55/2. san, s., and chaloupka, f. j. (2016). ‘the impact of tobacco expenditures on spending within turkish households’. tobacco control, 25(5), pp.558-563. seatca. ‘cycle of poverty in tobacco farming’: tobacco cultivation in southeast asia. bangkok: southeast asia tobacco control alliance, 2008. shafey, o., eriksen, m., ross, h., and mackay, j. (2009). ‘the tobacco atlas’. 3rd. atlanta, ga: american cancer society, world lung foundation. stigler, g. j. (1954). ‘the early history of empirical studies of consumer behavior’. journal of political economy, 62(2), pp.95-113. sufia, s., khan, a. a., and ijaz, s. (2003). ‘patterns of tobacco use in pakistan’. pak oral dental j, 23(1), pp.45-50. van walbeek, c. (2002). ‘recent trends in smoking prevalence in south africa-some evidence from amps data’. south african medical journal, 92(6), pp.468-472. vermeulen, f. (2003). ‘do smokers behave differently? a tale of zero expenditures and separability concepts’. economics bulletin, 4(6), pp.1-7. volinia, s., calin, g. a., liu, c. g., ambs, s., cimmino, a., petrocca, f., ... and prueitt, r. l. (2006). ‘a microrna expression signature of human solid tumors defines cancer gene targets’. proceedings of the national academy of sciences, 103(7), pp.2257-2261. wang, h., sindhelar, j. l., and busch, s. h. (2006). ‘the impact of tobacco expenditure on household consumption patterns in rural china’. social science and medicine, 62(6), pp.1414-1426. world health organization, and research for international tobacco control. (2008). ‘who report on the global tobacco epidemic, 2008: the mpower package’. world health organization. world health organization. (2002). ‘tobacco or health: a global status report’, country profiles by region. 1997. world health organization. (2008). ‘tobacco and poverty in the philippines’. national tobacco control team, department of health, manila college of public health, university of the philippines, manila philippine college of medical researchers foundation, inc., and the tobacco free initiative of the world health organization. xin, y., qian, j., xu, l., tang, s., gao, j., and critchley, j. a. (2009). ‘the impact of smoking and quitting on household expenditure patterns and medical care costs in china’. tobacco control, tc-2008. yao, t., huang, j., sung, h. y., ong, m. k., mao, z., jiang, y., ... and max, w. (2015). ‘determinants of smoking-induced deprivation in china’. tobacco control, 24(suppl 4), pp.iv35-iv39 journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 17 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x do the electricity price shocks influence the sectoral production and kse100 index in pakistan? an ardl structural breaks approach abbas khan1*, muhammad yar khan2 and abdul qayyum khan3 1 department of management sciences, comsat university islamabad wah campus 2 department of management sciences, comsat university islamabad wah campus 3 department of management sciences, comsat university islamabad wah campus abstract this research investigates the long-term cointegration of electricity price with sectoral production and equity market in pakistan. fourteen major industrial sectors and the kse100 index is taken into consideration to determine the relationship. literature in this regard is available but this research is distinct from previous literature for it tests the sectoral production and equity market relationship with electricity price change in pakistan. monthly data from 1st jan 2011 till 31st dec 2019 is taken for fourteen sectors from the sources of quantum index pakistan bureau of statistics (pbs) and for kse100 index from (www.investing.com). an auto regressive distributed lag (ardl) model and bound test for multiple structural breaks has been applied. it is found that almost the production of all industrial sectors and kse100 index stock prices are adversely affected by the electricity price shocks both in long-term and short-term. the study suggests that management should implement a moderate monitory policy that is neither more expansionary nor contractionary. the government should provide incentives to those who successfully control energy wastage. a mixed kind of energy policy is recommended with higher weightage to the development of renewable energies to reduce foreign exchange outflow with imported furnace oil. this study is limited to the sectoral production and equity market of pakistan. a cross-sectional research is encouraged to compare the connection between major energy costs and macroeconomic variables in different countries. keywords electricity price change, sectoral production, kse100, economic growth, ardl, pakistan. jel classification q21, d13, e23, o4 * abbaskhanswabi@gmail.com abbas khan, muhammad yar khan and abdul qayyum khan 18 1. introduction the world bank enterprise survey 2015 states that 45.3% of the total firms in pakistan have identified electricity as the top obstacle for the business sector in pakistan (bank, 2015). like other developing countries, the shortfall and higher cost of electricity will affect the economic activities. at average, in south asia each firm is facing a load shedding of 5.3 hours out of 24 hours while in pakistan the average load shedding faced by each firm is 13.2 hours out of 24 hours (grainger & zhang, 2017, 2019). currently, 50 million people have no access to electricity while others in access are facing regular load shedding. about 75% of the firms in pakistan have pointed out electricity as a major barrier to the production growth as shown in figure 1. pakistan is reckoned on 115 out of 137 countries for its reliable source of electricity in the world (schwab, 2018). in pakistan the distortion of power sector costs 7% of the total gdp, which equals to $18 billion a year. the report analyzes the power supply cycle including power generation, and supply to the users. this distortion is caused by poor infrastructure, faulty metering and theft cases which increases load shedding and per unit cost. consequently, the businesses collapse and the units of production are reduced (the world bank, 2013). therefore, electricity sector reform should be the top priority for the government of pakistan to quickly yield major economic gains which will directly increase the firm’s productivity and reliability. it also reduces the cost of production and 𝐶𝑂2 emissions (grim et al., 2020). figure 1 proportion of industries recognizing electricity as a major obstacial to production. the early classical study considered energy as the fundamental factor of industrial production. the study identified that the output cost varies with the cost of input. the classical theory proposed that additional to energy, the labor and capital are other major input costs of production (kümmel, 1982). the concept of the classical theory is contradicted by stating that journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 19 the importance of energy as input is increased with technological advancement (jorgenson, 1984; rosenberg, 1983). the significance of energy consumption is increased with the technological advancement in the industrial production. the pollution taxes and environmental control system also discourages the consumption of oil and gas. thus, the dependency of electricity as a major input source for production is increased (ayres et al., 2013; guo et al., 2019; wu et al., 2019). the link between the cost of energy and ip is one of the most important subjects for the economic policy makers. most of the researchers and policy makers focused on the cause and effect of overall production, gdp and electricity consumption. recently, the research aimed to investigate the electricity shortage effect on industrial production in pakistan (grainger & zhang, 2019). according to the third quarterly report of the state bank of pakistan for the year 2018-2019, the inflation rate increased from 6% to 8.5% in the third quarter of 2019. the cost push factors of inflation in the energy sectors are petrol, gas and electricity. it has risen up the consumer price index and increased the cost of production (sbp, 2019). in pakistan, mostly the impact of electricity shortage is taken as a proxy with overall industrial production (a. ali et al., 2019; grainger & zhang, 2017; jamil & ahmad, 2010; yasmin & qamar, 2013). traditionally, crude oil price, natural gas price and cpi are used to quantity the aggregate output of pakistan. specifically, the researchers focused on aggregate industrial output rather than sectoral production. furthermore, the planning commission of pakistan (pcp) in 2019 reported 2/3 of the electricity generated from high cost thermal power plants which has negatively affected the economy (government, 2020). previous studies usually investigated the casual connection between energy costs and gdp. there is a gap to quantity the impact of electricity price shock and the output of different industrial sectors in pakistan. this research objective is to expand the understanding by investigating the short-run and long-run connection between electricity price change and production growth in different industrial sectors of pakistan. the sector specific impact of electricity price change is significant for various reasons. firstly, the impact of the electricity price change is not similar for all sectors. the sector sensitivity to electricity price changes should be asymmetric because all the sectors might not be exposed to the electricity price change. the sector sensitivity depends on utilization of electricity as a major or a minor input source. secondly, by adding the electricity input cost to the final goods available in market may reduce production cut offs. measuring sectoral production sensitivity to electricity price changes is more explanatory than gdp. thirdly, the industries may switch from high cost to low cost energy input. this behavioral heterogeneity of the industries will identify the sensitivity to electricity price shocks. this study analyzes various industrial sectors in pakistan thus helping the policy makers and concerned reader and researcher to get comprehensive information about a relationship with electricity prices. it can help the government authorities to take well informed decisions for the betterment of effected industrial sectors by identifying cheaper sources of energies, power subsidies, awareness of energy savings and tax relief. the paper consists of: section-2 review of literature, section-3 on data collection and methodology, section-4 as empirical findings and discussion, section-5 on main findings lastly, section-6 as conclusion. abbas khan, muhammad yar khan and abdul qayyum khan 20 2. review of literature: the study of the connection between energy prices and the sector specific industrial production is not common. although, previous literature mainly focused on the casual connection between the consumption of energy and growth in the economy (bildirici et al., 2012; destek & aslan, 2017; dogan, 2015; ghali & el-sakka, 2004; ozturk, 2010; polemis & dagoumas, 2013; wolde-rufael, 2014; wu et al., 2019). a sector specific relationship with electricity consumption is carried out between 1993–2006 and 1993-2011. the findings proposed that the irregular upsurge in the consumption of electricity is due to the structural changes in the production line (blignaut et al., 2015; inglesi-lotz & blignaut, 2011). electricity consumption is usually more reactive to gdp (bildirici et al., 2012; ciarreta & zarraga, 2010b, 2010a; ghali & el-sakka, 2004). using the panel data to analyze the connection between consumption of electricity and growth in the economy of 12 european countries, but still the sectoral investigation is not emphasized (ciarreta & zarraga, 2010a). the relationship between consumption of electricity, cost of electricity and aggregate production between 1960-2010 is investigated. the study found a unidirectional causal relationship between electricity prices and gdp (jamil & ahmad, 2010). the short-run and long-run effect of electricity consumption and its determinants is investigated. the results suggested a short-term price adjustment strategy adopted by pakistan is not efficient. rather, pakistan should improve the utilization of power generation plants to reduce the cost of electricity available to the industries or to reuse the oil and gas resources for generating electricity. otherwise, a shortage will widely increase the cost of electricity available to the industries (alter & syed, 2011). another study analyses the connection between consumption of electricity and increase in economy. the study determines a positive bi-directional association among the two variables. it is further suggested to cover the increasing demand of electricity pakistan needs an effective power generation policy and cost control (shahbaz & lean, 2012). the economy and firm’s demand for electricity between 1998-2008 is investigated. the findings illustrates a negative relationship with 1% increase in the cost of electricity will lead to approximately -0.58% reduction in the electricity demand across the firms (amjad chaudhry, 2010). indeed, the islamabad chamber of commerce and industry in 2013 reported summer as the worst season for the public and commercial sectors of pakistan with a maximum average of 1618 hours load shedding all over the country (magnet, 2013). the shortage of electricity causes a decrease in the production, increase unemployment and closure of the industries. the study investigated canada, ecuador, norway and south africa (fei et al., 2014), bangladesh (masuduzzaman, 2013), nigeria (danmaraya & hassan, 2016; polemis & dagoumas, 2013), south africa (amusa et al., 2009; bildirici et al., 2012), china (zhang et al., 2017) found a positive connection among consumption of electricity and growth in the economy. another study in brazil, china, russia, india and south africa (khobai, 2018), pakistan (balcilar et al., 2019), usa (alola & yildirim, 2019) supports energy as a main driver of the economy. the literature has the opinion that the cost of electricity and consumption have a significant role in the industrial production. this implies that higher electricity prices have a negative impact on the consumption of electricity and production growth. other studies by (capros et al., 2016; gonese et al., 2019) examined the effect of electricity and gas prices on sectoral production in the european union (eu) and south africa. the findings suggest that journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 21 electricity is an essential part of the production having a significant effect on the output growth of different sectors in both eu and south africa. the shortage of electricity will increase the input cost by using diesel generators in producing electricity. it will reduce the capital available for productive use leading to decrease in the output. more critically, if the alternative source of electricity generation is not available during shortages it may lead to shut down industries, spoil of useful raw materials and labor productivity (allcott et al., 2016). continuous electricity shortage compels firms to outsource electricity which will increase the output cost (fisher-vanden et al., 2015). due to higher electricity cost firms avoid using energy intensive technology. adopting this strategy leads to decrease in the long term productivity growth (abeberese, 2017). another research examines the casual connection between growth in gdp and demand of electricity. it concludes bi-directional association between change in gdp and consumption of electricity (faisal et al., 2017). a comprehensive overview of various studies in china spanning from 1978 to 2016 concluded a significant relationship between electricity consumption and growth in the economy (zhang et al., 2017). another study investigates the casual relationship between electricity demand and gdp. a long-term bi-directional interdependency is found between both variables, but the relationship is insignificant in the short-run (hasan et al., 2017). the urbanization and electricity consumption are investigated in pakistan considering technology and transformation. it is found that both variables have a unidirectional positive impact on electricity consumption (shahbaz et al., 2017). the energy insecurity is explained as the unavailability of energy at reasonable prices. a study analyzed the energy security of pakistan in four dimensions as accessibility, availability, affordability, and applicability. the results found that the pakistan economy is continuously insecure over the last five years. it is recommended that pakistan should move toward green energy and advance metering systems (s. malik et al., 2020). the energy security is investigated between gdp growth and electricity consumption of the developing economy in south asia. the results reviled there is no longterm connection between consumption of energy and aggregate output. a 1% increase in the population will increase the consumption of electricity by 4.16%. it is suggested to use large scale hydropower to improve energy efficiency and climate control in the developing country of south asia like nepal (paija, 2019). additionally, there is a significant effect of energy costs on macro-economic variables (taghizadeh-hesary et al., 2015), energy insecurity also has great effect on finish goods and food prices (taghizadeh-hesary et al., 2019). energy price shocks influencing different macroeconomic factors i.e., gdp, rate of interest, rate of inflation, foreign exchange rate, human development, stock and bond prices, portfolio optimization and business cycle (ahmed et al., 2018; marza & daly, 2018; naser, 2019; nazlioglu et al., 2019; pönkä & zheng, 2019; s. sarwar et al., 2019; waheed et al., 2018; wesseh & lin, 2018). historically, the industrial production considered to be positively correlated with stock market return. it is further concluded that the growth in the industrial production largely represent the price movement of the stock market (fama, 1981, 1990). recently, a large proportion of researches constituted a literature which investigates the relationship between stock market and industrial production in different countries and scenarios. a study examined the cointegration between the stock price and industrial production, supply of money and foreign exchange rate by using long-term cointegration bound test. the results found a long-term cointegration between the stock price abbas khan, muhammad yar khan and abdul qayyum khan 22 movement and all the macroeconomic variables (bekhet & matar, 2013). there are many studies investigating the causality between equity market return and macroeconomic variables but the causality in the context of non-linear situation is investigated in china. it is concluded that the causality still exists between the stock price and macroeconomic variable in non-linear condition (borjigin et al., 2018). it is further concluded that industrial production and inflation rate performs a vital part in the equity return volatility during long-run and short-run prospects (engle et al., 2013). it is further added that macroeconomic fundamentals are playing an important role in speculation of stock market return (girardin & joyeux, 2013). another study has taken the industrial production and long-term interest rate as a factor that influences the european stock price movement. the finding revealed that the weight has clearly moved from interest rate to industrial production. it is concluded that ip has a greater impact of stock prices in comparison to interest rate (peiro, 2016). pakistan stock exchange (pse) is established in 2016 by merging three different stock markets i.e., karachi, lahore, and islamabad. it is regulated by security exchange commission of pakistan (secp). kse100 includes the top 100 multiple sector companies enlisted in pse based on higher market capitalization. like other countries, in pakistan energy is the major driver of the economy. in pakistan most electricity is generated using thermal power plant (solangi et al., 2018; zameer & wang, 2018). pakistan is facing a shortfall of 65000 mega walt (mw) due to load shedding. to fulfill the required demand of electricity in pakistan by using imported oil from gulf countries. it is concluded that pakistan’s economy is exposed to energy price shocks like other developing countries (wakeel et al., 2016). in the context of pakistan, previous literature found a negative connection of energy cost with economy. the relationship of individual sector contributing to the economy is not yet identified. all industrial sectors are individually contributing an important part in the economic development of a country. the study will provide new foresight by finding out the effect of electricity cost on sectoral production of pakistan. this study will help energy policy makers to identify the most sensitive sectors to electricity price shocks. further, it will grab the attention of policy makers to make a sophisticated energy policy for all affected industrial sectors. additionally, the results will help the investors to identify the energy price relationship with stock market return in pakistan. further, the association between ip and equity market return will be cross validate. 3. data collection this study considers 14 industrial sectors that are critically important for the gdp of pakistan. these industrial sectors include automobiles (cars, cars parts and lubricants), chemicals (refineries, petrochemicals, metallurgical and mineral based products), food, beverages and tobacco (cigarettes, foods sub products), iron & steel products (construction materials), coke petroleum products (oil and gas products, distribution services, alternative energy resources), paper board (raw paper, packing, plastics and construction materials), pharmaceuticals (medicines and surgical products, health care and biotechnological products), rubber products (general use rubber products for home and commercial use), nonmetallic mineral products (cement, ceramics, glass, and lime), textile ( animal wool and silk, cotton, flax and bamboo, glass fiber and synthetic materials), fertilizers (agricultural products), electronics (house hold equipment and heavy duty machinery), leather products (cloths, shoes, bags etc.), journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 23 wood products (furniture and fixtures), engineering products (construction and material, manufacturing equipment). additionally, the kse100 index that comprises of the top 100 companies from all the 14 industrial sectors of pakistan. large-scale manufacturing data is available on the quantum index (qi). qi measures output and structural changes of large-scale manufacturing industries. it provides data regarding production, raw material, contribution to gdp, fix assets and large-scale manufacturing taxes. it also provides data regarding new industrial development and production. in terms of data this study uses monthly data during 1st jan 2011 and 31st dec 2019 (zhang et al., 2017). all the industrial sector data is available in pakistan bureau of statistics (pbs). qi is calculated at constant factor cost of year 2005-2006 with help of laspeyer’s formula (biggeri et al., 2017) as equation 1 base year 2005-2006. 𝑄𝐼(𝑃𝑀 ) = 𝑀(𝑛) 𝑀(0) ∗ 100 eq 1 equation 2 calculates monthly growth rate electricity relative to base year cpi 2007-2008: 𝐶𝑃𝐼(𝐸𝑀 ) = 𝑀(𝑛) 𝑀(0) ∗ 100 eq 2 equation 3 calculates the log return of the closing prices. 𝐾𝑆𝐸100(𝑅𝑚 ) = 𝐿𝑛 ( 𝑝(𝑛) 𝑝(0) ) eq 3 where, in equation 1 the “𝑃𝑀” represents a large-scale manufacturing industry, “𝑀(𝑛)” is the real output for the current month and “𝑀(0)” is the real output of the base year 2005-2006 and in equation 2 “𝐸𝑀” represents electricity prices, “𝑀(𝑛)” is the current month cpi and “𝑀(0)” is the previous month cpi. the data for the control variables like government expenditure, money in circulation, foreign direct investment and wholesale prices are extracted from the statistics of pakistan’s economy report, 2018 available on website of state bank of pakistan (sbp) (yasmeen et al., 2019). kse100 index monthly data is extracted from (www.investing.com). in equation 3 the 𝑅𝑚 is the log stock return, 𝑝(𝑛) is the current price, 𝑝(0) is the previous month price, and ln is a natural log (hanif, 2020). further, adding the control variable like government expenditure, money in circulation, foreign direct investment and wholesale prices can improve the relationship between electricity price and sectoral production. higher electricity prices lead to increase the production cost which has a direct impact of the whole sale prices of the unit produced which creates an inflationary situation in the country. it depresses the saving power of the public and increases the money supply in the country. money supply has a positive effect on the growth of industrial production (ip). foreign direct investment (fdi) also has a positive effect on the industrial production. but the increase in the energy prices may lead to decrease in the fdi due to increase in the cost of production. on other hand, pakistan is generating electricity mostly form imported furnace oil but due to decrease in the pakistani rupee comparatively with dollar the energy prices will increase which also leads to increase in the production cost and decrease the overall production of the industry. due to circular debts many industries are tax default which leads to decrease the government tax revenue and increase the budget deficit. it leads to shutdown the production of different industries in pakistan (yasmeen et al., 2019). the list of variables is mentioned in table 1. abbas khan, muhammad yar khan and abdul qayyum khan 24 table 1: acronyms of the variables acro full title elep electricity prices fdi foreign direct investment gxp government expenditure mc money in circulation wpi wholesale price index aut automobiles chemicals (cars, cars parts and lubricants) che chemicals foo food, beverages tobacco (cigarettes, foods sub products iro iron steel products (construction materials) cok coke petroleum products (oil and gas products, distribution services, alternative energy resources pap paper board (raw paper, packing, plastics and construction materials) pha pharmaceuticals (medicines and surgical products, health care and bio technological products) rub rubber products (general use rubber products for home and commercial use) non nonmetallic mineral products (cement, ceramics, glass, and lime tex textile (animal wool and silk, cotton, flax and bamboo, glass fiber and synthetic materials) fer fertilizers (agricultural products) ele electronics (household equipment and heavy-duty machinery) let leather products (cloths, shoes, bags etc.) woo wood products (furniture and fixture) eng engineering products (construction and material, manufacturing equipment) kse100 karachi stock exchange top 100 firms ardl autoregressive distributed lag adf augmented dickey–fuller pp phillips–perron cusum cumulative sum cusumsq cumulative sum of squares 4. methodology this study investigates the effect of electricity price changes on industrial sector’s production using multifactor nonlinear regression analysis considering the open economy industrial sector (is) function for sectoral production. it determines the effect of electricity price changes in sectoral production. each model includes electricity price as an explanatory variable. to increase the model goodness of fit the study considers other four explanatory variables i.e., government expenditure (gxp) on projects, money in circulation (mc) due to general public investment, foreign direct investment (fdi) and wholesale prices (wpi) (bohi, 2017; jo, 2014; yasmeen et al., 2019). this research utilized auto regressive distributed lag (ardl) model to investigate the impact of electricity prices on sectoral production in pakistan (akadiri et al., 2019; shin & smith, 2001). the ardl model is gaining increasing popularity because of high potential and journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 25 less glitches connected with it in comparison with other cointegration models (m hashem pesaran, 1997; m hashem pesaran & shin, 1998; yasmeen et al., 2019). the eagle granger method is used to examine the connection between two variables and for more than two variables then the johansen cointegration is used (econometrics, 2015; engle et al., 1987; johansen, 1988). vector auto regressive (var) model has certain shortcomings. it is applied only when there is a large sample size used in the study and the var model prerequisite is that all the variables must be stationary at the same level (johamen & jtiselius, 1990). in comparison to var the ardl model has some additional benefits i.e., ardl can be used for small sample size, if the variables are stationary at level or at first order difference or a mix of both while johansen cointegration the variables must be in similar order difference (mohammad hashem pesaran & pesaran, 1997; shin & smith, 2001). ardl allow variables with optimal lags while it is not allowed in other conventional cointegration models. by applying bound test the ols model is transformed to error correction model (ecm). ecm helps further in adjusting the long-run and short-run relationship without losing the long-run information (laurenceson & chai, 2003). as discussed, earlier ardl approach cannot be functional if there are second order difference i (2) stationary variables. for testing the stationarity of the variables with i(0) and i(1), the adf and pp tests are used to test 𝐻0 of a unit root (dickey & fuller, 1979; phillips & perron, 1988). for testing the unit root most of the studies used adf and pp tests. in time series data, due to structural breaks the adf and pp have low power of finding unit root therefore the multiple structural break test is used (bai & perron, 2003; balcilar et al., 2017; smith et al., 2019a). further, ardl approach has two steps. firstly, to test the 𝐻0 of no long run cointegration between the variables. by using the f-statistics value the existence of cointegration is confirmed than the study further interprets the coefficients for long-run and short-run. the ardl model generates the lower bound i(0) and upper bound i(1) critical values. the f-value greater than upper bound i(1) it means long-run cointegration exists in the relationship and the 𝐻0 is rejected. in contrast, if the f-statistics is lesser than lower bound i(0) we can accept the 𝐻0 and ardl approach cannot be applied. the results are inconclusive to apply ardl in case the f-value is in between lower and upper bound. in 2017, the bound test is found significant for small sample size (ahmed et al., 2018; garg & prabheesh, 2020). the usefulness of small sample size i.e., 30 to 80 observations is supported (narayan, 2007). this study uses the smaller sample size based on the methodology of (yasmeen et al., 2019). by applying the ardl model for different industries the ecms are calculated as following. ∆𝐴𝑈𝑇𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝛿2∆𝐹𝐷𝐼𝑡−1 + 𝛿3∆𝐺𝑋𝑃𝑡−1 + 𝛿4∆𝑀𝐶𝑡−1 + 𝛿5∆𝑊𝑃𝐼𝑡−1 + 𝑑𝑢𝑚𝑚𝑦2018 + 𝑑𝑢𝑚𝑚𝑦2019 + 𝜇𝑡 eq 4 ∆𝐹𝑂𝑂𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝛿2∆𝐹𝐷𝐼𝑡−1 + 𝛿3∆𝐺𝑋𝑃𝑡−1 + 𝛿4∆𝑀𝐶𝑡−1 + 𝛿5∆𝑊𝑃𝐼𝑡−1 + 𝑑𝑢𝑚𝑚𝑦2017 + 𝜇𝑡 eq 5 ∆𝐼𝑅𝑂𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝛿2∆𝐹𝐷𝐼𝑡−1 + 𝛿3∆𝐺𝑋𝑃𝑡−1 + 𝛿4∆𝑀𝐶𝑡−1 + 𝛿5∆𝑊𝑃𝐼𝑡−1 + 𝑑𝑢𝑚𝑚𝑦2017 + 𝑑𝑢𝑚𝑚𝑦2018 + 𝜇𝑡 eq 6 abbas khan, muhammad yar khan and abdul qayyum khan 26 ∆𝐶𝑂𝐾𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝛿2∆𝐹𝐷𝐼𝑡−1 + 𝛿3∆𝐺𝑋𝑃𝑡−1 + 𝛿4∆𝑀𝐶𝑡−1 + 𝛿5∆𝑊𝑃𝐼𝑡−1 + 𝑑𝑢𝑚𝑚𝑦2013 + 𝑑𝑢𝑚𝑚𝑦2019 + 𝜇𝑡 eq 7 ∆𝑃𝐴𝑃𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝛿2∆𝐹𝐷𝐼𝑡−1 + 𝛿3∆𝐺𝑋𝑃𝑡−1 + 𝛿4∆𝑀𝐶𝑡−1 + 𝛿5∆𝑊𝑃𝐼𝑡−1 + 𝑑𝑢𝑚𝑚𝑦2017 + 𝑑𝑢𝑚𝑚𝑦2019 + 𝜇𝑡 eq 8 ∆𝑃𝐻𝐴𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝛿2∆𝐹𝐷𝐼𝑡−1 + 𝛿3∆𝐺𝑋𝑃𝑡−1 + 𝛿4∆𝑀𝐶𝑡−1 + 𝛿5∆𝑊𝑃𝐼𝑡−1 + 𝑑𝑢𝑚𝑚𝑦2017 + 𝑑𝑢𝑚𝑚𝑦2019 + 𝜇𝑡 eq 9 ∆𝑅𝑈𝐵𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝛿2∆𝐹𝐷𝐼𝑡−1 + 𝛿3∆𝐺𝑋𝑃𝑡−1 + 𝛿4∆𝑀𝐶𝑡−1 + 𝛿5∆𝑊𝑃𝐼𝑡−1 + 𝑑𝑢𝑚𝑚𝑦2016 + 𝑑𝑢𝑚𝑚𝑦2018 + 𝜇𝑡 eq 10 ∆𝑁𝑂𝑁𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝛿2∆𝐹𝐷𝐼𝑡−1 + 𝛿3∆𝐺𝑋𝑃𝑡−1 + 𝛿4∆𝑀𝐶𝑡−1 + 𝛿5∆𝑊𝑃𝐼𝑡−1 + 𝑑𝑢𝑚𝑚𝑦2011 + 𝑑𝑢𝑚𝑚𝑦2018 + 𝜇𝑡 eq 11 ∆𝑇𝐸𝑋𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝛿2∆𝐹𝐷𝐼𝑡−1 + 𝛿3∆𝐺𝑋𝑃𝑡−1 + 𝛿4∆𝑀𝐶𝑡−1 + 𝛿5∆𝑊𝑃𝐼𝑡−1 + 𝑑𝑢𝑚𝑚𝑦2015 + 𝑑𝑢𝑚𝑚𝑦2018 + 𝜇𝑡 eq 12 ∆𝐹𝐸𝑅𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝛿2∆𝐹𝐷𝐼𝑡−1 + 𝛿3∆𝐺𝑋𝑃𝑡−1 + 𝛿4∆𝑀𝐶𝑡−1 + 𝛿5∆𝑊𝑃𝐼𝑡−1 + 𝑑𝑢𝑚𝑚𝑦2013 + 𝑑𝑢𝑚𝑚𝑦2019 + 𝜇𝑡 eq 13 ∆𝐸𝐿𝐸𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝛿2∆𝐹𝐷𝐼𝑡−1 + 𝛿3∆𝐺𝑋𝑃𝑡−1 + 𝛿4∆𝑀𝐶𝑡−1 + 𝛿5∆𝑊𝑃𝐼𝑡−1 + 𝑑𝑢𝑚𝑚𝑦2013 + 𝜇𝑡 eq 14 ∆𝐿𝐸𝑇𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝛿2∆𝐹𝐷𝐼𝑡−1 + 𝛿3∆𝐺𝑋𝑃𝑡−1 + 𝛿4∆𝑀𝐶𝑡−1 + 𝛿5∆𝑊𝑃𝐼𝑡−1 + 𝑑𝑢𝑚𝑚𝑦2016 + 𝑑𝑢𝑚𝑚𝑦2019 + 𝜇𝑡 eq 15 ∆𝐸𝑁𝐺𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝛿2∆𝐹𝐷𝐼𝑡−1 + 𝛿3∆𝐺𝑋𝑃𝑡−1 + 𝛿4∆𝑀𝐶𝑡−1 + 𝛿5∆𝑊𝑃𝐼𝑡−1 + 𝑑𝑢𝑚𝑚𝑦2013 + 𝑑𝑢𝑚𝑚𝑦2019 + 𝜇𝑡 eq 16 ∆𝐾𝑆𝐸100𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝛿2∆𝐹𝐷𝐼𝑡−1 + 𝛿3∆𝐺𝑋𝑃𝑡−1 + 𝛿4∆𝑀𝐶𝑡−1 + 𝛿5∆𝑊𝑃𝐼𝑡−1 + 𝜇𝑡 eq 17 journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 27 in equation 4,” 𝛽0” is a constant and “𝛽1 𝑡𝑜 𝛽5” are utilized for error correction in the model. the dummy variables are used after applying the structural break bound test (bai & perron, 2003; yasmeen et al., 2019). the “∆” and “𝜇𝑡” represent the white noise error term. the long run association among the variables is represented by “𝛿1 𝑡𝑜 𝛿5”. the ardl model estimates “(𝑛 + 1)𝑘” times regression to get optimal lags length criteria. where “𝑛” is maximum number of lags and “𝑘” is the number of variables under investigation. the ardl model is applied to check the long-term cointegration among the variables by using wald fstatistics. the null hypothesis is no long-term cointegration which is “𝐻0 = 𝛿1" 𝑡𝑜 "𝛿5 = 0”. the alternative hypothesis “𝐻𝑎" is long-term cointegration among the variables. equation 5 to 17 follow the same explanation. by evaluating long run cointegration by using f-statistics value and applying the bound test for structural breaks and finding long term coefficients in the above model the study further finds the short-term coefficients using the model below. ∆𝐴𝑈𝑇𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 𝑛1𝐸𝐶𝑇𝑡−𝑖 + 𝑑𝑢𝑚𝑚𝑦2018 + 𝑑𝑢𝑚𝑚𝑦2019 + 𝜇𝑡 eq 18 ∆𝐹𝑂𝑂𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝑛2𝐸𝐶𝑇𝑡−𝑖 + 𝑑𝑢𝑚𝑚𝑦2017 + 𝜇𝑡 eq 19 ∆𝐼𝑅𝑂𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝑛3𝐸𝐶𝑇𝑡−𝑖 + 𝑑𝑢𝑚𝑚𝑦2017 + 𝑑𝑢𝑚𝑚𝑦2018 + 𝜇𝑡 eq 20 ∆𝐶𝑂𝐾𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝑛4𝐸𝐶𝑇𝑡−𝑖 + 𝑑𝑢𝑚𝑚𝑦2013 + 𝑑𝑢𝑚𝑚𝑦2019 + 𝜇𝑡 eq 21 ∆𝑃𝐴𝑃𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝑛5𝐸𝐶𝑇𝑡−𝑖 + 𝑑𝑢𝑚𝑚𝑦2017 + 𝑑𝑢𝑚𝑚𝑦2019 + 𝜇𝑡 eq 22 ∆𝑃𝐻𝐴𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝑛6𝐸𝐶𝑇𝑡−𝑖 + 𝑑𝑢𝑚𝑚𝑦2017 + 𝑑𝑢𝑚𝑚𝑦2019 + 𝜇𝑡 eq 23 ∆𝑅𝑈𝐵𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝛿1∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛7𝐸𝐶𝑇𝑡−𝑖 + 𝑑𝑢𝑚𝑚𝑦2016 + 𝑑𝑢𝑚𝑚𝑦2018 + 𝜇𝑡 eq 24 ∆𝑁𝑂𝑁𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝑛8𝐸𝐶𝑇𝑡−𝑖 + 𝑑𝑢𝑚𝑚𝑦2011 + 𝑑𝑢𝑚𝑚𝑦2018 + 𝜇𝑡 eq 25 ∆𝑇𝐸𝑋𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝑛9𝐸𝐶𝑇𝑡−𝑖 + 𝑑𝑢𝑚𝑚𝑦2015 + 𝑑𝑢𝑚𝑚𝑦2018 + 𝜇𝑡 eq 26 abbas khan, muhammad yar khan and abdul qayyum khan 28 ∆𝐹𝐸𝑅𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝑛10 𝐸𝐶𝑇𝑡−𝑖 + 𝑑𝑢𝑚𝑚𝑦2013 + 𝑑𝑢𝑚𝑚𝑦2019 + 𝜇𝑡 eq 27 ∆𝐸𝐿𝐸𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝑛11𝐸𝐶𝑇𝑡−𝑖 + 𝑑𝑢𝑚𝑚𝑦2013 + 𝜇𝑡 eq 28 ∆𝐿𝐸𝑇𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝑛12 𝐸𝐶𝑇𝑡−𝑖 + 𝑑𝑢𝑚𝑚𝑦2016 + 𝑑𝑢𝑚𝑚𝑦2019 + 𝜇𝑡 eq 29 ∆𝐸𝑁𝐺𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝑛13𝐸𝐶𝑇𝑡−𝑖 + 𝑑𝑢𝑚𝑚𝑦2013 + 𝑑𝑢𝑚𝑚𝑦2019 + 𝜇𝑡 eq 30 ∆𝐾𝑆𝐸100𝑡 = 𝛽0 + ∑ 𝛽1𝑖 ∆𝐸𝐿𝐸𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽2𝑖 ∆𝐹𝐷𝐼𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽3𝑖 ∆𝐺𝑋𝑃𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽4𝑖 ∆𝑀𝐶𝑡−1 + 𝑛 𝑖=1 ∑ 𝛽5𝑖 ∆𝑊𝑃𝐼𝑡−1 𝑛 𝑖=1 + 𝑛14𝐸𝐶𝑇𝑡−𝑖 + 𝜇𝑡 eq 30 equation 18 represents the short run cointegration among the ivs and dvs. the “𝐸𝐶𝑇” is the error correction term. 𝐸𝐶𝑇 is used when there is abnormality in the data, showing how long it will take to get back into its normal position in the long-term. “𝑛1” is the coefficient of the 𝐸𝐶𝑇. the dummy variables are used due to structural breaks in the data. the same explanation is followed by equation 19 to 31. the model stability of both short-term and longterm cointegration is tested by using cusum and cusumq (evans, 1974). 5. results: this portion of the study explains the unit root testing, structural breaks in the data, model fitting, auto correlation and heteroscedasticity of the data. in table 2 provides the results of adf and pp unit root test to investigate the stationarity at i(0) and i(1). the results confirmed that all the variables are stationary at level and first order difference and a mix of both. the prerequisite of the ardl has been accepted which means all the variables are stationary at level or first difference. table 2: unit root test variables adf pp i(0) i(1) i(0) i(1) ele -6.1652*** (1) -10.6589*** (1) -6.1516*** (1) -36.2876*** (1) elep -10.5650*** (1) -9.8471*** (1) -10.5716*** (1) -104.0480*** (1) eng -2.3713 -11.1627*** (2) -3.6598 -23.1542*** (2) fer -3.3128 -6.7570*** (2) -2.5401 -6.5013*** (2) iro -3.0672 -10.2214*** (2) -3.0758 -10.2287*** (2) let -3.9652*** (1) -10.3248*** (1) -3.9652 -23.2421*** (1) non -3.6235*** (1) -13.8405*** (1) -3.3212*** (1) -16.3095*** (1) rub -3.5172*** (1) -13.5994*** (1) -3.4465*** (1) -13.6553*** (1) tex -3.6453*** (1) -13.4525*** (1) -3.5125*** (1) -14.0943*** (1) aut -0.4199 -10.2469*** (1) -0.41998 (1) -10.2469*** (1) che -2.4073 -14.0132*** (1) -3.0274*** (1) -16.261*** (1) cok -4.6401*** (2) -10.4101*** (2) -4.6198*** (2) -16.2024*** (2) foo -1.7829*** (2) -13.9188*** (2) -2.2189 -13.9077*** (2) journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 29 pap -3.2660*** (0) -12.3287*** (0) -3.0935*** (0) -13.3710*** (0) pha -3.5810***(0) -7.4020***(0) -6.9604***(0) -31.9163***(0) fdi -12.2934*** (0) -11.0515*** (0) -12.4329*** (0) -61.1246*** (0) mc -2.5674 -4.6406*** (0) -2.4973 -4.7361*** (0) wpi -2.5731 -9.0003*** (0) -2.8641 -8.9984*** (0) gxp -2.8790 -3.7623*** (0) -2.3481 -5.6263*** (0) kse100 -11.0121***(0) -9.1917***(0) -11.0387***(0) -72.8086***(0) test critical value 1% level -3.494378 -3.494378 -3.493129 -3.493747 5% level -2.889474 -2.889474 -2.888932 -2.8892 10% level -2.581741 -2.581741 -2.581453 -2.581596 note: all the regressions are based on trend and intercept, the optimal lags are selected using schwarz bayesian criterion (sci ), () is used for optimal lags, *** is p-value less than 0.05 means the null hypothesis of unit root is rejected. critical value at 1%. as discussed previously, the findings of unite root testing are ambiguous due to structural changes in the time series data. following the procedure used previously by applying bai– perron multi structural breaks test (bai & perron, 2003; balcilar et al., 2017; smith et al., 2019b; yasmeen et al., 2019). before exploring the short-term and long-term cointegration it is important to find structural changes in the data. table 3 is providing the summary of multiple structural breaks in the data under study. it is observed that all the variables include structural breaks. for aut, che, cok, ele, eng, fer, foo, non, pap and pha with one structural break, one dummy variable is added. for iro, let, rub and tex with more than one structural break, multiple dummy variables are added. lastly the kse100 index has no structural breaks in the time series data. structural breaks improve the model stability and provide more significant results for interpretation. table 3: bai–perron structural breaks in the data variables schwarz* criterion lwz* criterion aut 2018m06 2018m06 che 2015m09 2015m09 cok 2013m10 2013m10 ele 2013m09 2013m09 eng 2013m12 2013m12 fer 2012m10 2012m10 foo 2017m10 2017m10 iro 2012m07, 2017m05, 2018m08 2012m07, 2017m05, 2018m08 let 2012m04, 2013m07, 2016m08 2016m08 non 2013m07, 2014m11, 2016m03, 2018m07 2015m11, 2018m07 pap 2017m08 2017m08 pha 2017m08 2017m08 rub 2012m06, 2013m10 2012m06 tex 2012m04, 2018m07 2012m04, 2018m08 elep 2013m12 2013m12 fdi 2014m10 2018m06 mc 2013m09 2013m09 wpi 2017m10 2017m10 gxp 2017m08 2017m08 kse100 n/a n/a ** bai–perron critical values. abbas khan, muhammad yar khan and abdul qayyum khan 30 in table 4 interpreting the f-value of bound test to validate the presence of long-term cointegration against the 𝐻0 of no long-term cointegration. different structural breaks are noticed in 2013, 2015, 2016 and 2017. the reasons are pakistan is producing 64% of electricity from thermal power plant but in 2013 due to shortage of oil and gas the country lost a potential of 3000mw of electricity generation. pakistan is using imported furnace oil and gas imported from other countries (ministry of finance, 2013). in 2015-2016 circular debt is the main reason of structural breaks when the power generating companies get defaulted and failed to pay dues to the oil and gas suppliers. the power generating companies are unable to generate enough sales due to inefficient power distributing companies like (discos). power distributing companies have no control on electricity theft cases, power distribution losses and low-cost tariffs. during 2015 and 2016 the circular debt has increased from 05 billion to 06 billion (tauhidi & chohan, 2020). in 2017 onward there is a 20% depreciation in the pakistan rupees against the u.s. dollar which made the import of furnace oil more expensive and result in decrease of foreign exchange reserves from $9.9 bn to $8.1 bn within four months (simon nicholas & buckley, 2018). in each model sectoral output, kse100 index is taken as a dependent variable and electricity prices, fdi, wpi, mc and gxp as iv. the dummy variables are also added because of structural breaks. the f-statistics of bound test are interpreted to verify the long-term cointegration in the data. in table 4 the f-value of all variables are higher than upper bound which means the longterm cointegration among the variables. variables bound test cointegration f-value i(0) i(1) remarks aut 9.5155 3.79 4.85 long term cointegration che 27.1777 3.79 4.85 long term cointegration cok 58.7201 3.79 4.85 long term cointegration ele 35.1667 3.79 4.85 long term cointegration eng 22.2469 3.79 4.85 long term cointegration fer 46.9500 3.79 4.85 long term cointegration foo 20.0303 3.62 4.16 long term cointegration iro 61.3658 3.62 4.16 long term cointegration let 5.00531 3.62 4.16 long term cointegration non 6.90343 3.62 4.16 long term cointegration pap 65.4195 3.62 4.16 long term cointegration journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 31 pha 49.3819 3.62 4.16 long term cointegration rub 117.4260 3.62 4.16 long term cointegration tex 07.7853 3.62 4.16 long term cointegration elep 06.3514 3.62 4.16 long term cointegration fdi 35.1667 3.62 4.16 long term cointegration mc 61.3658 3.62 4.16 long term cointegration wpi 65.4195 3.62 4.16 long term cointegration gxp 117.4260 3.62 4.16 long term cointegration kse100 24.0643 2.56 3.49 long term cointegration note: if the value of f-stat is > than 4.16 the long-term cointegration exist, if the f-stat value is < 3.62 short-term cointegration exists and if the value is between 3.62 and 4.16, the model is inconclusive. this study has taken the sectoral production to check the long-term and short-term relationship with electricity price change in pakistan. the table 5 provides the results of longterm and short-term cointegration for all industrial sectors. in the context of pakistan, electricity is one of the major input source to produce output (yasmeen et al., 2019). it is very important to investigate the fluctuation of energy prices and its impact on the economy. the study investigated the impact using ardl models. the result in table 5 indicates that all the ivs have significant and long-term negative relationship with electricity cost. the negative relationship indicates that all the sectoral productions are exposed to the electricity price shocks. these negative results have some serious consequences. on the supply side all sectors are highly dependent on electricity price shocks. the operations of all industrial sectors are highly dependent on energy prices and negatively affect the production growth and profitability (zameer & wang, 2018). on demand side, the industries and households also affect the electricity price shocks. it increases the expenditure and reduces the purchasing power of the public which leads to reduce the unnecessary purchases and increase savings. as a result, it decreases the aggregate demand of the finished products and cutoff on the productions of the industries. electricity price is playing one of the vital positions in the growth of pakistan’s economy. in pakistan electricity generation is extremely in need of imported furnace oil (zameer & wang, 2018). preceding researches focused on the impact of oil price in industrial sectors of pakistan (yasmeen et al., 2019). another study examines the impact of oil price variation on monetary policy (a. malik, 2008). other economic variable like inflation and interest rates are checked with oil price shocks and found a positive long run association between rate of inflation and interest rate (k. malik et al., 2017). the impact of oil price shocks on the stock exchange is also investigated (najaf & najaf, 2016; waheed et al., 2018). recently, the connection between oil price shocks and trade deficits is investigated and found a positive relationship between increase in oil price and trade deficit (ahad & anwer, 2020). additionally, the electricity prices are extremely in need of oil prices abbas khan, muhammad yar khan and abdul qayyum khan 32 and suffering from a severe crisis in the last two periods. this study focuses on electricity prices and sectoral production growth in pakistan. the study on individual industrial sector is rare. all the industrial sectors have a negative significant connection with electricity price shocks in long run and short run except coke petroleum products (oil and gas products, distribution services, alternative energy resources), engineering products (construction and material, manufacturing equipment), nonmetallic mineral products (cement, ceramics, glass, and lime) and paper board (raw paper, packing, plastics and construction materials) that have an insignificant short-term relationship with electricity price shocks. table 4: long-run and short-run coefficients using ardl models. variable ardl long-term (p-value) ardl short-term (p-value) aut -0.9467(0.0000) -0.2494(0.0391) che -1.2416(0.0000) -0.4307(0.0000) cok -1.6416(0.0000) -0.3935(0.0779) ele -1.8475(0.0000) -0.3802(0.0024) eng -0.9827(0.0000) 0.0172(0.7167) fer -0.6732(0.0000) 0.3267(0.0013) foo -1.3687(0.0000) -0.3687(0.0075) iro -0.0757(0.0041) 0.9242(0.0000) let -0.3193(0.0000) 0.6727(0.0000) non -1.3263(0.0000) -0.3263(0.0619) pap -1.1882(0.0000) -0.1882(0.1975) pha -1.5588(0.0000) -0.5588(0.0007) rub -0.1220(0.0021) 0.5762(0.0000) tex -0.2362(0.0002) 0.6423(0.0000) kse100 -1.0829(0.0000) -0.0018(0.0100) note: () is the p-value which is significant at 1%, 5% and 10%. 6. robustness test of the models: the diagnostic investigation is applied to check the stability of the estimated models. following tests are used: breauch–godfrey lm test to check serial correlation with the 𝐻0 of no serial correlation, jarque-bera test to check the validity of 𝐻0 is normally distributed, breusch-pagan-godfrey test for heteroskedasticity in the model. in table 5, the findings of the lm test provide the information about the serial correlation. it is determined that the p-value is greater than 5%, which accept the 𝐻0 of no serial correlation in all the variables. the value of 𝑅2 is very high which confirms the appropriateness of the models. the results of breusch– pagan–godfrey heteroscedasticity test accepts the null hypothesis of homoscedastic for all the residuals. the results of jarque–bera test are insignificant which accept the null hypothesis of normality. after applying the diagnostic test for model’s stability. the cusum and cusumq tests are applied to test the coefficient’s constancy (brown et al., 1975; khan et al., 2020; yasmeen et al., 2019). the results of cusum and cusumq in figure 1 to 14 demonstrates all the coefficients are stable and within the boundaries of 5% significance level. summing-up all the results the models used in the study are stable, residual normally distributes and homoscedastic, having no auto correlation and free of errors. it is derived that the association between sectoral production growth and price of electricity is justified which can be interpreted and used for future policy implications. table 5: diagnostic test results for all models variables lm test: 𝑅2 (f-stat) heteroscedasticity normality journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 33 aut 0.067951(0.7943) 0.92 10.00663(0.0008) 0.131973(0.3884) 65.7885(0.3210) che 0.022234(0.1428) 0.81 0.991776(0.0000) 6.011307(0.4219) 92.1418(0.1641) cok 3.36237(0.1862) 0.96 2.722422(0.0000) 1.127679(0.0614) 56.272(0.1821) ele 7.981973(0.0568) 0.95 1.856718(0.0000) 10.69740(0.0982) 41.6351(0.1321) eng 1.813207(0.6121) 0.82 10.34549(0.0000) 0.007015(0.1486) 31.0100(0.1131) fer 0.879715(0.6441) 0.81 2.962061(0.0106) 0.001733(0.9677) 85.7640(0.0841) foo 3.171740(0.2048) 0.87 3.429027(0.0025) 2.156707(0.9507) 46.6993(0.0731) iro 0.128622(0.7199) 0.94 326.4993(0.0000) 12.41041(0.0596) 219.7135(0.1531) let 1.240154(0.5379) 0.89 24.11097(0.0000) 23.46367(0.3818) 44.9160(0.32110 non 2.632324(0.2682) 0.95 2.694657(0.0136) 11.25665(0.1278) 60.7815(0.2001) pap 2.353056(0.3083) 0.79 1.364547(0.0000) 1.714686(0.6337) 44.9160(0.1321) pha 7.092175(0.2880) 0.88 15.15688(0.0000) 15.96735(0.1200) 81.2668(0.1231) rub 0.079018(0.9613) 0.90 66.39520(0.0000) 67.75367(0.8654) 44.3294(0.3214) tex 0.156421(0.1564) 0.73 28.08270(0.0000) 8.548380(0.2006) 37.4473(0.2131) kse100 0.4261(0.5155) 0.80 03.1029(0.0037) 0.003415 (0.6848) 0.0413(0.9795) note: the p-values are in closed in (), lm test for serial correlation, if the p-value > 0.05 the 𝐻0 of no serial correlation is accepted. jarque–bera test if p-value > 0.05 the 𝐻0 of normality is accepted, breusch–pagan–godfrey heteroscedasticity test. if critical value > 0.05 the 𝐻0 is the data is homoscedastic. figure 2 cusum and cusumq for automobiles industry model abbas khan, muhammad yar khan and abdul qayyum khan 34 figure 3 cusum and cusumq for chemical industry model. figure 4 cusum and cusumq for coke petroleum products industry model. figure 5 cusum and cusumq for electronics industry model. journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 35 figure 6 cusum and cusumq for engineering products industry model. figure 7 cusum and cusumq for fertilizers industry model. figure 8 cusum and cusumq for food, beverages tobacco industry model. abbas khan, muhammad yar khan and abdul qayyum khan 36 figure 9 cusum and cusumq for iron steel products industry model. figure 10 cusum and cusumq for leather products industry model. figure 11 cusum and cusumq for nonmetallic mineral products industry model. journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 37 figure 12 cusum and cusumq for paper board industry model. figure 13 cusum and cusumq for pharmaceuticals industry model. figure 14 cusum and cusumq for rubber products industry model. abbas khan, muhammad yar khan and abdul qayyum khan 38 figure 15 cusum and cusumq for textile industry model. figure 16 cusum and cusumq for kse100 model. 7. conclusion: the growth of sectoral production and electricity price fluctuation is investigated in the context of pakistan. the single sector investigation is insufficient to provide full knowledge about the economy. the impact of electricity prices on sectoral production is more beneficial than the aggregate level of production. the study utilized the multifactor ardl approach to investigate the long-term connection between cost of electricity and growth in sectoral production. the results found all the sectors have a long-term negative relationship with electricity price shocks. however, two of the sectors have an insignificant short-term relationship with electricity prices. the electricity prices affect all the economic sectors both on supply and demand side. the electricity price shocks increase the production cost and decrease production of goods supply. the increase in electricity prices reduces the income level and reduce the overall demand for consumption of goods produced. the coke, petroleum products and electronics industries are most negatively affected by electricity price shocks. the coke, petroleum industry produces the alternative sources of energies in pakistan and mostly the electricity generated is dependent on fossil fuels and the price relationship go parallel (rehman & deyuan, 2018). source of energy is the main driver behind every industry and if production journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 39 hampers consequently the whole economy is disturbed. the electronics industry production is affected due to demand driven link with higher electricity prices. if the electricity prices go up the demand of heavy-duty machinery and household equipment goes down which will lead to decrease in the production of this industry (yasmeen et al., 2019). the production of pharmaceutical industry, chemical industry, textile industry and leather industry have decreased due to higher cost and unavailability of power supply (m. s. ali et al., 2017). the agriculture sector of production has decreased due to increase in irrigation cost that ultimately affect the production of fertilizer industries (shahbaz, 2015). the findings are consistent with previous study as the kse100 index has a significant negative relationship with electricity price shocks (s. sarwar et al., 2018). in pakistan mostly electricity is generated by using furnace oil but due to increase in the cost of oil leads to increase in the cost of electricity which is utilized by machineries to produce food items and non-food item rubber, automobile parts, non-metallic and paper products industries. it enforces the industries to cutoff on production (m. n. sarwar et al., 2020). various factors are investigated in relationship with the production of cement and steel industries in pakistan. increase in the electricity price is weighted 70% in relationship with cement and steel production (ahmad et al., 2018). the results of the study match the current state of pakistan’s economy. it is inferred that additional to oil price the electricity price is another barrier to the industrial production growth and higher returns on equity. the study proposed different policy implications for government of pakistan. the results revealed the upsurge in electricity prices disturb the growth of sectoral production which also have a negative impact on equity market return. the study suggests a moderate monetary policy to overcome the negative impact of high electricity price. monetary policy should be neither higher expansionary (leading to inflationary situation) nor higher contractionary (causing growth reduction). the solution is temporary for the short-term in case the electricity prices are persistently high. in long-term the involvement of the government is increased. the following strategies may be taken by the government to overcome the negative impact of electricity price on sectoral production and equity market growth. save energy programs may be initiated by the government to bring awareness in the general public on how to reduce energy wastage and the government should provide incentives to the general public who successfully manage to reduce electricity wastages. (hille et al., 2019). private sectors should be encouraged to invest in renewable energy projects and government should offering tax reliefs to boost the usages of biofuels and renewable energies. to minimize electricity losses government should upgrade electricity transmission lines and power grids. strict custom duties shall be applied on the import of heavy electronic appliances and the import duties on solar cells and other equipment shall be reduces. further, the government should use a mix kind of energy policy with greater weightage to renewable energies. government should invest in exploring new reservoirs of crude oil and natural gas. it will reduce the cost of oil import and the outflow of foreign currency for generating electricity (hanif, 2020). consequently, increasing the supply of energy to the industries and protects from electricity price shocks (simon nicholas & buckley, 2018). it is proposed that government should increase investment in electricity generation projects to provide lower cost electricity input to the industries. considering the findings, the study provides some suggestions to policy maker to control negative impact of electricity price shocks on sectoral production. the study is limited to electricity prices which can be further extended abbas khan, muhammad yar khan and abdul qayyum khan 40 in relationship with other major energy prices i.e., crude oil and natural gas. a cross-sectional study may be encouraged to identify the generalized impact of energy prices on macroeconomic variable in developed and developing countries with advanced renewable energy and shale gas production like u.s. journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 41 reference abeberese, a. b. (2017). electricity cost and firm performance: evidence from india. 99(december), 839–852. https://doi.org/10.1162/rest ahad, m., & anwer, z. (2020). asymmetrical relationship between oil price shocks and trade deficit: evidence from pakistan. the journal of international trade & economic development, 29(2), 163– 180. ahmad, m. h., badrashi, y. i., ahad, m. z., khan, z., & khan, f. a. (2018). factors affecting material management in construction industry of khyber pakhtunkhwa pakistan. int j adv res sci, eng technol, 5(1), 7249–7258. ahmed, k., ahmed, n., & ramzan, m. (2018). decomposing the links between oil price shocks and macroeconomic indicators : evidence from saarc region. resources policy, march, 0–1. https://doi.org/10.1016/j.resourpol.2018.03.001 akadiri, s., alola, s. and, williams, a. a. and o., etokakpan, g. and, & udom, m. (2019). the role of electricity consumption , globalization and economic growth in carbon dioxide emissions and its implications for environmental sustainability targets. science of the total environment, 708, 134653. https://doi.org/10.1016/j.scitotenv.2019.134653 ali, a., imtiaz, m., & others. (2019). effects of pakistan’s energy crisis on farm households. utilities policy, 59, 100930. ali, m. s., zaigham, s., & others. (2017). energy crisis and comparative advantage industries: empirical evidence from the pakistan economy. global economics review, 2(1), 42–48. allcott, h., collard-wexler, a., & o’connell, s. d. (2016). how do electricity shortages affect industry? evidence from india. american economic review, 106(3), 587–624. https://doi.org/10.1257/aer.20140389 alola, a. a., & yildirim, h. (2019). the renewable energy consumption by sectors and household income growth in the united states. international journal of green energy, 1–8. alter, n., & syed, s. h. (2011). an empirical analysis of electricity demand in pakistan. international journal of energy economics and policy, 1(4), 116–139. amjad chaudhry, a. (2010). a panel data analysis of electricity demand in pakistan. the lahore journal of economics, 15(special edition), 75–106. https://doi.org/10.35536/lje.2010.v15.isp.a5 amusa, h., amusa, k., & mabugu, r. (2009). aggregate demand for electricity in south africa: an analysis using the bounds testing approach to cointegration. energy policy, 37(10), 4167–4175. ayres, r. u., den bergh, j. c. j. m., lindenberger, d., & warr, b. (2013). the underestimated contribution of energy to economic growth. structural change and economic dynamics, 27, 79– 88. bai, j., & perron, p. (2003). critical values for multiple structural change tests. the econometrics journal, 6(1), 72–78. https://doi.org/10.1111/1368-423x.00102 balcilar, m., bekun, f. v., & uzuner, g. (2019). revisiting the economic growth and electricity consumption nexus in pakistan. environmental science and pollution research, 26(12), 12158– 12170. balcilar, m., bouri, e., gupta, r., & roubaud, d. (2017). can volume predict bitcoin returns and volatility? a quantiles-based approach. economic modelling, 64(march), 74–81. https://doi.org/10.1016/j.econmod.2017.03.019 bank, w. (2015). pakistanenterprise survey 2013. 1–182. bekhet, h. a., & matar, a. (2013). co-integration and causality analysis between stock market prices and their determinates in jordan. economic modelling, 35, 508–514. abbas khan, muhammad yar khan and abdul qayyum khan 42 biggeri, l., ferrari, g., & zhao, y. (2017). estimating cross province and municipal city price level differences in china: some experiments and results. social indicators research, 131(1), 169–187. bildirici, m. e., bakirtas, t., & kayikci, f. (2012). economic growth and electricity consumption: auto regressive distributed lag analysis. journal of energy in southern africa, 23(4), 29–45. blignaut, j., inglesi-lotz, r., & weideman, j. p. (2015). sectoral electricity elasticities in south africa: before and after the supply crisis of 2008. south african journal of science, 111(9–10), 1–7. bohi, d. r. (2017). energy price shocks and macroeconomic performance. routledge. borjigin, s., yang, y., yang, x., & sun, l. (2018). econometric testing on linear and nonlinear dynamic relation between stock prices and macroeconomy in china. physica a: statistical mechanics and its applications, 493, 107–115. brown, r. l., durbin, j., & evans, j. m. (1975). techniques for testing the constancy of regression relationships over time. journal of the royal statistical society: series b (methodological), 37(2), 149–163. capros, p., paroussos, l., charalampidis, i., fragkiadakis, k., karkatsoulis, p., & tsani, s. (2016). assessment of the macroeconomic and sectoral effects of higher electricity and gas prices in the eu: a general equilibrium modeling approach. energy strategy reviews, 9, 18–27. ciarreta, a., & zarraga, a. (2010a). economic growth-electricity consumption causality in 12 european countries: a dynamic panel data approach. energy policy, 38(7), 3790–3796. ciarreta, a., & zarraga, a. (2010b). electricity consumption and economic growth in spain. applied economics letters, 17(14), 1417–1421.` danmaraya, i. a., & hassan, s. (2016). electricity consumption and manufacturing sector productivity in nigeria: an autoregressive distributed lag-bounds testing approach. international journal of energy economics and policy, 6(2), 195–201. destek, m. a., & aslan, a. (2017). renewable and non-renewable energy consumption and economic growth in emerging economies: evidence from bootstrap panel causality. renewable energy, 111, 757–763. dickey, d. a., & fuller, w. a. (1979). distribution of the estimators for autoregressive time series with a unit root. journal of the american statistical association, 74(366a), 427–431. dogan, e. (2015). the relationship between economic growth and electricity consumption from renewable and non-renewable sources: a study of turkey. renewable and sustainable energy reviews, 52, 534–546. econometrics, a. (2015). co-integration and error correction: representation, estimation, and testing. applied econometrics, 39(3), 106–135. engle, granger, r. f. and, & wj, c. (1987). co-integration and error correction : representation , estimation , and testing. econometrica: journal of the econometric society, 55(2), 251–276. engle, r. f., ghysels, e., & sohn, b. (2013). stock market volatility and macroeconomic fundamentals. review of economics and statistics, 95(3), 776–797. evans, m. (1974). techniques for testing the constancy of regression relationships over time. 1962, 149–163. https://doi.org/10.1111/j.2517-6161.1975.tb01532.x faisal, f., tursoy, t., & ercantan, o. (2017). the relationship between energy consumption and economic growth: evidence from non-granger causality test. procedia computer science, 120(december), 671–675. https://doi.org/10.1016/j.procs.2017.11.294 fama, e. f. (1981). stock returns, real activity, inflation, and money. the american economic review, 71(4), 545–565. fama, e. f. (1990). stock returns, expected returns, and real activity. the journal of finance, 45(4), journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 43 1089–1108. fei, q., rasiah, r., & leow, j. (2014). the impacts of energy prices and technological innovation on the fossil fuel-related electricity-growth nexus: an assessment of four net energy exporting countries. journal of energy in southern africa, 25(3), 37–46. fisher-vanden, k., mansur, e. t., & wang, q. j. (2015). electricity shortages and firm productivity: evidence from china’s industrial firms. journal of development economics, 114, 172–188. https://doi.org/10.1016/j.jdeveco.2015.01.002 garg, b., & prabheesh, k. p. (2020). testing the intertemporal sustainability of current account in the presence of endogenous structural breaks: evidence from the top deficit countries. economic modelling, december 2019. https://doi.org/10.1016/j.econmod.2020.04.007 ghali, k. h., & el-sakka, m. i. t. (2004). energy use and output growth in canada: a multivariate cointegration analysis. energy economics, 26(2), 225–238. girardin, e., & joyeux, r. (2013). macro fundamentals as a source of stock market volatility in china: a garch-midas approach. economic modelling, 34, 59–68. gonese, d., hompashe, d., & sibanda, k. (2019). the impact of electricity prices on sectoral output in south africa from 1994 to 2015. african journal of economic and management studies, 10(2), 198–211. government, p. (2020). planning commission report, ministry of planning development & reform government of pakistan. https://www.pc.gov.pk/uploads/vision2025/pakistan-vision-2025.pdf grainger, c. a., & zhang, f. (2017). the impact of electricity shortages on firm productivity: evidence from pakistan. the world bank. grainger, c. a., & zhang, f. (2019). electricity shortages and manufacturing productivity in pakistan. energy policy, 132, 1000–1008. grim, r. g., huang, z., guarnieri, m. t., ferrell, j. r., tao, l., & schaidle, j. a. (2020). transforming the carbon economy: challenges and opportunities in the convergence of low-cost electricity and reductive co2 utilization. energy and environmental science, 13(2), 472–494. https://doi.org/10.1039/c9ee02410g guo, z., zhang, x., wang, d., & zhao, x. (2019). the impacts of an energy price decline associated with a carbon tax on the energy-economy-environment system in china. emerging markets finance and trade, 1–14. hanif, m. (2020). relationship between oil and stock markets: evidence from pakistan stock exchange. international journal of energy economics and policy, 10(5), 150. hasan, a., zaman, a., sikder, z. i., & wadud, a. (2017). the dynamics of electricity consumption , energy use and gdp in bangladesh. 65. hille, s., weber, s., & brosch, t. (2019). consumers’ preferences for electricity-saving programs: evidence from a choice-based conjoint study. journal of cleaner production, 220, 800–815. inglesi-lotz, r., & blignaut, j. n. (2011). south africa’s electricity consumption: a sectoral decomposition analysis. applied energy, 88(12), 4779–4784. jamil, f., & ahmad, e. (2010). the relationship between electricity consumption, electricity prices and gdp in pakistan. energy policy, 38(10), 6016–6025. jo, s. (2014). the effects of oil price uncertainty on global real economic activity. journal of money, credit and banking, 46(6), 1113–1135. johamen, s., & jtiselius, k. (1990). maximum likelihood estimation and inference on cointegration— with appucations to the demand for money. 52(2), 169–210. johansen, s. (1988). statistical analysis of cointegration vectors. journal of economic dynamics and abbas khan, muhammad yar khan and abdul qayyum khan 44 control, 12(2–3), 231–254. jorgenson, d. w. (1984). the role of energy in productivity growth. the energy journal, 5(3), 11–26. khan, m. k., khan, m. i., & rehan, m. (2020). the relationship between energy consumption, economic growth and carbon dioxide emissions in pakistan. financial innovation, 6(1), 1–13. khobai, h. (2018). electricity consumption and economic growth: a panel data approach for brazil, russia, india, china and south africa countries. international journal of energy economics and policy, 8(3), 283–289. kümmel, r. (1982). the impact of energy on industrial growth. energy, 7(2), 189–203. laurenceson, j., & chai, j. c. h. (2003). financial reform and economic development in china. edward elgar publishing. magnet, b. (2013). a monthly magazine of islamabad chamber of commerce & industry. malik, a. (2008). crude oil price, monetary policy and output: the case of pakistan. the pakistan development review, 47(4ii), 425–436. malik, k., ajmal, h., & zahid, m. u. (2017). oil price shock and its impact on the macroeconomic variables of pakistan: a structural vector autoregressive approach. malik, kz, ajmal h, zahid m u.," international journal of energy economics and policy, 7(5), 83–92. malik, s., qasim, m., saeed, h., chang, y., & taghizadeh-hesary, f. (2020). energy security in pakistan : perspectives and policy implications from a quantitative analysis. energy policy, 144(june 2019), 111552. https://doi.org/10.1016/j.enpol.2020.111552 marza, m., & daly, s. s. (2018). impact of oil price fluctuations on human development: a standard study of iraq. 5, 396–399. https://doi.org/https://doi.org/10.32861/jssr.spi5.396.399 masuduzzaman, m. (2013). electricity consumption and economic growth in bangladesh: cointegration and causality analysis. research study series no. fdrs, 2, 2013. ministry of finance. (2013). pakistan economic survey 2012-2013 chapter-14 energy. 187–188. http://www.finance.gov.pk/survey/chapters_13/14-energy.pdf najaf, r., & najaf, k. (2016). an empirical study on the dynamic relationship between crude oil prices and pakistan stock market. j. account mark, 5(194), 2. narayan, p. k. (2007). the saving and investment nexus for china : evidence from cointegration tests the saving and investment nexus for china : evidence from cointegration tests. october 2012, 37– 41. https://doi.org/10.1080/00036840500278103 naser, h. (2019). oil price fluctuation, gold returns and inflationary pressure: an empirical analysis using cointegration approach. applied economics and finance, 6(2), 71. https://doi.org/10.11114/aef.v6i2.4054 nazlioglu, s., gormus, a., & soytas, u. (2019). oil prices and monetary policy in emerging markets: structural shifts in causal linkages. emerging markets finance and trade, 55(1), 105–117. https://doi.org/10.1080/1540496x.2018.1434072 ozturk, i. (2010). a literature survey on energy--growth nexus. energy policy, 38(1), 340–349. paija, n. (2019). energy security , electricity , population and economic growth : the case of a developing south asian resource-rich economy. energy policy, 132(july 2018), 771–781. https://doi.org/10.1016/j.enpol.2019.05.054 peiro, a. (2016). stock prices and macroeconomic factors: some european evidence. international review of economics & finance, 41, 287–294. pesaran, m hashem. (1997). the role of economic theory in modelling the long run. the economic journal, 107(440), 178–191. journal of applied economics and business studies, volume. 5, issue 1 (2021) 17-46 https://doi.org/10.34260/jaebs.512 45 pesaran, m hashem, & shin, y. (1998). an autoregressive distributed-lag modelling approach to cointegration analysis. econometric society monographs, 31, 371–413. pesaran, mohammad hashem, & pesaran, b. (1997). working with microfit 4.0: interactive econometric analysis;[windows version]. applied economics and finance, 6(2). phillips, p. c. b., & perron, p. (1988). testing for a unit root in time series regression. biometrika, 75(2), 335–346. polemis, m. l., & dagoumas, a. s. (2013). the electricity consumption and economic growth nexus: evidence from greece. energy policy, 62, 798–808. pönkä, h., & zheng, y. (2019). the role of oil prices on the russian business cycle. research in international business and finance, 50(january), 70–78. https://doi.org/10.1016/j.ribaf.2019.04.011 rehman, a., & deyuan, z. (2018). investigating the linkage between economic growth, electricity access, energy use, and population growth in pakistan. applied sciences, 8(12), 2442. rosenberg, n. (1983). the effects of energy supply characteristics on technology and economic growth. energy, productivity, and economic growth. cambridge, ma: oelgeschlager, gunn and hain. sarwar, m. n., hussain, h., & maqbool, m. b. (2020). pass through effects of oil price on food and nonfood prices in pakistan: a nonlinear ardl approach. resources policy, 69, 101876. sarwar, s., shahbaz, m., anwar, a., & tiwari, a. k. (2019). the importance of oil assets for portfolio optimization: the analysis of firm level stocks. energy economics, 78, 217–234. https://doi.org/10.1016/j.eneco.2018.11.021 sarwar, s., waheed, r., amir, m., khalid, m., & others. (2018). role of energy on economy the case of micro to macro level analysis’’. economics bulletin, 38(4), 1905–1926. sbp. (2019). the state of pakistan ’ s economy: the third quartely report. http://www.sbp.org.pk/reports/quarterly/fy19/third/qtr-index-eng.htm schwab, k. (2018). the global competitiveness report 2018. world economic forum, 9–14. shahbaz, m. (2015). measuring economic cost of electricity shortage: current challenges and future prospects in pakistan. shahbaz, m., chaudhary, a. r., & ozturk, i. (2017). does urbanization cause increasing energy demand in pakistan? empirical evidence from stirpat model. energy. https://doi.org/10.1016/j.energy.2017.01.080 shahbaz, m., & lean, h. h. (2012). the dynamics of electricity consumption and economic growth: a revisit study of their causality in pakistan. energy, 39(1), 146–153. shin, y., & smith, r. j. (2001). bounds testing approaches to the analysis. 326(february 1999), 289–326. https://doi.org/10.1002/jae.616 simon nicholas, & buckley, t. (2018). pakistan’s power future report 2018. institute for energy economics and financial analysis, december. http://ieefa.org/wpcontent/uploads/2018/11/pakistans-power-future_december-2018.pdf smith, s. c., timmermann, a., & zhu, y. (2019a). variable selection in panel models with breaks. journal of econometrics, 212(1), 323–344. https://doi.org/10.1016/j.jeconom.2019.04.033 smith, s. c., timmermann, a., & zhu, y. (2019b). variable selection in panel models with breaks. journal of econometrics, 212(1), 323–344. solangi, tan, y. a. and, khan, q. and, mirjat, m. w. a. and, ahmed, n. h. and, & ifzal. (2018). the selection of wind power project location in the southeastern corridor of pakistan: a factor analysis, ahp, and fuzzy-topsis application. energies, 11(8), 1940. https://doi.org/10.3390/en11081940 abbas khan, muhammad yar khan and abdul qayyum khan 46 taghizadeh-hesary, f., rasoulinezhad, e., & yoshino, n. (2019). energy and food security : linkages through price volatility. energy policy, 128(august 2018), 796–806. https://doi.org/10.1016/j.enpol.2018.12.043 taghizadeh-hesary, f., yoshino, n., hossein, m. m., & farboudmanesh, r. (2015). response of macro variables of emerging and developed oil importers to oil price movements. 7860(october). https://doi.org/10.1080/13547860.2015.1057955 tauhidi, a., & chohan, u. w. (2020). the conundrum of circular debt. the world bank. (2013). pakistan country profile 2013. in world bank enterprice surveys. https://www.enterprisesurveys.org/en/data/exploreeconomies/2013/pakistan waheed, r., wei, c., sarwar, s., & lv, y. (2018). impact of oil prices on firm stock return : industrywise analysis. empirical economics, 55(2), 765–780. https://doi.org/10.1007/s00181-017-1296-4 wakeel, m., chen, b., & jahangir, s. (2016). overview of energy portfolio in pakistan. energy procedia, 88, 71–75. https://doi.org/10.1016/j.egypro.2016.06.024 wesseh, p. k., & lin, b. (2018). exchange rate fluctuations, oil price shocks and economic growth in a small net-importing economy. energy. https://doi.org/10.1016/j.energy.2018.03.054 wolde-rufael, y. (2014). electricity consumption and economic growth in transition countries: a revisit using bootstrap panel granger causality analysis. energy economics, 44, 325–330. wu, c.-f., wang, c.-m., chang, t., & yuan, c.-c. (2019). the nexus of electricity and economic growth in major economies: the united states-india-china triangle. energy, 188, 116006. yasmeen, h., wang, y., zameer, h., & solangi, y. a. (2019). does oil price volatility influence real sector growth? empirical evidence from pakistan. energy reports, 5, 688–703. yasmin, b., & qamar, w. (2013). the role of power generation and industrial consumption uncertainty in de-industrialising pakistan. the pakistan development review, 517–534. zameer, h., & wang, y. (2018). energy production system optimization : evidence from pakistan. renewable and sustainable energy reviews, 82(march 2016), 886–893. https://doi.org/10.1016/j.rser.2017.09.089 zhang, c., zhou, k., yang, s., & shao, z. (2017). on electricity consumption and economic growth in china. renewable and sustainable energy reviews, 76, 353–368. here comes the title of the paper journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 55 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x cross-cultural issues and human resources management in bric countries muhammad aiman awalluddin*1 1 phd scholar, sakarya university, turkey abstract the brics market is a significant economic block, accounting for over 40% of the world population and over 20% of global gdp. international companies are eager to do business. cross-cultural issues, however, can be barriers in human resource management (hrm) and slow the process to achieve credibility among local employees, particularly if western approach is adopted among local employees. this study focuses on exploring cross-cultural problems in emerging markets, such as brazil, russia, india, and china, or infamously known as bric, using hofstede 's expertise to recognise local values that influence hrm periscope and recommend hrm approaches that can be taken to meet local employee standards. from the study, multinational companies are recommended to use the rbv strategy to achieve competitive advantage and local integration. by highlighting the descriptive analysis of the systematic literature approach and using web tool research, a critical assessment of findings shows that each bric country has its specific culture and challenge embedded in local culture. therefore, fully understanding local values are very useful in coping with contrasts and spot opportunities. keywords bric, crosscultural, human resource management, national dimension, competitive advantage jel classification m12, m14, m54 1. introduction with the rapid increase of the globalization phenomenon, the business has moved beyond country boundaries, which saw significant emergence of multi-national * muhammed.awaluddin@ogr.sakarya.edu.tr muhammad aiman awalluddin 56 companies that operated beyond their homeland. labor forces become more diverse, and hrm managers are facing growing problems. with the rising importance of emerging economies in the global business climate, the management of cross-cultural ties faces more challenge. growing rivalry both domestically and globally has led to a greater focus on human resource management. for example, recent hrm developments, such as cultural disparities in the workplace, cross-cultural business conditions, global talent management and a growing emergence of virtual international teams, have made management of the workforce more complex. these developments have made a manager's job even more complicated than ever. problems of cultural adaptation have become inevitable with the implementation of global business guidelines. cross-cultural analysis explores parallels and cultural differences. crosscultural studies analyse and compare hrm activities in the sense of hrm, including the selection, training, assessment and retention of workers from all cultures and nations (sander, a.cogin & bainbridge, 2014). cross-cultural management and ethics are used to avoid and address current issues (greblikaite and daugeliene, 2010). hrm’s diversity will draw on cultural diversity in the company and prevent cross-cultural misunderstandings (fujimoto, bahfen, fermelis and härtel, 2007). besides, each country has its institution, values, beliefs, social structure, and norms, all of which have a direct impact on hrm practice. the success of hrm practice in western counterparts may become inconsistent in the eastern counterpart due to cultural differences. therefore, cross-cultural management and ethics are growing due to several obvious reasons, mainly due to economic globalization, which driven business organizations are seeking out a new market to maximize their profit. rapidly growing and volatile economics of certain asian and latin american countries had become a new promise land for the capitalism based firm. the promise of a vast potential market that has the characteristic of a developed market offers business firm potentially high growth performance, which saw lots of them pouring money into the emerging market. the objective of this study to investigate the cross-cultural issues in the emerging market, such as brazil, russia, india, and china, or infamously known as bric using hofstede's insight, which may opposite with western hrm approach. it aims to seek better understanding and provide hindsight on cross-cultural management to interested parties such as academicians, hr managers, and corporate bodies. journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 57 2. cross-cultural management the origin of the problem, confusion and disagreement in social participation is often regarded as cultural differences. the question of cultural distance constitutes a barrier and, consequently, difficulties in working and doing business across national boundaries, thereby raising the gap between persons, groups and organizations (shenkar, 2001). many problems in regional and global business contexts have been addressed in terms like "foreign," "costing of unfamiliarity" and "institutional holes" (stahl, miska, lee & de luque, 2017). the main causes of incompatibility, discord and confrontation in the business community were gaps, distances and diversity. a crosscultural definition is a type of interactivity between different cultural groups, which incorporates various cultural inputs to resolve differences that can hamper communication, taking into account interaction between people of different cultural backgrounds during organizations. (thomas d. c., 2008). the challenges of crossculture have created problems associated with the most appropriate business arrangements and human resource practice. the very conflict raised by this emphasizes the importance of ongoing examination of cross-cultural management practices to encourage the firm to establish business and human resource practices that, while globally strategic, are also sympathetic to specific markets (fish, 1994). generalizing by assuming one culture suits another culture is irrelevant; hence particular culture requires particular settings to be tended. diversity-oriented hrm is essential to improve social justice through the creation of an organization –, in which no-one is privileged or disadvantaged. it will incorporate and align diversity across all the organizational systems and identify critical gaps in the promotion of social justice and integration. organizational targets can be achieved by understanding and redesigning diversity issues related to each traditional practice , in order to attract and conserve diverse people (meena & vanka, 2017). table 1: significant of cross-cultural management studies authors hrm practice/theory significant (andreassi, lawter, brockerhoff, & j. rutigliano, 2014) (mihaela, 2014) (chen, chao, xie, & tjosvold, 2018) hofstede's cultural dimension • the significant relationship between job characteristic on job satisfaction • impact on leadership style • recognizing eastern values can manage conflict cooperatively (sakikawa, chaudhuri, & arif, 2017) high-performance work practice • better performance varies in culture muhammad aiman awalluddin 58 (gahlawat & kundu, 2018) contingency approach (selection and compensation) • positive impact on firm performance, in compare to universality practice in western countries (latukha & malko, 2019) recruitment, training and development, motivation and performance management • commonwealth of independents state had different hr practice in compare with western countries (abugre, 2018) cross-cultural communication • different communicative behavior between western and sub-saharan africa (ying zhang, wei, & zhou, 2017) conflict management • conflict management and the role of cultural (weber, sadri, & gentry, 2018) (malik & singh, 2017) leadership • culture impacts on the relationship between a leader's diversity belief and performance rating. • transformational leadership affect the socialization processes of cultural minorities (ashta, stokes, & hughes, 2018) change management • significant of spiritual value in change management source : author cross-cultural management problems are becoming very important for a manager to address the complex problems arising from cultural differences. cross-cultural management describes organizational behavior, in which the organizational behavior of individual countries and religions is compared and how the interactions of coworkers, managers, managers, customers, suppliers and alliances partners are to be improved. culture has often been understood for more than two decades as a fixed attitude and behavioral pattern, based on the hofstede (1980) cultural dimension of transcultural management. geert hofstede’s approach has dominated cross-cultural management studies and is also a reflection of the general trend in cross-cultural research (stahl & tung, 2015). although some criticisms of the study of the cultural dimension are still used in the intercultural management (shenkar, 2001; zhu & bargiela-chiappini, 2013). the recent articles which started to look at the critical aspect of cross-cultural management such as in table 1 and situational culture learning (sicula) (zhu & bargiela-chiappini, 2013), enriching cross-cultural management studies and close the gaps in this field. 3. bric markets the bric market is an acronym for the emerging economy of brazil, russia, india, and china, expected to be the fastest growing economies in the world. by 2050, it has been estimated that these economies will be richer than any of the main economies of the g-7 (us, japan, germany, france, italy, canada and the uk). it is also has been highlighted by (keukeleire & hooijmaaijers, 2014) a shift in the negotiating power. the journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 59 bric countries postulated to become a new aggressive world demand growth and spending power (wilson & purushothaman, 2003) and also perceived to become future world power (hart & jones, 2010). in 2011, the term was modified by included south africa. since 2009, the four-member have been increasingly working together as an alliance, and several prominent studies have been conducted by highlighting the economics, foreign direct investment, political convergence and importance of these countries (armijo, 2007; gattai & natale, 2016; glosny, 2010; rodriguez-arango & gonzalez-perez, 2016; a. singh & singh, 2017). the brics are seen as a market-changing world as two of its members having a population of more than 2.7 billion people or 36.28% of the world population. together, with the five countries, the brics countries make up a sizeable portion of the world's population with an estimation of over 3 billion people (42% of the global population) (united nations, 2017). moreover, they economically incorporate approximately a third of the global ppp gross domestic product, which is the same as the g-7. the brics accounted for approximately 56 percent of global gnp (2005 pp) growth during 20082017 and is projected to continue taking into account more than half of global economic growth by 2030. the brics are emerging as gravity of international financial system, and will possibly remain the core source of economic dynamism. the brics 'increasing contribution to the global economy and the rising value of brics economic relationships generates the potential for better share of the economy and reduces inequality in the global economic climate. 4. conceptual framework 4.1. hofstede cultural dimension cross-cultural studies have been considered increasingly important for human resource management. cultural values can be the most important explanations of and influence employee behavior. national culture has long been recognized as a critical characteristic underlying systematic differences in management behavior (steenkamp, 2001). national cultural norms and beliefs that influence the actions, perception and characteristics of people are highly influential. in cultural implications, geert hofstede coined the national cultural model or the hofstede national culture model (hofstede, 1980). the original hofstede model presented in that paper included four elements of national culture study and understanding: individualism-collectivism, masculinityfemininity, power distance and avoidance of uncertainties. but thereafter, two more dimensions were added: the long-term orientation and indulgence axis. the hofstede cultural classification is according to (kirkman, lowe, & gibson, 2006), which has muhammad aiman awalluddin 60 influenced thousands of empirical studies and is the best structure for national culture in business literature. the importance of the cultural component of hofstede lies in the consideration of a specific country’s behaviors and standards. when a country's quirks are not correctly recognized, it will present the critical challenges to the manager. potential differences that may arise from cultural aspects should be appropriately understood and integrated within the corporation. by acknowledging cultural differences, any tense and dissatisfaction among employees can be reduced. a studies done by (andreassi, lawter, brockerhoff, & j. rutigliano, 2014; eskildsen, kristensen, & gjesing antvor, 2010) show that national culture impacted job satisfaction among employees. the cultural dimensions may also explain about leadership (mihaela, 2014), work-balanced and conflict (farivar, cameron, & yaghoubi, 2016), the linkage between business strategies, cultures and compensation (verma & sharma, 2019), and moral reasoning in making the decision (wilhelm & gunawong, 2016). 4.2. human resources management in cross-culture context hr traditionally described staff management within the company. hr tasks fall under the following scopes: recruiting and selection, training and development, rewards and benefits, motivation and retention. also, hrm is a systematic approach to handling people and work culture. efficient human resource management allows an organization’s employees to contribute efficiently to the overall objectives and priorities of the organization. organizational goals can be accomplished if hrm operates according to local standards and values. for example, cross-cultural management in china involves preparation, intercultural communication systems and consideration of unified organizational culture (zhang, 2009). the strong relationship between people and an organization’s goals means hrm’s significance in helping workers achieve full individual growth, productive jobs, and linking workers to employers. hrm further harmonizes employee-employee relationships and efficient human resource models as opposed to tangible assets not influenced by values and social norms. hrm functions also play major roles from the selection and recruiting and boarding process to ensure that workers are satisfied within the company. according to (adler, 1997) managerial principles influence any aspect of organizational behavior, which has a profound effect on administrative costs. in addition, the cultural gap between home and host country will significantly influence the collection, training and performance management system (dong & liu, 2010). this situation suggested that, through practice or action, cross-cultural management should be the key priority for the hrm functions. journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 61 people have always been difficult, and handling people has always been tricky. effective hrm is seen as implementing best practices; unsuccessful hrm is a result of failing to enforce best practices (pfeffer, 1998). contingency school, however, suggests combining hrm systems and facilities with certain essential factors (gibb, 2001). defenders of universalism note that hr activities should be viewed as a collection of hrm practices that can enhance efficiency regardless of location and scale. critics also contend that the "best practice" paradigm is based on the substantial variance in hrm activities that relay tangible and intangible assets and appear to disregard employee preferences (armstrong, 2012). the principle of best practice seems arbitrary as what is best for one country does not apply to another. on the other hand, the approach to contingency seems appropriate in handling hrm because theory suggests there is no right way to manage the organization. is it a principle frequently seen as too concerned with questions about which contingencies to be concerned should "strategic fit" or "climate suit" be the primary concern when deciding the form of hrm systems and services to be adopted? while the "best match" appears more practical, it tends to be static and does not take into account other factors affecting hrm approaches, such as cross-cultural difficulties and organizational priorities (armstrong, 2009). perhaps "resources-based approach" to maintain market competitiveness by creating specific internal tools, concentrating on human capital and optimizing employee skills, and identifying cultural differences as the best approach to managing cross-cultural employees. this approach allows the company to be seen as a special, nuanced, intangible and competitive resource package, not through its product market activities (saá-pérez & garcía-falcón, 2002). (leavy, 1996) believed that these tools were significantly below the surface of the highest strategic importance — for example, community, including traditions, norms , beliefs, etc. (schein, 1985). 5. methodology this study, based on the desk research approach, literally focuses on systematic review. data has been extracted from a secondary source such as books, articles, and formal reports. the literature substantially centered on peer-reviewed academic journals, such as the social science citation database and the scopus database, as they are frequently used in the academic world. because of the complexity in examining culture and business relationships, tedious extraction has been made from various fields of studies ranging from business management, culture, gender, psychology, education, and science. in the beginning, the research was focused on articles from 2010-2019. however, to support the arguments, the studies have been extended from 1970. muhammad aiman awalluddin 62 however, this paper still maintained on the latest findings and evidence to keep it abreast and up to date. also, due to inconsistency in results of cultural dimension by past researchers on bric countries, a specific tool for data analysis such as hofstede insight (https://www.hofstede-insights.com) has been used to organize data analysis. as a means to obtain corresponding data, the search keywords are set out, and a possible combination that related to culture and management will be sorted out. the first part of table 1 shows the number of articles that have been obtained by the corresponding keywords in the sakarya university (sau) library database. while table 2 presents several popular articles and the elimination of articles that mismatched with the aims of this study. the provided tables are a representation of the challenge and rigorous searching applied to attain related articles that match the purposes of the paper. due to the bulk of results finding from the database, associated studies will be sorted out, scan and filtering before selected for the study. the articles will be scanned based on their related title to decide whether they are relevant or not. skimming will determine whether the article contains information needs for this study. next, the articles will be subjected to the filter process to determine whether articles to be accepted or rejected. consequently, a careful selection of articles that meet the research objectives will be selected. table 2: search results search results in sau library database combination keywords results some relevant articles culture + management 25 000 articles n/a cross + culture + management 181 articles 65 articles cross culture + management + psychology 20 articles 15 articles cross culture + management + education 243 articles 43 articles hofstede cultural dimension + management 30 articles 20 articles cross cultural + human resource management 106 articles 55 articles table 3: some accepted and elimination article some accepted articles authors content matched objective (mendes et al., 2017) remuneration behaviour in brazil (bjorkman et al., 2008) human resources in india and china (arapova, 2017) cultural differences in russia and western some elimination articles authors content mismatched objective journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 63 (liu & wu, 2018) quality culture management model of chinese manufacturing enterprise (taylor et al., 2008) impacts of organizational culture, hrm and top management orientations (rode, huang & flynn, 2016) cross-cultural examination and organizational commitment 6. findings 6.1. bric and hofstede cultural dimension hofstede's cultural factor theory explains the influence of a society's culture on its members' beliefs, and how these beliefs contribute to people's actions. the clash between western and asian cultures, especially on the social background, should be the central focus for western corporations that operate/or want to run into asian countries. the closeness of culture among asian countries facilitating asian corporations settling among them comfortably in comparison to western corporations. for example, japan is closer culturally to india than the united states, and japanese practices are more consistent with developing countries through cultural proximity than those of the us. (wasti, 1998). according to (maharjan & sekiguchi, 2016), japan and india are collectivist and hierarchical societies; both of them are collectivist societies, while the us more towards individualism. to identify the cultural dimension in bric countries, recent data studies will be presented. according to (minkov et al., 2017), on the scores of the collectivism/individualism dimension, it is found that brazil, russia, china, and india are collectivist societies. while a study done by (beugelsdijk & welzel, 2018; lee park & paiva, 2018) indicate that brazil has shifted toward individualism, while there is a consistent result for russia, china and india for collectivism. further analysis has been adopted from hofstede insight (https://www.hofstede-insights.com) to analyze the cultural dimension among bric members. table 4: summary of cultural dimension hofstede cultural dimension brazil russia india china power distance index (pdi) high high high high uncertainty avoidance index (udi) high high low low individualism/collectivism index (ind) collectivism collectivism collectivism collectivism masculinity/femininity index (mas) femininity femininity masculinity masculinity long-term/short term orientation index (lto) short term long term long term long term indulgence/restrained index (ii) indulgence restrained restrained restrained source: author muhammad aiman awalluddin 64 6.1.1 power distance power distance index considers the tolerance to inequalities and powers. if there's a high-power gap (brazil, russia, china, and india), it suggests a society recognizes inequity and power disparities, promotes bureaucracy, and people are inclined to display high respect for rank and status. employees in high-power-distance culture expect managers to behave as a strong leader and become dissatisfied with decision delegating leaders. for example, chinese business organizations still have traditional paternalistic leadership (cheng, chou, wu, huang, & farh, 2004; yan zhang, huai, & xie, 2015). the case also applies to india, where its historical religious and cultural growth derives high power distance. the caste system and seniority establish a strongly hierarchical relationship, mixing conventional "yes boss" and silent features (managers disallowing their subordinates to ask questions or express their views). in india, this made decentralization and delegation incompatible (mathew & taylor, 2018). conversely, if there is a low power gap, a culture promotes flat organizational structure, collective decision-making, and participatory style management. 6.1.2 individualism for individualism and collectivism, the dimension explains about "i" or "we" attitude in achieving the goal. it is about working as a team (brazil, russia, china and india) or individually. in china context, collectivist behaviour may serve as a shield to save one's face (moser, migge, lockstroem, & neumann, 2011) and not to blame other people in front of anybody else (cardon, 2009). for russia, the collectivist traits can be traced from social roots of zadruga (clan), mir (commune) and artel (association), they do appreciate consensus in settling dispute arising, joint responsibility and support substantial intergroup mutual assistance (balykina, 2013). 6.1.3 uncertainty avoidance avoidance of vulnerability is a degree to which instability and uncertainty are accepted and discussed. high uncertainty (brazil and russia) reveals that there is no tolerance to confusion, ambiguity and risk-taking. strict rules, laws, policies and norms are introduced and applied in the sense of brazil to mitigate or decrease this degree of vulnerability (cornacchione & klaus, 2017). however, low avoidance of uncertainty (india and china) implies a strong tolerance for confusion, ambiguity and risk-taking. a piece of evidence from german managers in china reported that the chinese have an enterprising attitude and a risk-taking disposition (moser et al., 2011). they were still able to carry out control tests to ensure that the job was done. this attitude illustrates journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 65 that chinese workers need no rigid guidelines and are eager to conquer confusion, able to drive and influence. 6.1.4. masculinity for masculinity (india and china), it asserts the following characteristic towards egoism, building wealth, independent, aggressive, and self-confident. femininity (brazil and russia), relationship-oriented, tolerance towards making a mistake, be humble, asking questions and weak, innocent, dependent, nurturing, and concern about the quality of life. the differential is about to live to work or work to live. this cultural dimension seems to be interrelated with the uncertainty avoidance dimension; values with low uncertainty tend to perceive being independent and risk-taker, which is generic with values in masculinity. as an illustration, it was found by (k. singh & agrawal, 2007) that working indian women possess masculine characteristic such as "defend my own beliefs," "independent", "have a leadership quality", and "willing to take a risk" which is identical with entrepreneurial mentality owned by chinese. this indicated that there is a link between masculinity and uncertainty avoidance. in contrast, feminism is quite intense among male workers in russia who are unable to perform their breadwinner's role, declining men's ability to financially support their families with their diminishing of masculinity (ashwin & lytkina, 2004). akin to those lack of congruence with "habitus," a suspension of their commitment to the "game" (work). on the other hand, men had lost their ability to be providers and hence became "weak" (davidenko, 2017). the lack of masculinity among russian can be explained due to strong collectivism, power distance and uncertainty avoidance due to phenomena and attitude of people, such as "wait and hope for protection from the chief" (being dependent), low personal responsibility (afraid being risk taker) (balykina, 2013). despite that, the advantage of feminism should not be undermined as people with this value possessing strong cooperation, mutual understanding, and interpersonal relations. this can be evidence from the soviet's time when there is a gender contract between males and females, where females working at home with paid labor (paid less than male) under a state policy (davidenko, 2017). 6.1.5. time orientation long-term orientation (russia, india, and china) emphasises patience, perseverance, and long-term development. building a trade or good relationship with a chinese employee requires a manager to recognise the importance of trust, dependability, integrity , and benevolence. "guan" means a hurdle gate, and "xi" refers to tie. a partnership is extended in almost the domain of life, from politics to business. muhammad aiman awalluddin 66 impacts of potentially important "guanxi" as chinese people practise it in their everyday interactions with others to create, sustain and improve their interpersonal and interorganizational relationship with different counterparts at different levels (styles & ambler, 2003; yen & barnes, 2011). ironically, long-term orientation made people reluctant to respond to change and under less pressure to act (gopalan & stahl, 1998). by comparison, short term orientation (brazil) focuses on quick results and short term success. 6.1.6. indulgence the last dimension, indulgence (brazil), indicates society enjoying life, having fun, and less strict rules or ethics on society norms. in brazil, due to former excessive legalism and formalism of brazil society inherited from its portuguese colonizers, it has changed the way of brazilian society's personalism. the introduction of jeitinho encourages brazilian to find solutions to a bureaucracy that often do not well (jackson & hogg, 2010). it is used in almost all manner to get things done promptly, escaping applicable rules and breaking the rules even aware it may causing harm to others (cornacchione & klaus, 2017) and also enable corruption. restraint (russia, india, and china) is a culture that suppresses and controls needs pleasure by strict social norms. the restraint norms between india and china are profoundly rooted in a collective culture and country where it is practiced. vedantic (vedanta) greatly influenced indian culture in ethics that regulated life based on unselfishness and fair sense regulation (malik & vijay, 2016). whereas in china, the philosophy of confucianism, now deeply rooted in chinese, urges people to control their feelings, avoid interpersonal disputes, and abolish human rights to preserve good relations with others (yip, 2003). restraint importance emphasizes individuals' submissiveness to maintain social equilibrium. in other words, people make a rational decision based on moral principles to protect the entire interest. 6.2. discussion organizational culture is a set of ideals, beliefs and expectations held by organizational participants describing how they deal with external and internal influences (hofstede, 1980; schein, 1985). in the context of the social realities and ethos of the country in which the company works, it also defines performance levels, prevailing environment and hrm policies (davenport, 2000; kumar & hillergersberg, 2000). applying universalistic hrm methods explicitly won't work. (javidan, dorfman, luque, & home, 2006) believed that there was no guarantee that appropriate management practices would succeed in one country. operating company does not fight journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 67 host country culture, but rather integrate into it. (sparrow, brewster, & harris, 2004) believe that international business is about finding ways to work together for people of different experiences or perspectives. as discussed above, hrm should use resourcebased view (rbv) to efficiently manage people. finding the obstacles facing foreign companies in india (holtbrügge, friedmann and puck 2010) will be reduced when shaping hrm practice on the basis of rbv. organizations have several specific resources which can only be built in-house (barney, 1991; grant, 1991). by emphasizing rbv, organizations can gain a competitive advantage, especially if they succeed in following the cultural differences in hrm practices. in the sense of helping companies to gain a competitive advantage, a positive culture (genç, 2013) should include all relevant elements in alignment, management style, business and environmental circumstances. although the decided people (becker & huselid, 1998) are a strategic advantage. the rbv approach helps hrm to strategize its hrm practice in order to improve business success. the relationships between the hrm process (selective workers, incentive pay, and training) on organizational performance have been positive (delaney & huselid, 1996; katou & budhwar, 2007). while the rbv approach has important ties with hrm performance, a foreign organization should concentrate on the social context and economic and political history of the host county. using the cultural component of hofstede, companies will consider the cultural differences before they start hrm practices. (kulkarni et al. 2010) considered that in a global subsidiary with complex cultural orientations it would be hard for international companies to incorporate domestic management strategies, such as incentive and evaluation schemes. organizations working in india have strongly recommended that the legal, political and socio-cultural aspects of india be taken seriously (björkman, budhwar, smale & sumelius, 2008). there are many difficulties in bric markets, and relations with locals and adaptation are a must. the social system of bric varies considerably from that of western societies, as almost all bric countries are not western countries. even if russians regard europe (western civilization), it remains foreign and retains from their counterparts a different philosophy and culture. for example, differences in mutual understanding, interpersonal communication, symbols and signals are different in the preparation and execution of organizational change between russia and the west (michailova, 2005). muhammad aiman awalluddin 68 6.2.1. staffing staffing can be described as one of the most important management functions. it includes filling up the vacancy role with the right workers at the right time. therefore, everything will happen properly. it is a fact that human resources are one of the highest in any organization because, through proactive actions of the human resources, all other resources, such as capital, equipment, machinery, etc. can be efficiently utilized in every organization. each individual must therefore be in the right place within the organization to get the right job according to their skills, talents, aptitudes and specializations, in order to help the organization achieve the predetermined goals by contributing 100% of the workforce correctly. it can also be assumed that workers are an integral feature in all business organizations. the role of the staffing can be defined as follows, the primary function of the staffing being to recruit skilled workers for various positions in the organization. the best employee is hired for the best role in the workplace. this results in optimum efficiency and better output. it helps to facilitate the efficient use of human capital in different ways. job satisfaction and morality of employees are improved by hiring the right individual. staffing helps ensure that human resources are best used. it provides the organization’s consistency and growth through development managers. in china, chinese companies incline to do more external hiring, emphasizing the one-to-one interview, and least use on reference practice (li, samolejová, čech, & lampa, 2016). personal references play a vital role at the time of hiring ( heneman, judge, & mueller, 2019; rao, 2009). in a high uncertainty avoidance culture, they do not trust oral information and seek more elaboration on written information. however, in low uncertainty avoidance, the information provided by the employee is being believed, and managers are being provided with a degree of freedom to decide on the best interest of the organization (hofstede, hofstede, & minkov, 2010). in india, organizations must be very flexible in terms of the employment contract. indians are very emotional, and employment is a matter of bonding (long time orientation). a company cannot fragment an individual from his/her family (feeling of collectivist), allow the employee to leave work early, but also has the liberty to call them on sunday to work (restraint). at the same time, indian women aspire to be in the top management position (gupta & bhaskar, 2016). in russia context, there is a legitimacy problem with the hiring process, not focus on job-related but emphasizing on the trust and loyalty of the employee (domsch, 2017). stressing on the trust and loyalty (hiring someone they journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 69 know) shows a strong affiliation towards collectivism and a high level of uncertainty avoidance in russian culture. the problem of legitimacy was supported by fey, engström & björkman (1999), who argue that establishing formal requirements is difficult for russian hrm and typically relies on their "good feelings" in the recruitment process. in brazil, similar problems in indian and russian seem common, where the recruitment process takes account of certain aspects, such as the collectivist culture and the hrm selection procedure. the relationship with wealthy relatives, college friends, country club members or birthplace regional connections is more important than talent because of jeitinho and the power of collectivism culture, to be promoted or chosen in organizational hierarchies. the staffing processes of many latin american companies often rely on social connections (abarca, majluf, & rodriguez, 1998; flynn, 1994). the blend of influences in brazilian culture is allegedly derived from spanish arrogance and portuguese laxity and plasticity, resulting in laxity, chaos, lack of organization, disorder and discipline (de hilal, 2006). the ambiguity of the job description, the multi-tasking and the organization of the working environment can also adversely affect the manner in which brazilians function. managers also have to understand the leisure time and the value of families, because brazilians are more likely to spend time with their families and to give up their jobs (short-time). 6.2.2. performance management performance management is the mechanism or method by which a company monitors and enhances its employee efficiency. performance management may be used by a company to track the success of an entity, a department, a team and a single level. the word most often applies to the success of individuals. popular elements of a method of performance management include: (a) feedback of results (b) one-on-one manager (c) reviews from peer (d) process success plans (e) setting and monitoring of goals (f) programs of incentives and recognition muhammad aiman awalluddin 70 in a traditional performance management framework, managers monitor and improve success with team members and report to the higher leadership team. in the past, the method of performance management also consisted of an annual assessment of results per employee. however, since research indicates that periodic reviews are increasingly successful, many organizations have abandoned the annual review model and implemented a system of daily manager’s checks and informal feedback meetings. such performance management can help workers understand and align themselves consistently with organizational priorities and objectives, making incremental changes over time rather than attempting to develop their career at year-end. there are several factors which determine and motivate performance in india, including ambition, rapid growth, and appreciation. as the working culture is strongly built on the paternalist, the boss is known as the dad. therefore, the 360 degree evaluation would not apply as an employee more likely to hear supervisor / manager direct input (gupta & bhaskar, 2016). this evidence was supported by (sripirabaa & krishnaveni, 2009), who found that external members can appraise employees at a particular time, but the employee may reject it during the performance process. in contrast, a 360-degree appraisal well accepted among chinese (paetzel, quast, aroonsri, surya, & hart-mrema, 2019). the organizations need to ensure the appraisal process take into consideration employee's consideration and provide transparent feedback towards an employee (hui & qin-xuan, 2009). chinese are more open to hearing feedback from the surrounding, although they are also based on paternalistic leadership culture. perhaps these traits contribute by ethical philosophy that already been embedded among chinese on the concept of openness in their culture, respect and obedience (lin, li, & roelfsema, 2018). in russia, however, the question of personnel formalization continues to prevail in performance management. russia has inadequate levels of structured performance evaluation frameworks in a report (gurkov & zelenova, 2011). employees then prohibit direct reference to the outcome of the latest performance review in their termination appeals and preclude pay negotiations. the performance control of employees according to (gurkov & zelenova, 2009) is focused on direct observations and the quality and quantity of work of the supervisor. this situation indicates high power distance in russian society, where the power wielder (manager/top management) has control over bottom employee and dislike to follow the rules and formalization. a study done in the brazilian research center shows that performance appraisal is essential; there is an integration between performance appraisal and strategic plans. performance appraisal 360 degree does not consider well accepted in which all the staff is appraised journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 71 by their superior (lourenço ubeda & cesar almada santos, 2007). however, in the recent study shows that hrm manager in brazcon argued that brazilian companies nowadays are being less paternalistic and more professional. they adopted the contemporary global practice, such as brazman used 360-degree appraisals, and brazcem used the balanced scorecard to manage performance appraisal (mellahi, frynas, & collings, 2016). 6.2.3 reward and compensation the administration of benefits is a strategic concern. compensation would include benefits for offering monetary compensation such as promotions, incentives and varying bonuses. organizations reward their workers as they achieve the objectives or objectives they have set for their employees. rewards can be non-monetary for two, for example a paid holiday. compensation is considered a factor in hygiene or motivation. in other words, if the monthly payment is fixed, a hygiene factor. it can be used for comparing or assessing or deciding if the company pays 'fair' prices. hrm will typically set up a compensation survey, which can provide important information on incentive issues. the other aspect is the advantage. the provision of benefits is another component of compensation. to make the reward a motivating factor, it must be variable. it varies proportionally with each other's outcome, target or goals. this reward is typically referred to as an incentive or incentive to increase productivity for workers. motifs are a variable expense for specific business or service processes which differ explicitly from firm operations to firm operations. hrm must therefore take reimbursement as a strategic issue. hrm must plan and develop a system of incentives & rewards to attract the right people to the right job. first, hrm would have to efficiently implement the system in order to ensure retention. finally but not least, the control of skills. compensation is one means of attracting and retaining talents. however, hrm also needs other talent management tools for highpotential, highly professional and technical staff, including evaluation and mapping. hrm would also improve the wages to encourage staff as team players, creative people and productive workers. hrm also needs to design and develop incentives based on employee profile and work nature. rewards must be realistic and useful for the beneficiaries. an example is that the winner can choose from a number of awards. this method would give them the impression that the reward is tailored as they strive to achieve the goals. the iphone or even the ipad is an example of a recent incentive item because these items draw current attention. muhammad aiman awalluddin 72 organizations should take additional care when working in russia, india and china in strategizing reward and compensation. in these countries, the incentive and rewards should form the basis of long-term guidance. in indian social structure you are forced to take on a stable job with good pay and potential for growth. for organizations lacking adequate social security and long-term job insecurity, indians expect a higher base salary (gupta & bhaskar, 2016). as professional managers in india increase, the country is no longer a cost arbitration for organizations. for china, chinese citizens, with a low level of prevention and entrepreneurial mindset, are a risk-taker and predict the size of the reward for their work. wang & wu (2017 ) found that equity-based rewards ultimately affect the company's risk-taking rate. in comparison, incentive and compensation seem to be backward in russia, as in other post-socialist countries. in russia there is a dominant view that paying salaries are wasted income (gurkov & zelenova, 2011). in his further research, the majority of companies refuse to spend money on education and personal development; they are also pursuing dumping practices by making oligopolistic deals in order to make efforts of staff and mid-size executives useless in the negotiating process compensation systems are quite successful and formalized for brazil. they have already adopted international hrm practices and match their pay table with the world consulting table (mellahi et al., 2016). finally, as a short type of people, performance pay in brazil is likely to be feasible as practice has been commonly used since the 1990s (mendes, lunkes, flach & kruger, 2017). 6.2.4 equal employment opportunities equal employment opportunities or eeo is a right for any employee during a recruiting process. it refers among other eligible characteristics to protect employees from discrimination based on ethnicity, religion, sex or national origin. employers cannot judge or determine on one of these bases whether an individual is fit for a job. these requirements must be complied with by federal, state and local governments, jobs agencies and most private businesses. eeo also protects employees at present against discrimination in terms of promotions, wages, benefits or discharge. the united states commission on equal employment opportunities tracks adherence and reacts to discovered violations. fair working opportunities are an occupational activity in which employees do not participate in work which is prohibited by statute. employers cannot discrimination on the basis of ethnicity, age, color, sex, faith, nationality and origin against an applicant or employee. the ultimate political purpose of the workplace discrimination legislation we are reviewing will be summarized in the expression "equal opportunities." journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 73 generally, these regulations do not aim to achieve fair results but are meant to ensure that all workers or work applicants have equal opportunities to compete in the labor market. in other words, the laws aim to level the playing field so that some groups of persons that have been discriminated against in the past are not treated unfavorably based on unique features which have little to do with being a competent employee. the equal opportunities for jobs illustrated this: 1. recruitment and selection requirements, appointing or engaging an individual as an employee; 2. employee promotion and transfer; 3. training and career growth for a career; 4. conditions of jobs or any other behavior relevant to workers. as an example in india, in article 15 and in article 16, the indian constitution encapsulates social security provisions for all the people of india. the ban on discrimination on the basis of faith, ethnicity, caste, sex or place of birth is imposed in article 15 — the provision of article 16 is that of equal opportunities in matters of public employment. in addition, article 16(2) states that no person shall be disqualified, or discriminated against in respect of any employment or state office for reasons relating solely to religion, race, caste, sex, descent, place of birth, place of residence or any one of them. these suggest that hrm must conduct vigilant staffing in india. any violation of the law as stipulated in indian legislation will result in a penalty that will harm the company's credibility and standing. 6.2.5 suggestion hrm approach for bric the study verified that hofstede dimension is still reliable in explaining crosscultural differences, especially in hrm functions. it is indicated that western hrm approach and argument of "one size fits all" is no longer valid within hrm dimensions. the idea of western approach which emphasizes on the importance of individualism in every aspect of hrm functions such as individual pay for performance, making the decision, and promotion, which rooted from the medieval culture of competitiveness of western lifestyle will bring negative impacts toward local bric hrm management. the results of comparison study among bric countries showed that all bric members support collectivism spirit, perhaps derived from a strong feeling of cohesiveness, ethicality and bounded by strong religious affiliation such as taoism in china, hinduism in india and zadruga in russia. the cultural of bric also indicated that they are inclined towards value/subjective, sense of feeling/humanism rather than being muhammad aiman awalluddin 74 objective/impartial, anti-humanism where it is more permeated among western. therefore, the paper insisted on several approaches that could be adapted to meet cultural expectations for hrm in bric countries. table 5: suggestion for hrm approach workplace culture • job sharing • reward for consensus decision/punishment for group. • respecting or rewarding local affiliation such as religious events/family matter/culture festivals. • showing compassion and sense of empathy. • reward those employees that being cohesive. staffing • hiring employee that possess strong moral obligation. • evaluating the level of responsibilities and cohesiveness of employee before hiring. • strong and details on loyalty career structure, (promotion, rewards for loyalty, career advancement, and retirement planning before boarding to ensure employee satisfaction and long term affiliation with the company). • emphasize on referral method in selecting future employees. • equal employment based on merit, but some of them incline for equitable (taken into consideration of background) rather than being too objective on hiring. performance management • applied paternalistic approach "father and son" appraisal, as bric employees have a strong relationship with the direct supervisor. • apply face to face feedback, rather than public feedback as bric against public shame. • avoid self-projection, as bric employees dislike being compared. rewarding • senior or loyal employees should highly be rewarded. • risk-taker should be appreciated with some incentives. • gender and race-based wage should be avoided as may create unease/conflict within the organization. • reward structure should be formalized and clear from ambiguity. 7. conclusion the study of the relationship between contextual properties of cultural dimensions and bric country hrm management permits the re-establishment of hrm studies. this study shows that local / foreign organizations, who wish to grow into the developing markets, need to understand the culture of local communities and emerging markets — the bric markets with their special mix of intra-country cultures. therefore, it is important to consider the society in order to fix similarities and identify opportunities. furthermore, in order for international organizations to become completely operational, they should create trust in local citizens, as the close relationship is a vital elements in bric markets. journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 75 on the other hand, hrm in bric is greatly influenced by the characteristics of the culture. however, some of the countries are showing towards the western dimension. for massive culture societies like indian and chinese, moral values and philosophy of old religions veda/hindu (india) and confucianism/taoism (chinese) play essential roles in their daily life. therefore, understanding a country's culture is a sign of respect and foster effective communication, a vital factor in business success. by integrating the business into the local culture, it will facilitate the institutionalization process of organizations within the society and increase its legitimation. this study tried to reexamine what needs to be taken and considered by foreign companies to succeed in bric markets. the findings reveal that although some bric markets shared similar dimensions of culture, it does not mean general practice will apply to all of them. as an example, indians and chinese are both strong paternalistic societies but require different methods and tools in assessing performance management due to differences in philosophy and beliefs they hold. this data analysis is hoping to assist researchers/industry players in having depth understanding and a clear view of managing hr and cultural differences in bric markets. a significant limitation of this paper is based solely on secondary data. future research in the area of integration of hrm (hr function and practice) and culture in this area might consider these things. references abarca, n., majluf, n., & rodriguez, d. (1998). identifying management in chile: a behavioral approach. international studies of management and organization, 28(2), 18–37. abugre, j. b. (2018). cross-cultural communication imperatives. critical perspectives on international business, 14(2/3), 170–187. https://doi.org/10.1108/cpoib-012017-0005 adler, n. j. (1997). international dimensions of organizational behavior (3. ed.). cincinnati ohio: south-western college publ. andreassi, k. j., lawter, l., brockerhoff, m., & j. rutigliano, p. (2014). cultural impact of human resource practices on job satisfaction. cross cultural management: an international journal, 21(1), 55–77. https://doi.org/10.1108/ccm-05-2012-0044 armijo, l. e. (2007). the brics countries (brazil, russia, india, and china) as analytical category: mirage or insight? asian perspective. lynne rienner publishers. https://doi.org/10.2307/42704607 muhammad aiman awalluddin 76 ashish malik, & vijay pereira. (2016). indian culture and work organisations in transition google books. new york: routledge: taylor and francis group. ashta, a., stokes, p., & hughes, p. (2018). change management in indo-japanese cross-cultural collaborative contexts. journal of organizational change management, 31(1), 154–172. https://doi.org/10.1108/jocm-05-2017-0201 ashwin, s., & lytkina, t. (2004). men in crisis in russia. gender & society, 18(2), 189–206. https://doi.org/10.1177/0891243203261263 balykina, g. (2013). cultural dimensions and modern russian business. ssrn electronic journal. https://doi.org/10.2139/ssrn.2342954 barney, j. (1991). firm resources and sustained competitive advantage. journal of management, 17(1), 99–120. https://doi.org/10.1177/014920639101700108 becker, b. ., & huselid, m. a. (1998). high performance work systems and firm performance. a synthesis of research and managerial application. research in personnel and human resources management, 16, 53–101. beugelsdijk, s., & welzel, c. (2018). dimensions and dynamics of national culture: synthesizing hofstede with inglehart. journal of cross-cultural psychology, 49(10), 1469–1505. https://doi.org/10.1177/0022022118798505 björkman, i., budhwar, p., smale, a., & sumelius, j. (2008). human resource management in foreign-owned subsidiaries: china versus india. the international journal of human resource management, 19(5), 964–978. https://doi.org/10.1080/09585190801994180 cardon, p. w. (2009). a model of face practices in chinese business culture: implications for western businesspersons. thunderbird international business review, 51(1), 19–36. https://doi.org/10.1002/tie.20242 chen, n., chao, m. c., xie, h., & tjosvold, d. (2018). transforming cross-cultural conflict into collaboration. cross cultural & strategic management, 25(1), 70– 95. https://doi.org/10.1108/ccsm-10-2016-0187 cheng, b.-s., chou, l.-f., wu, t.-y., huang, m.-p., & farh, j.-l. (2004). paternalistic leadership and subordinate responses: establishing a leadership model in chinese organizations. asian journal of social psychology, 7(1), 89–117. https://doi.org/10.1111/j.1467-839x.2004.00137.x cornacchione, e., & klaus, l. (2017). ethical business culture in brazil. in ethical business cultures in emerging markets (pp. 7–8). cambridge university press. https://doi.org/10.1017/9781316225165.004 journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 77 davenport, t. a. (2000). mission critical: realizing the promise of enterprise systems. boston, ma: harvard business school press. davidenko, m. (2017). searching for lost femininity: russian middle-aged women's participation in the post-soviet consumer culture. journal of consumer culture, 146954051771402. https://doi.org/10.1177/1469540517714021 de hilal, a. v. g. (2006). brazilian national culture, organizational culture and cultural agreement. international journal of cross cultural management, 6(2), 139–167. https://doi.org/10.1177/1470595806066325 delaney, j. t., & huselid, m. a. (1996). the impact of human resource management practices on perceptions of organizational performance. academy of management journal, 39(4), 949–969. domsch, m. e. (2017). human resource management in russia (1st ed.). routledge. dong, k., & liu, y. (2010). cross‐cultural management in china. cross cultural management: an international journal, 17(3), 223–243. https://doi.org/10.1108/13527601011068333 eskildsen, j., kristensen, k., & gjesing antvor, h. (2010). the relationship between job satisfaction and national culture. the tqm journal, 22(4), 369–378. https://doi.org/10.1108/17542731011053299 farivar, f., cameron, r., & yaghoubi, m. (2016). work-family balance and cultural dimensions: from a developing nation perspective. personnel review, 45(2), 315–333. https://doi.org/10.1108/pr-09-2014-0196 fey, c., engström, p., & björkman, i. (1999). doing business in russia: effective human resource management practices for foreign firms in russia. organizational dynamics, 28(2), 69–79. fish, a. (1994). the importance and relevance of cross‐cultural management. crosscultural management: an international journal, 1(1), 35–38. https://doi.org/10.1108/eb008375 flynn, g. (1994). hr in mexico: what should you know. personal journal, 73(7), 34– 42. fujimoto, y., bahfen, n., fermelis, j., & härtel, c. e. j. (2007). the global village: online cross‐cultural communication and hrm. cross cultural management: an international journal, 14(1), 7–22. https://doi.org/10.1108/13527600710718804 muhammad aiman awalluddin 78 gahlawat, n., & kundu, s. c. (2018). progressive human resource management and firm performance. international journal of organizational analysis, ijoa-052017-1159. https://doi.org/10.1108/ijoa-05-2017-1159 gattai, v., & natale, p. (2016). new facts about bric multi-nationals: from macro to micro evidence (pp. 61–92). https://doi.org/10.1108/s1745886220160000011007 genç, k. y. (2013). culture as a strategic resource for organizations and an assessment on cultures of turkish large firms. procedia social and behavioral sciences, 75, 525–531. https://doi.org/10.1016/j.sbspro.2013.04.057 gibb, s. (2001). the state of human resource management: evidence from employees' views of hrm systems and staff. employee relations, 23(4), 318–336. https://doi.org/10.1108/eum0000000005685 glosny, m. a. (2010). china and the brics: a real (but limited) partnership in a unipolar world. polity, 42(1), 100–129. https://doi.org/10.1057/pol.2009.14 gopalan, s., & stahl, a. (1998). application of american management theories and practices to the indian business environment: understanding the impact of national culture. american business review, 16(2), 30–41. grant, r. m. (1991). the resource-based theory of competitive advantage: implications for strategy formulation. california management review, 33(3), 114–135. https://doi.org/10.2307/41166664 greblikaite, j., & daugeliene, r. (2010). the growing need of cross-cultural management and ethics in business institute of europe, kaunas university of technology. european integration studies, (4), 148–152. gupta, s., & bhaskar, a. u. (2016). doing business in india: cross-cultural issues in managing human resources. cross cultural and strategic management, 23(1), 184–204. https://doi.org/10.1108/ccsm-09-2014-0112 gurkov, i., & settles, a. m. (2012). a dominant archetype of the hrm system in russian industrial companies in post-recession times. ssrn electronic journal. https://doi.org/10.2139/ssrn.2039033 gurkov, i., & zelenova, o. (2009). human resource management in russia (pp. 150– 188). gurkov, i., & zelenova, o. (2011). human resource management in russian companies. studies of mgt. & org, 41(4), 66–80. https://doi.org/10.2753/imo0020-8825410404 journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 79 hart, a. f., & jones, b. d. (2010). how do rising powers rise? survival, 52(6), 63– 88. https://doi.org/10.1080/00396338.2010.540783 herbert g. heneman iii, judge, t. a., & kammeyer-mueller, j. (2019). staffing organizations (9th ed.). new york: mc graw hill. hofstede, g. (1980). culture's consequences: international differences in workrelated values. beverly hills, ca: sage. hofstede, g., hofstede, g. j., & minkov, m. (2010). cultures and organisations: software of the mind (3rd ed.). new york, ny: mc-graw hill. hofstede insight. (2019). hofstede insight: compare country. retrieved april 1, 2019, from https://www.hofstede-insights.com/product/compare-countries/ holtbrügge, d., friedmann, c. b., & puck, j. f. (2010). recruitment and retention in foreign firms in india: a resource-based view. human resource management, 49(3), 439–455. https://doi.org/10.1002/hrm.20353 hui, l., & qin-xuan, g. (2009). performance appraisal: what's the matter with you? procedia earth and planetary science, 1(1), 1751–1756. https://doi.org/10.1016/j.proeps.2009.09.268 jackson, r. l., & hogg, m. a. (2010). encyclopedia of identity. sage. javidan, s., dorfman, p. w., luque, m. s., & house, r. j. (2006). "in the eye of the beholder: cross cultural lessons in leadership from project globe. academy of management perspective, 69–90. katou, a. a., & budhwar, p. s. (2007). the effect of human resource management policies on organizational performance in greek manufacturing firms. thunderbird international business review, 49(1), 1–35. https://doi.org/10.1002/tie.20129 keukeleire, s., & hooijmaaijers, b. (2014). the brics and other emerging power alliances and multilateral organizations in the asia-pacific and the global south: challenges for the european union and its view on multilateralism. jcms: journal of common market studies, 52(3), 582–599. https://doi.org/10.1111/jcms.12102 kirkman, b. l., lowe, k. b., & gibson, c. b. (2006). a quarter century of culture's consequences: a review of empirical research incorporating hofstede's cultural values framework. journal of international business studies, 37(3), 285–320. https://doi.org/10.1057/palgrave.jibs.8400202 kulkarni, s. p., hudson, t., ramamoorthy, n., marchev, a., georgieva-kondakova, p., & gorskov, v. (2010). dimensions of individualism-collectivism: a muhammad aiman awalluddin 80 comparative study of five cultures. current issues of business and law, 5. https://doi.org/10.5200/1822-9530.2010.03 kumar, l., & hillergersberg, j. van. (2000). "erp experiences and evolution. communications of the ac, 16, 258–265. latukha, m., & malko, k. (2019). westernization or localization of human resource management practices in cis countries? evidence from kazakh firms. international journal of emerging markets, ijoem-05-2016-0122. https://doi.org/10.1108/ijoem-05-2016-0122 leavy, b. (1996). on studying leadership in the strategy field. the leadership quarterly, 7(4), 435–454. https://doi.org/10.1016/s1048-9843(96)90001-6 lee park, c., & paiva, e. l. (2018). how do national cultures impact the operations strategy process? international journal of operations & production management, 38(10), 1937–1963. https://doi.org/10.1108/ijopm-03-20170145 li, j., samolejová, a., čech, m., & lampa, m. (2016). comparison of hrm practices between chinese and czech companies. perspectives in science, 7, 2–5. https://doi.org/10.1016/j.pisc.2015.11.002 lin, l., li, p. p., & roelfsema, h. (2018). the traditional chinese philosophies in intercultural leadership. cross cultural & strategic management, 25(2), 299–336. https://doi.org/10.1108/ccsm-01-2017-0001 lourenço ubeda, c., & cesar almada santos, f. (2007). staff development and performance appraisal in a brazilian research centre. european journal of innovation management, 10(1), 109–125. https://doi.org/10.1108/14601060710720573 m. armstrong. (2009). armstrong's handbook of human resource management practice (11th ed.). london: kogan page. m. armstrong. (2012). armstrong's handbook of human resource management practice (12th ed.). london: kogan page. maharjan, m. p., & sekiguchi, t. (2016). the influence and effectiveness of us-style and japanese-style hr practices on indian firms: a conceptual perspective.south asian journal of human resources management, 3(1), 58– 74. https://doi.org/10.1177/2322093716642117 malik, a. r., & singh, p. (2017). transformational leadership and cultural minorities: a conceptual model. european business review, 29(5), 500–514. https://doi.org/10.1108/ebr-12-2015-0181 journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 81 mathew, s., & taylor, g. (2018). power distance in india. cross cultural & strategic management, ccsm-02-2018-0035. https://doi.org/10.1108/ccsm-02-20180035 meena, k., & vanka, s. (2017). developing an empirical typology of diversity-oriented human resource management practices. journal of management development, 36(7), 915–929. https://doi.org/10.1108/jmd-02-2016-0031 mellahi, k., frynas, j. g., & collings, d. g. (2016). performance management practices within emerging market multi-national enterprises: the case of brazilian multinationals. the international journal of human resource management, 27(8), 876–905. https://doi.org/10.1080/09585192.2015.1042900 mendes, a., lunkes, r. j., flach, l., & kruger, s. d. (2017). the influence of remuneration on the behavior of hospital employees in brazil. contaduría y administración, 62(1), 207–221. https://doi.org/10.1016/j.cya.2016.04.005 michailova, s. (2005). contrasts in culture: russian and western perspectives on organizational change. the academy of management executive (1993-2005). academy of management. https://doi.org/10.2307/4165688 mihaela, h. (2014). a synergistic approach of cross-cultural management and leadership style. journal of international studies, 7(2), 106–115. https://doi.org/10.14254/2071 minkov, m., dutt, p., schachner, m., morales, o., sanchez, c., jandosova, j., mudd, b. (2017). a revision of hofstede's individualism-collectivism dimension. cross cultural & strategic management, 24(3), 386–404. https://doi.org/10.1108/ccsm-11-2016-0197 moser, r., migge, t., lockstroem, m., & neumann, j. (2011). exploring chinese cultural standards through the lens of german managers: a case study approach. iimb management review, 23(2), 102–109. https://doi.org/10.1016/j.iimb.2011.02.002 o'neill, j. (2001). building better global economic brics. retrieved from https://www.goldmansachs.com/insights/archive/archive-pdfs/build-betterbrics.pdf (2.2.2019) paetzel, d. w., quast, l. n., aroonsri, p., surya, m., & hart-mrema, t. s. (2019). managerial behavior and performance in china, the uk, and the usa. european journal of training and development, ejtd-09-2018-0094. https://doi.org/10.1108/ejtd-09-2018-0094 pfeffer, j. (1998). the human equation. boston, ma: harvard business school press. muhammad aiman awalluddin 82 rao, p. (2009). the role of national culture on mexican staffing practices. employee relations, 31(3), 295–311. https://doi.org/10.1108/01425450910946488 redd, s. g., cui, z., guzman, m., jayadev, a., kaul, i., mckinley, t., & dos, p. s. (2017). the role of brics in the world economy & international development. retrieved from https://reddytoread.files.wordpress.com/2017/09/brics-2017.pdf (2 february 2019) rodriguez-arango, l., & gonzalez-perez, m. a. (2016). giants from emerging markets: the internationalization of bric multinationals (pp. 195–226). https://doi.org/10.1108/s1745-886220160000011011 saá-pérez, p. de, & garcía-falcón, j. m. (2002). a resource-based view of human resource management and organizational capabilities development. the international journal of human resource management, 13(1), 123–140. https://doi.org/10.1080/09585190110092848 sakikawa, t., chaudhuri, k., & arif, n. a. (2017). a study on management practices and manufacturing performance in india and malaysia. journal of asia business studies, 11(3), 278–295. https://doi.org/10.1108/jabs-07-2015-0115 sander, k., a. cogin, j., & bainbridge, h. t. j. (2014). research methods for human resource management (1st ed.). new york: routledge. schein, e. h. (1985). organizational culture and leadership third edition (3rd ed.). san francisco: jossey-bass. shenkar, o. (2001). cultural distance revisited: towards a more rigorous conceptualization and measurement of cultural differences. journal of international business studies, 32(3), 519–535. https://doi.org/10.1057/palgrave.jibs.8490982 singh, a., & singh, m. (2017). a revisit to stock market contagion and portfolio hedging strategies. international journal of law and management, 59(5), 618– 635. https://doi.org/10.1108/ijlma-03-2016-0026 singh, k., & agrawal, p. (2007). masculinity and femininity among working women in indian urban culture. journal of human ecology, 21(2), 135–138. https://doi.org/10.1080/09709274.2007.11905963 sparrow, p. s., brewster, c., & harris, h. (2004). globalizing human resource management. london: routledge. sripirabaa, b., & krishnaveni, r. (2009). performance management systems in an indian manufacturing sector. management research news, 32(10), 942–952. https://doi.org/10.1108/01409170910994150 journal of applied economics and business studies, volume. 4, issue 3 (2020) 55-84 https://doi.org/10.34260/jaebs.433 83 stahl, g. k., miska, c., lee, h.-j., & de luque, m. s. (2017). the upside of cultural differences. cross cultural & strategic management, 24(1), 2–12. https://doi.org/10.1108/ccsm-11-2016-0191 stahl, g. k., & tung, r. l. (2015). towards a more balanced treatment of culture in international business studies: the need for positive cross-cultural scholarship. journal of international business studies, 46(4), 391–414. https://doi.org/10.1057/jibs.2014.68 steenkamp, j. e. m. (2001). the role of national culture in international marketing research. international marketing review, 18(1), 30–44. https://doi.org/10.1108/02651330110381970 styles, c., & ambler, t. (2003). the coexistence of transaction and relational marketing: insights from the chinese business context. industrial marketing management, 32(8), 633–642. https://doi.org/10.1016/j.indmarman.2003.06.004 thomas d. c. (2008). cross-cultural management : essential concepts 2nd (2th ed., p. 327). california: thousand oaks: sage publications, inc. united nation. (2017). world population 2017. retrieved from www.unpopulation.org. (2.2.2019) verma, p., & sharma, r. r. k. (2019). the linkages between business strategies, culture, and compensation using miles &amp; snow's and hofstede culture framework in conglomerate firms. benchmarking: an international journal, bij-06-2017-0153. https://doi.org/10.1108/bij-06-2017-0153 wang, d., & wu, d. (2017). equity incentive and risk taking: evidence from china. nankai business review international, 8(1), 80–99. https://doi.org/10.1108/nbri-10-2016-0034 wasti, s. a. (1998). cultural barriers in the transferability of japanese and american human resources practices to developing countries: the turkish case. the international journal of human resource management, 9(4), 608–631. https://doi.org/10.1080/095851998340928 weber, t. j., sadri, g., & gentry, w. a. (2018). examining diversity beliefs and leader performance across cultures. cross cultural & strategic management, 25(3), 382–400. https://doi.org/10.1108/ccsm-11-2016-0200 muhammad aiman awalluddin 84 wilhelm, w. j., & gunawong, p. (2016). cultural dimensions and moral reasoning: a comparative study. international journal of sociology and social policy, 36(5/6), 335–357. https://doi.org/10.1108/ijssp-05-2015-0047 wilson, d., & purushothaman, r. (2003). dreaming with brics: the path to 2050. retrieved from https://www.gs.com (4.2.2019) yen, d. a., & barnes, b. r. (2011). analyzing stage and duration of anglo-chinese business-to-business relationships. industrial marketing management, 40(3), 346–357. https://doi.org/10.1016/j.indmarman.2010.08.003 yip, k. (2003). traditional confucian concepts of mental health: its implications to social work practice with chinese communities. asia pacific journal of social work and development, 13(2), 65–89. https://doi.org/10.1080/21650993.2003.9755929 zhang, j. j. (2009). study of the cross-cultural management in internationalized enterprise. digest of management science, 47(4), 47. zhang, y., huai, m., & xie, y. (2015). paternalistic leadership and employee voice in china: a dual process model. the leadership quarterly, 26(1), 25–36. https://doi.org/10.1016/j.leaqua.2014.01.002 zhang, y., wei, x., & zhou, w. (2017). an asymmetric cross-cultural perspective on the mediating role of conflict management styles in expatriation. international journal of conflict management, 28(5), 592–616. https://doi.org/10.1108/ijcma-06-2016-0052 zhu, y., & bargiela-chiappini, f. (2013). balancing emic and etic: situated learning and ethnography of communication in cross-cultural management education. academy of management learning & education, 12(3), 380–395. https://doi.org/10.5465/amle.2012.0221 journal of applied economics and business studies, volume. 4, issue 1 (2020) 1-28 https://doi.org/10.34260/jaebs.411 1 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x educational inequality and inclusiveness: the case of khyber pakhtunkhwa, pakistan tahira tauheed 1 & muhammad nasir 2*1 1 assistant professor of economics, incharge department of economics, lahore college for women university, lahore 2 senior research economist, pakistan institute of development economics (pide), quaid-i-azam university campus, islamabad abstract it is argued that masses in pakistan are excluded from the mainstream progress of education resulting in social unrest and adverse state of human development. this paper examines prevailing inequality in and exclusion from education in the khyber pakhtunkhwa province of pakistan, and provides an empirical base for designing an appropriate policy framework to mitigate the underline issues. towards this end the household-based education and inequality-adjusted education indices are derived using foster-lópez-calva-székely (fls) methodology at the provincial, and district levels from the most recent pakistan social and living standard measurement (pslm) survey 2014-15. the provincial analysis is elaborated at the urban and rural regions as well. the distribution of these indices across households are utilized to measure inequality and inclusiveness coefficients by employing “atkinson’s inequality measure” and “sixty percent of median as threshold of exclusion,” respectively. at the district level the impact of economic, social, demographic, and locational factors on inclusiveness of education are also investigated using linear regression. the results demonstrate that kpk households reside on average in low category of actual education level experiencing high interregional and intra-regional disparities and exclusions. at the district levels, the inequalities in educational achievement and exclusions are even more pronounced, indicating that aggregated analyses suppress the intra-regional disparities and segregations. based on these findings, it is asserted that investment in social infrastructure specifically educational, health, and law and order facilities, development of agriculture sector, and eradication of gender discrimination, are important factors to promote inclusive education in the province. keywords inclusive education, household-based education index, inequality coefficient, inclusion coefficient, socioeconomic determinants jel classification c00, c21, i24, i29 1. introduction inclusive education is listed among the top four sustainable development goals (sdgs) in a recent unseco report on sustainable development. it is also considered highly influential towards attaining other sdgs that are aimed at eradicating poverty by 2030 1*corresponding author: nasir84@pide.org.pk mailto:nasir84@pide.org.pk tahira tauheed & muhammad nasir 2 (unesco, 2017). inclusive education could be derived from the definition of inclusive development by rauniyar & kanbur (2010). it is a broad concept that includes enhancement of education level coupled with its equitable distribution in all segments of society, especially the deprived ones. to ensure inclusive education is a world-wide challenge. a substantial proportion of the world population is excluded from the arena of educational achievements (unesco, 2017). masses in pakistan are also suffering from educational exclusion resulting in social unrest and non-coherence (burki, et al., 2015; undp pakistan, 2016; undp, 2016). due to its complex socio-political structure khyber pakhtunkhwa (kpk), the northernmost province of pakistan, faces this challenge more hardly than most of its provincial counterparts (gouleta, 2015). to raise the level of inclusive education, a comprehensive and regionally integrated plan is required. the foremost steps in this regard are to evaluate the present status of education at all possible administrative levels and to identify the factors influencing inclusive education. the existing literature is unable to serve this purpose adequately. based on aggregated data, a few studies including jamal (2016) and pakistan national human development index report (nhdir) (2017) elaborate on the potential level of education (represented by education indices) at the national, provincial, and district levels1. however, these studies are unable to provide information about the distributional aspects of educational development including disparities and inclusion, especially at the intra-district level. there is hardly any study that empirically analyze the inclusiveness of education in kpk. some studies analyze the education system of kpk. zia uddin and tahir (2014) has investigated the primary data on the public-school monitoring in kpk. ahmed & rashid (2018) in a research study has analyzed the performance of public schools in districts swat and lower dir of kpk. gouleta (2015) analyzed kpk’s educational assessment policies and practices. a comprehensive measurement of inclusive education and analysis of its determinants remains limited for kpk. health and education are central in building people’s physical and mental capacities, and hence any serious inequality of opportunity in these areas will aggravate inequality in their future (kato, 2014). keeping in view the significance of the subject matter and the research gap, this study undertakes the task of executing a household-based analysis of educational achievements of kpk at the provincial and district levels. the data used in this study is taken from the latest pakistan social and living standard measurement (pslm) survey 2014-15. the analyses are further extended to urban and rural regions at the provincial levels. this study also examines factors (including economic, social, demographic, and locational factors) influencing inclusiveness of education at the district level. the three underlying aspects of inclusive education: education enhancement, inequality reduction and inclusion of the marginalized are covered in this study by measuring and analyzing household-based education index (iie), inequality 1 these studies are mentioned specifically as these are utilizing same measure of human development (hdi) and same data source (pslm 2014-15) as the present study. journal of applied economics and business studies, volume. 4, issue 1 (2020) 1-28 https://doi.org/10.34260/jaebs.411 3 adjusted education index (iie), loss due to inequality, and coefficient of inclusiveness. the classical linear regression model (clrm) is employed for analyzing the determinants of all aspects of inclusive education. kpk is the northernmost province of pakistan with a privileged geostrategic position and abundant natural and human resources. majority of the population in kpk resides in rural area and its main economic sources are forestry and agriculture. however, the current level of human development reveals that it could not exploit its advantageous position and abundant resources successfully to construct a shared and resilient society. a number of challenges including influx of afghan refugees, high incidence of terrorist activities, and a very complex and diversified socio-political structure could be accounted for its low performance (gouleta, 2015). the picture of education in kpk presents a gloomy outlook. according to pakistan education statistics (2017) kpk currently has about 2.38 million children of the ages of 5 to 16 that are out of schools. economic survey of 2016-17 reports that the adult literacy rate in kpk is 53 per cent since 2012, depicting that 47 per cent of adults in the province are illiterate. according to world development bank, in 2015 kpk is well ahead of islamabad capital territory, all four provinces, and other regions in the country in term of gross intake rate (gir) (hunter, 2020). it shows that effort has been started, however, a lot has to be done. as pakistan nhdr (2017) reports for the same year, the achievement level of kpk in terms of its provincial and district level educational indices depicts that it is just ahead of fata and balochistan and falls substantially below the national level, azad jammu & kashmir, islamabad, and punjab. the national level and kpk’s education indices are 0.538 and 0.49 and these lie in medium and low medium categories, respectively . kpk’s overall development level including the standard of living, health, and education reveals that it falls in medium category and is slightly ahead of national level achievements. a comparison of development in three dimensions of human development exhibits that education is the least achieved dimension in kpk. the consequences of low education manifest themselves in lower achievements in other social and economic dimensions. education can drastically change the growth and development cycle of a region, as established by the east asian countries during the 1990s (najam & bari, 2017). according to 18th amendment to the constitution of pakistan the education has become a provincial subject. therefore, the provinces should form statutes and articulate educational policies that guarantee the best education system. despite the fact that government of kpk spends substantially higher on education than any other province in the country and it has passed “khyber pakhtunkhwa right of children to free and compulsory education act 2017”, province is unable to meet the millennium development goals (mdgs) in education due to high prevalence of extreme poverty (gouleta, 2015). to emerge as a stable society tahira tauheed & muhammad nasir 4 in the twenty-first century and to get advantage of its increasing population, kpk must provide a skilled and educated workforce. to achieve this end, serious efforts are required, and a careful examination of present status of inclusive education is necessary to understand the state of the world today. furthermore, based on these analyses it is imperative to design a new development framework for the future. the foremost step in this regard is to formally and methodically assess the existing status of education and level of its inequality & inclusion. the next step is to inspect the factors influencing inclusive education so that appropriate policies and action plans could be designed. there are three main segments of this study. first is the empirical assessment of existing level of education and prevailing inter-regional and intra-regional educational inequalities. to achieve this end, household-based ie and iie for the year 2014-15 are constructed at the provincial and district levels, using data from pslm (2014-15) on the lines proposed by alkire and foster (2010). regional indices (rural and urban) are also constructed at the provincial level. the atkinson’s measure of inequality and loss in educational achievements due to this inequality are also estimated. the second segment supplements this analysis by examining the profile of inclusive education at provincial and district levels using a unified measure of inclusion, ‘coefficient of inclusive education’. it is based on distribution of households’ education indices computed in the first segment. rural and urban coefficients of inclusiveness are also estimated at the provincial level. the third segment comprises of empirical analysis of the prerequisites of inclusive development. to determine the proximate factors that influence inclusive education, district-wise education indices, inequality coefficient and coefficients of inclusive development are regressed on various economic, social, demographic, and locational factors considered to be influential for inclusive development in literature. the major contributions of this study are: first, it is a leading study in estimating education index at the household level (the smallest possible unit for which required data is available in pakistan). this study is also credited for being pioneer in constructing household-based provincial and district level education and inequality adjusted education indices for kpk. all the previous works on this subject involve aggregated data at a certain level that suppress the inter-regional variations. several factors that play a vital role in raising disparities at micro level have remained unaddressed. second, for the first-time across households’ inequalities in education indices at the provincial (overall, rural, and urban) and district levels are estimated for kpk. third, to our knowledge, it is a seminal work that calculates a unified measure of inclusive education. as a unified measure of inclusive education, coefficient of inclusiveness is an efficient tool for the analysis of its dynamics and determinants. forth, this study investigates and identifies major economic, social, demographic, and locational determinants of inclusive education at the district level. journal of applied economics and business studies, volume. 4, issue 1 (2020) 1-28 https://doi.org/10.34260/jaebs.411 5 this study is also important in the wake of adoption of sdgs and vision 2030 by pakistani government, and devolution resulting from pakistan's 18th constitutional amendment. the district level study of inclusive development and its determinants would assist local and provincial governments in identifying areas and sectors that require greater attention, enabling them to allocate resources accordingly. last but not the least, this study is also expected to generate dialogue and further research to deepen the understanding of the dynamics and key drivers of inclusive education in kpk. the rest of the paper is organized as follows: section 2 presents the data sources and research methods utilized in this work. this section also describes the procedures to estimate education indices and their distributional inequality. section 3 provides analyses of the estimated education indices and their inequality coefficients. section 4 outlines the estimation of inclusiveness coefficients of education and presents its analysis . section 5 presents the analyses of determinants of inclusive education. finally, section 6 highlights the conclusions of the study and lays out recommendations for policy and future research. 2. data and research methodology the main data utilized in this study is taken from the latest pakistan social and living standard measurement (pslm) survey 2014-15. it is a district as well as provincial and national level representative survey which covers 78635 households. (pakistan bureau of statistics, 2016). most of the district-wise data for the determinants of inclusive education is collected from various publications of pakistan bureau of statistics for the years 2014 and 2015. the data about education and health institutions, total area, forest area, cultivated area, road length, registered factories, police stations, and reported crimes for year 2014-15 is collected from kpk development statistics 2015 and 2017. data about population and sex ratio is collected from pakistan census 2017 as these figures are close approximates for year 2014-15. for detail description of data see table a.1 and a.2 in appendix. the general methodology utilized here to construct the household-based education index is taken mainly from lopez-calva & ortiz-juarez (2011). technical notes for human development reports (2014; 2015) are consulted for details of index construction, inequality measurement, and loss due to inequality. traditional component of education index are adult literacy and enrollment indicators. however, the household-based calculation of enrollment imposes the problem of missing data in households without children, as enrollment depends on the presence of individuals of school going age. in this work education index is calculated by replacing enrollment with a continuous variable capturing the years of schooling for individuals of or above the age of 7 (the age required to complete the first year of primary education). using this variable, missing values are avoided by imputing the household i’s average schooling to children below the age of 7, under the assumption that children could tahira tauheed & muhammad nasir 6 achieve at least such average over the course of their lives (lopez-calva & ortiz-juarez, 2011). household’s achievements are normalized to a score between 0 and 1 using extreme values across country, called the domestic goalposts. hence education indices are contextualized regarding domestic goalposts to consider the national realities and priorities. domestic goalposts provide a realistic assessment of the relative educational progress made by households and districts in kpk. information about schooling years and adult literacy of a household’s members are collected from section ‘c’ of pslm 2014-15. to construct schooling index of a household, firstly for each household member of age 7 years or above, an indicator of the years of schooling is computed and is compared it with a minimum value of zero and a maximum value that depends on age. for instance, a 7-year aged person must have 1 year of schooling as maximum; an 8-year aged person must have 2 years of schooling as maximum, and so on up to a maximum of 18 years of schooling which corresponds to individuals aged 24 or above. if a person aged 7 has 2 or more years of schooling, the value would be fixed up to 1; if a person aged 8 has 3 or more years of schooling, it would be fixed up to 2, and so on. the schooling index for individual j in household i (𝑆𝑐𝑖𝑗 ) is calculated by normalizing his/her schooling years as: 𝑆𝑐𝑖𝑗 = 𝑆𝑐𝑗−𝑆𝑐𝑚𝑖𝑛 𝑆𝑐𝑚𝑎𝑥−𝑆𝑐𝑚𝑖𝑛 -----------------------------------------(1) with 𝑆𝑐𝑗 being the observed years of schooling for individual j, and 𝑆𝑐𝑚𝑖𝑛 and 𝑆𝑐𝑚𝑎𝑥 the reference values. the average of the individual indices is calculated and imputed to children aged below 7 years. the schooling index for household i (𝑆𝑐𝑖) is the average of schooling for all the individuals in that household. in the case of adult literacy, if an adult with or above the age of 15 declared to be able to read and write in any language with comprehension a short simple statement on his/her everyday life, he/she is considered as literate (anon., 1997). hence, the adult literacy index is denoted as the proportion of population aged 15 years and older who can read and write with understanding in any language. household literacy rate (𝑙𝑖) is then calculated as: 𝐿𝑖 = 1 𝑇 ∑ 𝑙𝑗 𝑚 𝑗=1 ------------------------------------------------(2) with t being the total number of adults in household i, m the total number of literate adults, and 𝑙𝑗 an indicator taking the value of 1 if the adult j is literate, and 0 otherwise. this rate is equivalent to the literacy index. the education index for household i ( 𝐸𝑖 ) is computed as weighted average of household’s adult literacy index and schooling index. the weights proposed and used by undp in human development reports 1991-1994 are 2/3 for literacy and 1/3 for schooling. using these weights education index is calculated as: journal of applied economics and business studies, volume. 4, issue 1 (2020) 1-28 https://doi.org/10.34260/jaebs.411 7 𝐸𝑖 = 2 3 𝐿𝑖 + 1 3 𝑆𝑐𝑖--------------------------------------------(3) to conduct a household-based analysis of educational progress across kpk districts, the first task is to construct education index at household level. at next stage this measure is used to analyze the aggregate level education at provincial and district levels. a household’s education index is composed of its adult literacy index and schooling index. information about adult literacy and schooling years of a household’s members are collected from section ‘c’ of pslm 2014-15. for schooling index, at first step data about years of schooling for individuals of or above the age of 7 is collected from three questions. first of these questions is, “what the highest class /level of education is completed?”. the answer to this question comprises twenty different categories (classes/levels) with specific value labels. the years of schooling are assigned to each class/level according to educational system prevailing in the country. at second step the schooling index for everyone of or above the age of 7 is calculated by normalizing his/her schooling years by using equation 1. the schooling indices of all the individuals in a household are averaged out to obtain a household’s schooling index (𝑆𝑐𝑖). to avoid the underestimation of index, all the zero values are replaced by 0.02 under the assumption that individuals have accumulated some learning and experience throughout their lives, regardless of if they have attended school or not. the value of 0.02 is selected arbitrarily keeping in view very low mean years of schooling index in pakistan i.e. 0.3133 according to human development report 2014. moreover, this number involves no truncation of the distribution as the smallest non-zero observed household’s schooling index equals 0.0253. for adult literacy index, information is collected from the question, “can this person read & write in any language with understanding?”. there is no missing response for this question in case of individuals of age 15 years or above. household literacy index (li) is derived by dividing number of adult literates in a household by its total number of adults and normalizing it by natural goal posts of 0 and 1. for the same reasons outlined in the case of schooling, a minimum level of 0.05 is attached instead of 0 in those households with all illiterate adults. this does not truncate the distribution as smallest observed non-zero adult literacy index is 0.0625. education index of a household (ei) is calculated as a weighted average of its schooling index and adult literacy index, assigning weights of 1/3 and 2 /3 respectively. quintiles based on households’ education indices are computed considering the sampling weights. households’ literacy indices, schooling indices, education indices and education quintiles would be provided on request. these education indices and quintiles are utilized for analysis of development in the dimension of education at the provincial and district levels. tahira tauheed & muhammad nasir 8 in the present study to estimate the provincial or district level education index and inequality adjusted education index the general means are utilized for aggregation of household education indices based on (foster, et al., 2005). the inequality in distribution of educational progress across households is captured by an inequality measure suggested by alkire & foster (2010). foster, lopez-calva, and szekely (2005) proposed the use of a general mean or equally distributed equivalent (ede) achievement level for aggregation of achievement (x) to account for inequality in progress/development. the generalized mean can be referred as μα (x), and for a population of size n it is commonly expressed as: --------------(4) where α may take any value in the interval (− ∞, + ∞) (foster, et al., 2013). the general means for α < 1 are generally interpreted as measures of social welfare. the foster-lópezcalva-székely (fls) class of indices satisfies all basic axioms of a welfare index including subgroup consistency and distribution sensitivity (alkire & foster, 2010; seth, 2009). atkinson used the parameter ε = 1 α ≥ 0 (α≤ 1) to index the class of edes; he interpreted ε as an inequality aversion parameter in the aggregation method of achievements (which he considered to be welfare (alkire & foster, 2010). the case of ε = 0 yields the index that is based on the arithmetic mean, which is insensitive to inequality in achievements. the value of ε=1 yields index which is obtained by the geometric mean to evaluate achievements. for ε > 0, the inequality adjusted index discounts for inequality within-dimension according to the level of inequality aversion indicated by its associated parameter ε. in this study the arithmetic mean of households’ indices is employed to obtain a provincial or district education index (𝐼𝐸) without accounting for inequality. it is given as: 𝐼𝐸 = (𝐸1 + 𝐸2 + ⋯ +𝐸𝑛)/𝑛 -------------------------------(5) where as 𝐸𝑖 is the ith household education index and n is the number of households. to obtain inequality adjusted education index ( 𝐼𝑖𝐸) at the provincial or district level the households’ education indices are aggregated by using geometric mean as: 𝐼𝑖𝐸 = √𝐸1 × 𝐸2 × ………….× 𝐸𝑛 𝑛 -------------------------(6) atkinson (1970) family of inequality measures is used to capture inequality in underlying distributions of education across households. atkinson measure of inequality with inequality aversion parameter ε=1 is employed in this research. it can be expressed as: ----------------------------------------------(7) 𝜇𝛼(𝑥) = 𝑥1 𝛼 + 𝑥2 𝛼 + 𝑥3 𝛼 ……… . +𝑥𝑛 𝛼 𝑛 1/𝛼 𝑖𝑓 𝛼 ≠ 0 (𝑥1 × 𝑥2 × …… . .× 𝑥𝑛) 1/𝑛 𝑖𝑓 𝛼 = 0 𝐴𝐸 = 1 − 𝐼𝑖𝐸 𝐼𝐸 journal of applied economics and business studies, volume. 4, issue 1 (2020) 1-28 https://doi.org/10.34260/jaebs.411 9 the inequality measure 𝐴𝐸 represents the share of per household educational achievement that is wasted because of inequalities in its distribution across households. it is regarded as the percentage loss in potential level of educational achievement or welfare arising due to inequitable distribution. to compute unified measure of inclusive development at the district and provincial levels method proposed by suryanarayana (2008) is adopted in this study. the underlying idea is that the growth process under review will be inclusive if it is beneficial for deprived sections of the society. to identify the deprived, this approach compares the achievement of individual units of the society (individuals/ households/ regions) relative to the average achievement of the society. the population having achievement below sixty percent of median achievement of the society is considered as deprived. the same approach is adopted to measure inclusiveness of educational achievement in this study. thus, the segment of population which is deprived of education is defined regarding a threshold of education index, specified as a function of median education index. the population (households) having 𝐸𝑖 below sixty percent of median 𝐸𝑖 is considered as deprived. the 60% of median, and 50% of the mean are two commonly used thresholds for relative income deprivation; the former measure is probably the most extensively used measure nowadays (townsend & kennedy, 2004). the advantage of this threshold is that it will not change by the rise of incomes in the deprived section unless they cross the median income. (mack, 2016; bradshaw & mayhew, 2011). in this study the application of this threshold is extended to measure the proportion of households deprived of education. the deprived of education proportion of population is given as: 𝜃 = 𝐹(𝛿𝐸0.5) = ∫ 𝑓(𝐸)𝑑𝐸 𝛿𝐸0.5 0 ---------------------------(8) where θ = incidence of the deprived (id), 0<δ< 1. the 𝐸0.5 represents the households’ median education index such that: ∫ 𝑓(𝐸)𝑑𝐸 𝐸0.5 0 = 1 2 = ∫ 𝑓(𝐸)𝑑𝐸 ∞ 𝐸0.5 -------------------------(9) the value of δ is kept 0.6. f is the cumulative distribution function and 𝑓(𝐸) is the density function of ‘𝐸𝑖’. some important features and implications are as follows: the value of θ lies in the open interval (0, 0.5). (i) θ tends to 0 implies bottom half of the distribution concentrates in the “inclusion zone”, given by [δe0.50, e0.50] tahira tauheed & muhammad nasir 10 (ii) θ approaches to 0.5 implies bottom half of the distribution concentrates in the “exclusion zone”, given by [0, δ ξ0.50]. assuming society consisting of a homogeneous group with heterogeneity in educational achievement across households, a “coefficient of inclusion” is defined by suitable standardization regarding its limits. inclusion coefficient (ic) denoted by ‘ψ’ is given as: ψ = 1 − 2∫ 𝑓(𝐸)d𝐸 𝛿𝐸0.5 0 ---------------------------(10) where 0 < ψ < 1. it has the following relevant properties: (i) the value of ψ tends to the value 0 (unity), when no (all) household (households) with relatively poor educational achievement is (are) participating and hence, benefiting from the mainstream educational progress, implying a state of perfect exclusion (inclusion) (ii) a value of ψ greater (less) than ½ , indicates a situation where the proportion of the bottom half of the population falling in the inclusion zone is greater (less) than the proportion in the relative deprivation-zone, implying a state of inclusion (exclusion). the economic and social welfare is not evenly distributed across regions in pakistan (jamal, 2016; undp pakistan, 2016), exhibiting a scenario of non-homogeneous society. consequently, inclusiveness of education in kpk is analyzed in two ways i.e. across the regions (inter-regions) and within the regions (intra-region). inter-regional inclusion is examined with reference to disparities in median levels education across regions. it is measured by closeness of regional median ( 𝐸0.5 𝑅 ) to national median 𝐸0.5 𝑀 (of the national/mainstream population). for a given 𝛿 such that 0 < 𝛿 < 1, there can be two scenarios: (i) 𝐸0.5 𝑅 < δ𝐸0.5 𝑀 implies exclusion of the specific region. (ii) 𝐸0.5 𝑅 ≥δ𝐸0.5 𝑀 implies inclusion of the specific region. intra-regional inclusion is examined in terms of inclusion coefficients (ics) defined with respect to regional as well as mainstream (provincial) median. intra-regional inclusion for any given region ‘i’ is measured with respect to either own median (𝐸0.5 𝑅 ) providing a measure of 𝛹𝑖 𝑅 (ic regional) or mainstream median (𝐸0.5 𝑀 ) providing a measure of 𝛹𝑖 𝑀 (ic mainstream). these two measures are distinct and different for situations when there is inter-regional exclusion; and converge with progressive inter-regional inclusion. ic regional (𝛹𝑖 𝑅) measures the extent of inclusion of the bottom half population of the region under review in its own progress. its limits and properties are the same as discussed for the inclusion coefficient of a homogeneous society. ic mainstream (𝛹𝑖 𝑀) measures the extent of inclusion of the population (laying below national median) of concerned region in the progress of the country/ society. the limits for ic mainstream (𝛹𝑖 𝑀) are as follows: journal of applied economics and business studies, volume. 4, issue 1 (2020) 1-28 https://doi.org/10.34260/jaebs.411 11 𝛹𝑖 𝑀= -1 implies exclusion of the entire region 𝛹𝑖 𝑀 = 1 implies inclusion of the entire region an important objective of this study is to analyze the determinants of inclusive education. the determinants of inclusive education are analyzed at district level by estimating regression models for three of its aspects i.e. level of educational achievement, its distributional inequality, and inclusion of marginalized in educational progress. to achieve this end kpk districts’ education index (ie), inequality coefficients (ae), icmainstream (𝛹𝑖 𝑀), and ic-regional (𝛹𝑖 𝑅) are regressed on various potential factors for inclusive education. the selection of these probable determinants is based on evidence from existing literature and availability of data at district level for kpk. the choice of variable is constrained severely due to data availability at district level. inclusive development is influenced by several diversified factors, however, the factors considered in this study are grouped in to four major categories, economic factors (ef); social factors (sf); demographic factors (df); and locational factors (lf). the generalized form of the model is given below: inclusive education= f (ef, sf, df, lf) ----------------------------(11) economic factors comprise of industrial development (measured by no. of registered factories per hundred thousand of population), agricultural development (measured by percentage of cultivated area), percentage of forest area, and level of physical infrastructure (measured by road density, airport, and railway station) at district level. social factors include public education and health facilities, and law & order condition at district level. public education facilities are proxied by district-wise total number of government schools, high schools, middle schools, primary schools, and colleges per hundred thousand population. district-wise number of government hospitals per hundred thousand population and number of beds in government health institutions per ten thousand population are utilized to proxy public health facilities. law and order facility are assessed by number of police stations per hundred thousand population at district level. demographic factors utilized in this study are the population density, ratio of male to female population (sex ratio), and urbanization (ratio of urban population to total population). locational factor include the dummy variable for divisional capital. the functional forms of the regression models are given as: ie = 𝛼1 + 𝛽1 ef + 𝛾1 sf + 𝛿1 df+ ξ1 lf + ε1 -----------------------(12) ae = 𝛼2 + 𝛽2 ef + 𝛾2 sf + 𝛿2 df+ ξ2 lf + ε2-----------------------(13) 𝛹𝑖 𝑀= 𝛼3 + 𝛽3 ef + 𝛾3 sf + 𝛿3 df+ ξ3 lf + ε3----------------------(14) 𝛹𝑖 𝑅= 𝛼4 + 𝛽4 ef + 𝛾4 sf + 𝛿4 df+ ξ4 lf + ε4----------------------(15) where ε1, ε2, ε2, and ε4 represents the random error terms in the models. tahira tauheed & muhammad nasir 12 in this study classical linear regression model (clrm) is employed to estimate the above stated equations individually. the sur could be an appropriate technique for the estimation of this system of equation. however, selection of clrm to estimate individual equation is based on the two reasons. first, greene (2005) argues if the set of regressors is same across the two (or more) dependent variables, the outcomes from sur will be identical to those from ols (greene, 2005). second, the sample size in this study is too small to restore reasonable degrees of freedom in estimating a system of equation with large number of coefficients. to produce robust estimates, possible violations of the assumptions of the clrm relevant for cross-sectional data are explored. shapiro-wilk test is utilized to check the normality of residuals since it is recommended the best choice for testing the normality of data by some researchers (thode, 2002). to deal with the possibility of heteroskedasticity, robust standard errors (heteroskedasticity-consistent) are utilized as it is a common and popular technique in this respect (berry, 1993). multicollinearity is tested by analyzing the variance inflation factor (vif) that is the most extensively used diagnostic for multicollinearity (allison, 2012). the data issues pose a serious limitation on testing endogeneity of the model as the appropriate instruments could not be found for the district level data. therefore, to establish the causality between inclusive education and its determinant is beyond the scope of this study. in this scenario the objective of this work is to identify the significant covariates of inclusive education. 3. analysis of education index and its inequalities for aggregated analysis of educational progress households’ education indices are classified in to five categories. these categories/classes of education index are very low, low, medium, high, and very high. the cut-off values for classes of education index are determined by five provincial quintiles of households’ education indices following the lines proposed by undp (2014). the various categories of household’s education index along with their cut off values are given in table 1. a simple comparison of lowest and highest cut off values reveals substantial disparities in education indices of households. the gap between maximum and minimum cut off for education index it is 0.74. the wide differences in education indices across and within quintiles show the prevalence of high educational disparities across households. table 1: categories of educational achievements with cutoff values categories of education index cutoff values (based on quintiles of ie) very low less than 0.19 low 0.19 to 0.50 medium 0.51 to 0.69 high 0.70 to 0.93 very high greater than 0.93 source: authors’ calculations journal of applied economics and business studies, volume. 4, issue 1 (2020) 1-28 https://doi.org/10.34260/jaebs.411 13 the estimates of education index (ie), inequality-adjusted education index (iie), and the estimated percentage losses due to inequalities in educational achievements across households (ae) in kpk and, its rural and urban regions are exhibited in figure 1. these estimates establish the incidence of low actual educational achievements in kpk and of high disparities across regions. the overall kpk education index reveals that kpk falls in low category of potential educational achievements. there is a substantial urban-rural educational disparity in kpk. the urban households’ education index lies in medium category, while rural index falls in low category of educational achievements. the urban households’ education index is 1.4 times higher than that of rural households. figure 1: provincial education indices and inequality measures (source: authors’ calculations) inequality-adjusted education index demonstrate that achievement level in education is affected considerably due to inequalities. the loss is substantial with a varied magnitude in different regions. at provincial level, this loss is estimated around 28 percent. the loss due to inequality is markedly high in rural kpk as compared to its urban counter parts. 0.4681 0.6068 0.4366 0.3372 0.4903 0.3097 0.2796 0.1920 0.2906 0.00 0.20 0.40 0.60 0.80 overall urban rural kpk education index inequality adjusted education index % loss due to inequality tahira tauheed & muhammad nasir 14 table 2: district-wise education indices and inequality measures district education index (ie) inequality adjusted education index (iie) % loss due to inequality (ae) rank ie rank iie change in rank due to inequality haripur 0.6552 0.5660 0.1362 1 1 0 karak 0.5650 0.4758 0.1580 3 2 1 malakand 0.5425 0.4429 0.1836 6 3 3 abbottabad 0.5970 0.4381 0.2662 2 4 -2 chitral 0.5345 0.4292 0.1970 7 5 2 peshawar 0.5462 0.4258 0.2204 5 6 -1 lower dir 0.5047 0.4203 0.1672 10 7 3 mansehra 0.5522 0.4113 0.2552 4 8 -4 lakki marwat 0.5057 0.3982 0.2124 9 9 0 bannu 0.4959 0.3948 0.2038 11 10 1 nowshera 0.5064 0.3871 0.2356 8 11 -3 mardan 0.4541 0.3331 0.2664 13 12 1 kohat 0.4669 0.3295 0.2943 12 13 -1 charsadda 0.4353 0.3214 0.2617 14 14 0 swat 0.4319 0.3197 0.2598 15 15 0 swabi 0.4306 0.3009 0.3014 16 16 0 tank 0.3807 0.2725 0.2843 18 17 1 d. i. khan 0.4121 0.2719 0.3401 17 18 -1 hangu 0.3569 0.2557 0.2836 20 19 1 upper dir 0.3584 0.2537 0.2921 19 20 -1 batagram 0.3341 0.2326 0.3039 21 21 0 buner 0.3044 0.2096 0.3115 22 22 0 shangla 0.2975 0.1924 0.3533 23 23 0 tor ghar 0.2211 0.1395 0.3694 24 24 0 kohistan 0.2152 0.1270 0.4099 25 25 0 source: authors’ calculations the district-wise education indices (ie), inequality-adjusted education indices (iie), and coefficients of inequality (ae) are reported in table 2. there is no district in kpk with potential education index above the medium level of education index or in very low category. the data reveals that potential education index for most of the kpk districts fall in low category, ten districts are in medium category. inequality adjustment pulls down all districts’ education indices to further in low category or in very low category. the estimated inequality coefficients validate the prevalence of wide disparities within districts. in ranking of both education indices with and without inequality adjustment; haripur is at the top. kpk’s capital peshawar is ranked at 4th place in terms of ie and with inequality adjustment it ranks at 8th place. it is the largest loss of rank among all districts of kpk indicating the intensity of educational inequality in peshawar. it lies in medium category of education and fall in low category with inequality adjustment. kohistan is at lowest rank of education in terms of both education indices ie and iie. it is preceded by tor ghar, shangla, buner and batagram. all these bottom ranked districts are in low category of education and first three of them come down to very low category after accounting for inequality. journal of applied economics and business studies, volume. 4, issue 1 (2020) 1-28 https://doi.org/10.34260/jaebs.411 15 analysis of inequality coefficients reveals that loss due to inequality ranges from the loss of almost 14 percent in haripur to the substantial loss of 41 percent in kohistan. in top five districts maximum loss due to in equality is around 27 percent, in contrast the minimum loss in bottom ten districts is around 41 percent. in general, the magnitude of disparities within districts and across districts rises with deterioration of education level. however, considerable disparities are observed in some top ranked districts too. the kpk districts exhibiting very low level of human development are situated in its north and south. it is observed that mostly districts that have natural resource endowment are in low or very low category of human development. in contrast the majority districts with better status of human development and low disparities are either centers of administration, or are home to small industries, or are hub of commerce and trade. it indicates the skewed utilization of public and private funds, underutilization and wastage of natural resources, and the ignored agriculture sector. there are some obvious socio-political factors that could be responsible for adverse human development status in certain regions. in kpk the terrorist activities, afghan refugees, and armed conflict specifically in southern districts are some of the probable reasons for poor human development situation (akbar, 2015; khattak, 2017; yousaf, 2013). to address these issues further research at regional levels is required. the distribution of districts in categories of education, according to their education index (ie) and inequality-adjusted education index (iie) is given in table 3. table 3: distribution of districts in categories of education index education category districts (according to ie) districts (according to iie) number percentage number percentage kpk very high 0 0 0 0 high 0 0 0 0 medium 9 36 1 4 low 16 64 22 88 very low 0 0 2 8 source: authors’ calculations the findings of this study are comparable with the facts presented by pakistan nhdr (2017), see table a.3 in appendix. according to this report most of the kpk districts’ education indices falls in low medium or low categories, the same result is established by the present study. majority of the districts ranks according to the education index (ie) is almost similar in both studies. the interesting comparison is between the education index (ie)-wise rank of districts cited by the nhdr and the inequality adjusted inequalityadjusted education index (iie)-wise rank given by the present study. the inequality adjustment changes the education ranking of the districts remarkably. for example, abbottabad that stands first in nhdr in terms of education index slides down to fourth tahira tauheed & muhammad nasir 16 place with inequality adjustment calculations in the present study. these findings urge for policy measures focused not only to elevate the average educational achievements but also to alleviate the educational inequalities simultaneously. a mapping of kpk districts’ inequality adjusted education and human development indices and their ranks respectively calculated by present study and pakistan njdr (2017) is cited in appendix table a.4. it reveals substantial differences in the level of education and overall human development level in majority of kpk districts. district karak with hdi ranking of 13 and iie ranking of 2 is showing the highest divergence. at the lowest end for the districts of shangla, tor ghar, and kohistan this deviation is minimal. the considerable inequalities in districts’ hdi and iie suggest that the dimensions of hdi are not perfect substitutes of each other. 4. inclusiveness analysis of education the third important aspect of inclusive education in a society i.e. inclusion of marginalized, is examined by utilizing distribution of households’ education indices. to determine the inter-regional inclusion status the median education index for each region is compared to sixty percent of national median education index. the value of national median education index estimated in this study is 0.5714. the comparison demonstrates that overall, urban, and rural regions in kpk are inclusive in mainstream education, see table 4. inclusiveness describes that at least some proportion of households in bottom half (below median education index) of these regions have education index that fall in inclusion zone. table 4: provincial level estimates of inter-regional inclusion/exclusion in terms of education index education median education index inter-regional inclusion/exclusion* kpk overall 0.4707 inclusion__ urban 0.6143 inclusion rural 0.4306 inclusion source: authors’ calculations *criterion for inter-regional inclusion is regional median>= (0.6*national overall median). table 5 displays inter-district inclusion/exclusion in terms of education index in kpk. out of 25 districts of kpk 18 are exhibiting inter-regional inclusion. hence, 74 percent and 28 percent of the total kpk districts demonstrates the inter-regional inclusion and exclusion, respectively. the districts with very low median education indices showing exclusion include hangu, upper dir, batagram, buner, shangla, tor ghar, and kohistan. it implies that education index of all households in these districts fall in exclusion zone. journal of applied economics and business studies, volume. 4, issue 1 (2020) 1-28 https://doi.org/10.34260/jaebs.411 17 table 5: district-wise estimates of inter-regional inclusion/exclusion in terms of education index district median education index inter-regional inclusion/exclusion* haripur 0.6788 inclusion abbottabad 0.6667 inclusion mansehra 0.5714 inclusion karak 0.5714 inclusion chitral 0.5556 inclusion malakand 0.5417 inclusion lower dir 0.5333 inclusion peshawar 0.5333 inclusion lakki marwat 0.5238 inclusion nowshera 0.5000 inclusion bannu 0.5000 inclusion mardan 0.4675 inclusion kohat 0.4630 inclusion swabi 0.4444 inclusion charsadda 0.4287 inclusion swat 0.4222 inclusion tank 0.3968 inclusion d. i. khan 0.3662 inclusion hangu 0.3353 exclusion upper dir 0.3333 exclusion batagram 0.2917 exclusion buner 0.2540 exclusion shangla 0.2500 exclusion tor ghar 0.1444 exclusion kohistan 0.1167 exclusion source: authors’ calculations intra-regional inclusiveness analysis at the provincial and district levels is presented in figures 2 and 3, respectively. the estimate of provincial ic establishes that 75 percent of the lower half of households (with education index below median) in kpk falls in exclusion zone of education, consequently the percentage of inclusion is markedly low. the substantial rural-urban disparities are evident in inclusiveness of education. in urban region mainstream inclusion is more than three times higher than its rural counterpart. in urban region and rural regions respectively, 59 percent and 18 percent of the bottom half of households lie in mainstream inclusion zone. the regional inclusion of rural region is less than two-third of that of the urban region. the ic-regional for rural households and urban households is 56 percent and 34 percent, respectively. tahira tauheed & muhammad nasir 18 figure 2: provincial level estimates of education’s regional and mainstream inclusion coefficients (source: authors’ calculations) the intra district inclusion analysis reveals that in kpk only 2 out of 25 districts are in state of inclusion with respect to ic-mainstream and seven districts with negative icmainstream exhibit nearly perfect exclusion. in kpk, haripur district has the highest level of mainstream inclusion and regional inclusion of 73 percent and 66 percent respectively. the second highest level of mainstream and regional inclusion is exhibited by district karak for which both measures have the same value of 0.58. provincial capital peshawar also exhibit mainstream exclusion with ic-mainstream at 0.43 and ic-regional at 0.50. the districts with the lowest level of inclusion with respect to mainstream education and regional education are tor ghar (-47 %) and kohistan (-45 %), respectively. it shows that education indices of almost 97 percent and 95 percent of households of tor ghar and kohistan falls in exclusion zone of education, respectively. 0.56 0.37 0.34 0.59 0.25 0.18 0.00 0.50 1.00 kpk urban kpk overall kpk rural inclusion coefficient regional inclusion coefficientmainstream journal of applied economics and business studies, volume. 4, issue 1 (2020) 1-28 https://doi.org/10.34260/jaebs.411 19 figure 3: district-wise regional and mainstream inclusion coefficients of education (source: authors’ calculations) 5. analysis of determinants of inclusive development analyzing the determining factors of inclusive education is a prerequisite to identify critical areas for optimal utilization of available resources (oluseye & gabriel, 2017). one of the objectives of this study is to identify the factors that could ensure and enhance the inclusive education across districts in pakistan. the district level diagnosis recognizes the most significant local factors that boost or hampers the inclusive education. this analysis provides a base to suggest appropriate policies. the methodology for the analysis is discussed in section 2. in this regression analysis dependent variables are the indicators of inclusive education estimated by this study and the independent variables are factors that influence these indicators. the individual regression models are estimated for indicators of inclusive education representing its three aspects including education index, inequality coefficient, and inclusiveness coefficients (mainstream & regional). a set of 18 variables mentioned in section 2 is initially selected to include in regression analysis. the descriptive statistics of these regression variables are cited in tables a.1 and a.2 in appendix. the descriptive analysis shows high variability in the regressors, it improves the precision with which the parameters are estimated (anon., 2016). the four regression models presented by equations 12-15 are estimated in this study by utilizing stata 13. the assumptions of clrm vital for cross-sectional data are identified -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 h a ri p u r k a ra k m a la k a n d c h it ra l a b b o tt a b a d l a k k i m a rw a t p e sh a w a r l o w e r d ir m a n se h ra b a n n u n o w sh e ra m a rd a n k o h a t s w a t s w a b i c h a rs a d d a t a n k d . i. k h a n h a n g u u p p e r d ir b a ta g ra m s h a n g la b u n e r k o h is ta n t o r g h a r part (a): kpk's districts regional inclusion coeffient mainstream inclusion coefficient tahira tauheed & muhammad nasir 20 and are taken care of to generate robust estimates of regression coefficients. the robust standard errors are used to address the probable prevalence of heteroscedasticity (williams, 2015). to estimate a parsimonious clrms for the education index, inequality coefficient, ic-mainstream, ic-regional are estimated by including different subsets of independent variables from initially selected 17 variables. to control for high multicollinearity and to restore degree of freedom final models are selected on the basis of vif values, overall significance of the model, and the model selection criteria of akaike and schwarz. in these selected models individual and mean vifs are less than 2.5 and residuals are approximately normal and homoskedastic. the estimated regression models are reported in table 6. the regression results demonstrate that being divisional headquarter, public investment in social infrastructure, eradication of gender bias, development of agriculture sector, and urbanization are the significant determinants of one or more aspects of inclusive education in districts of kpk. the public expenditure on health and education have a significant impact on all aspects of inclusive education. the impact of physical infrastructure ( road density) is found to be statistically insignificant for all aspects of inclusive education. according to analysis the inequality coefficients of districts which are divisional headquarters are significantly high. it implies that in administrative centers the proportions of households excluded from mainstream of educational achievement is higher than other districts. the regression results provide an empirical evidence for the strong role of social infrastructure to ensure a higher level of inclusive education. it is asserted by regression finding that total number of government /government high schools (an indicator of public education facilities) has a substantial and statistically significant effect on all indicators of inclusive development. the substantial effect of public spending on education provide an empirical evidence for multiplicity of benefits that investment in education yields (mitra, 2011). this finding leads to the policy recommendation of keeping education at the highest priority in development agenda. the regression results revealed that impact of number of hospital beds (an indicator of public health facilities) is statistically significant on all aspects of inclusive education in kpk districts. these findings are in line with the existing studies that witness the positive significant role of public health facility (james, 2016). the police stations as institutions for safeguarding law and order in the society are also found to be a significant factor of mainstream inclusive education. a noticeably clear picture of relationship between education status and sex ratio (gender discrimination) is portrayed from present analysis. the negative impact of high sex ratio on educational achievements leads to lower inclusive education. it signifies that female inclusiveness is a key prerequisite for inclusive education. it is depicted by regression results that cultivated area has a statistically significant positive effect on education index. a larger percentage of cultivated area leads to higher educational achievement. journal of applied economics and business studies, volume. 4, issue 1 (2020) 1-28 https://doi.org/10.34260/jaebs.411 21 table 6: regression models for determinants of inclusive education regressand regressor education index inequality coefficient icmainstream ic-regional district hq -0.0211 (0.0320) 0.0460** (0.0186) -0.0659 (0.0641) -0.0479 (0.0315) urbanization 0.0038* (0.0020) -0.0005 (0.0010) 0.0074 (0.0045) 0.0007 (0.0017) sex ratio -0.0046* (0.0024) 0.0004 (0.0014) 0.0034 (0.0064) 0.0002 (0.0028) govt. schools 0.0018** (0.0009) _______ _______ _______ govt. high schools _______ -0.0145*** (0.0043) 0.0782*** (0.0160) 0.0212** (0.0096) beds in govt. health institutes 0.0168** (0.0069) -0.0090** (0.0041) 0.0351*** (0.0132) 0.0146** (0.0070) cultivated area 2.6688** (1.1347) -0.3584 (.6693) 1.7009 (2.7991) 0.8350 (1.2751) road density -0.0013 (0.0014) (-0.0010) (0.0009) 0.0022 (0.0038) 0.0011 (0.0015) police stations -0.0398 (0.0465) 0.0398 (0.0247) -0.2345** (.1028) -0.0454 (0.0523) constant 0.5892 (0.2584) 0.3693 (0.1530) -0.8752 (0.7379) 0.1414 (0.3021) r-squared 0.7215 0.7540 0.8055 0.6365 f value 12.77*** 15.55 *** 21.41*** 11.37*** rmse 0.0738 0.0413 0.1726 0.0816 note: ***, **, * indicate 1%, 5% and 10% significance level. robust standard errors are cited in parenthesis. 6. conclusions and recommendations this paper provides an empirical study of the education inclusiveness in the province of khyber pakhtunkhwa (kpk) in pakistan based on data from the most recent pakistan social and living standard measurement (pslm) survey 2014-15. it finds that in general, the magnitude of disparities and exclusion within districts and across districts in kpk rises with deterioration of education status. however, considerable disparities and/or exclusion in some top ranked districts are also observed. it is noted that mostly districts that are in low or very low category of education index are rich in any one or more than one natural resource endowment including minerals, forests, and cultivable lands. in contrast the majority districts with better status of education, low disparities, and high inclusion are either centers of administration, or home to small industries, or hub of commerce and trade. it points out tahira tauheed & muhammad nasir 22 to the skewed utilization of public and private funds, underutilization and wastage of natural resources, and the ignored agriculture sector. to achieve a higher level of inclusive education, the regions with different status of education and different hindering factors require different strategies. the policies must be formulated keeping in view all the aspects of inclusive education. the status of inclusive education of a region must be one of the criteria of allocating public funds. in the regions that are at a very low level of education, it is more important to get educational growth acceleration, as the inclusivity of education may have to come later. for the regions with higher level of education accompanied with high disparities and exclusions, a progressive taxation policy would be more effective. for the regions exhibiting high level of education and lower inclusion, it is vital to facilitate the emergence of inclusive institutions. the education policies must be designed that target the balance of educational achievements in urban and rural areas of kpk. a comparison of kpk districts’ human development estimates from pakistan nhdr (2017) and inequality adjusted education level estimated by present study suggests that the dimensions of human development are not perfect substitutes of each other. these results suggest formulating public policies that focus on balanced development in all the dimensions of human development. it is recommended to allocate reasonable percentage of provincial and local bodies budget for the enhancement of quantity and quality of social infrastructure ( specifically basic education, health, and law and order). based on highly significant negative impact of sex ratio (utilized as an indicator of gender discrimination) on inclusive education it is recommended to formulate effective policies for elimination of gender bias at all levels. the increasing trends of urbanization in kpk and the empirical evidence of its positive significant impact on inclusive education in this study imply that policies must be formulated to control these factors so that their negative effect could be avoided in the long run as well. as a major proportion of population in kpk is dependent on agriculture sector, its development must be at top priority in inclusive education agenda. a policy framework must be designed for uplifting and mechanizing agriculture sector. to reform the police department is also one of the important implications of the present analysis. there are some major limitations of this study that must be acknowledged. some important economic factors of inclusive development could not be included in the analysis due to unavailability of data at the district level. for many indicators the available data is not standardized across provinces. it is suggested to formulate policy for collection and standardization of data about macroeconomic indicators at provincial, districts, and subdistrict level. journal of applied economics and business studies, volume. 4, issue 1 (2020) 1-28 https://doi.org/10.34260/jaebs.411 23 this study provides an empirical analysis of the existing status of inclusive education and its determinants in kpk in the best possible way. however, a great deal of additional research is required in this direction. some recommendations for future research are the natural extensions of this study. first is the dynamic and comparative static analysis of inclusive education in kpk and other provinces of pakistan utilizing different rounds of pslm-hies. the second is to estimate the inequality coefficient and coefficient of inclusion by utilizing various non-conventional values of risk aversion parameter and threshold for deprivation and compare its findings with that obtained by conventional measures utilized in the present study. third is to include the qualitative aspects of education in its index. the forth is to utilize various measures of inequality such as gini coefficient in addition to the atkinson’s inequality index to measure the inequalities and compare the results. to recommend policies in accordance to each specific region and administration level, it is suggested for future studies to investigate the factors of inclusive development that could not be covered adequately in the present research specifically the institutions, economy, local customs and traditions, and geography (spatial analysis). the literature highlights some socio-political factors that could be responsible for adverse inclusive education status in certain regions (as indicated by the present study) of kpk. it is recommended for further research to execute case studies for specific regions to explore the impact of these factors on inclusive development. references ahmed, a. & rashid, s., 2018. citizen report card study education sector swat and lower dir khyber pakhtunkhwa , karachi: transparancy international pakistan. akbar, a., 2015. over 600,000 afghan refugees in kp. [online] available at: https://www.dawn.com/news/1167103 alkire, s. & foster, j., 2010. designing the inequality adjusted human development index (hdi), oxford: oxford poverty and human development initiative. allison, p., 2012. when can you safely ignore multicollinearity?. [online] available at: https://statisticalhorizons.com/multicollinearity anon., 1997. national literacy policies pakistan. [online] available at: http://www.accu.or.jp/litdbase/policy/pak/ atkinson, a. b., 1970. on the measurement of inequality. journal of economic theory 2(3), pp. 244-263. berry, w. d., 1993. understanding regression assumptions. london: sage publication, inc.. bradshaw, j. & mayhew, e., 2011. the measurement of extreme poverty in europian union. york: european commission. tahira tauheed & muhammad nasir 24 burki, a. a., memon, r. & mir, k., 2015. multiple inequalities and policies to mitigate inequality traps in pakistan, lahore: oxfam. foster, j. e., lópez-calva, l. f. & székely, m., 2005. measuring the distribution of human development: methodology and an application to mexico. journal of human development. foster, j., seth, s., lokshinl, m. & sajaia, z., 2013. a uniform approach to measuring poverty and inequality theory and practice. washington, dc: world bank. gouleta, e., 2015. educational assessment in khyber pakhtunkhwa pakistan’s north-west frontier province: practices, issues, and challenges for educating culturally linguistically diverse and exceptional children. global education review, 2(4), pp. 19-39. greene, w. h., 2005. econometric analysis. 5th ed. s.l.:pearson education. jamal, h., 2016. quantifying sub-national human development indices from household survey data, karachi: social policy and development centre. james, c., 2016. health and inclusive growth: changing the dialogue. s.l.:world health orgnization. kato, h., 2014. executive summary. in: perspectives on the post-2015 development agenda. tokyo: japan international cooperationagency research institute, pp. 1-15. khattak, d., 2017. reviewing pakistan's anti-terror fight, 3 years after the peshawar school attack. [online] years-after-the-peshawar-school-attack/ lopez-calva, l. f. & ortiz-juarez, e., 2011. https://www.springer.com: springer science+business media b.v. 2011 . mack, j., 2016. income threshold approach. [online] available at: http://www.poverty.ac.uk/definitions-poverty/income-threshold approach mitra, d., 2011. pennsylvania’s best investment: the social and economic benefits of public education. [online] available at: https://www.elcpa.org/wpcontent/uploads/2011/06/bestinvestment_full_report_6.27.11.p df najam, a. & bari, f., 2017. pakistan human development index report, 2017, islamabad: undp, pakistan. oluseye, i. c. & gabriel, a. a., 2017. determinants of inclusive growth in nigeria: an ardl approach. american journal of economics, pp. 97-109. pakistan bureau of statistics, 2016. pslm 2014-15, islamabad: pakistan bureau of statistics. rauniyar, g. & kanbur, r., 2010. conceptualizing inclusive development: with application to rural infrastructure and development assistance. ithaca(newyork): department of applied economics and management, cornell university, ithaca, newyork. seth, s., 2009. inequality, interactions, and human development. s.l.:oxford poverty & human development initiative (ophi). suryanarayana, m. h., 2008. what is exclusive about 'inclusive growth'?. economic & political weekly, 25 october, pp. 93-101. thode, h. c., 2002. testing for normality. new york: marcel dekker, inc.. http://www.poverty.ac.uk/definitions-poverty/income-threshold journal of applied economics and business studies, volume. 4, issue 1 (2020) 1-28 https://doi.org/10.34260/jaebs.411 25 townsend, i. & kennedy, s., 2004. poverty: measures and targets. [online] available at: http://www.parliament.uk uddin, z. & tahir, m., 2014. monitoring of the delivery of educational services: a case study of sovernment schools in khyber pakhtunkhwa. the dialogue, volume ix, no. 3, pp. 271284. undp pakistan, 2016. development advocate pakistan, volume 3, issue 2, inequality: missing from the public agenda, islamabad: undp pakistan. undp, 2014. technical notes. [online] available at: http://hdr.undp.org/sites/default/files/hdr14_technical_notes.pdf undp, 2015. technical notes. [online] available at: http://hdr.undp.org/sites/default/files/hdr2015_technical_notes.pdf undp, 2016. human development report 2016: human development for everyone, new york: undp. unesco, 2017. a guide for ensuring inclusion and equity in education. [online] available at: https://www.european-agency.org/sites/default/files/news/newsfiles/unesco%20guide.pdf [accessed april 2020]. williams, r., 2015. heteroskedasticity. [online] available at: https://www3.nd.edu/~rwilliam/stats2/l25.pdf yousaf, n., 2013. kyber pukhtunkhawa's sad south. [online] available at: https://www.dawn.com/news/1029031 https://www.european-agency.org/sites/default/files/news/news-files/unesco%20guide.pdf https://www.european-agency.org/sites/default/files/news/news-files/unesco%20guide.pdf tahira tauheed & muhammad nasir 26 appendix table a.1: descriptive statistics of variables in regression analysis of inclusive development variable mean minimum maximum s.d. education index (ie) 0.4442 0.2152 0.6552 0.1119 inequality coefficient (ae) 0.2627 0.1362 0.4099 0.0666 ic-mainstream 0.1950 -0.4699 0.7327 0.3131 ic-regional 0.4094 0.2095 0.6642 0.1083 forest density (percentage) 33.4716 0.0000 86.946 25.739 population density (per sq. km) 676.3255 30.1254 3396.24 678.29 urban population (percentage) 13.0295 0.0000 46.1468 11.168 sex ratio (male to female) 102.2356 92.6900 124.360 5.5109 total no. of govt. schools 104.7996 30.8497 181.509 34.034 no. of govt. primary schools 87.4001 23.9630 150.017 29.933 no. of govt. middle schools 9.7419 3.6073 19.0003 3.1487 no. of govt. high schools 7.6576 2.9172 16.3179 2.8099 no. of colleges 1.8197 0.0000 2.8317 0.7508 no. of hospitals 0.6377 0.0000 1.7503 0.3589 no. of beds in govt. health institutions 5.2704 0.0000 12.4336 2.8171 cultivated area 0.0311 0.0000 0.0736 0.0201 no. of factories 6.0789 0.0000 21.5085 6.5550 road density (per sq. km) 29.7069 5.5793 71.3936 15.955 no. of police stations 1.1245 0.5927 2.6824 0.4616 note: number of schools, hospitals, factories and police stations are reported as per hundred thousand of population, and cultivated area is reported as a percentage of total area. table a.2: frequency distribution of categorical determinants of inclusive development variable frequency relative frequency divisional headquarter no 20 80 yes 5 20 railway station no 11 44 yes 14 56 airport no 4 16 yes 21 84 journal of applied economics and business studies, volume. 4, issue 1 (2020) 1-28 https://doi.org/10.34260/jaebs.411 27 table a.3: comparative analysis of education indices of kpk districts district inequality adjusted education index (iie) education index (ie) present study present study pakistan nhdr 2017 haripur 0.5660 0.6552 0.6367 karak 0.4758 0.5650 0.5567 malakand 0.4429 0.5425 0.5750 abbottabad 0.4381 0.5970 0.6400 chitral 0.4292 0.5345 0.5500 peshawar 0.4258 0.5462 0.5833 lower dir 0.4203 0.5047 0.5033 mansehra 0.4113 0.5522 0.5500 lakki marwat 0.3982 0.5057 0.5117 bannu 0.3948 0.4959 0.5133 nowshera 0.3871 0.5064 0.5033 mardan 0.3331 0.4541 0.5117 kohat 0.3295 0.4669 0.4950 charsadda 0.3214 0.4353 0.4683 swat 0.3197 0.4319 0.4600 swabi 0.3009 0.4306 0.4817 tank 0.2725 0.3807 0.4000 d. i. khan 0.2719 0.4121 0.4033 hangu 0.2557 0.3569 0.3850 upper dir 0.2537 0.3584 0.3600 batagram 0.2326 0.3341 0.3533 buner 0.2096 0.3044 0.3717 shangla 0.1924 0.2975 0.3083 tor ghar 0.1395 0.2211 0.2483 kohistan 0.1270 0.2152 0.2483 source: authors’ calculations and national human development index report (2017) tahira tauheed & muhammad nasir 28 table a.4: a mapping of kpk districts' human development indices & education indices district present study pakistan nhdr 2017 difference in iie & hdi ranks inequality adjusted education index (iie) rank hdi rank haripur 0.5660 1 0.732 3 -2 karak 0.4758 2 0.615 13 -11 malakand 0.4429 3 0.69 6 -3 abbottabad 0.4381 4 0.761 1 3 chitral 0.4292 5 0.674 8 -3 peshawar 0.4258 6 0.756 2 4 lower dir 0.4203 7 0.6 15 -8 mansehra 0.4113 8 0.676 7 1 lakki marwat 0.3982 9 0.577 17 -8 bannu 0.3948 10 0.613 14 -4 nowshera 0.3871 11 0.697 5 6 mardan 0.3331 12 0.703 4 8 kohat 0.3295 13 0.666 9 4 charsadda 0.3214 14 0.65 11 3 swat 0.3197 15 0.618 12 3 swabi 0.3009 16 0.654 10 6 tank 0.2725 17 0.459 21 -4 d. i. khan 0.2719 18 0.496 20 -2 hangu 0.2557 19 0.594 16 3 upper dir 0.2537 20 0.375 23 -3 batagram 0.2326 21 0.505 19 2 buner 0.2096 22 0.528 18 4 shangla 0.1924 23 0.438 22 1 tor ghar 0.1395 24 0.24 24 0 kohistan 0.1270 25 0.229 25 0 source: authors’ calculations and pakistan national human development index report (2017) journal of applied economics and business studies, volume. 6, issue 2 (2022) 95-104 https://doi.org/10.34260/jaebs.626 95 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x government trust as a significant determinant of tax compliance: a study of pakistan nabila saddaf1, muhammad tariq mehmood2 , adnan pitafi3 & shahbaz husain4 1 phd scholar, federal urdu university of arts, science and technology, islamabad 2 hod, department of economics, federal urdu university of arts, science and technology, islamabad 3 assistant professor, mehran university of engineering and technology, jamshoro 4 national university of science and technology, islamabad abstract tax compliance behavior of individuals is ascertained in this research. among the variables which affect the tax compliance behavior, government trust and tax justice perception are studied. 300 individuals were taken as sample from pakistan to study the relationship between variables of government trust, tax compliance and tax justice perception. data is obtained through questionnaire and analyzed using reliability test, correlation and regression analysis. the results of the study showed that tax justice perception and government trust are the major determinants of tax compliance behavior in pakistan. tax justice perception plays the role of mediator through government trust to enhance tax compliance. study implicates some useful suggestion that government should always put efforts in gaining public trust by increasing public services, making tax system understandable for everyone by making documentation easier so that more and more people can obey the rules which can increase compliance level. keywords compliance, government trust, tax justice perception jel classification h29, k34 1. introduction taxes are the major components of public expenditure in any economy either it is developing or developed. tax revenues constitute the highest part in public revenues. tax revenue of pakistan in 2021 was 5.7% (tax report by fbr 2021). for the execution of public services tax revenues are very important in a country like pakistan. nabila saddaf, muhammad tariq mehmood, adnan pitafi & shahbaz husain 96 government and tax authorities collect taxes from individuals according to rules and regulations for use in different public services. this cause decrease in income of individuals owing to which some obey the rules as a duty but some try to evade and do not fulfill the duty as a responsible citizen. tax compliance is the fulfillment of responsibilities and duties as a taxpayer. compliance towards tax is basically the obedience of taxpayers towards the taxation authorities and laws. tax compliance can also be termed as the voluntary compliance behavior of tax payers towards the laws and duties. taxpayers trust towards laws and government shows that their contribution will be used appropriately in public and state services which will increase their obedience towards the system and increase their compliance behavior. if individuals believe that the government will work for public interests, that their trust in the others and state is common and there is fair system for all then individuals comply and cooperate even when they can choose evasion (levi, 1998). when taxpayers believe that tax or the tax system is unfair even there is no change in tax rate then they will evade tax. unfair system will enhance tax evasion more than the increase in tax rate (etzioni, 1986). for understanding tax compliance behavior, relation between government/tax authorities and taxpayers is very important (torgler, 2012 and smith, 1992). government and tax authorities impose different type of penalties, audits and fine to ensure that every citizen is compliant enough. these strategies always may not work so authorities should have to consider non-economic variables along with economic variables to bridge the gap between compliance and noncompliance behavior of individuals. noneconomic variables such as ethics, demographic factors, tax awareness, government trust, perception of tax justice, effective use of tax revenues etc. play dominant role in tax compliance (andreoni et al., 1998 and alm et al., 2012). many studies claim that noneconomic variables are important in determining tax compliance such as tax justice perception (saad, 2011) and taxpayers’ trust in authorities (scholz and lubell, 1998; van dijkeand verboon, 2010 and feld and frey, 2002). still there is need of research to study the direct and indirect effects of tax justice perception and government trust on tax compliance. tax justice perception is an important psychological determinant of tax compliance. opinions and thoughts of taxpayers show the taxpayers feelings regarding tax system and regulations. if government makes efforts to improve public services then taxpayers perception of fairness can be improved which leads towards tax compliance. adam smith was also of the view that major principle of better tax collection is justice in the taxation system. is perception towards justice will increase then compliance will automatically enhance. there should be certainty in economy’s taxation system so that there is less and less tax evasion. journal of applied economics and business studies, volume. 6, issue 2 (2022) 95-104 https://doi.org/10.34260/jaebs.626 97 government trust and tax justice perception is different for different societies depending on their culture. tax justice perception increases with the trust in government, which directly influence tax compliance behavior positively. this study basically will ascertain that government trust helps in increasing tax justice perception which will affects the tax compliance of individuals. tax compliance behavior of individuals is studied in this study and variables like government trust and tax justice perception, affecting this behavior will be examined. 2. literature review karakostas and zizzo (2016) suggested that capability of an organization depends on compliance which is the fundamental part of the way in which organizations operate. they conducted an experiment that shows that compliance towards the sign or indication is powerful instrument. in their experiment sixty to seventy percent of the people comply with the indications given to them. compliance to authority is essential either it is public policy or the managerial system. even in the absence of any incentives, individuals comply with the indications given. alm et al., (2010) conducted lab experiments to test the efficiency of service programs for improving compliance of tax. subjects of the experiment earn income and decide to report or underreport it to the specific authority. their results showed that filing income and compliance of tax depends upon uncertainty. if liability is uncertain then subjects will report low tax. tax systems need to be improved for proper compliance and if proper services to taxpayers are provided then compliance rate can be increased. zeng (2014) studied the compliance behavior of tax payers. he surveyed students of economics and business from canada who are taxpayers and know about tax. he found that system of tax is the most important factor of behavior regarding tax compliance. his study investigated that members of the survey are dependent upon their experience of filing and gender. he found that penalty and audit does not affect the behavior of reporting tax. but the existing literature gave three factors that affect the compliance behavior that are expenditures of government, tax system and execution system. helhel and ahmed (2014) surveyed tax payers through likert scale questionnaire from capital city of yemen, sana. results of the survey showed that compliance rate is low because of unfair tax system and high tax rates. results showed the low compliance rate because of no return as services or public goods. age is not an important factor of compliance but gender is because females from the survey are more compliant than males but on the other hand old age group was more compliant than the young age group. alm et al., (1992) explained the reason of paying taxes. they questioned the statement that why people pay taxes even they can evade and have opportunity in the form of incentives. nabila saddaf, muhammad tariq mehmood, adnan pitafi & shahbaz husain 98 then they tested their question and gave the reason of paying taxes. they conducted an experiment from the undergraduate students of colorado university. the results of the experiment show that individuals comply to taxes not because they know that evasion is wrong or they have incentive to evade. tax compliance depends upon the behavior of individual. some pay taxes because they value the public good which they have from taxes finance, some pay because of the probability of audit. ullah et al., (2018) examined the impact of tax fairness system on behaviors regarding compliance in pakistan. they carried primary study to find the impact of compliance behaviors on tax. their study showed that individuals pay unfair amount of tax or evade tax due to tax system. individuals are of the view that if tax system will be transparent and fair enough then there will be fair amount of increase in tax revenue. tax justice is one of the main psychological determinants of tax compliance (andreoni et al. 1998). when taxpayers cannot get services according to what they pay as taxes is because according to them the system is unfair. tax payer’s perception regarding justice is directly affected by the increase in public services (roberts, 1994). tax compliance can be increased by a fair tax system (spicer and becker, 1980). justice is a perception which can enhance compliant behavior by fairer and unbiased tax system (reuben and winden, 2010). as a fair tax system would propose compliance so taxpayer perception about the system is important. the government's taxation policy is defined by the perception of taxpayers (gerbing, 1988). if the system is unfair then taxpayers will avoid and evade taxes (spicer and becker, 1980). an increase in the tax justice perception leads to an increase in the tax compliance (forest and sheffrin, 2002). 3. data analysis technique 3.1 data collection and variable scales data of the research is calculated through questionnaire from major cities of pakistan in 2021. population of the study includes students and tax payers. 500 individuals were selected for the sample but 350 questionnaires were received with complete information so sample of the study reduced to 350. 71% of the population in the survey is male and 29% are female. 55% of the sample is married and 45% is unmarried. qualification wise 36% are having bachelors, 39% are of masters and 25% are having post graduates qualification. 46% are of age 31 to 40, 54% are above 40. the variables used in the study were created from different scales. main variables of the study were tax compliance, government trust, tax justice perception and demographic characteristics. questionnaire form includes questions regarding these variables. likert scale journal of applied economics and business studies, volume. 6, issue 2 (2022) 95-104 https://doi.org/10.34260/jaebs.626 99 was used for measuring the questions of the survey. 4. results and discussions 4.1 internal consistency and reliability test answers about the variables in the study were used and exploratory factor analysis (efa) was conducted with variance maximization analysis. variance maximization analysis is shown in table 1. tax justice perception is factor 1 which accounts for 68.380% accumulating large part of the whole variance. government trust is the second factor which accounts for 22.611% of total variance. 3rd factor that is tax compliance, variance explaining percentage was 9.009%. cronbach’s alpha test was conducted to evaluate the reliability of scales. the coefficient of reliability value is higher than the acceptance value which shows that all the variables of the study have internal consistency. the kaiser-meyer-olkin (kmo) was used to measure the sample size adequacy for factor analysis by comparing the magnitude of observed and partial correlation coefficients (norusis, 1993). the kmo value obtained with efa was 0.688 which shows that variables included in the scale are suitable for efa (sharma, 1996). the results of bartlett test show chi-square value of 290.613 and p value is less than 0.000 showing relation in the variables. table 1: reliability results variables cronbach α variance explanation ratio (%) tax justice perception 0.753 68.380 government trust 22.611 tax compliance 9.009 total variance explanation rate 100 kmo criteria value = 0.688 barlett test: chi square = 290.613 and p value < 0.000 table 2 below gives the descriptive stats and correlation results of the variables of the study. there seems decomposition validity among the variables because cronbach alpha value is bigger than the correlation values (gaski, 1984). variables are having significantly positive relation with one another and tax compliance also at 1% significance level. table 2: descriptive stats and correlation matrix correlation values are significant at 0.01 levels. mean sd tax justice perception government trust tax compliance tax justice perception 0.588 0.328 1 government trust 0.37 0.484 0.334* 1 tax compliance 0.538 0.371 0.539* 0.684* 1 nabila saddaf, muhammad tariq mehmood, adnan pitafi & shahbaz husain 100 4.2 regression analysis in regression analysis we had defined the variable as independent, dependent and mediator variable. in our analysis tax compliance is the dependent variable, government trust is independent and tax justice perception is the mediator which has direct and indirect effect on the dependent variable tax compliance. we will test two hypotheses from the regression analysis. h1: trust in government of individuals is associated to their tax compliance behaviour. h2: trust in government of individuals is associated to their tax justice perception. regression analyses are in table 3. according to model 1 results model is statistically significant showing anova p value .000. tax justice perception and government trust are statistically significant. r square is showing 0.858 value which shows the power of explanation of the model 1 studied. demographic variables; marital status, gender and age are showing insignificant results which is according to the theory. results of model 2 show that model is statistically significant as p value of anova is .000. r square is also showing good fit result with value of 0.751. government trust, marital status and gender are showing positive and statistically significant results with tax compliance. model results reveal that trust in government is the main determinant of tax compliance which is why hypothesis 1 of the study is accepted. results of model 3 shows r square of 0.631 which is considered as good fit and model is also statistically significant with anova p value of .000. this model shows that government trust, marital status and gender are statistically significant and had positive effect on perception of tax justice. trust in government is related to tax justice perception so our second hypothesis is also accepted. table 3: regression analysis independent variables model 1 tax compliance b model 2 tax compliance b model 3 tax justice perception b tax justice perception 0.523 .000 government trust 0.445 .000 .700 .000 .489 .000 marital status 0.057 .121 .198 .000 .268 .000 gender -0.020 .502 .119 .002 .267 .000 age 0.019 .390 .044 .124 .049 .176 r2 0.858 0.751 0.631 journal of applied economics and business studies, volume. 6, issue 2 (2022) 95-104 https://doi.org/10.34260/jaebs.626 101 the mediation analysis is done in spss by linear regression analysis and to test the significance of indirect effect we had used online sobel test calculator. the mediation analysis results are in table 4. table 4: tax justice perception as a mediator effect b sig. total 0.867 .000 direct .730 .000 indirect 0.418 .000 tax compliance behavior of individuals is analysed in the study and variables affecting tax compliance; trust in government and tax justice perception are examined. hypotheses are tested through 300 sample population. findings from the analysis reveal that government trust has positively significant relationship on tax justice perception and tax compliance. another finding shows that tax justice perception plays a mediator role in the relationship between tax compliance and government trust. findings from the analysis have some theoretical implications. findings of the study reveal that demographic variable age has no effect on the tax compliance. this result of the study is consistent with song and yarbrough, (1978). marital status is related to the gender in the data so we consider just gender for the effect. gender has positive and significant effect on tax justice perception because tax justice perception is a behavior which varies from person to person. this result is consistent with d’attoma et al., (2017) and kaslunger et al., (2010). another finding of the analysis revealed that tax justice perception has statistically positive relationship with tax compliance. this shows that as tax justice perception will increase, the tax compliance level will be increased. this result of the study is consistent with spicer and becker, (1980); harris, (1989); roberts, (1994); saad, (2011). on the other hand research finding of the study also shows that tax justice perception arbitrate on the relationship between trust in government and tax compliance. the indirect effect of trust in government on tax compliance is positively significant and this effect is through the tax justice perception. the result is in line with the results of jimenez and iyer (2016). another finding of the analysis showed that government trust has statistically significant and positive relation with tax compliance which shows that as the trust in government of the individuals will increase, there will be positive increase in tax compliance. this finding of the study is in line with hammer et al., (2009); kogler et al., (2013); gangl et al., (2013). study analyses reveal that government trust has significant and positive relation with the tax justice perception which shows that as the government trust of individuals will be increased, the level of perception regarding tax justice will be enhanced as well. this nabila saddaf, muhammad tariq mehmood, adnan pitafi & shahbaz husain 102 consequence of the study is consistent with jimenez and iyer (2016), brashear et al, (2005) and holtz and harrod (2008). most of the literature review showed that justice affects trust like kogler et al., (2013) but present study according to pakistan reveal that trust makes the justice. this is due to the structure of the society of pakistan because studies already revealed in literature that behaviors of individuals are associated with cultural aspects (rozin, 2003). any society’s relationship with government is according to the values of society and it varies across societies. expectations from the government vary across societies depending upon the cultural differences. in the societies where there is trust in government, tax justice perception is higher and positive in those societies. so tax justice perception and government trust vary across societies. 5. conclusion the important source of revenues for any government is the tax revenue. if the tax compliance will increase then it will lead to an increase in government’s income. therefore the variables which help government in affecting tax compliance should be considered while making strategies and tax policies. present study results are helpful for the governments like pakistan in improving tax compliance. government trust has significantly positive relation with tax compliance which can be used for taking measures which will improve the trust level of citizens. when government trust level is enhanced then tax justice perception will be increased which leads to higher tax revenues by increased compliance. decision of taxpayers depends upon the perception about government and tax authorities (alm et al., 2012). so if the tax laws and tax system is unbiased and fair then the compliance level can be increased in an economy. government should always put efforts in gaining public trust by increasing public services, making tax system understandable for everyone by making documentation easier so that more and more people can obey the rules. references alm, j et al., (1992). why do people pay taxes? journal of public economics 48, 21-38. alm, j., et al., (2010). taxpayer information assistance services and tax compliance behavior. journal of economic psychology, 31(4), 577-586. alm, j., kirchler, e., muehlbacher, s., 2012. combining psychology and economics in the analysis of compliance: from enforcement to cooperation. econ. anal. policy 42 (2), 133–151. andreoni, j., erard, b., feinstein, j., 1998. tax compliance. j. econ. lit. 36 (2), 818–860. brashear, t.g., manolis, c., brooks, c.m., 2005. the effects of control, trust, and justiceon journal of applied economics and business studies, volume. 6, issue 2 (2022) 95-104 https://doi.org/10.34260/jaebs.626 103 salesperson turnover. j. bus. res. 58, 241–249. d’attoma, j., volintiru, c and steinmo, s., 2017. willing to share? tax compliance and gender in europe and america. research & politics 4(2). etzioni, a 1986, ‘tax evasion and perceptions of tax fairness: a research note’, journal of applied behavioral science, vol. 22, no. 2, pp. 177-185. feld, l.p., frey, b.s., 2002. trust breeds trust: how taxpayers are treated. econ. gov. 3 (2), 87–99. forest, a., sheffrin, s.m., 2002. complexity and compliance: an empirical investigation.natl. tax j. lv (1), 75–88. gangl, k., muehlbacher, s., hofmann, e., kogler, c., 2013. “how can i help you?” perceived service orientation of t ax authorities and t ax compliance. public financ.anal. 69 (4), 487–510. gaski, j.f., 1984. the theory of power and conflict in channels of distribution. j. mark.48, 9–29. gerbing, m.d., 1988. an empirical study of taxpayer perceptions of fairness phd thesis. the university of texas at austin, austin. gilligan, g., richardson, g., 2005. perceptions of tax fairness and tax compliance in australia and hong kong: a preliminary study. j. financ. crime 12 (4), 331–343. hammar, h., jagersb, s.c., nordblom, k., 2009. perceived tax evasion and the importance of trust. j. socio-econ. 38, 238–245. harris, t.d., 1989. the effect of type of tax knowledge on individuals’ perceptions of fairness and compliance with the federal income tax system: an empirical study.ph.d. thesis. university of south carolina, south carolina. helhel, y and ahmed, y (2014). factors affecting tax attitudes and tax compliance: a survey study in yemen. european journal of business and management volume 6, no. 22. holtz, b.c., harold, c.m., 2008. when your boss says no! the effects of leadership style and trust on employee reactions to managerial explanations. j. occup. organ.psychol. 81, 777–802. jimenez, p., iyer, g.s., 2016. tax compliance in a social setting: the influence of social norms, trust in government, and perceived fairness on taxpayer compliance. adv.account. 34, 17–26. karakostas. a and zizzo. d. (2016). compliance and the power of authority. journal of economic behavior and organization. uk: elsevier b.v. kastlunger, b., dressler, s., kirchler, e., mottone, l and voracek, m., 2010. sex differences in tax compliance: differentiating between demographic sex, gender role orientation and prenatal masculinization (2d: 4d). journal of economic nabila saddaf, muhammad tariq mehmood, adnan pitafi & shahbaz husain 104 psychology 31(4) kogler, c., batrancea, l., nichita, a., pantya, j., belianin, a., kirchler, e., 2013. trust and power as determinants of tax compliance: testing the assumptions of the slippery slope framework ın austria, hungary, romania and russia. j. econ. psychol. 34,169–180. levi, m.,1998. a state of trust. in: v. braithwaite and m. levi, eds. trust and governance. new york: russell sage foundation, pp. 77-101. norusis, m.j., 1993. spss for windows: professional statics, release 6.0. spps inc., chicago. reuben, e., winden, f.v., 2010. fairness perceptions and prosocial emotions in the power to take. j. econ. psychol. 31, 908–922. roberts, m.l., 1994. an experimental approach to changing taxpayers’ attitudes towardsfairness and compliance via television. j. am. tax. assoc. 16, 67–86. rozin, p., 2003. five potential principles for understanding cultural differences in relation to individual differences. j. res. personal. 37, 273–283. saad, n., 2011. fairness perceptions and compliance behaviour: taxpayers judgments in self-assessment environments. phd thesis. university of canterbury, new zealand. scholz, j.t., lubell, m., 1998. trust and taxpaying: testing the heuristic approach to collective action. am. j. political sci. 42 (2), 398–417. sharma, s., 1996. applied multivariate techniques. john wiley and sons inc, new york. smith, k w 1992, ‘reciprocity and fairness: positive incentives for tax compliance’, in slemrod, j (ed.), why people pay taxes: tax compliance and enforcement, university of michigan press, ann arbor, mi, pp. 223-250. song, y., yarbrough, t.e., 1978. tax ethics and taxpayer attitudes: a survey. public adm.rev. 38, 442–453. spicer, m.w., becker, l.a., 1980. fiscal inequity and tax evasion: an experimental approach. natl. tax j. 33 (2), 171–175. torgler, b 2012, ‘tax morale, eastern europe and european enlargement’, communist and post communist studies, vol. 45, nos. 1-2, pp. 11-25. ullah, f et al., (2018). impact of tax fairness system on individual compliance behaviors in pakistan. city university research journal, volume 8(1), 47-57. van dijke, m., verboon, p., 2010. trust in authorities as a boundary condition to procedural fairness effects on tax compliance. j. econ. psychol. 31 (1), 80–91. zeng, t., (2014). what affects tax compliance behaviour: a survey study. ipedr volume 69.5. journal of applied economics and business studies, volume. 6, issue 2 (2022) 33-54 https://doi.org/10.34260/jaebs.623 33 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x an exploration of women owned home-based business through institutional theory lens: a case of peshawar, pakistan madiha gohar1*1& ayesha abrar2 1 assistant professor, department of management and hr at nust business school, national university of sciences and technology, islamabad 2 assistant professor, department of management and hr at nust business school, national university of sciences and technology, islamabad abstract this paper aims to explore the dynamics of women owned homebased businesses in the normative context of peshawar. it helps to add to our understanding of how women create and manage their ventures and get legitimacy in a context where they live like secondclass citizen with no rights. the study also highlights the impact of these ventures on their lives and household. a qualitative approach was used to explore the creation and management of women owned home-based businesses in the context of peshawar. 20 women entrepreneurs were selected through purposive sampling technique. in-depth interviews were conducted to understand the lived experiences. data was analyzed using thematic analysis. we have built on institutional theory to understand the creation and management of women owned home-based enterprises, which brought about unique insights into this ever-prevailing phenomenon. the findings of this study reveal that given the distinctive context of peshawar, the role responsibilities of women are shaped through multiple socio-cultural and religious interpretations. compliance to which, defines the grounds of negotiation for adopting an entrepreneurial career. this makes the venture creation a bounded keywords home-based business, institutional theory, negotiation, legitimacy, empowerment jel classification d10, l26, m13 1* madiha.gohar@nbs.nust.edu.pk mailto:madiha.gohar@nbs.nust.edu.pk madiha gohar & ayesha abrar 34 phenomenon and highly dependent on the familial approval for which norms and values within normative context are negotiated. with familial approval, a home based women entrepreneur (hbwe) ensures her belongingness to the family mores. a legitimacy of hbwe is contingent upon the trust of the family members as an entrepreneur and home maker. at this level, she represents as role model by sharing control which enhances the cultural acceptability of home-based businesses (hbbs). this study has portrayed transition of hbwe through compliance, negotiating, belongingness, legitimacy, emancipation and empowerment. 1. introduction the rising importance of home-based businesses (hbbs) can be seen in the context of high growth small business sector in most of the developed world that has favored the creation of home-based businesses. 50% to 60% of small businesses are home-based in uk and most of other oecd countries (mason et al., 2011; hastings and anwar, 2019). hbbs have significantly raised self-employment and in some cases employment opportunities (burgess and paguio, 2016) that can be tapped as avenue for local economic growth (walker, 2003; barrett, et al., 2008). small clientele, poor turnover, maximum working hours, low hourly wages, limited sectoral choices, challenges in acquiring legitimacy and market credibility, childcare and home duties are all considered to be the antecedents of hbbs (thompson et al., 2009). having said that, the number of such firms is gradually increasing, and now that internet and many platforms for conducting business are available, people prefer to run their online businesses from their homes (although a minority yet). despite hbbs being a viable business alternative with economic potential and social significance (sayers, 2010) it remains missing in majority of the small business research (pratt, 2008), partly due to the debate if these businesses fulfill the criteria to be entrepreneurial and partly because these ventures are hidden. existing literature has two opposing perspectives to hbbs. it is in one-way situated around childcare and domestic responsibilities and the promotion of such hobby businesses is not encouraged. alternatively, hbbs are seen as ‘micro incubators’ for the nascent entrepreneurs who can experiment and buildup their venture before taking it out to commercial places (earles and lynn, 2006; phillips 2002; mason et al., 2011). we believe that hbbs are a strategic approach to economic empowerment while staying socially legitimate. journal of applied economics and business studies, volume. 6, issue 2 (2022) 33-54 https://doi.org/10.34260/jaebs.623 35 it is interesting to note that majority of hbbs are women owned with the purpose of achieving better work-life balance (breen, 2010; price et al., 2019). other reasons cited in the literature are socio-cultural dynamics that has resulted in the upsurge of female entrepreneurship (berke 2003), insufficient single earning (edwards and field-hendrey 2002) desire for better lifestyle (culkin and smith 2000), less start-up capital (loscocco and smith-hunter, 2004; thompson et al., 2009), greater flexibility and convenience (berke 2003; smith, 2000) and lower psychological pressure in working from home attract women to hbbs. thompson et al., (2009) assert that hbbs are found by people who want to test the water of self-employment for a shorter period, or those who would like to convert their ‘hobby’ into business where income remains secondary concern with meager growth aspirations. in developing countries huge proportion of economic activity is concentrated in the informal sector (anderson et al., 2013) mainly home-based and invisible (lent et al., 2019) largely owned by women (williams et al., 2017) hidden within the context of patriarchy (carter and shaw, 2006). most research on hbbs is carried out in developed context; less is known about developing/islamic societies (write and hitt, 2017). it would be interesting to explore ‘how’ the process of women home-based venture creation unfolds in the patriarchal, normative and resources constraint context of peshawar. for this research we take institutional theory to understand what motivates women in peshawar to start a homebased venture and how they successfully create and operate a venture in the normative context. this research adds to the growing literature on home-based business and women entrepreneurship. by drawing on a contextual and institutional perspective to women owned home-based businesses we are not only adding to wider body of entrepreneurship research but also presenting a unique understanding of this growing sector of the small business research from a more informed perspective. recent studies have called for understanding the heterogeneity of female business owners in new, under-researched and everyday contexts and spaces, home has largely been neglected in the spatial/locational studies of the women businesses (henry et al., 2019; price et al., 2019). the paper begins with a discussion of context that makes the phenomenon unique and worth exploring. after a review of literature on hbbs we present institutional theory asa lens to inform the discussion and draw conclusions. female entrepreneurship in peshawar female labour force participation in pakistan is just 22% of the total population . the lower level of resource investment in a female child when coupled with the ideology of madiha gohar & ayesha abrar 36 pardah and izaat becomes the basis for gender discriminations in all walks of life and a tool to limit mobility of women (bhattacharya, 2014) which discourages them making critical decisions at their own (roomi and harrison, 2010).women entrepreneurship is not approved in pakistani society (manzooret al., 2014). however, a change has been observed in this trend too and women participation has increased over time either due to economic needs or to achieve a noble socio-economic status (firdouse, 2007; roomi et al., 2018). however, it is equally important to note that the society is diverse and there are opposing perspectives regarding status of women pertaining to the rich diversity in social practices across all four provinces of pakistan. khyber pukhtoonkhwa (kp) is one of the provinces of pakistan, with relatively conservative cultural norms, patriarchy and strong gender discrimination. gender and religion are among the organizing principals of society. patriarchal beliefs are embedded and transmitted through local traditions and culture that predestines the social value and gender roles. predominantly there exists a strong divide of production and reproduction activities that limit women to private sphere of life and men are to be the breadwinners in public domain. this is one of the major reasons of reduced female labor force participation. culture of kp is guided by the unwritten code of honor called paukhtoonwali (mohammad et al.,2016). a pukhtoon at all times is expected to guard his land, family, women and property from external invaders. among one of 4 tenants of pukhtoonwali that is directly connected to women is “namus” (women’s sexual honor). sexual integrity and chastity of women in family is extremely important as it reflect honor of entire family. in order to protect honor of the family, women are kept in household, and they are not allowed to mingle with unrelated men. an attractive feature of pukhtoon culture is the joint family system. economically the pukhtoon household is a single unit, where males of the family mainly share the expenses. women have limited say indecision-making process in household particularly the ones related to financial affairs of the household. women are highly appreciated when carrying their ‘feminine identity’ and perform their housebound role. women especially those in rural areas of kp hardly find any job opportunities and if there is any a few might be ‘allowed’ to take them. patriarchy has reduced women’s rights and control over economic resources and their mobility and access to education (mahbub ul haq, 2000). the notion of purdah also presents challenge to mobility of women, which has largely been used to keep them home, or they are not allowed for unaccompanied travel. however, the interpretation of purdah varies in different social classes and family backgrounds. in some families traveling alone is not a big issue, while in families with conservative norms women avoid going out of home journal of applied economics and business studies, volume. 6, issue 2 (2022) 33-54 https://doi.org/10.34260/jaebs.623 37 unnecessarily. workingwomen are considered ‘second class citizens’, or women with ‘masculine identities’, hence, face social disapprobation. their independence and agency are interpreted as breaking the social norms (jalal-ud-din and khan, 2008). it is evident that career choices of women in general and in the kp in particular, are shaped by the complex intertwinement of gender, familial attitudes and socio-cultural norms. it would be interesting to explore what goes on within household that lead women decide about starting a business? how women ‘navigate through’ the normative context in creating and sustaining of home-based businesses? how do they negotiate the power structure within the household to get access to resources? how do they establish their belongingness even after starting a venture? how is this entire process shaped and how it unfolds? in order to understand that it is important to explore what hbb is? the next section will define hbb and will briefly review the extant research on the topic. home-based business(s) conceptually a home-based business is defined as any business entity engaged in selling products or services into the market operated by a self-employed person (entrepreneur), with or without employees, that uses residential property as a base from which the operation is run (mason et al., 2011; bin dahari et al., 2019). existing research on home-based business is highly descriptive with limited theoretical insights (anwar and daniel, 2017). we find studies that lists characteristics of home-based businesses (mason et al., 2011), reasons for starting home-based business (phillips, 2002; walker and webster, 2004), work–life and family issues (berke, 2003; fitzgerald and winter, 2001), growth prospects of these ventures (meisam et al., 2017; breen, 2010), employment generation (earles and lynn, 2006) and role of gender in opting for home as a business location (nansen et al., 2010). hbbs are heterogeneous and cannot be limited to any particular business area especially in this digital era (sayers, 2010), however, the unique characteristics of almost all the hbbs is their small turn over (mason et al., 2011), slow growth prospects (breen, 2010) and invisibility (masonet al., 2011). existing research has highlighted that fact the hbbs face the issue of legitimacy and credibility (wynarczyk and graham, 2013). walker (2003) demonstrates that hbbs are not only legitimate businesses but have higher growth potential. hbb is very popular among women across the globe, justifying the great interest of scholars in hbbs from a ‘gender’ perspective (breen, 2010). women opt for home-based business due to non-availability or high opportunity cost of alternatives. women owned hbbs are financially marginal as they work part time and opt for something that is currently trending (loscocco and smithhunter, 2004; thompson et al, 2009; bin dahari et al., 2019). breen (2010) suggest some ‘push’ factors for women self-employment are desire for madiha gohar & ayesha abrar 38 challenge, better work environment, independence and meaningful work. push factors mentioned here reflect the influence of local environment (and the prevailing norms) on employment choices of women (minniti and naudé, 2010), hence understanding that local context of high significance. gender roles are embedded in specific contexts and may stipulate entrepreneurial behaviour (welter et al., 2014). understanding of these external and social factors is important because these factors shape behavioral outcomes. by adopting institutional approach, we shall be able to highlight the normative context to understand women’s actions and their context, we can offer an indepth explanation of women owned home-based businesses and their ability to negotiate the startup in the patriarchal context. institutional approach to home-based businesses institutions are defined as cultural-cognitive, normative and regulative systems that offer meaning and stability to social life (scott, 2008). cultural-cognitive institutions are about shared understandings, beliefs, assumptions and ideologies at a wider level. these systems shape the understanding of how a society works and the role each gender plays in it. normative systems define the patterns of social life by describing social norms, values and behavior and stipulate the legitimate means to achieve the desired behaviours. these normative expectations are constrainers or enablers of individuals’ social behaviour. to ensure conformity to the normative systems, the regulative systems offer rewards and punishments. in short, to attain social rewards, avoid sanctions and maintain legitimacy, individuals must abide by shared norms and values (scott, 2008). individuals’ choices are informed by injunctive norms (which involve the perception of the right thing to do) and descriptive norms (which involve the description of others' behavior) (schultz et al., 2007). when a society admires women in their domestic role, the entrepreneurial activity is automatically low; as women perceive it as something ‘undesirable’ from a societal perspective (i-e descriptive norms disapprove any such activity). in the patriarchal society of peshawar, women are expected to abide to their ‘feminine identity and roles’ where their wishes are subjugated to the normative context of household. in such circumstances it would be interesting to investigate what leads to the formation of these home-based businesses and what comes in the process of taking a masculine identity in a highly normative context and creating a legitimate venture. for a business to be successful, it must be legitimate (stringfellow et al,. 2014) to get access to resources (adler and kwon 2002) and enhance credibility (bensemann et al. 2018). legitimacy is “a generalized perception or assumption that the actions of an [enterprise] are journal of applied economics and business studies, volume. 6, issue 2 (2022) 33-54 https://doi.org/10.34260/jaebs.623 39 desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions” (suchman, 1995: 574). research has highlighted two main perspectives of legitimacy. one suggests being legitimate, organizations tend to conform to the socially established standards of appropriateness (mizruchi and fein,1999) due to external pressures and attain legitimacy. strategic perspective suggests, organizations and organizational actors are always involved in legitimacy seeking behaviors (pfeffer, 1981: 9). suchman (1995) listed 3 strategies to build legitimacy. first, to conform to standards set by environment, second, to select an environment where the actor gets support in performing the action, while the third, and manipulate the existing structures in way that new standards of legitimacy can be created. in the context of entrepreneurship zimmerman and zeitz (2002: 423) added fourth category creation“the creation of the social context, rules, norms, values, beliefs, models, etc.” women home-based businesses and their legitimation strategy also present an interesting area to be explored in the existence of both external forces to abide to the normative context and (internal) presence and enactment of individual agency to become an entrepreneur. we build on lent et al., (2019) conceptualization of legitimacy proposing legitimacy as interplay of both traditional and strategic perspective “to (understand) external and internal influences on the legitimation process…. their nature, scope and interaction” (348). it would be interesting to see how women conform, select, manipulate (suchman, 1995) and create environmental structures (if they do) (zimmerman and zeitz, 2002: 423) to be legitimate in their entrepreneurial endeavors. so, an institutional framework with lent et al., conceptualization of legitimacy provides an interesting lens to explore the invisible presence of women hbbs. 2. research methodology an interpretivist stance approach is adopted for this research because the main aim is to get an in-depth understanding of the institutional dynamics at play shaping women entrepreneurship in peshawar (de-bruin et al., 2007b; mcgowan et al., 2012). interpretivism is “associated with the philosophical position of idealism, and is used to group together diverse approaches, including social constructivism, phenomenology and hermeneutics; approaches that reject the objectivist view that meaning resides within the world independently of consciousness” (collins, 2010: 10). through interpretive philosophy, the social world is interpreted, understood, experienced, produced and/or constituted with the underlying assumption that the world is complex and multilayered and textured (mason, 2002). interpretivist research philosophy suggests that researchers are social actors hence it is important for them to appreciate differences between individuals (saunders et al., 2012). madiha gohar & ayesha abrar 40 interpretive approach seeks to interpret events and phenomena in terms of how the people concerned perceive and understand their own experience (patton, 2002). the reality in this case is understood to be subjective and it is function of interpretation (burrell & morgan, 1979).for contextual understanding it is pertinent to opt for a qualitative approach (burg et al., 2020), both of the authors belong to the same culture and they prefer to indulge in conversation with women entrepreneurs, resulting in better understanding of the researcher’s interest. on the practical side, there is no such database prevalent about women entrepreneurs and their invisible home-based enterprises that can be utilized for data collection (roomi et al, 2018). 2.1 data collection in order to understand the creation and management of women owned enterprises in the context of peshawar, 20 women entrepreneurs selected through purposive sampling technique (marshall, 1996). we also requested the interviewees if they could snowball us to some other women entrepreneurs. as in qualitative research information-rich cases that can reflect upon experiences to gain an in-depth understanding of the phenomenon (creswell and creswell, 2017) hence, a small sample size is appropriate. women running their hbbs for at-least 03 years either in manufacturing or service sectors were contacted for the present study. after taking their consent, face-to-face indepth interviews were conducted with a focus on the entrepreneurial activity of the participants with a reference to the normative context of its occurrence. interview provides thick description that helps in reaching the essence of the experience (eisenhardt and graebner, 2007). medium of communication was adopted as per comfort of the respondent; hence, interviews were conducted either in pushto (local language) or urdu (national language). interviews were digitally recorded with the consent of the respondents also some observations were made. the length of the interviews varied from an hour to an hour and half. 2.2 data analysis after intelligent verbatim transcription of the interviews, thematic analysis was carried out. thematic analysis is one of the widely used qualitative data analysis tools (huberman and miles, 2002).it provides an opportunity to examine the perspective of various research participants that are helpful in generating unanticipated insights (nowell et al., 2017). thematic analysis helped us summarize insightful patterns from the large dataset to understand the entrepreneurial processes of whbes in their normative context. there is no consensus on how to carry out thematic analysis, for this research steps suggest bybraun and clarke (2006) were followed. journal of applied economics and business studies, volume. 6, issue 2 (2022) 33-54 https://doi.org/10.34260/jaebs.623 41 thematic analysis: authors the analysis resulted in the following themes 1. the normative context of action 2. towards familial and social acceptance 3. turning the table/ taking control 2.3 demographic profile majority of the respondents (75%) were initially married with children. out of married, 02 were divorced and 01 was widowed. those who were single used to live with their parents. majority of venture were in feminine businesses like beauty salon, boutique, stitching/embroidery, jewelry designing while couple of them were furniture designing, stage decoration/event management and tent service etc. most of the entrepreneurs (70%) started their ventures out of necessity while the remaining (30%) were opportunity entrepreneurs. those with beauty salon used to have either another woman as an employee or someone from the household used to help them in their business activity. 3. findings 3.1 the normative context of action in traditional pakistani society, family is the fundamental social unit that provides identity, resources, and safety to its members while also influencing their life opportunities and choices. surprisingly, women were well aware of the privileges granted to them by islam; they were aware that they were not expected to provide for their families if the male members were there. they were grateful for the freedom to create their own business and handle their own finances. they understood, however, that in order to be acceptable, they had to adhere to conventional segregationist conventions and do business in places that were acceptable to the family. through their logics and explanations, women were able to negotiate their startup (both necessity and opportunity). women who were given the freedom to run their own businesses mentioned two key factors that aid them in negotiating. one was the financial assistance they would provide, and the other was the 'home-based' business. “my husband lost his job and was unable to find another one. we were financially constrained; meeting ends was becoming possible while having young children. leaving home and children unattended was not an option, so the only choice i was left with, was to do something from home, so that this sphere of life is not disturbed” (hbwe 6) level 01 familiarization with data level 02 creation of initial codes and categories level 03 reviewing themes searching for themes level 04 level 05 defining and naming themes level 06 reporting themes madiha gohar & ayesha abrar 42 it can be seen that in some cases the venture becomes ‘the only’ source of income. hbwe 5 mentioned the disproportionate increase in prices and lower levels of earnings, which are not sufficient to meet the expenses. however, for prosperous entrepreneurs their business was a luxury or productive use of time. hbwe 03 stated "i am an excellent artist and craftsperson, and my business is doing well." i make a good living, which i put towards my own costs rather than putting money into the family's coffers. but i'm also taking care of my household duties; my house is clean, my children are well fed and behaved, so if you have the skills and the time, there's no harm in starting something from home; look it is at the cost of nothing." in pukhtoon culture women are made to believe that they can have a life of their ‘own’ choice but after marriage, they will be free to take their decisions. with husband’s support women were confidently starting and managing their enterprises. women are the primary caregivers in the home; launching a business would not free them of their responsibilities. hbwes were proud of their status as moms, and they valued the role of women in the home and in running a business. "she (woman) has a significant role; women have a significant impact on the household; they are the ones who make or break a household," says hbwe4. women's home-based businesses revolved around traditional feminine pursuits, limiting their ability to influence their responsibilities. her business ideas and decisions are ingrained in her family. she must ensure that her business does not interfere with her household responsibilities and obligations as a wife and mother, and she must be able to negotiate as a result. 3.2 towards familial and social acceptance the necessity of belonging to a culture and family was emphasized by female entrepreneurs. the family, in particular, was shown as the focus of life. they are not allowed to do anything that may bring shame to the family. they established the business with the family's blessing and support, which helped them gain acceptability and respect. "when i was about to start the business, my mother and mother-in-law were very upset and opposed it because they were afraid that it would disrupt my children and my household, and that their education would suffer as a result. so, as of now, i have started the venture and it has been 5 years, but i have to make sure that everything is running smoothly; my first priority is the household."(whbe 8) women choose home-based business strategically to ascertain belonging to the families, and if they are not forced to, they establish a business after their children enters school and they have more free time (whbe, 02 and 05). another entrepreneur (19) states that her journal of applied economics and business studies, volume. 6, issue 2 (2022) 33-54 https://doi.org/10.34260/jaebs.623 43 husband had turned down her opportunity to work as a coach in a vocational centre, which would have given her access to the mainstream market, but this required her to leave her home for set hours, which was not acceptable. in pukhtoon few professions are considered appropriate for women because society consider it respectable like education, medicines and trade (from and within home). interestingly, the definition of 'respectable' differs based on one's caste or culture. one of the interviewees talks about her experience breaking into the beauty industry and the challenges she experienced. “…this occupation is associated with barbering, therefore they used to joke about what people would think if a girl from the syed family worked as a barber”(whbe 15). the quote symbolises the environment for entrepreneurship; belonging to a particular cast limits industrial options and raises questions about authenticity. women from other family origins, on the other hand, do not have this problem eg (whbe 05). so there is social as well as familial perception of the various professions. being home-based is essentially necessity pushed which has its own pros and cons for women and their ventures. “…my children are my main priority in life, as is appropriately running the home and caring for your children. that is also an industry...their future is being shaped, as are their personalities. that's why, despite the fact that i was having a great time at the newspaper, i decided to leave. then i thought i needed a vacation because, as you know, i've never been home during meal times and now i'm constantly home, living out my dream of being active while also caring for children, which is quite relaxing”(whbe 20). on the plus side, along with childcare and home obligations, businesspeople who were attached to their traditional identities found sanctuary and psychological cover in the fact that they are with their children and capable of caring for them. on the other hand, opportunity entrepreneurs were feeling overwhelmed by the reality that running a business from home results in a smaller market share and fewer opportunities for expansion. being an entrepreneur is not a justification to avoid doing household chores; rather, in some cases, using a woman's business identity to belittle her is a taunt. “….my in-laws would mock me that "oh yes, she is a businesswoman, why will she be performing such home activities?" if there was a delay in cooking lunch” (whbe13). such family members' behaviour causes psychological problems for women, who do not take pride in their business identity and instead feel terrible for not being able to effectively execute domestic responsibilities. however, this is the case in houses where the powerful family figure is unhappy with the business or simply wants to keep their women under control by pulling their legs. madiha gohar & ayesha abrar 44 getting support in both material and non-material ways is another key part of belonging. women, particularly males, require financial and emotional support from their families, which they can only receive if they follow the norms set down for them. “…i never thought of it that way. …whether its my money or is his money. i suppose family or husband has to give because if i was not working before, like i was not on job that i was being paid for, which i could have usedfor business, so yes they have to provide me”. interestingly pukhtoon women entrepreneurs think it is the responsibility of male members of family to help them financially. because they rely on their families for entrepreneurial capital, they are cautious in their company decisions and choose something with a minimal risk of failure. furthermore, it was discovered that families only invest in an enterprise if they believe it is credible. empirical results show that family is not only the main source of financial capital but it provides strong support in managing and running business and household. entrepreneurs are vigilant of their families’ role and they highly appreciate the role played by their parents, husband and children in particular. an entrepreneur shared her response to a critic “i am not doing any sin. my husband is sitting [with me] and i am doing it with his permission, if he has no objection you are not allowed to object, just go away and mind your own business” (whbe 6) her husband's trust and support gave her the fortitude to face and respond to the criticism. it is important to highlight, however, that this assistance is contingent on women's capacity to plan their operations around the family's preferences. the way a society encourages women to start businesses reflects social concerns about belonging; in the case of hbwe, this may be observed in the women's business choices. they chose businesses that are primarily referred to be feminine and that may be run from home with minimal interaction from the opposing gender. not only did women seek approval from their families, but they also sought approval from their peers. they were able to achieve social belonging by replicating and reproducing established conventions of what constitutes a woman's 'entrepreneurial' status. whbe stated that their ventures are in sectors where there are fewer chances of mixing with people of the other gender, and that if it is necessary for the business, they maintain proper purdah. however, this implication might be interpreted in a variety of ways. “purdah means to cover yourself as much as possible when you're a muslim, so while i'm here, it's my house, the way i live with my children, and those that come are all ladies; gents are not allowed. i can wear jeans, shorts, skirts, or pakistani dresses at home, but when i go out, i have to wear a full abaya with hijab because it gets you respect.” journal of applied economics and business studies, volume. 6, issue 2 (2022) 33-54 https://doi.org/10.34260/jaebs.623 45 (whbe11) observing pardah not only give social approval but leads to legitimacy, men of the surrounding households would allow their women to visit that business if they perceive that whbe is observing purdah. again this observance can be symbolic in many cases to get legitimacy. purdah introduces a new concept of constrained mobility: women in pukhtoon culture are expected to stay at home or be escorted by a male family member when venturing out. this expectation of being accompanied has a severe impact on women entrepreneurs, who must wait for the availability of 'someone' to accompany them even if they need supplies. however, after a few years in business, one of the entrepreneurs noted that they have gained confidence and are able to deal with external factors, thus they were allowed to go alone or with another woman or child in numerous circumstances. 3.3 turning the table: taking control women entrepreneurs generally believed that they do not have control over several aspects of their lives, however, after successfully running their ventures now they have started experiencing positive changes on personal and family level. the majority of the entrepreneurs were necessity entrepreneurs, so their ventures have brought about financial ease. they were able to provide better meals and educational possibilities for their children as a result of their efforts. children's abilities symbolise the mothers' abilities in pukhtoon society, thus if they act badly or do poorly in school, it is the mother's fault. women were ecstatic that working from home allows them to teach their children and demonstrate that they are "excellent" mothers. for opportunity entrepreneurs’ entrepreneurship is way to become independent, confident and stay close to their children while being at work. further at family level women felt that after financial independence they are heard in the household. “that is me; when you have authority and are self-sufficient, you gain respect and your viewpoint is valued. my husband now asks me how to do this or that, and i'm stumped. as a result, i do not always impose my opinion, but it is important. as a result, they follow my recommendations” whbe 10 explained the decision-making process in her home. this is also related to the issue of self-assurance. entrepreneurs stated that they are confident in expressing their opinions and arguing for their rights, but that they prefer to engage with important parties rather than confront them. this ability, however, is not solely dependent on the venture; upbringing and education also play a role in building women's confidence. madiha gohar & ayesha abrar 46 women feel empowered with their ventures, and they are content with their contribution to their household and family. related to financial independence, most of the entrepreneurs mentioned they need not suppress their desires for themselves or their children anymore. whbe (6) stated “the confidence you gain as a woman should be earned... you will not ask your spouse for money every time you want something; if you want to give something to your children, do so with your own money; if you want to give something to your husband, do so with your own money. it makes me happy”. women used to look at their male custodians for their desires, which have now been met on their own. an interesting point was observed here that some women felt that they need to take permission from their male guardians for spending the money (whbe 3, 8,11, 14, 18) while those with some level of education and exposure were aware of their rights and control over their earned income. turning tables from the perspectives of the wider family can be seen from the fact that venture gave them confidence to stand for themselves. successful running of venture has resulted in elevated status in wider family, and they are seen as role model. (whbe 13) “after seeing me, my other sister-in laws who are not educated they started following me one started beauty salon, the other did boutique. why? they follow me as role model now”. women who once faced familial objection have started getting appreciation by brining positive change to household and now were earning praise and applause. in pukhtoon tradition, keeping in touch with extended relatives is extremely essential, and women have deliberately taken advantage of this. as a result, the businesses were not only satisfying social and familial obligations, but also engaging in some word-of-mouth marketing. whbe (11) spoke of it as “i manage extended family relationships, which is critical in keeping the mother-in-law pleased and ensuring that you receive the clients. i 'manage' it, and if i have to go, i will do so in the evening. i was able to strike a balance (business and relationships)”. though it is time consuming to leave the household and attend variousoccasions, but they strategically use this to promote their ventures. the ventures of these women had an emancipatory impact on them and their households. “…it is very important for yourself personally, because when you earn, you may have the experience that you think of yourself as very strong and confident; however, if your husband gives you in lacs (millions), you may not feel that happiness and satisfaction, because when your husband gives you money, he also asks you about your expenditure, and you are not answerable to anyone for your money; you can spend it wherever you journal of applied economics and business studies, volume. 6, issue 2 (2022) 33-54 https://doi.org/10.34260/jaebs.623 47 want.…”(whbe19). this can be seen for both necessity and opportunity entrepreneurs, they were happy and passionate about their venture, they wanted to grow large subject to cultural and financial restrictions. they were aware of the fact that their survival and progress is linked to help and support from family and approval from society. to make sure that they get it, they were tactfully molding the norms without overtly confronting. 4. discussion the life of muslim women living in an islamic society is entirely different from their counterparts elsewhere (dupree and gouttierre 1997). khalid and frieze (2004: 293) note, “although there are common features among these women, the realities of their lives vary greatly from group-to-group,” as “islam provides a general framework for life, with a range of individual options”. our research aims to better understand how women-owned hbbs are established and run in a strongly normative muslim environment shared meaning of reality embedded in institutional patterns carried by whbe in kp: authors findings illustrate the intertwinement of women owned home-based businesses with the normative context of its occurrence through three themes of the normative context of action, the desire and efforts to accomplish social and familial acceptance or the creation and running of venture which gave them power to turn the table as a result. using these themes as lens to understand the dynamics of these home-based enterprises it is clear that the enterprising ability of a woman entrepreneur in peshawar is significantly shaped by her ability to negotiate various roles and identities. the strongly patriarchal structure of peshawar embeds madiha gohar & ayesha abrar 48 female agency in household and renders her dependent on the family for taking any role other than the traditional one. in order to turn entrepreneurial intentions into actions women in peshawar have to negotiate the gendered identities, household consensus and their relational embeddedness to bypass the relational and structural barriers. findings depicts that they have negotiated their agency tactfully by linking the motivation to enterprise with financial needs of the family and have reduced objection of family by strategically selecting home as business location. most of the women overtly expressed about the knowledge of their right to business given by islam, by narrating the example of hazrat khadija who was an entrepreneur and wife of the holy prophet muhammad (pbuh) but alongside they were feeling religious and social pressures to be stay at home mom. the findings are aligned to previous studies such as essers and benschops (2009), al-dajani and marlow (2010) and roomi et al., (2018).we can conclude that women use religion to get some space to exercise their agency while negotiating their normative context. however, pukhtoon culture is more rigid, knowing their right to work, family’s willingness is kind of prerequisite. not only male members of the household but permission of elderly women e.g. mother or mother in law is equally significant. these women have the ability to influence operations of the ventures by exercising their powers in the form of increasing domestic workload or withdrawing support in childcare and domestic responsibilities. belongingness is among core social motivations that underpin much social behaviour (fiske, 2018). following family’s will be an effort to accomplish belongingness as an entrepreneur by pukhtoon women, nonconforming can have serious consequences in the form of breaking up of family ties and relations. close connection to family members opens door for financial and non-financial support from the family (williams and horodnic, 2016). pukhtoon women do not have any capital of their own that can be used as investment and even if they do they cannot use it without taking family in confidence (gohar et al, 2018). women wanted to be accepted and appreciated with this new identity rather than being shunned. women were able to get legitimacy as an entrepreneur by responding to family values, cushioning on family’s needs and operating as per social expectation. to get social acceptance and respect as entrepreneurs the social actors are supposed to abide to conventional norms and beliefs (scott, 2008; taliss, 2019). entrepreneurs utilize various strategies to gain social legitimacy like “conforming to the rules of the cultural order; selecting an environment(s) whose constituents will support current practices; manipulating the current environment by attempting to make changes that would achieve consistency journal of applied economics and business studies, volume. 6, issue 2 (2022) 33-54 https://doi.org/10.34260/jaebs.623 49 between an organization and its environment and creating a new social context by creating new rules, norms, values, scripts beliefs, models….” (lent et al., 2019: 18), however, in the present study we found that it was more of an effort to accomplish belongingness through legitimacy practices. they started home-based not because they wanted to be legitimate first, but they wanted to stay connected and be a part of the family. keeping with the domestic identities, taking on another one becomes a double-edged sword sometimes, which adds pressure to women’s lives. however, their ability to add to finances gave them power to turn the table in their favor. women were not powerful and resourceful enough to create a new social context; hence, their limited agency was utilized in a way that they could get maximum of both worlds by taking family on board. their ventures have not only added to the family’s wallet but have brought about positive changes to their personal wellbeing. 5. conclusion this research is an effort to understand the home-based businesses of women entrepreneurs in the normative context of peshawar, pakistan. we have employed institutional theory to understand how gender and entrepreneurship co-exist in a highly normative, conservative and traditional context. to the best of our knowledge institutional theory has not been used so far in the hbb literature, hence, we bring novel insights to understand the intricacies of hbbs in the context of peshawar. in the context of peshawar being home-based is a necessity pushed, as well as a strategic choice. it has its pros like little start-up cost, being available for domestic tasks, help in childcare, however, it does have the cons associated to it in the form of small customer base, less creative and smaller growth chances. women in peshawar do not start home-based business keeping in mind all these pros and cons, to them being home-based is an effort to accomplish belongingness, which later lead to attaining legitimacy and support. studies using institutional theory have not used the perspective of belongingness; rather they assumed that following norms is to gain legitimacy (lent et al, 2019). we believe that individuals cannot be isolated from the social structure and the concept of belongingness connects individual (woman) to the social (wider culture) through family. women in the sample were adhering to the norms of the family because they wanted to stay connected and be a part of them. bell (1999) highlights belongingness as performative, created and recreated through different practices. these hbwes were performing their belongingness by making sure they follow the normative construct of family and entrepreneurial identity does not overshadow their conventional gender identity. these entrepreneurs were responsible for family care and madiha gohar & ayesha abrar 50 other domestic responsibilities and were supposed to manage a constant shift and balance of domestic and business life, both of which are located within the family (hamilton, 2006). the choices made in the process are complex and dynamic whereby women make sense of work and family throughout. women in order to start venture have to negotiate various social barriers in the form of gendered attitudes and practices (fielden and dawe, 2004). empirics show that women in peshawar strategically adjust gender norms than conforming to it. the adherence and disobedience produce tension, which motivates social actors to innovate but within the confines of patriarchal structures. women felt empowered and emancipated by being able to take control of some aspects of their lives after getting some financial independence. the respondents expressed that their self-employment has resulted in autonomy and assertiveness (al-dajani and marlow, 2010) may it be on the very basic level, but they do have a say (gohar et al., 2018). the hbbs of women are not to challenge the norms of pukhtoonwali rather they epitomize the navigation and negotiation of gendered identities within the ‘protected enclave of domestic sphere’ (al-dajani and marlow, 2010:13) with some support from male members of family. women in kp never aimed to replace the prevailing narrative rather they through their entrepreneurial practices were (re)creating, negotiating and maintaining the gender, hence these enterprises were not seen as a threat to the wider normative context, but as an effort to accomplish belongingness in both public and private sphere of life. references al-dajani, h., & marlow, s. (2010) ‘impact of women’s home-based enterprise on family dynamics: evidence from jordan, international small business journal, 28(5), pp. 470-486. anderson, a., harbi, s. e., & brahem, m. (2013) ‘enacting entrepreneurship in ‘informal’ businesses’, the international journal of entrepreneurship and innovation, 14(3), pp. 137-149. anwar, m. n., & daniel, e. m. (2017) ‘ethnic entrepreneurs and online home-based businesses: an exploratory study’, journal of global entrepreneurship research, 7(1), pp. 6-19. bell, v. (1999) ‘performativity and belonging: an introduction’. theory, culture & society, 16(2), pp.1–10. bensemann, j., warren, l., & anderson, a. r. (2018) ‘entrepreneurial engagement in a depleted small town: legitimacy and embeddedness’, journal of management & organization, pp. 1–17 early cite. berke, d. l. (2003) ‘coming home again: the challenges and rewards of home-based selfemployment, journal of family issues, 24(4), pp. 513-546. journal of applied economics and business studies, volume. 6, issue 2 (2022) 33-54 https://doi.org/10.34260/jaebs.623 51 bhattacharya, s. (2014) ‘status of women in pakistan’, journal of the research society of pakistan, 51(1), pp. 179-211. bin dahari, z., abu bakar, a.r. and al-gosaibi, s., (2019) ‘key success factors of homebased business among female entrepreneur in saudi arabia’, asia pacific journal of business review, 3(2), pp.43-66. braun, v., clarke, v., hayfield, n., & terry, g. (2019), ‘reflecting on reflexive thematic analysis’, qualitative research in sport, exercise and health, 11(4), pp. 589-597. breen, j. (2010) ‘gender differences in home-based business ownership’, small enterprise research, 17(2), pp. 124-136. burgess, s. and paguio, r. (2016) ‘examining ict application adoption in australian home-based businesses: an innovation-decision process approach’, journal of enterprise information management, 29(2), pp. 276-299. collins, h. (2010) ‘creative research: the theory and practice of research for the creative industries’ ava publications creswell, j.w. and creswell, j.d., (2017) ‘research design: qualitative, quantitative, and mixed methods approaches’ , sage publications, london. culkin, n., & smith, d. (2000) ‘an emotional business: a guide to understanding the motivations of small business decision takers’, qualitative market research, 3(3), pp. 145-157. de-bruin, a., brush, c. g., & welter, f. (2007) ‘advancing a framework for coherent research on women's entrepreneurship, entrepreneurship theory and practice, 31(3), pp. 323-339. earles, w., & lynn, r. (2006) ‘futures of home-based businesses: a literature review’, australasian journal of business and social inquiry, 4, pp. 13-38. edwards, l. n., & field-hendrey, e. (2002) home-based work and women’s labor force decisions, journal of labor economics, 20(1) pp. 170-200. eisenhardt, k. m., & graebner, m. e. (2007) ‘theory building from cases: opportunities and challenges’, academy of management journal, 50(1) pp. 25-32. essers, c., & benschop, y. (2009) ‘muslim businesswomen doing boundary work: the negotiation of islam, gender and ethnicity within entrepreneurial contexts’, human relations, 62(3) pp. 403-423. fielden, s.l. and dawe, a. (2004) ‘entrepreneurship and social inclusion’, women in management review, 19(3), pp. 139-142. firdouse, a. (2007), ‘social status of rural and urban working women in pakistan’: a comparative study. phd dissertation. fiske, s.t., (2018) ‘social beings: core motives in social psychology’, john wiley & sons, usa. fitzgerald, m. a., & winter, m. (2001) ‘the intrusiveness of home-based work on family life’, journal of family and economic issues, 22(1), pp. 75-92. gohar, m., basit, a. and abrar, a., (2018) ‘growth aspirations of pukhtoon women entrepreneurs in the household context, fwu journal of social sciences, 12(1), madiha gohar & ayesha abrar 52 pp. 180-189. goheer, n. a. (2003) ‘women entrepreneurs in pakistan’ geneva: international labour organization, pp. 1-43. hamilton, e. (2006) ‘whose story is it anyway? narrative accounts of the role of women in founding and establishing family businesses’, international small business journal, 24(3) pp. 253-271. hastings, d., & anwar, n. (2019) ‘home-based businesses and the factors affecting their collaborative propensity’ , available at https://www.researchgate.net/publication/337224487_homebased_businesses_and_the_factors_affecting_their_collaborative_propensity huberman, m. and miles, m.b., (2002) ‘the qualitative researcher's companion. sage, london. jalal-ud-din, m., & khan, m. (2008) ‘socio-economic and cultural constraints of women in pakistan with special reference to mardan district, nwfp province’ , sarhad journal of agriculture, 24(3), pp. 485-493. lent, m., anderson, a., yunis, m. s., & hashim, h. (2019) ‘understanding how legitimacy is acquired among informal home-based pakistani small businesses’, international entrepreneurship and management journal, 15(2), pp. 341-361. littlejohn, s.w. & foss, k.a. (2009) encyclopedia of communication theory, vol.1, sage publication. loscocco k. and smith-hunter, a (2004) ‘women home‐based business owners: insights from comparative analyses’, women in management review, 19(3), pp. 164-173. manzoor, raheel, s., & jalil, m. f. (2014) ‘psychological capital intervention towards women entrepreneurship development and entrepreneurial orientation’, international journal of business management and administration, 3 pp. 22–28. marshall, m. n. (1996) ‘sampling for qualitative research” family practice, 13(6), pp522526. mason, c., carter, s., & tagg, s. (2011) ‘invisible businesses: the characteristics of homebased businesses in the united kingdom’ regional studies, 45(5), pp. 625-639. mcgowan, p., redeker, c.l., cooper, s.y. and greenan, k., (2012) ‘female entrepreneurship and the management of business and domestic roles: motivations, expectations and realities’, entrepreneurship & regional development, 24(1-2), pp.53-72. meisam modarresi, zahra arasti, kambiz talebi, maghsoud farasatkhah. (2017) ‘growth barriers of women-owned home-based businesses in iran: an exploratory study’, gender in management: an international journal, 32(4), pp.244-267. minniti, m., naudé, w. (2010) ‘what do we know about the patterns and determinants of female entrepreneurship across countries?, the european journal of development research, 22, pp. 277–293. journal of applied economics and business studies, volume. 6, issue 2 (2022) 33-54 https://doi.org/10.34260/jaebs.623 53 mizruchi, m. s., & fein, l. c. (1999) ‘the social construction of organizational knowledge: a study of the uses of coercive, mimetic, and normative isomorphism’, administrative science quarterly, 44(4), pp. 653–683. mohammad, f., thomas, a. r., & tabassum, i. (2016) ‘honor, revenge in sociogeographic space of pashtuns, pakistan journal of criminology, 8(3), pp. 74-90. nansen, b., arnold, m., gibbs, m., & davis, h. (2010) ‘time, space and technology in the working‐home: an unsettled nexus’, new technology, work and employment, 25(2), pp. 136-153. nasrullah, a. m., &dickson, k. (2011) ‘empowerment of rural women, decent employment and micro enterprise development programs of ngos in bangladesh, international journal of afro-asian studies, 2(2), pp. 30-40. pfeffer, j. (1981) ‘management as symbolic action: the creation and maintenance of organizational paradigms’. in l. l. cummings & b. m. staw (eds.), research in organizational behavior. 13, pp. 1– 52. greenwich: jai press. phillips, b. d. (2002) ‘home-based firms, e-commerce, and high-technology small firms: are they related?, economic development quarterly, 16(1), pp. 39-48. pratt, j. h. (2008) ‘revisioning the home-based business: an exploratory model of homebased location choice, in united states association for small business and entrepreneurship. conference proceedings in united states association for small business and entrepreneurship pp. 212-218. price, victoria, sian and reuschke, darja (2019) ‘performance and innovation in womenowned home-based businesses: exploring the gender-location link, institute for small business and entrepreneurship (isbe) 2019: space – exploring new frontiers and entrepreneurial places, united kingdom. 14 15 nov 2019. raghuvanshi, j., agrawal, r. and ghosh, p.k.,(2017), ‘analysis of barriers to women entrepreneurship: the dematel approach’ the journal of entrepreneurship, 26(2), pp.220-238. roomi, m. a & harrison, p. (2010) ‘behind the veil: women‐only entrepreneurship training in pakistan’, international journal of gender and entrepreneurship, 2(2), pp. 150-172. roomi, m. a., rehman, s., & henry, c. (2018) ‘exploring the normative context for women’s entrepreneurship in pakistan: a critical analysis’, international journal of gender and entrepreneurship, 10(2), pp. 158-180. saunders, m., lewis, p. & thornhill, a. (2012) “research methods for business students” 6th edition, pearson education limited sayers, j. g. (2010) ‘home-based businesses in the city, small enterprise research, 17(2), pp. 165-176. schultz, p. w., nolan, j. m., cialdini, r. b., goldstein, n. j., & griskevicius, v. (2007) ‘the constructive, destructive, and reconstructive power of social norms’ psychological science, 18(5), pp. 429-434. scott, w. r. (2008), institutions and organizations. ideas, interests and identities. sage madiha gohar & ayesha abrar 54 publications london. smith, c.r. (2000) ‘managing work and family in small “copreneurial” business: an australian study, women in management review, 15(5/6), pp. 283-289. stringfellow, l., shaw, e., & maclean, m. (2014) ‘apostasy versus legitimacy: relational dynamics and routes to resource acquisition in entrepreneurial ventures’, international small business journal, 32(5), pp 571–592. suchman, m. c. (1995) ‘managing legitimacy: strategic and institutional approaches’, academy of management review, 20(3), pp 571–610. thompson, p., jones-evans, d., & kwong, c. (2009) ‘women and home-based entrepreneurship: evidence from the united kingdom’, international small business journal, 27(2), pp. 227-239. van burg, e., cornelissen, j., stam, w. and jack, s., (2020) ‘advancing qualitative entrepreneurship research: leveraging methodological plurality for achieving scholarly impact’, entrepreneurship theory and practice, pp. 1–18. walker, e. (2003) ‘home-based businesses: setting straight the urban myths, small enterprise research, 11(2), pp. 35-48 walker, e. and webster, b. (2004) ‘gender issues in home‐based businesses’, women in management review, 19(8), pp. 404-412. williams, c. c., & horodnic, i. a. (2016) ‘cross-country variations in the participation of small businesses in the informal economy, journal of small business and enterprise development, 23(1), pp. 3-24. williams, c. c., martinez–perez, a., & kedir, a. m. (2017) ‘informal entrepreneurship in developing economies: the impacts of starting up unregistered on firm performance’, entrepreneurship theory and practice, 41(5), 773-799. wright, m. and hitt, m.a., (2017) ‘strategic entrepreneurship and sej: development and current progress’, strategic entrepreneurship journal, 11(3), pp.200-210. wynarczyk, p., & graham, j. (2013) ‘the impact of connectivity technology on homebased business venturing: the case of women in the north east of england’, local economy, 28(5), pp. 451-470. zimmerman, m. a., & zeitz, g. j. (2002) ‘beyond survival: achieving new venture growth by building legitimacy’, the academy of management review, 27(3), pp 414–431 journal of applied economics and business studies, volume. 6, issue 2 (2022) 1-14 https://doi.org/10.34260/jaebs.621 1 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x do energy prices drive inflation in south africa? an unrestricted var approach marius masoga1, thabang kumalo2, bekithemba qeqe3 & baneng naape4*1 1 department of economics, university of limpopo, south africa 2 department of economics, university of limpopo, south africa 3 department of economics, university of fort hare, south africa 4 department of economics, university of the witwatersrand, south africa abstract this study was aimed at investigating the effects of energy prices on inflation in south africa. this was achieved by means of econometric analysis and annual time series data spanning from 1994 to 2020. the findings revealed that electricity tariffs and petrol prices exhibit a positive effect on inflation in south africa. on the contrary, interest rates and the exchange rate as control variables were found to have a negative effect on the inflation rate. the impulse response function indicated that inflation responds positively to innovations in petrol prices, electricity tariffs and money supply in the medium and long term but responds negatively to innovations in interest rates and exchange rates. further to this, the variance decomposition function revealed that variations in inflation are largely explained by its own innovations and partially explained by innovations in petrol prices, money supply and interest rates. lastly, the granger causality analysis showed no evidence of causality between inflation and explanatory variables during the specified period. given that energy prices place upward pressure on inflation, the study recommends a review of the current fuel and electricity tariff structure to provide the much-needed relief to households and businesses. keywords energy prices, inflation, south africa, var approach jel classification b22, b27, c13, e31, q4 1 *banengnaape@gmail.com mailto:banengnaape@gmail.com marius masoga, thabang kumalo, bekithemba qeqe & baneng naape 2 1. introduction and background the covid19 pandemic and the corresponding volatility in commodity markets, particularly the energy markets, has renewed interest in the dynamic link between energy price movements and economic performance (kenny, 2004). the cost of living in south africa is undoubtedly skyrocketing. for example, in recent months, energy prices have accelerated markedly and placed upward pressure on inflation. according to statssa (2021) petrol prices were increased by 40.5% in december 2021 as compared to 2020. the price of petrol for the first time in history has hit the r20 bar, as the price of inland 95-octane petrol raised to r20.29 per litre in december 2021. oil prices have also been on the rise, while the south african rand against the us dollar has been relatively stable. electricity tariffs are no exception. in the past 13 years, electricity prices increased by 307%, surpassing inflation. state utility “eskom” has applied for an electricity price increase of 20.5% for the 2023 financial year, which was effective on the 1st of april 2022. meanwhile, on 5 march 2021, the national energy regulator of south africa (nersa) approved a hike of 15.6% for individual households and 17.8% for municipalities, which was thereafter implemented on the 1st of april 2021 and 1st july 2021 respectively (nersa, 2021). according to langager (2013), inflation can be defined as the general rise in prices in general. a major concern for policy makers worldwide is the rising energy prices and subsequent high inflation rates which erode economic growth and domestic investment. rizvi and sahminan (2020) argue that the worst is yet to come for private households in terms of price increases for natural gas and electricity. energy prices were the main drivers of inflation in 2021. from an economic perspective, higher energy prices have the potential to slow economic recovery (jackson, 2005). energy prices are a vital instrument in determining inflation and are one of the most salient costs for households and businesses. a question remains as to whether the recent energy price movements are permanent or transitory. the price movements can be expected to have bigger implications for growth and inflation given that higher energy prices might have an effect on supply chains, profit margins and the likelihood for pass through to consumer prices and the wage bargaining process (nkomo, 2005). not with standing, the energy price dynamics play an important role in the context of climate transition and in terms of the commercial viability of longer-term investments. given the current situation, there is a strong need for the transition from fossil fuels to renewable energy sources such as solar, hydro and wind energy. nuclear energy has likewise yielded positive returns in countries that have invested in it, although it requires huge capital injections and technical know-how (agama, 2003). the relationship between energy prices and inflation is of paramount importance for the journal of applied economics and business studies, volume. 6, issue 2 (2022) 1-14 https://doi.org/10.34260/jaebs.621 3 south african economy (devpura and narayan, 2020). notably, in terms of income, south africa is the most unequal country in the world. ultimately, this implies that inflation disproportionately affects the poor population that does not have enough income to keep up with increases in energy and consequently food prices (nert, 2008). furthermore, south africa is an oil importing country and as a result, the country is exposed to external shocks of rising energy prices. the south african government since 1994, has been firm about balancing macroeconomics in order to attract investors, reduce the budget deficit and fight inflation through interest rates (aej, 2005). spalding-fecher (2002; 5) notes that “concerns around energy security led the apartheid government to develop a synthetic fuel programme to meet demand and to lessen the country’s dependence on energy imports”. imported crude oil accounts for 22% of primary energy used mainly by the transport sector. south africa’s massive investment in the 1960s and 1970s in coal-fired power plants left the national utility with large excess capacity in the 1980s and 1990s (nersa, 2002). the excess capacity has helped to keep electricity prices low, but it is now practically exhausted (eskom 2000). the energy sector in south africa has been at the heart of policy debates. this is because, energy generation and distribution are crucial for the well-functioning of the economy. sustained economic growth is vital to job creation and foreign direct investment. it is for this reasons that this study aims to examine the effect of energy prices on inflation in south africa. earlier studies (e.g., rangasamy, 2017; madito and odhiambo, 2018; rafiq, 2014) have isolated energy variables (i.e., electricity tariffs and oil price) to assess the drivers of inflation, a gap which this study aims to fill. throughout the world, energy prices are on the increase which directly impacts input prices, transport fair and other aspects of the supply chain leading to an increase in inflation. the paper has addressed the phenomena of inflation in such a hard time and is the need of the day to investigate such a topic in order to identify and understand the policy events that resulted in an increase in energy prices globally. 2. literature review 2.1 theoretical literature theoretical considerations form an integral part of this study to capture views from different schools of thought, such as the keynesian, monetarists, post-keynesian, and new classical school of thought. as one of the key macroeconomic variables in south africa (and the global economy alike), inflation is normally monitored closely by policymakers, notably the central banks. in monetary economics, as baumann, rossi and volkmannx (2021) indicated, inflation is an important monetary phenomenon. according to monetarists, inflation is induced by an increase in the quantity of money at a faster rate than output growth marius masoga, thabang kumalo, bekithemba qeqe & baneng naape 4 (i.e., an excess of money stock over the rate of output growth) in the economy (moosa, 1982) and (branson, 1975). an important implication of the monetarist model stated by the latter author is that the national rate of inflation ultimately converges to the world inflation rate. the analysis of inflation as a monetary phenomenon does not specify the underlying behaviour of individuals that coincide with the inflation rate (colander, 1992). nevertheless, growth in money stock can be used to explain the consumers behaviour and the transmission to inflation. in this instance, the level of interest rate manipulated by central banks is a good starting point. using the south african reserve bank (sarb) as an example, a reduction of repo rate result in greater amount of goods and services bought by individuals and businesses in the economy, and ultimately cause inflation. the keynesian principle pertaining to “demand-pull inflation” (i.e., when aggregate demand surge because of increased consumption expenditure) is a pertinent referral to corroborate the views thereof. energy products like electricity and fuel are amongst the mostly purchased consumer goods and services (as per statistics reported in the earlier section of this study) that drives inflation on the demand side. moreover, the post-keynesian hongkil (2020) indicated that us inflation was more driven by supply-side variables in the past than demand-side variables recently. the views of neoclassical economics reverberate the monetarists theory that attribute inflation rate to excessive growth in exogenous money supply. in controversy, the postkeynesian economic theory considers money supply to be an endogenous variable (susjan and lah, 1997). therefore, it appears that inflation can emerge from either demand-side or supply-side depending on the outcome of key macroeconomic variables. as far as this study is concerned, the extent to which energy prices drives inflation in south africa depends on the data observed and methodology applied, while noting empirical results conversed in the next subsection. 2.2 empirical literature this section discusses the findings of studies pertaining to inflation rate driven by energy prices. considering the different sources of energy (such as, oil and electricity), scholars have scrutinised various determinants of inflation. for instance, madito and odhiambo (2018) examined the following factors that determine the movement of inflation; import prices, government expenditure, inflation expectations, labour costs, exchange rates and gross domestic product. moreover, rangasamy (2017) analysed the effect of petrol price movements on the rate of inflation in south africa, using mid-1970 data as the implied starting point. as a result, granger causality tests and the autoregressive distributed lag journal of applied economics and business studies, volume. 6, issue 2 (2022) 1-14 https://doi.org/10.34260/jaebs.621 5 approach to co-integration testing revealed that a surge in the price of petrol had a significant bearing on the outcomes of inflation in south africa. according to madito and odhiambo (2018), inflation is positively determined by government expenditure, import prices and labour costs. these results are based on quarterly data spanning from 1970q1 to 2015q4, discovered by the error correction model. nonetheless, when the intensity of oil production process diminishes amid lower inflation environment, the influence of oil price shocks on inflation proved to be lacklustre (balcilar, uwilingiye and gupta, 2018). in their study, the authors thereof have applied the time series data from january 1922 to july 2013 to examine the nexus between oil price and rate of inflation in south africa. the results obtained through the symmetric and asymmetric dynamic conditional correlation garch revealed a positive link between inflation rate and oil price. however, such correlation was less robust owing to the commitment by south african reserve bank to stabilize inflation when external shocks hit the economy. using a time-varying vector autoregression with stochastic volatility, rafiq (2014) examined the impact on uk inflation resulting from oil price shocks. to estimate the results, the incorporated data covered the period from 1980s to 2013. the findings demonstrated that the oil price shock led to a significant surge of inflation rate in the uk. in their study pertaining to oil shocks and the impact on the south african economy, chisadza, dlamini, gupta and modise (2016) found a short-lived significant impact of shock to oil supply on the inflation rate. therefore, in the absence of monetary policy reaction, disturbances on the supply lead to a short-term increase in the inflation rate domestically. the methodology applied by the study thereof is called “a sign restriction-based structural vector autoregressive (var)”, and the ultimate result emphasized the need to understand the source of the variations in oil price. baghestani (2014) examined whether inflation expectation by consumers in the us are useful to accurately predict the direction of adjustments in energy prices. using the period 1987 – 2012, a forecasting model was employed to generate the one-two-three quarter ahead random walk forecast of crude oil heating oil and gasoline. as a result, it was determined that expectations may contain predictive information that is useful for crude oil, heating oil and gasoline. thus, the direction of change in energy prices was accurately predicted by change in expectations on inflation. meanwhile, it is vital to consider the spread of energy prices to high food prices, resulting in inflation. as revealed by spectral bc causality test, the energy price index and food price indexes have a bidirectional causality (kirikkaleli and darbaz, 2021). against this background, oscillations in energy prices have become a pertinent factor that determines the performance of the macroeconomy. to understand the roots and transmission marius masoga, thabang kumalo, bekithemba qeqe & baneng naape 6 mechanism leading to devastation on the overall economy, this study scrutinizes the sources of primary energy (namely, oil, petrol, and electricity prices). therefore, intensifying the existing literature pertaining to energy prices and transmission to inflation rate in south africa. as per the discussion of the empirical literature, most studies like rangasamy (2017); madito and odhiambo (2018); rafiq (2014), have isolated energy variables (i.e., petrol price and oil price) to assess the impact on inflation. nevertheless, this study combines a few selected variables (specified in the methodology section) to serve as the basis for the significance of the study, using unrestricted var approach. this study is important as policy guidance amid high volatility and unstable inflation rate that present greatest risk of uncertainty in the economy. 3. methodology the study followed a quantitative research approach by making use of modern econometric techniques. the study employed annual time series data spanning from 1994 to 2020. the data was collected from reliable databases including the department of energy and mineral resources, st louis federal reserve and world development indicators. in line with chisadza et al., (2016), our empirical model can be expressed in linear form as: 𝑦𝑡 = 𝛽0 + 𝛽1𝐴𝐸𝑇𝑡 + 𝛽2𝐵𝑃𝑃𝑡 + 𝛽3𝐼𝑁𝑇𝑡 + 𝛽4𝐸𝑋𝑅𝑡 + 𝛽5𝑀𝑆𝑡 + 𝜀𝑡 where: y_t is the dependant variable represented by the consumer price index aep is the average electricity tariffs bpp is the basic petrol price int is the nominal interest rates exr is the nominal exchange rate ms is the broad money supply ε_t is the error correction term the study employed an unrestricted vector autoregression approach consisting of the following steps: in the first step, the descriptive analysis and augmented dickey fuller stationarity test were executed as pre-estimation tests. following this, the standard vector autoregression model was estimated to examine the effect of explanatory variable namely: electricity tariffs, petrol prices, interest rates, money supply and nominal exchange rate on the dependant variable. the variance decomposition and impulse response were later performed to evaluate the response of the inflation rate to shocks in each of the explanatory journal of applied economics and business studies, volume. 6, issue 2 (2022) 1-14 https://doi.org/10.34260/jaebs.621 7 variables. furthermore, the granger causality analysis was executed to determine if energy prices cause inflation in south africa. 4. results and discussion this section provides a detailed discussion of the summary of findings obtained from the econometric analysis. this includes the descriptive analysis, stationarity analysis, standard var estimation, impulse response, variance decomposition and granger causality analysis. a summary of descriptive statistics is provided in table 1 below. table 01: descriptive statistics inflation electricity tariffs petrol prices exchange rate money supply interest rate mean 5.94 41.06 386.33 8.62 11.78 5.48 median 5.64 19.80 372.83 7.55 11.90 4.81 maximum 10.99 110.93 1136.58 16.45 20.70 12.6 minimum 1.43 10.32 50.34 3.55 1.72 2.31 std. dev. 2.14 33.82 277.93 3.54 5.82 2.78 skewness 0.25 0.78 0.69 0.62 0.02 1.21 kurtosis 3.04 2.11 2.99 2.43 1.83 3.53 jarque-bera 0.28 3.59 2.12 2.08 1.55 6.95 probability 0.87 0.17 0.346 0.35 0.46 0.03 observations 27 27 27 27 27 27 source: authors estimates electricity tariffs averaged 41.1 between 1994 and 2020 while basic petrol prices averaged 386.3 during the same period. inflation on the other hand, averaged 5.9 between 1994 and 2020 with a range between 1.4% and 10.9%. petrol prices have proven to have a relatively higher standard deviation (277) followed by electricity tariffs (33.8), money supply (5.8), exchange rate (3.5), interest rates (2.78) and inflation (2.13). in general, higher standard deviations indicate that the data points are spread out while lower values of standard deviation indicate that the data points are closer to the mean. in respect of the skewness of the data, the data appears to be slightly skewed to the right. the variables were examined for unit root. the results are summarized in table 2 below. table 2: stationarity analysis variable level 1st difference 2nd difference outcome inflation -3.55* -5.45* d(0) electricity tariffs 4.15 -2.15 -6.99* d(2) petrol prices -1.76 -6.37* d(1) marius masoga, thabang kumalo, bekithemba qeqe & baneng naape 8 exchange rate 0.06 -3.64* d(1) interest rates -2.11 -5.96* d(1) money supply -2.78** -5.37* d(0) source: author’s computations the adf stationarity test revealed that inflation and money supply are stationary at level while petrol prices, exchange rate and interest rates were found to be stationary after first differencing. electricity tariffs on the other hand, were found to be stationary after second differencing. table 3 below provides a summary of findings from the var estimation. the akaike information criterion (aic) recommended one lag. table 3: var estimation inflation petrol price electricit y tariffs exchange rate interest rate money supply inflation (-1) 0.41 -0.25 0.04 -0.03 0.17 -0.54 (0.19) (0.14) (0.03) (0.06) (0.17) (0.28) [ 2.09] [-1.79] [ 1.18] [-0.48] [ 1.00] [-1.96] petrol price (-1) 0.08 0.81 0.09 0.06 -0.45 -0.41 (0.20) (0.14) (0.03) (0.07) (0.18) (0.28) [ 0.41] [ 5.71] [ 2.76] [ 0.97] [-2.51] [-1.43] electricity tariffs (1) 0.03 0.07 0.93 0.09 0.22 -0.12 (0.23) (0.16) (0.04) (0.07) (0.20) (0.32) [ 0.13] [ 0.46] [ 24.05] [ 1.24] [ 1.08] [-0.38] exchange rate (-1) -0.49 -0.13 -0.07 0.74 -0.03 0.12 (0.36) (0.25) (0.06) (0.12) (0.32) (0.51) [-1.37] [-0.51] [-1.14] [ 6.38] [-0.08] [ 0.24] interest rate (-1) -0.29 -0.15 -0.01 0.13 0.04 -0.46 (0.23) (0.16) (0.04) (0.07) (0.21) (0.32) [-1.25] [-0.95] [-0.34] [ 1.81] [ 0.16] [-1.41] money supply (-1) 0.21 0.02 -0.03 0.02 0.13 0.23 (0.16) (0.11) (0.03) (0.05) (0.14) (0.22) [ 1.36] [ 0.17] [-1.31] [ 0.34] [ 0.98] [ 1.02] c 1.41 1.78 -0.05 -0.28 2.78 5.89 (1.17) (0.83) (0.20) (0.38) (1.05) (1.66) [ 1.20] [ 2.16] [-0.23] [-0.74] [ 2.64] [ 3.54] source: author’s computations the findings from the var estimation indicate that petrol prices, electricity prices and money supply have a positive effect on the inflation rate. the rationale is that, increase in the journal of applied economics and business studies, volume. 6, issue 2 (2022) 1-14 https://doi.org/10.34260/jaebs.621 9 basic petrol price raises the fuel pump price and ultimately, public and private transport costs. transports costs are factored into the cost of production and thus, increases in the price of fuel will presumably result in an increase in the total cost of production. also, certain manufactures involved in the production of iron and steel, aluminium, chemicals and cement, make use of fuel. the same rationale applies to increases in electricity tariffs. electricity forms a significant operational cost in most, if not all, industries. thus, increases in electricity tariffs place upward pressure on production costs. this translates into increases in food prices. the combination of food and transport inflation leads to the rise in prices in general and consequently, headline inflation. these findings are in line with rafiq (2014), rangasamy (2017) and balcilar et al., (2018) who analysed the impact of oil price shocks on inflation. the exchange rate and interest rates were found to have a negative effect on inflation. this is because central banks make use of interest rates to curb inflation in the economy. when interest rates rise, the cost of borrowing increases, thus inducing debtors to make less use of credit facilities in fear of incurring higher interest charges. this results in less money circulation and credit purchases. the appreciation of the exchange rate negatively influences the domestic inflation rate. the reasoning is that, when the exchange rate increases, the exporting country becomes less competitive in the global market, holding other factors constant. as such, the quantity of goods and services demanded by foreign consumers decreases due to relatively higher prices as a result of the appreciation of the exchange rate. in contrast, money supply was found to move in the same direction as the inflation rate. theoretically, an increase in the quantity of money in circulation triggers the demand for goods and services. when the increased demand for goods and services is not met by an equivalent increase in production, this leads to inflation. following this, the impulse response function was executed to determine the response of inflation to shocks in explanatory variables. the findings are illustrated in figure 01. marius masoga, thabang kumalo, bekithemba qeqe & baneng naape 10 figure 01: impulse responses of inflation -.4 -.2 .0 .2 .4 .6 1 2 3 4 5 6 7 8 9 10 accum response of inf lation to inf lation -.4 -.2 .0 .2 .4 .6 1 2 3 4 5 6 7 8 9 10 accum response of inf lation to petrol prices -.4 -.2 .0 .2 .4 .6 1 2 3 4 5 6 7 8 9 10 accum response of inf lation to electricity tarif f s -.4 -.2 .0 .2 .4 .6 1 2 3 4 5 6 7 8 9 10 accum response of inf lation to exchange rate -.4 -.2 .0 .2 .4 .6 1 2 3 4 5 6 7 8 9 10 accum response of inf lation to interest rates -.4 -.2 .0 .2 .4 .6 1 2 3 4 5 6 7 8 9 10 accum response of inf lation to money supply accumulated response to cholesky one s.d. (d.f. adjusted) innovations ± 2 s.e. source: author’s computations the response of inflation to its own innovations was found to be positive over the entire period. this implies that previous values of inflation inform current inflation rates. the response of inflation to a one standard deviation in petrol prices and electricity tariffs was found to be negative in the short term and positive in the medium and long term. the negative response of inflation in the short run might be due to economic lags. nonetheless, the response of inflation to innovations in the exchange rate and interest rate was found to be negative across the entire period, while on the contrary, the response of inflation to innovation in money supply was found to be positive in the short and long run. the estimated model was assessed for autocorrelation and heteroskedasticity. table 4: residual diagnostics autocorrelation heteroskedasticity lags lm-stat prob chi-sq. df. prob. 1 45.99 0.16 254.69 252 0.44 source: author’s computations the results indicated that the model is free from autocorrelation and heteroskedasticity. this is because, the probability values in both the autocorrelation and heteroskedasticity test were above 5%, thus confirming that the estimated model is free from spurious regression. the variance decomposition test was likewise performed to measure forecast errors of each variable in relation to its own shock. the results are summarised in table 5. journal of applied economics and business studies, volume. 6, issue 2 (2022) 1-14 https://doi.org/10.34260/jaebs.621 11 table 5: variance decomposition of inflation period s.e. inflation petrol price electricit y tariffs exchang e rate interest rate money supply 1 0.38 100.00 0.00 0.00 0.00 0.00 0.00 2 0.41 91.58 0.55 0.20 0.00 3.33 4.35 3 0.42 86.96 0.52 0.24 0.04 7.49 4.75 4 0.43 85.17 1.41 0.42 0.25 8.19 4.56 5 0.44 83.99 2.42 0.46 0.45 8.05 4.62 6 0.44 83.54 2.88 0.45 0.54 7.97 4.63 7 0.44 83.43 2.98 0.48 0.56 7.93 4.61 8 0.44 83.40 2.99 0.53 0.56 7.91 4.61 9 0.44 83.35 2.98 0.61 0.56 7.90 4.59 10 0.45 83.23 3.04 0.69 0.56 7.89 4.59 source: author’s computations the results from the variance decomposition indicate that in the short run (1-3 years), variations in the inflation rate are largely explained by its own shocks while 7.5% and 4.7% of the variations are explained by shocks in interest rates and money supply, respectively. in the medium term (4-6 years), variations in inflation are explained by 83.5% of its own shocks, 2.9% by shocks in petrol prices, 7.9% by shocks in interest rates and 4.6% by shocks in money supply. shocks in the exchange rate and electricity tariffs are found to have little to muted effects on variations in the inflation rate. in the long term (7-10 years), variations in the inflation rate remain largely explained by its own innovations and partially by innovations in money supply (4.6%), interest rates (7.9%) and petrol prices (3.0%). it is worth noting however that that the share of interest rate shocks in inflation appear to be increasing at a decreasing rate in the medium to long term. the granger causality test was executed to determine the direction of causality between the dependant and explanatory variables. the results are given in table 6. table 6: var granger causality/block exogeneity wald tests dependent variable: inflation excluded chi-sq. df prob. petrol price 0.17 1 0.68 electricity tariffs 0.02 1 0.89 exchange rate 1.89 1 0.16 interest rate 1.57 1 0.21 money supply 1.84 1 0.17 all 6.71 5 0.24 source: author’s computations, asterisks ** indicate statistical significance at the 5% level the results in table 6 indicate no evidence of causality between the dependent variable marius masoga, thabang kumalo, bekithemba qeqe & baneng naape 12 and explanatory variables. this implies that past values of money supply, interest rates, electricity tariffs, exchange rates and petrol prices do not contain information that helps predict current and future values of inflation in the given period of study. the null hypothesis of no causality is thus accepted at the 5% significance level against the alternative hypothesis of causality. 5. conclusion and recommendations the primary goal of this study was to investigate the effects of energy prices on inflation in south africa. this was achieved by means of econometric analysis comprising of the vector autoregression model, impulse response function, variance decomposition and granger causality analysis. the study made use of annual time series data spanning from 1994 to 2020. the findings revealed that electricity tariffs and petrol prices exhibit a positive effect on inflation in south africa. on the contrary, interest rates and the exchange rate as control variables were found to have a negative effect on the inflation rate. the impulse response function indicated that inflation responds positively to innovations in petrol prices, electricity tariffs and money supply in the medium and long term but responds negatively to innovations in interest rates and exchange rates. further to this, the variance decomposition function revealed that variations in inflation are largely explained by its own innovations in the short, medium and long term and partially explained by innovations in petrol prices, money supply and interest rates. lastly, the granger causality analysis showed no evidence of causality between the inflation rate and explanatory variables. given that energy prices place upward pressure on inflation, the study recommends a review of the current fuel and electricity tariff structure to provide the much-needed relief to households and businesses. references aej (2005). who’s ahead in the htr race – china or south africa? african energy journal 7 (3): pp. 12-15. africa, a. (2003). demand side management in south africa. electricity supply industry 1: pp.44–46. agama energy. (2003). employment potential of renewable energy in south africa. a study commissioned by sustainable energy and climate change partnership, a project of earth life africa johannesburg, in partnership with wwf, denmark. johannesburg, seccp. baghestani, h. (2014). inflation expectation and energy price forecasting. organization of the petroleum exporting countries. published by john wiley & sons ltd. baumann, p.f.m., rossi, e. and volkmann, a. (2021). what drives inflation and how? evidence from additive mixed models selected by caic. swiss national bank working papers, 12(-), pp.1–57. balcilar, m. uwilingiye, j. and gupta, r. (2018). dynamic relationship between oil journal of applied economics and business studies, volume. 6, issue 2 (2022) 1-14 https://doi.org/10.34260/jaebs.621 13 price and inflation in south africa. the journal of developing areas, 52(2), pp.73–93. branson, w.h. (1975). monetarist and keynesian models of the transmission of inflation. american economic review, 65(2), pp.115–119. chisadza, c., dlamini, j., gupta, r. and modise, m.p. (2016). the impact of oil shocks on the south african economy. energy sources part b: economics, planning and policy, 11(8), pp.739–745. colander, d. (1992). a real theory of inflation and incentive anti-inflation plans. american economic review, 82(2), pp.335–340. devpura, n., & narayan, p. k. (2020). hourly oil price volatility: the role of covid19. energy research letters, 1. 13683. https://doi.org/10.46557/001c.13683 gelos, g., & ustyugova, y. (2017). inflation responses to commodity price shocks– how and why do countries differ? journal of international money and finance, 72, pp.28-47 hongkil, k. (2020). a missing element in the empirical post keynesian theory of inflation – total credits to households: a first-differentiated var approach to u.s inflation. journal of post-keynesian economics, 43(4), pp.640–656. jackson, g. (2005). oil prices and the inflation myth. available at http://www.brookesnews.com/051010oil.html accessed on 22 february 2022 kenny, a. (2004). energy supply in south africa’ in energy for sustainable development: south african profile. energy research centre, university of cape town. kirikkaleli, d. and darbaz, i. (2021). the causal link between energy price and food price. energies, 14(-), pp.2–13. lengager, c. (2013). what is the relationship between oil prices and inflation? available at http://www.investopedia.com/ask/answers/06/oilpricesinflation.asp. accessed on 20 february 2022 madito, o. and odhiambo, n.m. (2018). the main determinants of inflation in south africa: an empirical investigation. organisations and markets in emerging economies, 9(2), pp.212–232. moosa, s. (1982). money, inflation, and the monetarist explanation: evidence from the post-war u.s experience. eastern economic journal, 8(2), 101–119. national electricity regulator south africa annual reports, 1995 – 2002. national electricity regulator south africa 2021, regulatory framework for the economic regulation of the electricity supply industry (esi) of south africa, discussion document by the nersa pricing & tariff department, pretoria nkomo, j.c, (2005). energy and economic development: challenges for south africa. energy research centre, university of cape town nkomo, j.c. (2006). the impact of higher oil prices on southern african countries. marius masoga, thabang kumalo, bekithemba qeqe & baneng naape 14 journal of energy in southern africa, 17(1), 10–17. rangasamy, l. (2017). the impact of petrol price movements on southern african inflation. journal of energy in southern africa, 28(1), 120–132. rizvi, s.r & sahminan, s. (2020). commodity price and inflation dynamics: evidence from brics. bulletin of monetary economics and banking, vol. 23 no. 4, 2020, pp. 485 – 500 p-issn: 1410 8046, e-issn: 2460 9196 spalding-fecher, r. (2002). energy sustainability indicators for south africa. energy and development research centre, university of cape town. susjan, a. and lah, m. (1997). inflation in the transition economies: the post-keynesian views. review of political economy, 9(4), 381–393. journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 59 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x productivity and export performance of emerging asian economies asma saeed1 & mehrin zaid ullah2 1 lecturer, department of economics, women university mardan (wum), pakistan. 2 lecturer, department of economics, mirpur university of science & technology (must), mirpur, aj&k, pakistan. abstract this study estimates the effect of productivity on export performance of selected sectors and countries. specifically, this study analyzes both theoretically and empirically, the nexus between productivity and export performance of selected emerging asian economies including china, india, indonesia, south korea, and japan. total factor productivity (tfp) is used as a proxy for productivity and measured using two different techniques, growth accounting framework and data envelopment analysis. the study uses auto regressive distributed lag technique to estimate the effect of productivity on export performance of agricultural and manufacturing sectors of emerging asian economies over the period of 1990 to 2016. results show that tfp significantly and positively affect the agricultural and manufacturing sectors’ export performance of selected emerging asian economies both in long and short run. hence, policies and economic strategies promoting adoption of advanced production technologies will increase exports of the emerging asian economies in the long run. keywords productivity, total factor productivity, export, growth accounting framework, data envelopment analysis jel classification o47, f1 1. introduction the factors responsible for the export performance are classified into two main categories by the literature. the first set of factors is concerned with the supply side factors such as total production capacity and productivity. productivity is the ratio of output(s) to input(s). it reflects technological progress and the efficiency of all factors of production (liao & liu, 2009). productivity can significantly contribute not only to the https://scholar.google.com.pk/citations?view_op=view_org&hl=en&org=8197556685281238691 asma saeed and mehrin zaid ullah 60 economic growth and investment but can also boost the export performance of a country (ahmad, ilyas, mahmood, & afzal, 2010; comin & gertler, 2006; david fadiran & akanbi, 2017; jienwatcharamongkhol, 2013). the literature has explained the nexus between productivity and export performance in at least two ways. first, productivity raises export performance via raising the efficiency of the production factors which lowers the cost of production in domestic market and this in turn increases the competitiveness of domestic product in the international market (morley & morgan, 2008). this relationship between productivity and export performance is commonly referred to as the productivity-led export growth hypothesis. melitz (2003) provides the theoretical framework for exploring this relationship using heterogenous firms, monopolistic competition, differentiated products and only one factor of production (i.e.labor). melitz (2003) argued that firms can self-select themseleves to exports markets only if they have high level of productivity. because firms with high level of productivity are able to afford trade costs. so, more producive firms can expand output and this in turn will lead to generate exportable surplus, higher revenues, lowering prices and higher profits. helpman, melitz, and yeaple (2004) and head and ries (2003) extended the the melitz (2003) framework. . helpman et al. (2004) consider firms’ foreign direct investment (fdi) in the analysis to extended melitz model. helpman et al. (2004) argue that more productive firms attract fdi which further extend their exports. while head and ries (2003) extend the melitz (2003) model by incorpoating heterogenity with respect to foreign countries. they states that less productive frims can take advatge of the low cost countries in terms of labor wages. therfore, head and ries (2003) pointed out that less productivie firms should invest in low cost countries, as it will reult in positive profits for less productive firms. alongside the less productive firms, high productive firms can also consider countries with low labor cost. the productivity-led export growth hypothesis has been supported by most of the empirical work. bernard and jensen (1999) showed that it is the productivity which leads to an increase in the export growth. thus, their findings accept the productivity-led export growth hypothesis. sharma and mishra (2009) also confirmed that productivity causes the export growth. second, the productivity-export performance relationship was further elaborated by the new endogenous growth model. this model provides a framework for linking trade and productivity growth to find out the way trade can increase the productivity growth of the host country through technology spillover effects. romer (1993) pointed out that trade causes productivity by increasing access to new ideas. according to him if a firm journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 61 invests in research and development, the resulting benefit may not be restricted only to this firm but also will spillover to other firms and trade partners and hence can increase their production. therefore, export-oriented knowledge leads to productivity. this relationship between exports and productivity is referred as learning by exporting hypothesis. grossman and helpman (1991) argued that firms’ exposure to foreign markets results in the knowledge spillover effect. the knowledge gained in this way can then be used by the exporting firms to further increase their level of productivity. this knowledge spillover effect will also help the firms in improving their design of the products and the process of production. so, the exporters can gain technological progress and knowledge through learning by exporting effect which ultimately increases the productivity of the firms. the learning by exporting hypothesis is also supported by clerides, lach, and tybout (1998). clerides et al. (1998) claimed that exposure to trade increases the firm’s productivity through transfer of knowledge from foreign buyers. they also claimed the firm’s productivity depends on his previous experience of the export participation. the findings of greenaway and kneller (2004) and baldwin and gu (2003) also revealed that productivity increases with the increasing participation in the foreign markets. baldwin and gu (2003) showed that productivity is positively associated to the economies of scale and product specialization. they also states that one percent increase of the firm’s participation in the export market increases the labor productivity by 6% and tfp by 2%. baldwin and gu (2003) also found that exporters are productive as compared to non-exporters. the second category of the factors responsible for the export performance is concerned with the demand side factors of export performance. it focuses on the world income, exchange rate and export prices as the main reasons for export demand. world income is expected to have a positive association with world’s export expansion. however, exchange rate and export prices have negative association with export demand. the association between world’s income, exchange rate, export prices and export demand has been confirmed by nadeem, azam, and islam (2012) and gururaj, satishkumar, and kumar (2016). in summary, there is a wealth of the literature that discusses the factors affecting the export performance. most of the studies on the nexus between productivity and export performance used ad hoc models without providing a theoretical framework to estimate the impact of productivity on export. the present study filled the gap by developing the theoretical framework by reconciling consumer and producer side which are obtained from the process of optimization, and then applied it to study the relationship between productivity and export performance of agricultural and manufacturing sectors in china, asma saeed and mehrin zaid ullah 62 india, indonesia, south korea, and japan. findings on both the productivity-led export growth hypothesis and learning by exporting hypothesis are mixed so far. similarly, the effect of demand and supply side factors on the export performance have been extensively studied. however, the evidence to support productivity as a major determinant of the export performance is limited in the sample countries. therefore, the present study is motivated by the lack of the theoretical model and the mixed results from empirical models investigating the nexus between productivity and export performance in the sample countries. the rest of the paper is arranged into four sections. section 2 provides summary of earlier studies. section 3 discusses theoretical model, data, and empirical model. section 4 compiles empirical results of the econometric model. while section 5 summaries the conclusion of the study. 2. literature review in this section, a review is provided on the nexus between export and total factor productivity. this review provides the theoretical as well as the empirical framework for the present study. it is discussed as follows 2.1 theoretical models of the nexus between productivity and trade the ricardo idea of comparative advantage can be traced back to the link between productivity and export performance. according to this theory, even though countries do not have an absolute advantage in the manufacture of goods, they may benefit from trade by specializing in the production of goods in which they have a productivity advantage. ricardo presented his theory based on the assumption of only one factor of production (labor) and constant return to scale in production. ricardo’s theory mainly focused on the trade in the homogenous products and the inter-industry trade of two goods between two countries. contrasting to ricardo’s theory of comparative advantage, the heckscher-ohlin (h-o) proposed the theory of the endowments of the production factors. h-o highlights that a country will produce and export the good that is produced from its relatively abundant factor. so according to this theory, the difference in the factor’s endowment is the major driver of trade between countries. like ricardo’s theory, this theory is also based on some assumptions such as the assumptions of two factors of production (labor and capital), constant returns to scale, and perfect competition. the h-o theory also viewed an inter-industry trade of two homogenous goods between two countries. journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 63 though the h-o theory focused on the differences in the factor’s endowment as the main cause of trade, however, this theory failed to elucidate some recent realities. the first is the non-existence of perfect competition in the real world as most of the production of goods takes place under imperfect competition. secondly, the goods are differentiated and not homogenous. thirdly, unlike the new trade theories, old trade theories assumed that goods were commonly produced under constant returns to scale. similarly, the old trade theories focused only on inter-industry trade and failed to explain the intra-industry trade. the work of krugman (1980) generally referred to as the new trade theories in the literature incorporated the monopolistic competition and increasing returns to scale into the old trade theories. in the monopolistic competition, goods are differentiated so firms can produce and export its unique variety of the goods. further, under increasing returns to scale, the average cost of production decreases as the production of goods increases. in the new trade theories, firms are assumed to be homogenous. melitz (2003), however, incorporated firms heterogeneity into the new trade theories. melitz (2003) assumed that firms are different from each other based on the differences in the level of productivity. krugman (1987) developed a model of productivity where it is assumed that there are two countries (i) domestic (d); and (ii) foreign (f) country. krugman (1987) postulates that the productivity of a domestic country increases with the exposure to trade with a foreign country over time. this implies that a firm’s productivity increases through learning by exporting. firm’s exposure to foreign trade results in the knowledge spillover effect. the knowledge gained in this way can then be used by the exporting firms to increase their level of productivity. melitz (2003) showed that the fixed cost of entry and productivity are the two major determinants of a firm’s exports. melitz (2003) criticised krugman for assuming homogenous firms within sectors. melitz (2003) showed that firms can export only if they have a high level of productivity. 2.2 empirical review of the factors influencing export there are numerous studies, which have empirically examined the impact of productivity and other macroeconomic variables on the export performance. 2.2.1 productivity and export performance a lot of research has discovered a link between productivity and export success that is both positive and significant. south africa, kenya, egypt, madagascar, asma saeed and mehrin zaid ullah 64 morocco, tanzania, mauritius, and zambia were among the african economies studied by balchin (2008), who looked at the influence of productivity on export performance. balchin (2008) used a panel data set from 2002 to 2005 to quantify the influence of productivity on export performance using a probit regression model. balchin (2008) discovered that the productivity coefficient is positive and significant at the 5% level of significance. this demonstrates that the better the productivity of african enterprises, the more likely they are to export on the global market. ricci and trionfetti (2011) examined the relationship between productivity, comparative advantage and export peformance. ricci and trionfetti (2011) found that productivity and comparative adavatage positively affects the firms’ export performance. faustino, lima, and matos (2012) conducted a study on exports of portugal to spain. specifically, faustino, lima, and matos (2012) studied the impact of productivity and innovation on export performance of portugal. they used panel data over the period 2004 to 2008, the pooled ols and gmm methods to empirically evaluate the impact of productivity and innivation on export performance. faustino, lima, and matos (2012) found that the expected positive sign of the coefficient of tfp is consistent and showing that an increase in productivity leads to an improvement in portugal’s exports to spain. so, a positive association is found between productivity and export performance of portugal. similarly, the coefficient of r&d expenditures is also positive and significant implying that innovation increases the export performance of portugal. amornkitvikai, harvie, and charoenrat (2012) investigated the nexus between productivity and the export decisions of firms in thailand. amornkitvikai et al. (2012) argued that a firm can become exporters only if they have a high level of productivity. the impact of productivity on export decisions of firms is tested by using cross-sectional data for the period 2007 and estimating probit and logit regression models. they also found that productivity proxied by the level of output per unit of labor has a positive impact on firms’ decision to export. reis and forte (2016) examined the impact of productivity on the export performance of firms with reference to portugal. using panel data for the period 2008 to 2010, they found that the coefficient of labor productivity is positive and significant implying that firms having a high level of productivity could increase its exports performance. journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 65 2.2.2 gdp and export performance there have been quite a few studies that estimated the relationship between the gross domestic product (gdp) and the export performance in various countries. chan and au (2007) estimated the determinants of textiles’ export of china. specifically, they studied the impact of gdp on china’s textile export. chan and au (2007) argued that raising gross domestic product (gdp) of the exporting country results in the expansion of production capacity and export supply. they used time series data for the period 1985-2004 and the gravity model for the estimation of results. they found that all the variables except distance are significant at 1 percent level of significance. the coefficient of china’s gdp is positive and significant 1 percent, showing that an increase in gdp increases the textile export of china. while the coefficient of exchange rate is negative and significant implying that an increase in the exchange value of china’s currency reduces exports. das (2007) studied the relationship between gdp and exports of china, india, and malaysia. he used secondary data covering the period 1995 to 2000. he showed that the coefficient of gdp is positive and significant implying that an increase in production results in improving the export performance of sample countries. while fdi had no impact on the export performance. srinivasan (2012) estimated the relationship between gdp and exports of india by using the ardl model. srinivasan (2012) concluded that gdp and world income is positively related to the exports of india. while the real exchange rate has a negative relationship with the export performance showing that appreciation of the exchange rate lowers exports of india. similarly, bhatt (2013) investigated the impact of gdp and fdi on exports of china. he used the vector autoregressive model and time series data for the period 1978 to 2009. he found that gdp and fdi have a positive impact on china’s exports showing that an increase in the gdp and fdi inflows raises the export performance of china in the long run as well as in the short run. jawaid, raza, mustafa, and karim (2016) analyzed the relationship between gdp and exports of pakistan. the study used time series data over the period 1974 to 2012. autoregressive distributed lag (ardl) model is employed for the estimation of results. jawaid et al. (2016) showed the the coefficient of gdp is positive and significant at 5 percent level of significance, implying that higher the gdp higher is the exports growth of pakistan both in the long run and short run as well. similarly, asma saeed and mehrin zaid ullah 66 fdi also has a positive impact on the export performance of pakistan in the long run as well as in the short run. potelwa, lubinga, and ntshangase (2016) analyzed the growth of south africa’s agricultural exports. this study used the gravity model for the period 2001-2014 to estimate the determinants of export growth. they concluded that gdp and population of south africa’s trading partners cause an increase in the agricultural export growth of south africa. the gross domestic product of south africa and export capacity also caused an increase in its export growth. epaphra (2016) used the ols method and concluded that gross domestic product, exchange rate, and trade liberalization positively affect the export in tanzania. 2.2.3 real exchange rate and export performance veeramani (2008) analyzed the impact of real exchange rate on exports of india using time series data for the period 1960 to 2007 and multiple regression model. veeramani (2008) found that exchange rate is negatively and statistically significantly associated to the export performance of india. veeramani (2008) also found real gross domestic product positively affect the exports suggesting that an increase in the real gdp increases the production and supply capacity of exports in india. chit (2008) studied the nexus between bilateral exchanger rate volatility and bilateral exports with reference to emerging east asian economies including china, malaysia, indonesia, thiland, and philpines. chit (2008) used panel data set for the period 1982 to 2005 and a gravity-type bilateral trade model. fixed effect model was employed for the estimation of results. chit (2008) found the coefficient of bilateral exchange rate is positive and significant at 5 percent level of significance showing depreciation of the exchange rate has the effect of increasing bilateral export performance. while the coefficient of bilateral exchange rate instability is negative and significant at 1 percent level of significance, inferring that instability in the bilateral exchange rate declined the export performance in the sample countries. similarly, the coefficient of gdp of importing country and the common border is positive and significant at 5 percent level of significance. this means that there is a positive relationship between the importing country’s income and sharing a common border. kafayat (2013) evaluated the association between the exchange rate volatility and export performance of pakistan. he argued that volatility of exchange rate can effect exports performance positively and negatively as well. if volatility of exchange rate is journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 67 triggered by an appreciation of domestic currency then it will reduce domestic exports by making it expensive in the international market. however, an appreciation of domestic currency will increase imports by lowering the value of imported goods. on the other side, if the volatility of the exchange rate is triggered by depreciation of domestic currency then it will positively affect export performance and will negatively affect imports by raising prices of imported goods. kafayat (2013) employed time series data for the period 1981 to 2011 and regression model to empirically test the impact of depreciation on export and imports performance of pakistan. kafayat (2013) found that the coefficient of the exchange rate is positive and significant at 1 percent level of significance. however, depreciation of the exchange rate positively affects imports of pakistan, which is not consistent with the prior expectations. this study concluded that the exchange rate volatility triggered by depreciation has improved the export performance and have no or very little impact on imports of pakistan. he suggested that boosting exports via a depreciation of the exchange rate is not a good tool. for sustained export growth, the dire need is to enlarge the production capacity. and only through massive production pakistan can increase its export supply in the international markets. hassan, hassan, and mahmood (2013) investigated the nexus between exchange and exports performance of pakistan. they postulated that the appreciation of domestic currency can increase the supply of export by reducing the price of imported inputs used in the production of goods and services. hence, the supply side effects of the exchange rate appreciation result in expanding the capacity to exports. however, the demand side effects of the exchange rate appreciation lower the demand for exports in the international markets. they further argued that depreciation of exchange rate will increase the demand for exports and decrease the supply of exports by raising the cost of imported inputs and hence lowering domestic production capacity. they employed secondary time series data and using the ardl model for the empirical investigation of the impact of the exchange rate and other macroeconomic variables on the exports performance of pakistan. they concluded that the coefficient of the exchange rate is positive and significant showing that depreciation of the exchange rate has a positive impact on the export performance of pakistan. this result confirmed the existence of a long run relationship between exchange rate and export performance of pakistan. the coefficient of gdp and trade liberalization were also positive and significant showing that an increase in gdp and trade liberalization leads to an increase in the export performance of pakistan both in the long run and short run as well. asma saeed and mehrin zaid ullah 68 genc and artar (2014) examined the long run relationship between exchange rate and export performance for 22 emerging economies. they employed panel dataset and ardl model for estimating the long run relationship. they deduced that the pooled mean group estimator (pmg) results confirmed the existence of a long run relationship between the exchange rate and export performance of the sample economies. this implies that the appreciation of the exchange rate is not beneficial for the export growth in the sample countries. falianty (2015) studied the relationship between exchange rate and exports in indonesia. the main objective of his study was to estimate the impact of exchange rate depreciation on the export performance of indonesia. falianty (2015) used time series data (1995-2014) and regression model for the estimation of results. this study found that the coefficient of the exchange rate is positive and significant implying that one percent depreciation of indonesia’s rupiah increases its export performance by 0.103 percent. this means that depreciation of the rupiah can improve the export performance of indonesia. 2.2.4 world income and export performance santos-paulino (2002) examined the factors responsible for the export performance of developing economies. they found results consistent with the prior expectations. the coefficient of world income is positive and significant at 10 percent level of significance, showing that the larger the world income larger is the export growth. trade liberalization is also found with a positive impact on the export growth of developing countries. while relative prices had a negative association with the export performance. gupta, raychaudhuri, and haldar (2015) studied the determinants of india’s exports. gupta et al. (2015) estimated the relationship between exports and its major determinants using industry level data. they found a positive link between world income and export showing that the increase in world demand for exports increases the exports performance of india. similarly, public expenditures on research and development activities are also positively linked. whlie foreign capital inflows and appreciation of the real exchange rate results in lowering the export performance of india. in summary, there are many studies that discuss the factors affecting the export performance. however, evidence to support productivity as a major determinant of the export performance in emerging asian economies is limited. the previous studies journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 69 have also ignored using multiple measurements of productivity and hence do not provide robust results. as a result, the findings of these studies are mixed. the present study fills the gap mentioned above by developing the theoretical model and then uses such model to study the relationship between productivity and export performance of agricultural and manufacturing sectors of selected emerging asian economies namely china, india, indonesia, south korea, and japan. further, this study measures productivity with more than one technique and hence provides robust estimates of the effect of productivity on export performance of the sample economies. 3. data and research methods 3.1.theoretical model the present study develops a theoretical framework based on the framework of melitz (2003). the theoretical framework is described in detail as follows: 3.1.1 consumers consider an open economy with monopolistic competition. the key agents in this economy are consumers and producers. producers produces varieties of product, q_(ja(m)), in sectors a and sector m demanded by identical consumers. assume, a representative consumer having the following constant elasticity of substitution (ces) preferences. u = (∑ q ja(m) ρn i=1 ) 1 ρ (1) where u represent a utility function, n is the number of varieties of product q_(ja(m)) , ρ is a measure of substitutability while sectors a and m represents agriculture and manufacturing sectors, respectively. assuming non-satiation, the representative consumer maximizes utility subject to income constraint, p_(ja(m)) q_(ja(m))=i where p_(ja(m)) is the price of commodity q_(ja(m)) and i is the income of consumers. maximization yield the following marshallian demand function: qja(m) = pja(m) −σ i p1−σ (2) where σ is the elasticity of substitution among varieties, q_(ja(m)) is the quantity demanded of commodity j, p_(ja(m)) is the price of commodity q_(ja(m)), i is total income of the consumer and p is the price index of the economy other than q_(ja(m)). 3.1.2 producers producers are heterogenous in terms of their level of productivity. they produce varieties of product q_(ja(m)) using labor l_(ja(m)) and capital k_(ja(m)) in sector a asma saeed and mehrin zaid ullah 70 and sector m . where sectors a and m represents agriculture and manufacturing sectors, respectively. each variety competes with all other varieties, and hence the firm is better off by selecting a variety that is not being produced by others, to avoid sharing demand for its variety. the production function of the producers is assumed to have increasing returns to scale (irs). qja(m) = φa(m)f(lja(m), kja(m)) (3) where φ is producers’ productivity in sector a and sector m. the objective of the producer is to maximize profit given a technology approximated by a cost function (cja(m)) as follows. cja(m) = wlja(m) + ikja(m) + qja(m) φa(m) (4) where w represent wage of labor, l is labor, i represent interest rate and k is capital. maximizing the producer’s profit (π = pjqj − qj φ − wlj − ikj) yields the equilibrium price, pj = 1 ρφ which is a function of productivity of producers. producer’s supply function in sector a and sector m is generated as under. qja(m) = φa(m)(σ − 1)(wlja(m) + ikja(m)) (5) at national level, income of a country is determined by the entire use of labor and capital and their respective rewards that is gdp = wlja(m) + ikja(m) in our case and hence, qja(m) = φa(m)(σ − 1)gdp . this illustrates that producer’s supply function of varieties of product qja(m) is increasing with level of productivity φa(m) and income of the country. 3.1.3 the productivity-trade nexus the productivity and trade nexus is developed by subtracting the consumer’s demand function from the producer’s supply function. eja(m) = φa(m)(σ − 1)gdp − pja(m) −σ i p1−σ (6) where eja(m) represents exports of varieties of product qja(m) in sector a and sector m and other variables are already defined. journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 71 3.2. data and econometric methodology 3.2.1 data this study is carried out for a panel of five countries which include china, india, indonesia, south korea, and japan by employing panel data over the period 1990 to 2016. the selection of sample is purely based on the data availability. the data used in this study are agriculture and manufacturing sectors’ export performance, total factor productivity, real gross domestic product (gdp) per capita, consumer’s demand in the exporting country, real exchange rate, cost to export per container and world income. the data for agriculture and manufacturing sectors exports are collected from the world trade organization (wto). while the data for gdp, consumer’s demand, exchange rate, cost to export and world income are mainly taken from wdi. the data for total factor productivity is computed by employing the growth accounting method and data envelopment analysis. the detail computation of productivity data is reported in appendix-a 3.2.2 econometric methodology this section has two subsections. in first subsection 3.2.1 the unit root analysis is discussed. while second subsection 3.2.2 discusses estimation techniques that are used for the estimation of results. 3.2.3 unit root analysis the first step in econometric analysis is to conduct the unit root analysis. the unit root analysis is a standard approach that determines the stationarity of time series data. the present study carried out both panel unit root analysis as well as country specific unit root analysis. the most widely employed tests for panel unit root analysis are lavin, lin and chu-t test, im, pesaran and shin w-stat test and adf-fisher chisquare test. while augmented dickey-fuller test is mostly employed for a country specific unit root analysis. in these tests the null hypothesis of unit root (nonstationarity) is tested against the alternative hypothesis of stationarity. so, if the null hypothesis is accepted this means that the variables are non-stationary. the problem of non-stationarity in a given series can be solved by taking the first difference of the series. after taking the first difference, if the test-statistics leads to the rejection of null hypothesis this means that the variable is now stationary and can be used for the estimation purpose. the results of panel unit root tests are reported in table 1. the results of these tests reveal that the study variables include unit root meaning that they are not stationary at level. in case of full sample, test statistics demonstrates that both agricultural and manufacturing sectors’ total factor productivity (tfp), consumer asma saeed and mehrin zaid ullah 72 ‘demand and world income are stationary at level while real gdp per capita, real exchange rate, cost to export and agricultural and manufacturing sectors exports variables are non-stationary at level and become stationary at first difference. the results of country specific unit root test statistics demonstrated in table 2 also indicates that variables are not level stationary but made stationary at first difference. 3.2.4 estimation methodology the panel unit root test unit results reveal that the study variables include unit root meaning that they are not stationary at level. similarly, the country specific unit root test results also demonstrate that the variables have a mixture of order of integration meaning that some variables are integrated of order one and some are integrated of order zero. based on the unit root test results, therefore this study uses the auto regressive distributed lag (ardl) model for the estimation of results. so, the export performance model specified in section 3.1 is re-specified to get a panel data ardl model. the panel ardl model is specified as follows. ∆lnexpit = α° + ∑ βi r k=1 ∆lnexpit−k + ∑ δi r k=1 ∆ lnstfpit−k + ∑ δi r k=1 ∆ lngdpit−k + ∑ δi r k=1 ∆ lnfceit−k + ∑ δi r k=1 ∆ lnrexit−k + ∑ δi r k=1 ∆ lnceit−k + ∑ δi r k=1 ∆lnwyit−k + ∅1lnexpit−1 + ∅2lnstfpit−1 + ∅3lngdpit−1 + ∅4lnfceit−1 + ∅5lnrexit−1 + ∅6lnceit−1 + ∅7lnwyit−1 + εit (7) where β_i and δ_i are short term coefficients, ∅_1…. ∅_7 are the long term coefficients, i is the ith country, t is time period for the study variables, exp is sectoral export measured as the value of exports in million dollars, stfp is sectoral productivity, gdp is real gross domestic product per capita of exporting country, fce is the final consumption expenditure which is a proxy for the consumer’s demand in the model, rex is the real exchange rate, ce is the cost to export per container and wy is the world income measured by the us gdp, ∆ is the first difference operator and ε is the error term. the decision between pooled mean group (pmg) and mean group (mg) model is made based on the hausman specification test. the model is estimated as pooled mean group (pmg) panel data model after conducting the hausman specification test. journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 73 table 1: panel unit root tests source: authors ‘own calculation variables lavin lin chu test im pesaran test fisher-adf-chi square test level first difference level first difference level first difference t-stat p-value t-stat p-value t-stat p-value t-stat p-value t-stat pvalue t-stat p-value agricultural export 0.218 0.413 -5.681 0.000 1.667 0.952 -5.377 0.000 2.983 0.982 43.867 0.000 manufacturing export 1.728 0.958 -8.663 0.000 1.311 0.905 -8.387 0.000 4.079 0.943 69.538 0.000 agricultural tfp -4.396 0.000 -----3.522 0.000 ----28.198 0.002 ---- agricultural malmquist total factor productivity index (atfpi) -10.518 0.000 -----9.745 0.000 ----82.647 0.000 ---- manufacturing tfp -6.259 0.000 -7.306 0.000 -----61.284 0.000 ---- manufacturing malmquist total factor productivity index (mtfpi) -4.434 0.000 ------4.675 0.000 ----40.905 0.000 ---- real gdp per capita 0.553 0.710 -6.161 0.000 2.210 0.986 -5.341 0.000 12.255 0.268 44.894 0.000 consumer’s demand 1.301 0.903 -7.793 0.000 -2.293 0.989 -8.778 0.000 12.718 0.239 73.353 0.000 real exchange rate -0.982 0.163 -8.441 0.000 -0.498 0.309 -7.736 0.000 9.089 0.523 68.761 0.000 cost to export 1.212 0.887 -7.363 0.000 2.341 0.991 -6.870 0.000 1.812 0.997 66.103 0.000 world income -2.365 0.009 -----2.863 0.002 ----23.247 0.009 ---- asma saeed and mehrin zaid ullah 74 table 2: county specific unit root test (adf test) note: ***, **, * denote 1%, 5% and 10% level of significance respectively. china india indonesia south korea japan l e v e l 1 st d if fe r e n c e l e v e l 1 st d if fe r e n c e l e v e l 1 st d if fe r e n c e l e v e l 1 st d if fe r e n c e l e v e l 1 st d if fe r e n c e agricultural export -1.298 -3.847** -1.415 -3.905** -1.533 -5.281*** -1.808 -3.686** -1.574 -5.000*** manufacturing export -1.711 -4.309** -1.731 -4.706*** -2.742 -4.809*** -2.168 -4.376** -2.154 -5.808*** agricultural tfp -3.426** ---3.487* ---3.848** ---3.549** ---3.962** -- atfpi -4.961*** ---7.532*** ---7.532*** ---3.557* ---5.147** -- manufacturing tfp -3.319** ---3.109 -5.222*** -4.337** ---7.734*** ---5.695** -- mtfpi -2.071 -6.519*** -4.196** ---6.494*** ---4.322*** ---3.790** real gdp per capita -3.102 -3.686** -5.667*** ---6.809*** ---1.912 -5.208*** -2.823 -3.648** consumer’s demand 1.547 -5.516*** -1.557 -4.900*** -3.612** ---2.402 -5.142*** -2.946 -4.381*** real exchange rate -2.563 -4.289** -2.162 -3.833** -1.725 -5.292*** -1.997 -4.424*** -3.716** -- cost to export -3.898** ---4.524** ---2.722 -4.438*** -5.135*** ---1.219 -5.447*** world income -3.296* ---3.296* -3.296* -3.296* ---3.296* -- journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 75 4. results and discussion the ardl estimated results are compiled in this section. this section has two subsections. in subsection 4.1 the ardl long run results are discussed. while second subsection 4.2 discusses the ardl short run results 4.1.ardl long run results for the nexus between productivity and export performance of emerging asian economies the pedroni’s cointegration test is employed to check the existence of long run relationship between productivity and export performance within the panel ardl framework. the results of ardl pedroni’s test of cointegration and bound test of cointegration are shown in table 3. the results of pedroni’s test of cointegration reveal that out of seven test statistics. table 3: pedroni’s co integration test note: authors’ own calculations the calculated values of four test statistics are more than two in absolute terms, so this indicates the presence of long run relationship between productivity, export performance and other variables of the study. while the bound test or joint f test is used to check the existence of long run relationship between productivity and export performance of the individual country within the ardl framework. the results of ardl bound test of cointegration are shown in table 4. similarly, for each country the calculated values of fstatistics are compared with the pesaran statistical table. as these f-statistics values are greater than the upper bound of the pesaran table, so this also indicates the presence of long run relationship between productivity, export performance and other variables of the study. test statistics agriculture sector export performance and tfp manufacturing sector export performance and tfp (i) (ii) (i) (ii) panel v -1.303 -1.286 -1.318 -1.103 panel rho -0.694 -0.742 -0.855 -0.441 panel t -6.859 -6.851 -6.592 -5.096 panel adf -5.291 -5.366 -6.065 -4.214 group rho -0.126 -0.075 -0.059 -0.469 group t -7.474 -7.249 -7.409 -5.179 group adf -4.024 -5.700 -6.161 -3.623 asma saeed and mehrin zaid ullah 76 table 4: ardl bound test of co integration between productivity and export performance of emerging asian economies note: authors’ own calculations the presence of long run relationship can also be checked via the negative and significant value of the error correction term (ect) which is the sign of long run relationship between the variables of the study. the cointegration equations for both agriculture and manufacturing sectors shows that the value of ect is negative and statistically significant which is the evidence of long run association between productivity and export performance of emerging asian economies. the value of ect is the speed of adjustment of short run disequilibrium towards the long run equilibrium. to inspect the impact of productivity on export performance of the emerging asian economies, pooled mean group (pmg) method is used after conducting the hausman specification test. the estimated results of long run coefficient for agriculture and manufacturing sectors are presented in table 5. the impact of productivity on export performance is estimated for full sample of five countries and then separately for china, india, indonesia, south korea, and japan. the impact of productivity on export performance is measured using two different specifications. in the first specification tfp is calculated using growth accounting technique and in the second specification tfp is estimated via the malmquist total factor productivity index (mtfpi) using data envelopment analysis (dea). the results show that both the agricultural and manufacturing sectors tfp have a positive and statistically significant impact on agricultural and manufacturing sector’s export performance of sample countries in all specifications which is consistent to the theory of productivity. it could be due to the reason that productivity improves the efficiency of factors of production and reduces the cost of production of domestic products, improving their competitive advantage in the worlds market (morley & morgan, 2008). countries agriculture sector manufacturing sector result f-bound test f-bound test (i) (ii) (i) (ii) china 12.635*** 7.736*** 23.808*** 28.739*** co integration india 10.757*** 8.249*** 23.478*** 15.708*** co integration indonesia 7.579*** 9.450*** 6.749*** 5.486*** co integration south korea 3.943** 3.679** 3.828** 16.517*** co integration japan 5.856*** 6.431*** 9.464*** 23.599*** co integration journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 77 table 5: ardl long run results for the nexus between productivity and export performance of emerging asian economies note: * p<0.10, **p<0.05, ***<0.01 (standard errors are in parentheses) variables agriculture sector manufacturing sector pmg 𝑴𝑮 pmg 𝑴𝑮 (i) (ii) (i) (ii) (i) (ii) (i) (ii) tfp 0.373*** (0.113) -0.893** (0.417) -0.717*** (0.093) -0.396* (0.229) - mtfpi -0.178*** (0.056) - 0.438** (0.198) -0.065** (0.033) -0.084** (0.038) real gdp per capita 1.497*** (0.433) 1.761*** (0.591) 5.212*** (1.616) 5.207*** (1.624) 2.110*** (0.531) 0.405** (0.203) 2.681** (1.364) 2.606** (1.139) consumer’s demand -0.080** (0.035) -0.029 (0.051) -6.209*** (2.327) -5.367*** (1.516) -0.136*** (0.049) -0.174*** (0.050) -0.805 (0.602) -0.269 (0.721) real exchange rate -0.027 (0.035) -0.089** (0.041) -0.002 (0.062) -0.028 (0.095) -0.054*** (0.021) -0.058* (0.031) -0.181** (0.072) -0.016 (0.086) cost to export -0.003 (0.049) -0.023 (0.044) -0.272** (0.129) -0.045 (0.232) -0.021 (0.051) 0.036 (0.039) -0.111 (0.077) 0.011 (0.187) world income 0.667*** (0.127) 0.433** (0.168) 0.373 (0.237) 0.560*** (0.209) 0.586*** (0.160) 0.453** (0.168) 1.043** (0.459) 0.561*** (0.197) constant -4.100*** (0.647) -3.498*** (0.249) -3.194 (2.253) -5.942*** (2.152) -2.367*** (0.163) -0.064 (0.061) -11.175 (13.368) -2.977*** (0.952) hausman test 0.827 0.554 ----0.463 0.856 ---- asma saeed and mehrin zaid ullah 78 with respect to other factors, the gdp per capita has a positive and statistically significant relationship with the export performance implying that economic growth is beneficial for the export performance of agricultural and manufacturing sectors in the sample countries. the positive association between export performance and gdp per capita results in surplus output and this surplus output can then be exported in international markets. various empirical studies also provide evidence that gdp increases the export performance (epaphra, 2016; nadeem et al., 2012; potelwa et al., 2016). consumer’s demand in the exporting country has a negative and statistically significant impact on export performance, implying that consumer’s demand has a towing effect on the export performance of sample countries through decreasing the exportable surplus. these findings are in line with studies that have find a negative relationship between consumer’s demand and export performance (boansi, odilonkounagbélokonon, & appah, 2014; rahmaddi & ichihashi, 2012). exchange rate has a negative and statistically significant impact on the export performance, implying that appreciation of exchange rate is not beneficial for the export performance in the sample countries. this confirms the traditional view that countries with a high exchange rate are not better in terms of their export performance. because high exchange rate makes domestic products expensive in the international markets via increasing the price of that products. these results are in accordance to previous studies that have concluded a negative association between exchange rate and export performance (kohler & ferjani, 2018; saqib & sana, 2012). cost to export has a negative and statistically significant impact on the export performance. while world income has a positive and statistically significant impact on the export performance, implying that an increase in income of trading partner economies increases the demand for domestic products in the sample countries. so, to improve its export performance policy makers of the sample countries should observe cyclical booms in economies of its trade partners. these results are line with study conducted by (nadeem et al., 2012). 4.2.ardl short run results for the nexus between productivity and export performance of emerging asian economies the ardl short run results are shown in table 6. like long run results, short run results indicate that agricultural and manufacturing tfp have a positive and significant association with the agricultural and manufacturing sectors export of emerging asian journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 79 economies. the coefficient of ect shows the speed of adjustment. the coefficient of ect is negative and highly significant which is also evidence of the existence of cointegration. the value of the ect shows that if there is any disequilibrium in the agricultural and manufacturing sectors export it will converge to the long run equilibrium at the speed of 82%, 93%, 91% and 86% respectively. the results of the individual country-wise for both long run and short run are presented in table 7, table 8, table 9 and table 10. the results of the individual country-wise also shows that improving productivity through increasing accesses to advance production technology is positively and significantly associated to the export performance of china, india, indonesia, south korea and japan. asma saeed and mehrin zaid ullah 80 table 6: ardl short run results for the nexus between productivity and export performance of emerging asian economies note: * p<0.10, **p<0.05, ***<0.01 (standard errors are in parentheses) variables agriculture sector manufacturing sector pmg 𝑴𝑮 pmg 𝑴𝑮 (i) (ii) (i) (ii) (i) (ii) (i) (ii) tfp 0.299** (0.158) -0.588* (0.326) -0.411*** (0.057) -0.235* (0.136) - mtfpi -0.137*** (0.047) -0.234** (0.092) -0.087* (0.052) -0.105** (0.051) real gdp per capita 1.329* (0.825) 0.362 (0.489) 2.859*** (1.083) 2.515** (1.267) 1.932*** (0.396) 0.932* (0.563) 2.442* (1.356) 3.279** (1.340) consumer’s demand -0.326 (451) -0.808 (1.294) -2.666*** (0.929) -2.704*** (0.844) -0.573 (0.628) -0.475 (0.386) -2.841** (1.243) -1.389 (0.957) real exchange rate -0.133* (0.074) -0.379*** (0.141) -0.039 (0.086) -0.064 (0.113) -0.392*** (0.106) -0.477*** (0.103) -0.504*** (0.133) -0.281 (0.189) cost to export -0.029 (0.055) -0.112* (0.061) -0.004 (0.039) -0.097** (0.051) -0.009 (0.047) -0.023 (0.029) -0.003 (0.029) 0.033 (0.029) world income 2.776*** (0.897) 2.442*** (0.542) 3.673* (2.168) 1.787 (1.447) 2.962*** (0.590) 2.268*** (0.557) 3.251*** (0.684) 1.682* (1.017) ecm -0.819*** (0.126) -0.928*** (0.044) -0.979*** (0.078) -0.971*** (0.022) -0.915*** (0.056) -0.858*** (0.042) -0.971*** (0.076) -0.739*** (0.150) journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 81 table 7: country-wise ardl long run results for the effect of productivity on agricultural exports performance note: * p<0.10, **p<0.05, ***<0.01 (standard errors are in parentheses) variables china india indonesia south korea japan (i) (ii) (i) (ii) (i) (ii) (i) (ii) (i) (ii) tfp 0.451*** (0.101) -1.911*** (0.583) -0.895** (0.390) -0.944*** (0.207) -0.168** (0.067) - mtfpi -0.325** (0.127) -1.059** (0.482) -0.297* (0.166) -0.407*** (0.068) -0.334** (0.137) real gdp per capita -0.347 (-0.637) -1.135 (1.098) 0.403 (0.375) 0.674** (0.303) 2.009* (1.072) 3.352*** (1.008) 7.460* (3.739) 10.758** (3.506) 1.321** (0.535) 3.669*** (1.074) consumer’s demand -0.109** (0.048) -0.948** (0.332) 11.953*** (2.595) 0.212 (2.111) -1.944* (1.037) -3.281*** (0.950) -2.857* (1.549) -6.747** (2.384) -0.398 (2.234) -6.265*** (1.764) real exchange rate -0.109* (0.059) -0.514** (0.232) -1.916*** (0.582) -0.323 (0.817) -1.029** (0.359) -1.370*** (0.408) -0.000 (0.286) 0.609 (0.306) -0.858*** (0.153) -0.774*** (0.189) cost to export -0.074*** (0.016) -0.001 (0.032) -0.078 (0.128) 0.402** (0.139) 0.116 (0.311) 0.605 (0.466) -0.639* (0.350) -0.041 (0.186) -0.018 (0.063) 0.037 (0.158) world income 0.621*** (0.076) 0.795*** (0.168) -0.325 (0.876) 0.305 (0.911) 2.074* (1.103) 3. 267*** (1.053) 0.093 (0.274) 0.413 (0.269) 0.026* (0.141) 0.236 (0.852) constant -0.617*** (0.726) -8.633*** (1.729) 1.672** (7.615) 0.068 (8.496) 13.867 (8.684) 26.095*** (7.819) 3.019 (3.497) -4.669 (3.297) -2.672 (1.594) -0.507 (1.100) serial correlation 1.202 (0.349) 1.077 (0.398) 0.554 (0.591) 0.265 (0.774) 1.564 (0.249) 1.361 (0.291) 1.250 (0.337) 1.865 (0.265) 0.496 (0.624) 0.251 (0.784) heteroskedasticity 1.296 (0.345) 0.779 (0.681) 1.836 (0.179) 0.686 (0.751) 1.351 (0.294) 1.391 (0.273) 1.425 (0.290) 0.188 (0.995) 1.841 (0.158) 0.610 (0.808) jerque bera normality test 1.637 (0.441) 2.190 (0.334) 0.025 (0.987) 4.468 (0.110) 0.185 (0.912 0.066 (0.967) 1.020 (0.600) 0.524 (0.769) 0.181 (0.913) 0.052 (0.974) functional form ramsey reset test 0.987 (0.346) 0.277 (0.614) 1.972 (0.184) 0.876 (0.376) 0.133 (0.721) 1.126 (0.306) 2.847 (0.145) 1.156 (0.342) 0.096 (0.763) 0.316 (0.260 asma saeed and mehrin zaid ullah 82 table 8: country-wise ardl short run results for the effect of productivity on agricultural exports performance note: * p<0.10, **p<0.05, ***<0.01 (standard errors are in parentheses) variables china india indonesia south korea japan (i) (ii) (i) (ii) (i) (ii) (i) (ii) (i) (ii) tfp 0.361** (0.123) -0.051* (0.027) -1.418*** (0.339) -0.525*** (0.114) -0.205** (0.087) - mtfpi -0.144*** (0.034) -0.340*** (0.100) -0.531*** (0.105) -0.212** (0.071) -0.130** (0.059) real gdp per capita 2.828*** (0.624) 1.308* (0.712) 0.976* (0.497) 2.100*** (0.499) -2.613 (1.583) 4.151*** (1.303) 6.107*** (1.273) 4.903*** (1.295) 2.893*** (0.636) 3.596*** (0.704) consumer’s demand -0.119 (0.073) -0.422*** (0.081) -3.321*** (0.997) -2.601** (0.904) -4.060*** (1.121) -2.451** (0.958) -4.393*** (0.864) -3.445*** (0.914) -0.434 (1.573) -4.592*** (1.273) real exchange rate -0.057 (0.106) -0.622** (0.108) -0.847*** (0.234) -0.904** (0.210) -1.145*** (0.197) -1.128*** (0.134) -0.263 (0.191) -0.102 (0.209) -0.323** (0.129) -0.148 (0.138) cost to export -0.067 (0.055) -0.063 (0.051) -0.059 (0.089) -0.047 (0.086) 0.249 (0.286) 0.302 (0.248) -0.035 (0.440) -0.147 (0.482) -0.145 (0.195) -0.015 (0.175) world income 1.490** (0.507) 3.169*** (0.515) 4.186*** (1.169) 3.581*** (1.124) 8.653*** (1.467) 6.497*** (1.161) 1.441* (0.761) -0.735 (0.758) 1.048 (0.631) 1.055 (1.553) ecm (-1) -0.896*** (0.179) -0.863*** (0.155) -0.824*** (0.117) -0.988*** (0.122) -0.745*** (0.136) -0.639*** (0.099) -0.963*** (0.191) -0.913*** (0.195) -0.701*** (0.073) -0.903*** (0.105) r-square 0.74 0.85 0.83 0.84 0.85 0.86 0.76 0.71 0.77 0.79 f-statistics (p-value) 7.183 (0.002) 14.962 (0.023) 5.791 (0.006) 5.678 (0.006) 2.583 (0.048) 3.137 (0.023) 2.505 (0.074) 2.313 (0.069) 2.865 (0.044) 2.438 (0.089) durbin. watson stat. 2.116 2.022 2.204 2.294 2.099 2.101 2.349 2.055 2.47 2.019 journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 83 table 9: country-wise ardl long run results for the effect of productivity on manufacturing exports performance note: * p<0.10, **p<0.05, ***<0.01 (standard errors are in parentheses) variables china india indonesia south korea japan (i) (ii) (i) (ii) (i) (ii) (i) (ii) (i) (ii) tfp 0.426*** (0.127) -0.657** (0.239) -0.516*** (0.139) -1.509*** (0.273) -0.790** (0.220) - mtfpi -0.423** (0.161) -0.074** (0.033) -0.141* (0.077) -0.909*** (0.231) -0.302* (0.167) real gdp per capita -1.621 (0.902) -3.669 (2.197) 0.152** (0.058) 0.151 (0.086) 1.211* (0.676) 2.555** (1.005) 4.061 (2.139) 3.416** (1.579) 3.026*** (0.542) 3.538** (1.331) consumer’s demand -0.711** (0.275) -0.883** (0.391) -0.419 (1.153) -4.158*** (0.878) -1.205* (0.658) -2.402** (0.922) -3.497** (1.492) -3.545*** (1.136) -1.455*** (0.270) -0.059 (0.247) real exchange rate -0.518** (0.235) 0.474 (0.288) -0.347* (0.171) -0.608*** (0.182) -0.286** (0.108) -0.591 (0.426) -0.450*** (0.085) -0.406** (0.194) -1.013*** (0.084) -1.021*** (0.225) cost to export -0.126*** (0.028) -0.315** (0.098) -0.099*** (0.022) -0.065** (0.028) -0.285 (0.276) -0.140 (0.292) -0.649** (0.182) -0.391** (0.176) 0.042 (0.149) 0.055 (0.175) world income 0.524*** (0.129) 1.368** (0.460) -0.151 (0.161) -0.052 (0.251) 1.014 (0.779) 2.022* (1.024) -0.109 (0.087) 0.304* (0.164) 1.521*** (0.236) 2.293 (1.374) constant -4.659*** (1.222) -12.429** (4.231) 1.345 (1.423) 0.206 (2.225) 11.819** (5.698) 22.333** (7.795) 5.437*** (1.144) -6.729** (2.559) 25.848*** (5.862) 1.399 (7.298) serial correlation 2.794 (0.128) 0.483 (0.258) 2.251 (0.113) 1.153 (0.362) 0.461 (0.639) 0.157 (0.856) 3.683 (0.156) 1.188 (0.338) 0.485 (0.647) 0.793 (0.474) heteroskedasticity 0.677 (0.757) 0.707 (0.737) 1.027 (0.502) 0.577 (0.831) 0.899 (0.546) 0.474 (0.897) 1.035 (0.524) 0.471 (0.883) 0.484 (0.891) 0.370 (0.940) jerque bera normality test 0.999 (0.606) 1.760 (0.414) 0.437 (0.803) 1.196 (0.571) 0.355 (0.837) 2.485 (0.288) 0.311 (0.855) 0.209 (0.901 0.631 (0.729) 0.384 (0.639) functional form ramsey reset test 2.028 (0.901) 0.110 (0.751) 0.487 (0.504) 0.110 (0.751) 0.487 (0.504) 0.133 (0.721) 0.271 (0.625) 1.972 (0.184) 3.064 (0.156) 0.072 (0.792) asma saeed and mehrin zaid ullah 84 table 10: country-wise ardl short run results for the effect of productivity on manufacturing exports performance note: * p<0.10, **p<0.05, ***<0.01 (standard errors are in parentheses) variables china india indonesia south korea japan (i) (ii) (i) (ii) (i) (ii) (i) (ii) (i) (ii) tfp 0.645*** (0.124) -0.615** (0.207) -0.514*** (0.085) -0.226** (0.097) -0.741*** (0.223) - mtfpi -0.363*** (0.111) -0.112* (0.059) -0.042** (0.019) -0.431*** (0.109) -0.308** (0.125) real gdp per capita 2.234*** (0.633) 3.020*** (0.821) 3.351*** (0.572) 0.786** (0.375) 0.425 (0.575) 3.416*** (0.923) 4.376*** (0.886) 4.112*** (0.732) 6.687*** (0.564) 5.726*** (0.667) consumer’s demand 0.121 (0.074) -0.045 (0.085) 0.322 (0.699) -0.680 (0.732) -2.726*** (0.580) -0.792 (0.718) -3.921*** (0.631) -4.022*** (0.493) -6.337*** (1.306) -5.623*** (1.481) real exchange rate -0.472*** (0.124) -0.875*** (0.158) -0.819** (0.152) -1.157*** (0.176) -0.384*** (0.052) -0.715*** (0.094) -0.421** (0.151) -0.559*** (0.129) -0.661*** (0.105) -0.661*** (0.101 cost to export -0.029 (0.053) -0.013 (0.065) -0.006 (0.069) -0.028 (0.073) -0.114 (0.153) 0.291 (0.191) -0.486 (0.357) -0.179 (0.280) 0.193 (0.128) 0.139 (0.125) world income 4.264*** (0.519) 4.182*** (0.659) 0.333 (0.861) 0.710 (0.868) 1.825** (0.669) 5.144*** (0.952) -0.854 (0.626) 1.004* (0.474) 0.478 (0.799) 1.643* (0.867) ecm (-1) -0.951*** (0.079) -0.931*** (0.229) -0.556*** (0.069) -0.502*** (0.071) -0.779*** (0.140) -0.533*** (0.082) -0.729*** (0.105) -0.885*** (0.084) -0.954*** (0.161) -0.713*** (0.173) r-square 0.90 0.86 0.81 0.81 0.86 0.82 0.87 0.92 0.98 0.86 f-statistics (p-value) 28.210 (0.000) 19.744 (0.000) 9.018 (0.001) 5.589 (0.005) 6.111 (0.000) 3.642 (0.014) 7.598 (0.009) 9.053 (0.000) 17.339 (0.000) 9.069 (0.000) durbin. watson stat. 2.075 2.294 2.11 2.308 2.255 2.239 2.014 2.061 2.401 2.187 journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 85 gdp per capita positively affect the export performance of all the six countries. the long run coefficient of gdp has not the expected positive sign for both agriculture and manufacturing sector export in china, however, this result is statistically insignificant. consumers’ demand has a negative impact on the export performance of both agriculture and manufacturing sectors export performance in all the countries. similarly, the coefficient of real exchange rate is negatively and statistically significant associated to the export performance for all the countries in both agriculture and manufacturing sectors. while cost to export has a negative impact and world income has a positive impact on the export of both sectors in the sample countries. the lower part of table 7 and table 9 presents results of diagnostic tests. the results of these tests reveal that the model has the desirable econometric properties and the model’s residuals are normally distributed and free from the problem of serial correlation and heteroskedasticity. it is also confirmed that the model has correct functional form. the model’s stability for agricultural and manufacturing sectors is tested through the ramsey’s reset stability test. the results of the test show that the functional form of the model is correct and overall, the model is stable as the value of fstatistics and their respective probabilities for the whole sample are greater than 5%. the cusum and cusum square test results of agriculture and manufacturing sectors for all specifications are given in the appendix-b. the results revealed that parameters for the two sectors are stable over the period of the study. 5. conclusions in this study, we estimate the relationship between productivity and export performance of china, india, indonesia, south korea, and japan. the empirical findings revealed that both agricultural and manufacturing productivity have a positive and statistically significant impact on the export performance of china, india, indonesia, south korea, and japan. similarly, world income is positively associated to the export performance while consumer’s demand, real exchange rate and cost to export have a negative impact on the export performance of emerging asian economies. these findings offer some policy implications to improve further export performance. policies and economic strategies should aim at investing in and adoption of advance production technology at firm level in the emerging asian economies. gdp growth is directly related to the export performance as it expands the production capacity and hence exportable surplus. therefore, keeping in mind this result the present study offer important policy implication that the objectives of government policies must be to ensure economic growth. asma saeed and mehrin zaid ullah 86 references ahmad, h. k., ilyas, m., mahmood, t., & afzal, m. (2010). exploring the effect of total factor productivity growth on future output growth: evidence from a panel of east asian countries. pakistan economic and social review, 105-122. amornkitvikai, y., harvie, c., & charoenrat, t. (2012). factors affecting the export participation and performance of thai manufacturing small and medium sized enterprises (smes). balchin, l. (2008). trade-related business climate and manufacturing export performance in africa: a firm-level analysis. journal of development perspectives, 4(1), 32-66. baldwin, j. r., & gu, w. (2003). export‐market participation and productivity performance in canadian manufacturing. canadian journal of economics/revue canadienne d'économique, 36(3), 634-657. bernard, a. b., & jensen, j. b. (1999). exceptional exporter performance: cause, effect, or both? journal of international economics, 47(1), 1-25. bhatt, p. (2013). china’s exports and foreign direct investment. applied econometrics and international development, 13(2), 183-196. boansi, d., odilonkounagbélokonon, b., & appah, j. (2014). determinants of agricultural export trade: case of fresh pineapple exports from ghana. journal of economics, management and trade, 1736-1754. chan, e. m., & au, k. f. (2007). determinants of china's textile exports: an analysis by gravity model. journal of the textile institute, 98(5), 463-469. chit, m. m. (2008). exchange rate volatility and exports: evidence from the asean-china free trade area. journal of chinese economic and business studies, 6(3), 261-277. clerides, s. k., lach, s., & tybout, j. r. (1998). is learning by exporting important? microdynamic evidence from colombia, mexico, and morocco. the quarterly journal of economics, 113(3), 903-947. coelli, t. j. (1996). measurement of total factor productivity growth and biases in technological change in western australian agriculture. journal of applied econometrics, 11(1), 7791. comin, d., & gertler, m. (2006). medium-term business cycles. american economic review, 96(3), 523-551. das, b. b. (2007). impact of foreign direct investment on the export performance of india, china and malaysia. asia pacific business review, 3(1), 107-114. david fadiran, & akanbi, o. a. (2017). institutions and other determinants of total factor productivity in sub-saharan africa. university of south africa,hazelwood, pretoria, south africa. douglas, c., laurits, c., & erwin, d. w. (1982). the economic theory of index numbers and the measurement of input, output, and productivity. econometrica: journal of the econometric society, 1393-1414. epaphra, m. (2016). determinants of export performance in tanzania. journal of economics library, 3(3), 470. falianty, t. (2015). ’exchange rate effect on indonesian export: comparisons of two crises episodes’. international journal of economic research, 6(6), 60-75. fare, r., grosskopf, s., lindgren, b., & roos, p. (1992). productivity changes in swedish journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 87 pharamacies 1980–1989: a non-parametric malmquist approach. journal of productivity analysis, 3(1-2), 85-101. faustino, h. c., lima, j. c., & matos, p. v. (2012). exports, productivity and innovation: evidence from portugal using microdata. genc, e. g., & artar, o. k. (2014). the effect of exchange rates on exports and imports of emerging countries. european scientific journal, esj, 10(13). greenaway, d., & kneller, r. (2004). exporting and productivity in the united kingdom. oxford review of economic policy, 20(3), 358-371. grossman, g., & helpman, e. (1991). innovation and growth in the global economy. cambridge: the mit press. hansen. gupta, s. d., raychaudhuri, a., & haldar, s. k. (2015). determinants of exports of information technology in india: an empirical analysis. south asia economic journal, 16(1), 64-81. gururaj, b., satishkumar, m., & kumar, m. a. (2016). analysis of factors affecting the performance of exports in india. international journal of agriculture, environment and biotechnology, 9(4), 613. hassan, m. u., hassan, m. s., & mahmood, h. (2013). an empirical inquisition of the impact of exchange rate and economic growth on export performance of pakistan. middle-east journal of scientific research, 14(2), 288-299. head, k., & ries, j. (2003). heterogeneity and the fdi versus export decision of japanese manufacturers. journal of the japanese and international economies, 17(4), 448-467. helpman, e., melitz, m. j., & yeaple, s. r. (2004). export versus fdi with heterogeneous firms. american economic review, 94(1), 300-316. jawaid, s. t., raza, s. a., mustafa, k., & karim, m. z. a. (2016). does inward foreign direct investment lead export performance in pakistan? global business review, 17(6), 12961313. jienwatcharamongkhol, v. (2013). effects of productivity and import on firm-level export. ratio working papers, paper(225). kafayat, a. (2013). effects of exchange rate instability on imports and exports of pakistan. acta universitatis danubius. œconomica, 10(2). khan, s. u. (2006). macroeconomic determinants of total factor productivity in pakistan. working paper no. 10. state bank of pakistan. kohler, a., & ferjani, a. (2018). exchange rate effects: a case study of the export performance of the swiss agriculture and food sector. the world economy, 41(2), 494-518. liao, h., & liu, x. (2009). export-total factor productivity growth nexus in east asian economies. applied economics, 41(13), 1663-1675. melitz, m. j. (2003). the impact of trade on intra‐industry reallocations and aggregate industry productivity. econometrica, 71(6), 1695-1725. morley, b., & morgan, w. (2008). causality between exports, productivity and financial support in european union agriculture. regional studies, 42(2), 189-198. nadeem, m. m., azam, m., & islam, r. (2012). an investigation of the various factors influence on exports. global journal of management and business research, 12(19). potelwa, x. y., lubinga, m. h., & ntshangase, t. (2016). factors influencing the growth of south africa’s agricultural exports toworld markets. european scientific journal, esj, 12(34). rahmaddi, r., & ichihashi, m. (2012). how do foreign and domestic demand affect exports asma saeed and mehrin zaid ullah 88 performance? an econometric investigation of indonesia’s exports. reis, j., & forte, r. (2016). the impact of industry characteristics on firms’ export intensity. international area studies review, 19(3), 266-281. ricci, m. l. a., & trionfetti, m. f. (2011). evidenceon productivity, comparative advantage, and networks in the export performance of firms: international monetary fund. rolf, f. (1988). fundamentals of production theory: lecture notes in economics and mathmetical system, . heidelberg: springer. rolf, f., shawna, g., mary, n., & zhongyang, z. (1994). productivity growth, technical progress, and efficiency change in industrialized countries. the american economic review, 66-83. romer, p. (1993). idea gaps and object gaps in economic development. journal of monetary economics, 32(3), 543-573. santos-paulino, a. u. (2002). trade liberalisation and export performance in selected developing countries. journal of development studies, 39(1), 140-164. saqib, n., & sana, i. (2012). exchange rate volatility and its effect on pakistan's export volume. advances in management and applied economics, 2(4), 109. sharma, c., & mishra, r. k. (2009). does export and productivity growth linkage exist. evidence from the indian manufacturing industry. shephard, r. w. (1970). theory of cost and production functions princeton university press. princeton, new jersey. solow, r. m. (1956). a contribution to the theory of economic growth. the quarterly journal of economics, 70(1), 65-94. solow, r. m. (1957). technical change and the aggregate production function. the review of economics and statistics, 312-320. srinivasan, p. (2012). exchange rate volatility and export growth in india: an empirical investigation. mpra paper. veeramani, c. (2008). impact of exchange rate appreciation on india's exports. economic and political weekly, 10-14. vikram, n., & ashok, d. (1993). a new database on physical capital stock: sources, methodology and results. revista de análisis económico, 8(1), 37-59. journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 89 appendix -a measures of total factor productivity researchers and policy makers measured tfp performance via different techniques. the present study uses two methods (1) growth accounting framework and (2) data envelopment analysis (dea) to estimate productivity performance measures. growth accounting framework (gaf) the growth accounting method follows the seminal work of solow (1956). this method involves the production function which is based on the assumption of constant returns to scale and the perfect competition in factor markets. symbolically this function is specified as follows: 𝑌𝑎(𝑚)𝑖𝑡 = 𝐴𝑎(𝑚)𝑖𝑡 𝐹(𝐾𝑎(𝑚)𝑖𝑡 , 𝐿𝑎(𝑚)𝑖𝑡 ) (𝐴. 1) where 𝑌𝑎(𝑚)𝑖𝑡 is output, 𝐴𝑎(𝑚)𝑖𝑡 is total factor productivity, 𝐾𝑎(𝑚)𝑖𝑡 𝑎𝑛𝑑 𝐿𝑎(𝑚)𝑖𝑡 are units of capital and labor inputs respectively. when taking logarithms on both sides of equation (a.1) , it takes the following form: lnya(m)it = lnaa(m)it + lnf(ka(m)it, la(m)it) (a. 2) in the next step differentiating equation (a.2) with respect to time, equation (a.3) is obtained as: 𝑌𝑎(𝑚)𝑖𝑡̇ 𝑌𝑎(𝑚)𝑖𝑡 = 𝐴𝑎(𝑚)𝑖𝑡̇ 𝐴𝑎(𝑚)𝑖𝑡 + 𝐴𝑡 . 𝐹𝑘. 𝐾𝑎(𝑚)𝑖𝑡 𝑌𝑎(𝑚)𝑖𝑡 . �̇�𝑎(𝑚)𝑖𝑡 𝐾𝑎(𝑚)𝑖𝑡 + 𝐴𝑡 . 𝐹𝑙 𝐿𝑎(𝑚)𝑖𝑡 𝑌𝑎(𝑚)𝑖𝑡 . �̇�𝑎(𝑚)𝑖𝑡 𝐿𝑎(𝑚)𝑖𝑡 (𝐴. 3) assuming that capital and labor markets are competitive, so the share of the marginal product of the factors are equal to their prices. where ∆y ∆k = a ∂y ∂k = afk = r is the marginal product of capital and ∆y ∆l = a ∂y ∂l = afl = w is the marginal product of labor and r and w are the prices of capital and labor respectively. therefore, equation (a. 3) takes the following form: 𝑌𝑎(𝑚)𝑖𝑡̇ 𝑌𝑎(𝑚)𝑖𝑡 = 𝐴𝑎(𝑚)𝑖𝑡̇ 𝐴𝑎(𝑚)𝑖𝑡 + 𝑟 𝐾𝑎(𝑚)𝑖𝑡 𝑌𝑎(𝑚)𝑖𝑡 . �̇�𝑎(𝑚)𝑖𝑡 𝐾𝑎(𝑚)𝑖𝑡 + 𝑤 𝐿𝑎(𝑚)𝑖𝑡 𝑌𝑎(𝑚)𝑖𝑡 . �̇�𝑎(𝑚)𝑖𝑡 𝐿𝑎(𝑚)𝑖𝑡 (𝐴. 4) now rearranging equation (a.4) solow residual can be obtained as follows: asma saeed and mehrin zaid ullah 90 𝐴𝑎(𝑚)𝑖𝑡̇ 𝐴𝑎(𝑚)𝑖𝑡 = 𝑌𝑎(𝑚)𝑖𝑡̇ 𝑌𝑎(𝑚)𝑖𝑡 − 𝑟 𝐾𝑎(𝑚)𝑖𝑡 𝑌𝑎(𝑚)𝑖𝑡 . �̇�𝑎(𝑚)𝑖𝑡 𝐾𝑎(𝑚)𝑖𝑡 + 𝑤 𝐿𝑎(𝑚)𝑖𝑡 𝑌𝑎(𝑚)𝑖𝑡 . �̇�𝑎(𝑚)𝑖𝑡 𝐿𝑎(𝑚)𝑖𝑡 (𝐴. 5) where 𝐴𝑎(𝑚)𝑖𝑡̇ 𝐴𝑎(𝑚)𝑖𝑡 shows the growth rate of the total factor productivity, 𝑌𝑎(𝑚)𝑖𝑡̇ 𝑌𝑎(𝑚)𝑖𝑡 is the growth rate of output, 𝐾𝑎(𝑚)𝑖𝑡 𝑌𝑎(𝑚)𝑖𝑡 is the ratio of stock of capital to output, �̇�𝑎(𝑚)𝑖𝑡 𝐾𝑎(𝑚)𝑖𝑡 is the growth rate of the stock of capital, 𝐿𝑎(𝑚)𝑖𝑡 𝑌𝑎(𝑚)𝑖𝑡 is the ratio of labor to output, �̇�𝑎(𝑚)𝑖𝑡 𝐿𝑎(𝑚)𝑖𝑡 is the growth rate of labor while 𝑟 and 𝑤 are the prices of capital and labor. as the capital’s stock data is not available, so the present study uses the following perpetual inventory method to gauge it . 𝐾𝑎(𝑚)𝑖𝑡 = (1 − 𝜎)𝐾𝑎(𝑚)𝑖𝑡−1 + 𝐼𝑎(𝑚)𝑖𝑡 (𝐴. 6) the data on initial stock of capital is gauged based on the following formula: 𝐾𝑎(𝑚)𝑖𝑡−1 = 𝐼𝑎(𝑚)𝑖𝑡−1 𝜎 + 𝑔𝑎(𝑚) (𝐴. 7) where k_(a(m)t) shows the stock of capital in the current period, i_(a(m)t) is the level of investment in the current period, σ is the rate at which the stock of capital depreciates and g_(a(m)) represents the growth rate of output. the present study used 4 percent depreciation rate of the stock of capital. vikram and ashok (1993) and khan (2006) also used 4 percent depreciation rate for the stock of capital. malmquist total factor productivity index (mtfpi) malmquist productivity index was first introduced by douglas, laurits, and erwin (1982) and then further developed by rolf (1988) and rolf, shawna, mary, and zhongyang (1994). coelli (1996) defined malmquist productivity index as the ratio of output distances from the production frontier. the malmquist productivity indices have several desirable advantages. they can breakdown productivity growth into total factor productivity change, technical change and efficiency change. the decomposition of productivity change provide insight into the sources of productivity performance. the data envelopment analysis (dea) presented by coelli (1996) is employed in the present study to compute output-based malmquist productivity journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 91 indices. rolf et al. (1994) pointed out that the malmquist productivity indices are based on distance functions that reflects patterns of production technology and requires only input and output data. rolf et al. (1994) also argued that distance functions can be utilized to identify the sources of productivity growth and to find whether change in productivity growth is due to the efficiency change or whether it is attributed to the technological change. shephard (1970) and rolf (1988) defined distance function on production technology s_t such that input x_t can produce output y_t. the distance function can be expressed as follows; do t (yt, xt) = min{(θ: yt, xt/θ)ϵst} (a. 8) this function measures the maximum proportional change in output that can be attained from a certain combination of inputs. when distance function d_(o )^t (y_t,x_t ) is equal to 1, it shows that (y_t,x_t ) is on the boundary of the production frontier. this implies that output is technically efficient. however, if the distance function d_(o )^t (y_t,x_t ) is less than 1 this implies that there is technical inefficiency. to show trends in productivity growth, the malmquist productivity index uses distance function with respect to time that are defined as; do t (yt+1, xt+1) = min{(θ: yt+1, xt+1/θ)ϵst} (a. 9) do t (yt+1, xt+1) measure maximum proportional change in output to confirm whether (yt+1, xt+1) is attainable, or not attainable in time period t. the maximum proportional change in output entailed to ensure the attainability of (yt, xt) in time period t + 1 that is represented by do t+1(yt, xt). according to rolf et al. (1994) the malmquist productivity index mpi in time period t is defined as: mpit = do t (yt+1,xt+1) do t (yt,xt) (a. 10) likewise, the productivity index at time period t + 1 is defined as follows; mpit+1 = do t (yt+1,xt+1) do t (yt,xt) (a. 11) so, according to douglas et al. (1982) the malmquist total factor productivity index can be defined in terms of the geometric mean of two malmquist indexes as defined above. mpio(ys, xs, yt, xt) = [ do t (yt+1,xt+1) do t (yt,xt) x do t+1(yt+1,xt+1) do t+1(yt,xt) ] 1 2 (a. 12) this index is based on the ratios of output distance functions to show changes in productivity growth over the time. a value of mpio > 1 shows growth in the overall asma saeed and mehrin zaid ullah 92 total factor productivity performance from period t to period t + 1. and a value of mpio < 1 shows that the overall productivity performance is declining over the period. by rewriting equation (a.12), mpi can be decomposed into its two main sources; technical efficiency that represent a shift towards the production frontier and technological change that represent a shift of the production frontier (fare, grosskopf, lindgren, & roos, 1992); 𝑀𝑃𝐼𝑜 ( 𝑦𝑡+1, 𝑥𝑡+1, 𝑦𝑡 , 𝑥𝑡 ) = 𝑑𝑜 𝑡+1(𝑦𝑡+1,𝑥𝑡+1) 𝑑𝑜 𝑡 (𝑦𝑡,𝑥𝑡) [ 𝑑𝑜 𝑡 (𝑦𝑡+1,𝑥𝑡+1) 𝑑𝑜 𝑡+!(𝑦𝑡+1,𝑥𝑡+1) 𝑋 𝑑𝑜 𝑡 (𝑦𝑡,𝑥𝑡) 𝑑𝑜 𝑡+1(𝑦𝑡,𝑥𝑡) ] 1 2 (𝐴. 13) where 𝑑𝑜 𝑡+1(𝑦𝑡+1,𝑥𝑡+1) 𝑑𝑜 𝑡 (𝑦𝑡,𝑥𝑡) is the change in technical efficiency between the current period 𝑡 and next period 𝑡 + 1 and the remaining part of this equation, [ 𝑑𝑜 𝑡 (𝑦𝑡+1,𝑥𝑡+1) 𝑑𝑜 𝑡+!(𝑦𝑡+1,𝑥𝑡+1) 𝑋 𝑑𝑜 𝑡 (𝑦𝑡,𝑥𝑡) 𝑑𝑜 𝑡+1(𝑦𝑡,𝑥𝑡) ] 1 2 , is the technical change between the current period 𝑡 and next period 𝑡 + 1. journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 93 appendix-b stability test results for first specification of the nexus between productivity and export performance of agriculture sector of emerging asian economies india -10.0 -7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 09 10 11 12 13 14 15 16 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 09 10 11 12 13 14 15 16 cusum of squares 5% significance south korea -10.0 -7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 09 10 11 12 13 14 15 16 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 09 10 11 12 13 14 15 16 cusum of squares 5% significance indonesia -10.0 -7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 08 09 10 11 12 13 14 15 16 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 08 09 10 11 12 13 14 15 16 cusum of squares 5% significance japan -8 -6 -4 -2 0 2 4 6 8 2011 2012 2013 2014 2015 2016 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 2011 2012 2013 2014 2015 2016 cusum of squares 5% significance china -8 -6 -4 -2 0 2 4 6 8 2010 2011 2012 2013 2014 2015 2016 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 09 10 11 12 13 14 15 16 cusum of squares 5% significance asma saeed and mehrin zaid ullah 94 stability test results for first specification of the nexus between productivity and export performance of manufacturing sector of emerging asian economies china -10.0 -7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 09 10 11 12 13 14 15 16 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 2010 2011 2012 2013 2014 2015 2016 cusum of squares 5% significance india -10 -8 -6 -4 -2 0 2 4 6 8 10 09 10 11 12 13 14 15 16 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 2011 2012 2013 2014 2015 2016 cusum of squares 5% significance indonesia -10.0 -7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 08 09 10 11 12 13 14 15 16 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 10 11 12 13 14 15 16 cusum of squares 5% significance south korea -8 -6 -4 -2 0 2 4 6 8 2011 2012 2013 2014 2015 2016 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 2011 2012 2013 2014 2015 2016 cusum of squares 5% significance japan -6 -4 -2 0 2 4 6 2013 2014 2015 2016 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 2013 2014 2015 2016 cusum of squares 5% significance journal of applied economics and business studies, volume. 5, issue 4 (2021) 59-96 https://doi.org/10.34260/jaebs.544 95 stability test results for second specification of the nexus between productivity and export performance of agriculture sector of emerging asian economies india -10.0 -7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 09 10 11 12 13 14 15 16 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 09 10 11 12 13 14 15 16 cusum of squares 5% significance south korea -8 -6 -4 -2 0 2 4 6 8 2012 2013 2014 2015 2016 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 2013 2014 2015 2016 cusum of squares 5% significance indonesia -10.0 -7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 09 10 11 12 13 14 15 16 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 09 10 11 12 13 14 15 16 cusum of squares 5% significance japan -6 -4 -2 0 2 4 6 2013 2014 2015 2016 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 2013 2014 2015 2016 cusum of squares 5% significance china -10.0 -7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 09 10 11 12 13 14 15 16 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 09 10 11 12 13 14 15 16 cusum of squares 5% significance asma saeed and mehrin zaid ullah 96 stability test results for second specification of the nexus between productivity and export performance of manufacturing sector of emerging asian economies china -8 -6 -4 -2 0 2 4 6 8 2011 2012 2013 2014 2015 2016 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 2011 2012 2013 2014 2015 2016 cusum of squares 5% significance india -8 -6 -4 -2 0 2 4 6 8 10 11 12 13 14 15 16 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 10 11 12 13 14 15 16 cusum of squares 5% significance indonesia -10.0 -7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 09 10 11 12 13 14 15 16 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 09 10 11 12 13 14 15 16 cusum of squares 5% significance south korea -10.0 -7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 08 09 10 11 12 13 14 15 16 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 08 09 10 11 12 13 14 15 16 cusum of squares 5% significance japan -8 -6 -4 -2 0 2 4 6 8 2011 2012 2013 2014 2015 2016 cusum 5% significance -0.4 0.0 0.4 0.8 1.2 1.6 2013 2014 2015 2016 cusum of squares 5% significance journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 113 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x a study of socioeconomic determinants of anemia among expecting mothers in khyber pakhtunkhwa pakistan faiza nawaz1, muska mukhtar2 & atta ur rahman*3 1research scholar at institute of management sciences peshawar, pakistan 2lecturer, institute of management sciences peshawar, pakistan 3associate professor, institute of management sciences peshawar, pakistan abstract one of the instrumental factors contributing towards the gdp of any country is human capital and for a progressive human capital, the main resource is physical and mental health. females constitute approximately 48% of population of pakistan, out of which a great number suffer from nutritional deficiency anemia at their reproductive age, which results in an unhealthy child. this study is designed to find the socioeconomic factors that lead to anemia among expecting women. the research design for this study is case control where data was collected from two groups; control group (nonanemic pregnant women) and case group (anemic pregnant women) registered in the basic health unit (bhu) from different areas of peshawar. collected data was analyzed using binary logistic regression. findings revealed that monthly income of household, independent members or number of employed members of the family, gap between children, non-staple food consumption and supplements continuity have significant inverse relationship with anemia, indicating that increase in mentioned variables will decrease the prevalence of anemia in pregnant women of rural areas. the study concluded that anemia being a common grievance amongst pregnant women is mostly caused by low level of monthly income, family size with lower number of employed members, low frequency of non-staple food consumption per week, lack of continuity of iron supplements and poor health seeking behavior. awareness regarding attainment keywords socioeconomic, anemia, logistic regression * attaurrahman@imsciences.edu.pk journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 114 of education irrespective of gender, government focus on health promotion interventions and more job opportunities can improve health conditions and socioeconomic status of women in khyber pakhtunkhwa. 1. introduction the word ‘anemia’ is derived from two greek words ‘an’ means ‘without’ and ‘haima’ means ‘blood’. it was called ‘anaimia’ meaning without blood till late 18th century. however, in the early 19th century via modern latin it came to be called as ‘anemia.’ red blood cells (also called erythrocytes or rbc’s) contains a special protein called ‘hemoglobin’ which helps to carry oxygen from the lungs to the rest of the body and then returns carbon dioxide from the body to the lungs so that it can be exhaled. when the level of hemoglobin decreases, it results in anemia. in other words, anemia diminishes the capacity of the blood to carry oxygen to different organs of human body due to which an individual faces pallor and weariness, (american society of hematology, 2017; world health organization, 1992). there are more than 400 types of anemia, which are divided into three groups. the first group of anemia indicates symptoms of blood loss. chronic blood loss anemia is most often the result of chronic gastrointestinal bleeding. it is due to conditions such as ulcers, hemorrhoids, gastritis (inflammation of the stomach) and cancer. similarly, another cause of blood loss is use of non-steroidal anti-inflammatory drugs (nsaids) such as aspirin or ibuprofen, which can cause ulcers and gastritis. menstruation is another potential cause of this issue, especially if menstrual bleeding is excessive. next group is of patients whose red blood cells may be faulty or decreased due to abnormal red blood cells or a lack of minerals and vitamins needed for red blood cells to work properly. conditions associated with these causes of anemia include the sickle cell anemia, iron-deficiency anemia, vitamin deficiency, bone marrow and stem cell problems, and other health conditions. the third group comprises of people suffering from hemolytic anemia. these inherited conditions include sickle cell anemia and thalassemia. it may also result from stressors such as infections, drugs, snake or spider venom, or certain foods. moreover, the toxins from advanced liver or kidney disease and inappropriate attack by the immune system (called hemolytic disease of the newborn when it occurs in the fetus of a pregnant woman) are also some of the prominent characteristics identified in the patients of this group. other symptoms may include vascular grafts, prosthetic journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 115 heart values, tumors, severe burns, exposure to certain chemicals, severe hypertension and clotting disorders. amongst the aforementioned three groups of anemia, iron deficiency anemia (ida) from the second group is extremely common in individuals and is the most prevalent cause of anemia worldwide. there are various causes of ida, but it occurs mostly due to lack of the mineral iron in the body due to insufficient nutrition intake (nhanes, 2012). it is affecting 50% of females at reproductive age globally and two-third of pregnant women of developing countries (world health organization, 2007, 2014; mehta, 2015). almost 50% of pakistani females at their reproductive age are suffering from anemia amongst which 21% of the anemic females are aged between 9-29 years and similar conditions are faced by the females of punjab, a province of pakistan (mazhar, 2015; akhtar. 2013). results of the studies conducted in our neighboring country, india, on prevalence of anemia among females, has similar outcomes (jawarkar, lokare & kizhatil, 2015). the factors that are considered as the determinants of anemia are age, social class, stress, menstrual blood loss, and dietary deficiencies. helminthic infection is another factor caused by different species of parasitic worms, mostly soil transmitted infections. anemia is mostly detected in the underweight females having body mass index (bmi) less than 18.5 kg/m2. hence, it shows that the nutritional status has significant relationship with anemia in the young females (jawed et al., 2017). micronutrients are very important for human body because human being need them to have strong mental and physical health, to fight with numerous diseases and produce strong and healthy children. in most cases, deficiency of micronutrient such as iron is caused by inadequate access to food and high burden of disease. moreover, improvement in the dietary intake and quality of food also plays a significant role in developing healthy human body. minerals and vitamins are required for body tissues and plays a vital role in human body growth and development (khan, 2015). multiple reasons and factors play a significant role in causing anemia. the existence of anemia in human body varies according to socioeconomic status, dietary deficiencies, different infectious diseases, cultural taboos, multiple pregnancies and genetic hemoglobin (ahmad et al., 2010). as there are multiple antecedents of anemia, nevertheless, common type of anemia is iron deficiency, which is nutritional deficiency disorder. although, nutritional deficiency anemia affect members of both sexes and all age groups however, the problem among women is more dominant. moreover, according to estimation nutritional deficiency anemia affects almost two-third of pregnant women in developing countries. the prevalence of anemia among pregnant women therefore journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 116 results in maternal morbidity and mortality, low birthrate and even neonatal death. (dharmalingam et al., 2010; world health organization, 1992) anemia is most prevalent public health problem and it is indicated mostly in poor nutrition and health. moreover, anemia is widespread in young females and pregnant women due to nutritional deficiency, poor health facilities, lack of awareness and education (world health organization, 2016). furthermore, in developing countries, poor pregnancy outcomes are mostly due to improper intake of iron rich food and supplements, which mostly occurs due to lower socioeconomic status and lack of awareness regarding basic nutritional requirement especially during pregnancy (united nation standing committee on nutrition unscn, 2004). past studies have indicated that improper nutritional intake is one of the main causes of ida and few studies had found it widely prevalent in low and middle-income countries. therefore, income status of the household is very important factor for availing the facility of iron rich food and supplements during pregnancy (united nation international children’s emergency fund/world health organization, 1999). low income reduces the iron intake and increase in the prices of meat and fresh fruits make them less affordable (bhargava et al., 2000). there is significant impact of low socioeconomic status on iron intake status as low household income (<us$ 116.7 per month) was associated with high prevalence of iron deficiency anemia (ngui, 2012). in pakistan, the prevalence of anemia among pregnant women is reported to be 29% in urban areas and 47% to 50% in rural areas (pakistan medical research council, 1998; lone et al., 2004). maternal anemia is one of the causes for prevalence of malnutrition in children under age of three years in the rural areas of peshawar. maternal anemia was observed in 72.8% mothers of malnourished children (gul & kibria, 2013). therefore, a case control study will help to find the impact of socio-demographic and socioeconomic factors on anemia among pregnant mothers. furthermore, significant relation of family planning, non-staple food and iron supplements intake with anemia needs to be more elaborated in pregnant women not only for mother’s health but for child’s health and development as well. consequently, findings of this study will indicate the factors causing anemia, which will play a major role in bringing awareness amongst the people of the area. moreover, it will help the organizations working in the field of health to work in these areas. it will also assist the practitioners by involving respective government departments with an aim to eliminate the causes of anemia in the region. journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 117 as maternal anemia is one of rising health issue therefore, the aim of this study is to find the impact of sociodemographic and socioeconomic factors on anemia among pregnant mothers. further, a case control study will help to analyze the impact of factors in more detail. the purpose of the study at hand was to investigate: • socioeconomic and sociodemographic determinants of anemia in expecting mothers. • the impact of family planning on anemia in pregnant women. • the impact of non-staple food and iron supplements intake status on anemia in pregnant women. 1.1 theoretical framework the analysis of this paper is grounded on the theoretical framework proposed by grossman (1972). the grossman model primarily focuses on the interrelationship between an individual health their socioeconomic status (e.g. wealth, wage and earning), their demand for consumption, as well as for medical goods and services, and lastly time investment in health. from a purely economic perspective, health is a capital that individuals used to produce output of ‘healthy time’, (grossman 1972, 2000). health is one of major determinant on the basis of which an individual allocate an amount of time that could be spend on both market and non-market activities. furthermore, health provides utility which an individual wants to maximize subject to individual recourses’ constraint. whereas health production function provides information about health status input and output over specified period, where each family member is producer of his own and other family members’ health (grossman, 1972, 2000). many studies contributes to the understanding of the socioeconomic inequalities in health and focused on finding the causal relationship between health and socioeconomic factors. these studies identify the underlying apparatus that explains the causal relationship between health and socioeconomic status such as employment and income (e.g. smith, 1999; garcia-gomez et all, 2013). whereas some studies intend to study relationship of health with socioeconomic characteristics particularly education (e.g. lleras-muney, 2005; conti, heckman and urzua, 2010; van kippersluis et all, 2011). however, much of the research have failed to explain the exact mechanism through which education leads to achieving a better health advantage. case and deaton (2005) note that “it is extremely difficult to untangle the links between work, earnings, health, and education, without some sort of guiding framework”. therefore, grossman model https://www.ncbi.nlm.nih.gov/pmc/articles/pmc3932058/#r70 https://www.ncbi.nlm.nih.gov/pmc/articles/pmc3932058/#r70 https://www.ncbi.nlm.nih.gov/pmc/articles/pmc3932058/#r28 https://www.ncbi.nlm.nih.gov/pmc/articles/pmc3932058/#r51 https://www.ncbi.nlm.nih.gov/pmc/articles/pmc3932058/#r51 https://www.ncbi.nlm.nih.gov/pmc/articles/pmc3932058/#r7 https://www.ncbi.nlm.nih.gov/pmc/articles/pmc3932058/#r77 https://www.ncbi.nlm.nih.gov/pmc/articles/pmc3932058/#r35 journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 118 (1972a;b) provides a solid foundation for the study of health inequalities across socioeconomic groups (galama and van kippersluis, 2013). furthermore, galama and van kippersluis (2018) presented a new model that incorporated a life-cycle model of different behaviors that explain a large portion of the observed disparities. this new model serves as a conceptual framework for the seshealth gradient as well as a more developed framework for the production of health. model includes socioeconomic determinants such as wealth, earnings, education, health investments and healthy and unhealthy consumption. based on the discussion above, conceptual framework to analyze the socioeconomic determinants of anemia among expecting mothers is grounded on health-capital theory. like other decision makers, pregnant woman (producer of health) makes decision within the constraint they have. health status (prevalence of anemia among pregnant women in our case) may influenced by socio economic factors. 𝐻 = 𝐹(𝑋) … … (1.1) where “h” is health status output and “x” is vector of different socio economic input factors. health output is proxy by prevalence of anemia among pregnant women while “x” can be divided into following sub factors. 𝐻 = 𝐹 (𝑀, 𝑆, 𝑍) … . . (1.2) where “m” is the vector of economic variables, “s” is a vector of social variables and “z” is the vector of environmental variables. the prevalence of anemia is also caused by other different factors such as dietary intake, level of income and education level because if the frequency of intake of iron food is twice or more per week before pregnancy, it increases the hemoglobin level. studies suggest that around two-thirds of the social gradient in health deterioration could be explained by working environment and lifestyle factors alone (borg and kristensen, 2000). therefore, nutritional status of pregnant women has significant relation with anemia. similarly, income level of household and education level of the mother is also significantly associated with anemia (ansari et al., 2008). the socioeconomic and demographic status has significant impact on anemia among pregnant women. as, high income level of house hold, education of pregnant women, improving the nutritional status in the form of consuming iron rich food and supplements can decrease the percentage of anemic women (rizvi,2012). the iron deficiency anemia widely persist in pakistan among pregnant women and children under 5 years of age and one of the https://www.ncbi.nlm.nih.gov/pmc/articles/pmc3932058/#r35 https://www.ncbi.nlm.nih.gov/pmc/articles/pmc3932058/#r36 journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 119 very prominent factor associated with anemia is monthly income of household which has significant relation with iron deficiency anemia (akhtar et al., 2013). past studies have highlighted the prevalence of anemia under the light of different factors. current study under the heading of socio-economic, family planning, sociodemographic and food and supplements analyzed its impact on prevalence of anemia in pregnant women of rural areas of peshawar. it accentuates that age, gap between children, total children, iron deficiency, non-staple food consumption and supplements continuity has direct effect on anemic and non-anemic condition of expecting women as these factors are directly linked with the human body. additionally, factors like education level of respondent, education level of spouse, family size, independent members, dependent members and monthly income has indirect effect on anemic condition of expecting women. education increases health consciousness and thus leads to an efficient use of medical and preventive care usage and time involvement in the production of health investment, (grossman 1972, 2000). increase or decrease in education level of respondent and her spouse, independent and dependent members and monthly income primarily have impact on awareness regarding gap between children, iron rich food and supplements and their purchasing power. it eventually has an impact on the condition of anemic and non-anemic pregnant women. therefore, assessment of these variables with anemia can become tool for the awareness among people. furthermore, if government implements progressive rules and measures for treatment it can lead to the goal of having healthy mothers and children. 2. methodology gul and kibria (2013) stated that there is 72.8% maternal anemia, which is one of the factors causing malnutrition in children in the areas of pishtakhara and sarband. therefore, study has been conducted in three rural areas namely pishtakhara, sarband and sangu in peshawar in order to find out the factors associated with anemia in the pregnant mothers. primary data based on cross sectional study design was collected through semi-structured questionnaire from 300 pregnant women, from november 2017 to february 2018. data was collected from those pregnant respondents who were registered as anemic and non-anemic in bhu of each rural area through semi-structured questionnaire from each respondent individually by researcher. 2.1 sampling design data of household size was collected from pakistan bureau of statistics, government of pakistan, census results 2017. these areas in terms of roads, streets, journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 120 sanitation system, drinking water system, schools and health seems to be underdeveloped. livelihood of people depends on daily wages. due to lower education level men work as laborers, rickshaws and taxi drivers, some depend on farming. moreover, people with education have jobs but due to lack of job opportunities have small shops. a multistage sampling technique was used. at first stage total sample size was calculated on the following assumptions: expected prevalence of using non-food items as a risk factor for anemia during pregnancy among women with moderate to severe anemia: 20.7% (ansari et al, 2008). expected prevalence of using non-food items as a risk factor for anemia during pregnancy among women with no anemia: 7.2% (ansari et al 2008). based on above findings about percentage of anemic and non-anemic pregnant women, openepi1 software calculated 300 sample size for the current study. at second stage through cluster sampling technique, which is a type of probability sampling technique, three clusters (three rural areas) were selected based on convenience, sangu, sarband and pishtakhara (gul and kibria, 2013). data was collected from the bhu of each rural area. the number of registered pregnant mothers was 144 in the bhu of sangu, 286 in the bhu of sarband and 336 in the bhu of pishtakhara from november 2017 to february 2018. hence, the total population in terms of registered pregnant mothers was 766. the total sample size 300 makes 39% of the total population size, according to the following formula: 𝑁𝑖 = 𝑛 𝑁 × 100 … … (2.1) where, ni = required sample size n = sample size n = total population mathematically, 𝑁𝑖 = 300 766 × 100 = 39% … … (2.2) 1openepi (open source epidemiologic statistics for public health, version 3.01). journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 121 therefore, from each cluster, 39% of the total registered pregnant mothers were selected, which was 56 out of 144, 112 out of 286 and 132 out of 336 registered expecting women in bhu of sangu, sarband and pishtakhara respectively. furthermore, the study has two groups; study group2 and control group3, which means that from total sample size 150 was our study group and 150 was control group. as dandajena (2013) has argued that equal representation of patients suffering from anemia and the one who is not suffering from this disease can help in understanding the impact of independent factors on the dependent variable more accurately. therefore, from the first cluster of 56 sample size data was collected from 28 anemic and 28 nonanemic expecting mothers from the second cluster of 112 sample size 56 anemic and 56 non-anemic were surveyed and similarly from the third cluster out of sample size of 132, data was collected from 66 anemic and 66 non-anemic pregnant mothers. 2.2 logistic regression to analyze the relationship between a dependent variable with one or more than one independent variable, methods of regression are used. among the methods of regression, logistic regression techniques are used when dependent variable is categorical. hence, binary (binomial) logistic regression is used when the dependent variable is binary/dichotomous, such as (use or not use, save or no save, presence or absence, success or failure etc.), (mccullagh, 1989). usually, the categories are coded as 0 and 1, where 0 is coded for failure/absence while 1 is coded for success/presence. to quantify the relationship between dependent variable which is of binary nature and the predictors (independent variables), method of odds ratios is used in logistic regression (kleinbaum, 2008). where, the odds of happening of an event (e.g. the event that y = 1) is defined as the ratio of the probability that the event will occur divided by the probability that the event will not occur. that is, the odd of event a is given by: 𝑂𝑑𝑑𝑠(𝐸) = 𝑃(𝐴) 𝑃(𝐴′) = 𝑃(𝐴) 1−𝑃(𝐴) … … (2.3) where a is an event defined as mothers diagnosed with anemia. the logistic model can be written as: 2 pregnant mothers registered as anemic, having hb less than 11gm/dl, in the bhu of sangu, sarband and pishtakhara 3 pregnant mothers registered as non-anemic, having hb equal to and greater than 11 gm/dl, in the bhu of sangu, sarband and pishtakhara. journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 122 𝑝 ( 𝜋 1 − 𝜋 ) = exp(𝛽0 + 𝛽1𝑥1 + … … … … … … + 𝛽𝑝𝑥𝑝) 1 + exp (𝛽0 + 𝛽1𝑥1 + … … … … … … + 𝛽𝑝𝑥𝑝) … … . (2.4) where π represents probability of occurrence with the value 1 while 1π for probability of failure with the value 0. in order to have linear combination of n explanatory variables, the logistic transform of the corresponding success probabilities pi follows as: 𝐿𝑛 ( 𝜋 1 − 𝜋 ) = (𝛽0 + 𝛽1𝑥1 + … … … … … … + 𝛽𝑝𝑥𝑝: 0 < 𝜋 < 1) … … (2.5) where β0 is intercept and β1,………,βp are regression coefficients of the explanatory variables. therefore, for the current study the following equation has been designed according to selected dependent and independent variables to find the impact of these independent variables on anemia which is the outcome variable of the study: 𝐿𝑜𝑔𝑖𝑡(𝐴𝑛𝑒𝑚𝑖𝑎 ) = 𝛽0 + 𝛽1𝐴𝑔𝑒𝑟𝑒𝑠 + 𝛽2𝐸𝑑𝑢𝑟𝑒𝑠 + 𝛽3𝐸𝑑𝑢𝑠 + 𝛽4𝐹𝑎𝑚𝑠 + 𝛽5𝐶ℎ𝑡 + 𝛽6𝐺𝑎𝑝𝑐ℎ + 𝛽7𝐼𝑛𝑑𝑚 + 𝛽8𝐷𝑒𝑝𝑚 + 𝛽9𝐼𝐻 + 𝛽10𝑁𝑆. 𝐹𝑜𝑜𝑑𝑓𝑟𝑒𝑞 + 𝛽11𝑆𝑢𝑝𝑝𝑐𝑜𝑛 + 𝑒 … … (2.6) whereas, the independent variables are age of respondent (pregnant mother), family size, independent members, dependent members, education level of respondent, education level of spouse, total number of children, gap between children in years, nonstaple food (beef, eggs, milk, fruits) intake frequency per week, supplements continuity status and monthly income of the house hold. moreover, the dependent variable in this study is anemia. 2.3 methods of estimation of logistic regression method of maximum like hood is the commonly used method to estimate the parameters of a logistic regression model. it is defined as the joint probability function of random variables specifically, suppose (𝑥1, 𝑥2, … … … , 𝑥𝑝) are p independent random observations correspond to the random variables (𝑋1, 𝑋2, … … … , 𝑋𝑝) the probability function of xi is: (𝑋 = 𝑥𝑖 ) = 𝜋𝑖 𝑦𝑖 (1 − 𝜋𝑖 ) 1−𝑥𝑖 , 𝑥𝑖 = 0 𝑜𝑟 1, 𝑖 = 1,2, … … … , 𝑛 … (2.7) the log likelihood functions as: journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 123 𝐿(𝛽0, 𝛽1, … , 𝛽𝑝) = ∑ 𝑦𝑖 (𝛽0 + 𝛽1𝑥1 + 𝛽2𝑥2 + … 𝑛 𝑖=0 + 𝛽𝑝𝑥𝑝) − ∑ ln {1 + exp (𝛽0 𝑛 𝑖=0 + 𝛽1𝑥1 + 𝛽2𝑥2 + … + 𝛽𝑝𝑥𝑝)} … . . (2.8) the most effective and well known newton raph sol iterative method can solve the equations. 2.4 test of goodness of model fit pseudo r2 for logistic regression in order to measure the predictive power of a model, r2 is used whereas it is used to measures the amount of variation in the dependent variable which is explained by the independent variable in a linear regression. the r2 for logistic regression is estimated by the cox and snell r2 computed as: 𝐶𝑜𝑥 & 𝑆𝑛𝑒𝑙𝑙 𝑃𝑠𝑒𝑢𝑑𝑜 𝑅2 = [ −𝐿𝐿0 − 𝐿𝐿𝑘 −𝐿𝐿0 ] 𝑛 2⁄ … . . (2.9) where ll0 is the log likelihood of the null model and llk is the log likelihood of the current model. this value cannot reach 1 and nagelkerke improved it to reach 1. the improved r2 is given by: 𝑁𝑎𝑔𝑒𝑙𝑘𝑒𝑟𝑘𝑒 𝑃𝑠𝑒𝑢𝑑𝑜 𝑅2 = 1 − [ −𝐿𝐿0 − 𝐿𝐿𝑘 −𝐿𝐿0 ] 𝑛 2⁄ 1 − [−2𝐿𝐿0] 𝑛 2⁄ … . . (2.10) where ll0 is the log likelihood of the null model and llk is the log likelihood of the current model (hosmer and lemeshow, 2000). hosmer lemeshow test hosmer and lemeshow test is used to find the overall goodness of fit. in order to find goodness of fit test divides subjects into deciles based on predicted probabilities and computes a chi square from observed and expected frequencies using these grouping strategies, the hosmer lemeshow goodness of fit of statistic is calculated by using the formula: 𝐻𝑜𝑠𝑚𝑒𝑟 − 𝐿𝑒𝑚𝑒𝑠ℎ𝑜𝑤 𝑡𝑒𝑠𝑡 = ∑ (0𝑖 − 𝑛𝑖 ′�̂�𝑖 ) 2 𝑛𝑖 ′�̂�𝑖 (1 − �̂�𝑖 ) 𝑔 𝑖=1 … . . (2.11) where, 𝑛𝑖 ′ represents the number of observations in the 𝑖𝑡ℎgroup, journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 124 0𝑖 is the observed outcomes in group 𝑖, given by: 0𝑖 = ∑ 𝑦𝑗 𝑐𝑖 𝑗=1 𝐶𝑖 denotes the number of covariate patterns in the in the 𝑖 𝑡ℎ group �̂�𝑖 is the estimated probability that an event outcome for group 𝑖 and 𝑔 is the number of groups. the likelihood ratio test in the relevance of overall statistical significance, the likelihood ration test (lr) test is performed by estimating two models and comparing the fit of one model to the fit of the other. the likelihood ratio test is performed to test the overall significance of all coefficients in the model on the basis of test statistic: 𝐺 2 = [(−2𝐿𝐿𝑛 𝐿0) − (−2𝐿𝑛𝐿0)] … . . (2.12) where, l0 is the likelihood of the null model and ln is the likelihood of the saturated model. the statistic g2 plays the same role in logistic regression as the numerator of the partial f test does in linear regression. under the global null hypothesis, 𝐻0, 𝛽1 = 𝛽2 = … … … = 𝛽𝑃 = 0 the statistic g2 follows a chi square distribution with p degrees of freedom and measures how well the independent variable affects the response variable. 3. results 3.1 descriptive analysis table 3.1 provides descriptive statistics for anemic and non-anemic respondents. maximum age of anemic respondent was 39 years while non-anemic respondent ages lie between 21 to 33 years. it is apparent from the table that total number of children, family size and dependent members of anemic pregnant mothers are greater from nonanemic pregnant mothers. moreover, gap between child and independent members of anemic pregnant mothers are less in number as compared to non-anemic pregnant mothers. journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 125 table 3.1: descriptive statistics for anemic and non-anemic respondent for closer inspection, chi-square test was carried out to find relation between categorical independent variables. pearson chi-square value for respondent education level is 32.795 with df (degree of freedom) 4, which is highly significant. it means that there is significant relationship between levels of education described in various categories towards lifestyle, which results in either anemic or non-anemic condition of respondent. along with being significant, pearson chi-square value 32.795, which is calculated value, is greater than tabulated value, against p value 0.05 and df 4, which is 9.49. it means that there is association between illiterate, primary, secondary, college and higher education level respondent with anemia. similarly, pearson value is also significant for education level of spouse, monthly income of household, non-staple food consumption per week and supplement continuity. pearson calculated value for spouse education level is 93.617, with df 4, which is greater than tabulated value 9.49. moreover, for monthly income of household pearson calculated value with df 5 is 94.632, which is greater than tabulated value 11.07. hence, the result exhibits that variation in amount of monthly income of household has significant relation and household with different financial status has different approach towards their health. moreover, non-staple food consumption pearson calculated value at df 3 is 283.704 which is greater than tabulated value 7.81 and supplements continuity pearson calculated value at df 1 is 113.072 which is greater than tabulated value 3.84 against p value 0.05. both of the results clearly show that respondent’s practice of consumption of healthy food and supplements continuity has significant role towards outcome of variables unit of measurement min max mean s.d anemic age of respondent years 17 39 26.24 3.49 total children no of children 0 11 5.73 2.48 gap between child years 1.0 2.0 1.13 0.26 family size no of members 7 31 22.78 5.02 independent members no of members 1 6 3.2 0.89 dependent members no of members 6 26 19.57 4.39 nonanemic age of respondent years 21 33 26.54 2.47 total children no of children 0.00 7.00 2.95 1.33 gap between child years 2.0 4.0 2.37 0.47 family size no of members 5 27 14.31 4.68 independent members no of members 1 9 4.23 1.62 dependent members no of members 4 19 10.07 3.57 journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 126 health condition. therefore, difference in status of consumption behavior is associated with anemia. table 3.2: chi-square test variable pearson chi-square df sig respondent education level 32.795 4 .000 spouse education level 93.617 4 .000 monthly income 94.632 5 .000 non-staple food consumption per week 283.704 3 .000 supplements continuity 113.072 1 .000 3.2 empirical results the results of binary logistic regression model are as follows: table 3.3: omnibus test of model coefficients step 1 chi-square df step 108.767 17 block 108.767 17 model 108.767 17 the value of omnibus tests of model coefficients is chi-square of 108.767, which is significant at 0.05. therefore, it can be concluded that adding the predictors to the model have significantly increased the ability to predict factors associated with anemia in pregnant mothers. table 3.4: model summary step -2 log likelihood cox & snell r square nagelkerke r square 1 78.362a .594 .739 the value of cox and snell 𝑅2and nagelkerke 𝑅2are 0.594 and 0.739, which are good enough. the cox and snell 𝑅2indicates that 59.4% of the variation in the dependent variable anemia was explained by the explanatory variables while nagelkerke 𝑅2, which is 0.739, indicates that 73.9% of the variability in the dependent variable anemia was explained by the explanatory variables. the p-value of hosmer and lemeshow test is 0.489>0.05, hence logistic regression model of this study is good fit. table 3.4: results of binary logistic regression analysis variables b (coefficient) s.e. wald df sig. exp(b) odd ratio 95% c.i.forexp(b) lower upper age of respondent -.367 .172 4.581 1 .132 .693 .495 .970 res edu ref: illiterate 3.516 4 .475 primary -.203 .729 1.752 1 .186 .381 .091 1.590 secondary -2.140 1.303 2.550 1 .110 .325 .008 1.628 journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 127 college -2.476 1.632 .475 1 .491 .117 .013 7.952 higher -3.069 3.761 1.657 1 .401 .487 .001 1.014 spouse edu ref: illiterate 13.287 4 .010*** primary -2.273 .866 6.890 1 .009*** .541 .019 1.562 secondary -2.937 .923 .475 1 .125 .426 .164 6.114 college -3.854 .915 .870 1 .351 .399 .071 2.560 higher -9.159 1.600 3.899 1 .048*** .103 .024 1.406 family size .534 .126 12.657 1 .000*** 1.564 1.222 2.000 total children 1.244 .351 6.609 1 .010*** 2.468 1.239 4.913 gap between children -2.141 .716 2.851 1 .009*** .350 .023 13.627 emp. family members -1.201 .123 2.659 1 .012*** .223 .060 1.557 unemp. family members .249 .125 3.944 1 .035*** 1.780 .610 1.997 monthly income 5000-15000 13.110 5 .022*** 15000-25000 -1.384 .995 11.565 1 .001*** .352 .005 .895 25000-35000 -3.016 .925 4.155 1 .042*** .144 .025 .930 35000-45000 -3.117 1.446 4.644 1 .031*** .085 .003 .754 45000-55000 -3.700 1.683 4.835 1 .028*** .052 .021 .669 55000-above -9.192 5.828 7.241 1 .032*** .034 .001 .475 non-staple food ref: none 10.847 3 .013*** daily -23.193 6.129 4.461 1 .024*** .213 .905 2.604 twice per week -21.893 5.457 4.162 1 .035*** .459 .121 7.205 once per week -20.180 5.132 2.438 1 .016*** .741 .061 4.350 supp. continuity ref: no -2.493 .879 8.055 1 .005*** .083 .015 .462 constant 1.657 2.057 .649 1 .421 .191 *= significant at 0.10 (10% significance level and 90% confidence interval) **= significant at 0.05 (5% significance level and 95% confidence interval) ***=significant at 0.01 (1% significance level and 99% confidence interval) the results reveal that age of respondent has positive but insignificant relation with anemia whereas, total children, family size and dependent members has positive and significant relation with anemia. moreover, education of spouse, gap between children, independent members, monthly income, non-staple food consumption per week and supplement continuity has negative and significant while respondent education have negative but insignificant relation with dependent variable anemia. variables such as age of respondent, family size, total children, gap between children, independent members and dependent members are not categorical. on the contrary, education of respondent, education of spouse, monthly income, non-staple food and supplement intake are categorical. following is the interpretation of the results of above table. journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 128 the log odd results showed that an additional year of age of respondent reduces the odd of being anemic by 0.307 times. while the log odd results of family size and total children exhibits that with every single increase in number of member odds of becoming anemic increases by 0.564 and 1.468 times respectively. in terms of gap between children odd ratio result shows that with every increased year as a gap between children odd of level of anemia among pregnant mothers decreases by 0.650 times. similarly odds of prevalence of anemia among pregnant mothers decreases by 0.777 times with every increase in the number of member being independent (employed member) whereas the odds of prevalence of anemia among pregnant mothers increases by 0.780 times with every increase in the number of member being dependent (unemployed member). according the above results of the table, respondents having education level up-to primary, secondary, college and higher level had 0.381,0.325,0.117 and 0.487 times the odds, respectively, of becoming anemic as compare to the respondents who were in the illiterate category. similarly, respondents whose spouse having education up to primary, secondary, college and higher level, had 0.541, 0.426, 0.399 and 0.103 times the odds, respectively, as the spouse of the respondents who were illiterate to get anemic. the respondents who were in 15000-25000, 25000-35000, 35000-45000, 45000-55000 and 55000-above income group had 0.352, 0.144, 0.085, 0.052, 0.021 and 0.001 times the odds, respectively, as compare to those respondents who were in 5000-15000 income group to expose to anemia. considering the effect of intake of non-staple food, respondents, having status of non-staple food intake on daily basis, twice per week and once per week had 0.213, 0.459 and 0.741 times odds, respectively, as to those who fall in category of none to face anemia. whereas, the respondents who use to keep continuity in consuming supplements had 0.083 times odds as compare to those who do not consume supplements on continuity basis to become anemic. 4. discussion the results based on the sample size used in the analysis in terms of anemic and non-anemic pregnant women’s non-staple food consumption per week and supplements continuity shows that lack of intake of iron rich food can strongly affect the level of hemoglobin, which in case of decrease can result in anemia in pregnant women. hence, prolonged negative iron balance due to insufficient dietary intake during pregnancy contributes in prevalence of anemia (world health organization, 2014). it is also evident from the literature that the use of iron rich food and multivitamins sprinkle are journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 129 very effective during pregnancy and they can be easily added to the routine diet (hardling et al., 2017; milman, 2011). according to the results, frequency of non-staple food consumption per week is zero among 85% anemic expecting females and just 15% anemic pregnant women consume non-staple food like beef, eggs, fruits and milk, once per week. on the contrary, 73% non-anemic respondents consume non-staple food twice per week. similarly, 12% anemic mothers keep continuity in supplements intake while 88% do not, while on the other side 73% non-anemic keep continuity in the intake of iron supplements. there can be many reasons behind poor status of iron rich nutritional intake but during the survey from sample size for current study examined that socioeconomic status, which plays a vital role as improvement in monthly earnings, can increase the power of affording the non-staple food and iron supplements. as one of the most important factor that determines the food choices a person can make is the purchasing power, which means that how much one can spend on buying food (marmot, 1984). similarly, jones et al., (1992) found that the children of the rich and middle class family had low prevalence of anemia as compared to those children who belongs to poor house hold because the children from rich and middle class family had better nutritional status. the results of socio-economic variables in terms of dependent, independent members and monthly income shows that non-anemic population size has greater number of independent members and higher monthly incomes as compared to anemic population size of the study. therefore, it can be stated that if the number of earning members from a family size increases, it will help in raising the monthly earnings and can even overcome the burden of bigger family size. low income can result in various outcomes in the form of cutting down the purchase of quality food and availing health care services, which can be an antecedent for prevalence of anemia during pregnancy (kramer, 1987). similarly, habte et al., (2013) concluded that if the households with higher family size have more than one employed person that boosts the economic support at house hold level and can improves food availability which can help in decreasing the level of anemia in pregnant females. further, the study at hand indicated significant results regarding family planning in terms of total number of children and gap between children. it revealed that the respondents with greater number of children are mostly anemic i.e. 13%, 9%, 3% and 1% have 8, 9, 10 and 11 children respectively. on the contrary, the percentage of nonanemic respondents having the aforementioned number of children is zero. similarly, the gap between children is also a serious issue as the results shows that only 19% and 3% anemic respondents keep 1.5 to 2 years gap between their children respectively where 77% have one-year gap between their children. however, the journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 130 minimum gap, which the non-anemic respondents keep in their children, starts from 2 years. therefore, it is clear from the findings that family planning can play an important role in improving the condition of anemia as a major health problem especially among pregnant women. results reveal that 7% and 2% non-anemic mothers also have 5 to 7 children respectively but there is at least two years gap between their children. hence, it indicated that if a family has up to 5 to 7 children, it can be less harmful for mother’s health if there is at least two years of gap between children along with proper intake of non-staple food and supplements continuity. results are related with the of the study of khaskheli et al., (2016), which concluded that approximately two-year gap between children can decrease the number of anemic mothers. findings of this study corroborate with ansari et al., (2008) and harding et al., (2017) which concluded that the increase in the size of family and number of children increases the chances of prevalence of anemia in pregnant women. independent variables like education level of respondent, which is insignificant whereas, education level of spouse has significant relation, but both have a negative and inverse relation with anemia. the descriptive analysis show that 81% of anemic pregnant women are illiterate while just 13% have primary, 5% have secondary and just 1% has college and higher level of education. on the contrary, 63% spouses of the respondents are illiterate where 11% has secondary, 16% have college and just 3% have higher education. the condition of non-anemic respondents and their spouses are different in terms of education level. the results indicate that 51% non-anemic respondents are illiterate while 37% has primary, 6% has secondary, 5% has college and 1% has higher education level. moreover, the results for spouses of the expecting women shows that only 11% spouses are illiterate while 20% has primary, 37% has secondary, 24% has college and 9% has higher education level. hence, the results show that education level is greater among non-anemic expecting women and in their spouses as compared to the anemic respondents and their spouses. however, during data collection it was observed that attainment of education for females and up to some level for males is not considered as priority due to poverty and other priorities, which mostly include necessities of daily life. better socio-economic status can improve nutritional status along with the provision of academic opportunities for females (feinstein et al., 2006). therefore, improvement in socio-economic status can improve the living standard of the expecting mothers in terms of good education, which can improve awareness regarding healthy life style and family planning. moreover, it can develop practice of journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 131 healthy nutrition in terms of intake of iron rich food and supplements, which can decrease the prevalence of anemia in expecting women up to certain level. 5. conclusion the study at hand analyzed the impact of demographic conditions, family planning decisions, socioeconomic status, non-staple food consumption and supplements intake on anemia in the pregnant women of three rural areas of peshawar, kpk. the study explored that anemia is a common disease undergone by most of the pregnant women. its most common cause is low level of monthly income, large family size with lower number of independent members while higher number of dependent members, low frequency of non-staple food consumption per week, lack of continuity of iron supplements and poor health seeking behavior. additionally family planning decision like keeping gap between children has also negative and significant relation with anemia. 6. policy recommendations the study at hand based on its findings and suggestions from the respondents indicate the following policy recommendations: 1. awareness regarding attainment of education irrespective of gender should be developed in these rural areas along with improvement in the quality of education in schools. since respondent education has negative relation with anemia, it indicates that increase in level of education of mothers will decrease the chances of anemia in them. similarly, it is evident from the significant results and negative relationship that increase in education of respondents’ spouse will help in decreasing the percentage of anemic females. 2. the health facilities conditions at basic health units (bhu) in the rural areas needs to develop in terms of basic health facilities. there shall be proper screening facilities available for anemia along with maintaining proper stock of supplements for expecting mothers. during the visit to bhus of all three rural areas, it was observed that there is no facility for screening of anemia and ambulance for emergency condition. moreover, the bhus received enough stock of supplements for whole month but during the first week administration used to come up with an excuse that they are out of stock resulting in discontinuity of supplements intake. hence, the results indicated a significant relation of continuity of supplements intake status with anemia. journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 132 3. steps for awareness regarding importance of family planning should be taken as the results of the study suggests that increase in the gap between children can mitigate the condition of anemia. 4. expecting mothers should be provided with free food vouchers by the government in order to facilitate quality food intake. 5. in order to improve the socioeconomic status of the households, more job opportunities should be provided as increase in monthly income and number of employed members of family has significant relation with anemia. moreover, it indicates that if government works on improving the employment status it will improve the income status of households, which can leads towards adopting healthy life style. references akhtar, s., ahmed, a., ahmad, a., ali, z., riaz, m., & ismail, t. (2013). iron status of the pakistani population-current issues and strategies. asia pacific journal of clinical nutrition, 22(3), 340–347. ahmad, n., kalakoti, p., bano, r., & aarif, s. m. (2010). the prevalence of anaemia and associated factors in pregnant women in a rural indian community. hindu, 208, 67-1. baig-ansari, n., badruddin, s. h., karmaliani, r., harris, h., jehan, i., pasha, o., … goldenberg, r. l. (2008). anemia prevalence and risk factors in pregnant women in an urban area of pakistan. food and nutrition bulletin, 29(2), 132–139. bhargava, a. (2000) modeling the effects of maternal nutritional status and socioeconomic variables on the anthropometric and psychological indicators of kenyan infants from age 0–6 months. am. j. phys. anthropol. 111: 89–104. case, anne, and angus deaton. 2005. "health and wealth among the poor: india and south africa compared." american economic review, 95 (2): 229-233. conti g, heckman jj, urzua s. the education-health gradient. american economic review papers and proceedings. 2010;100:234–238. journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 133 dharmalingam, a., navaneetham, k., & krishnakumar, c. s. (2010). nutritional status of mothers and low birth weight in india. maternal and child health journal, 14(2), 290–298. dandajena, g. (2013). factors associated with chronic malnutrition in mazowe district, mashonaland central province, 2012. fedrick, j., & anderson, a. b. (1976). factors associated with spontaneous pre‐term birth. bjog: an international journal of obstetrics & gynaecology, 83(5), 342350. feinstein, l., & duckworth, k. (2006). development in the early years: its importance for school performance and adult outcomes [wider benefits of learning research report no. 20]. centre for research on the wider benefits of learning, institute of education, university of london. galama tj, kapteyn a, fonseca r, michaud pc. a health production model with endogenous retirement. health economics. 2013 forthcoming. galama tj, van kippersluis, (2018). a theory of socio-economic disparities in health over the life cycle, the economic journal, 129 (january), 338–374 garcia-gomez p, van kippersluis h, o’donnell o, van doorslaer e. long term and spillover effects of health shocks on employment and income. journal of human resources. 2013 forthcoming. government of pakistan & unicef pakistan. (2011). national nutrition survey pakistan 2011. gul, r., & kibria, z. (2013). prevalence and predeterminants of malnutrition in children under 3 years of age in the two. grossman, michael, 1972, the demand for health: a theoretical and empirical investigation, nber occasional papers 119. habte-tsion, h. m., liu, b., ge, x., xie, j., xu, p., ren, m., ... & chen, r. (2013). effects of dietary protein level on growth performance, muscle composition, blood composition, and digestive enzyme activity of wuchang bream (megalobrama amblycephala) fry. the israeli journal of aquaculture-bamidgeh, ija_65. haniff, j., das, a., onn, l. t., sun, c. w., nordin, n. m., rampal, s., ... & zaher, z. m. m. (2007). anemia in pregnancy in malaysia: a cross-sectional survey. asia pacific journal of clinical nutrition, 16(3), 527-536. harding, k. l., aguayo, v. m., namirembe, g., & webb, p. (2017). determinants of journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 134 anemia among women and children in nepal and pakistan: an analysis of recent national survey data. maternal and child nutrition, (november 2016), 1–13. jawed, s., tariq, s., tariq, s., & kamal, a. (2017). frequency of nutritional anemia among female medical students of faisalabad. pakistan journal of medical sciences, 33(2), 398–403. jawarkar, a. k., lokare, p. o., kizhatil, a., & jawarkar, j. a. (2015). prevalence of anemia and effectiveness of iron supplementation in anemic adolescent school girls at amravati city (maharashtra). journal of health research and reviews, 2(1), 7. jonas, o., roder, d., & chan, a. (1992). the association of low socio-economic status in metropolitan adelaide with maternal demographic and obstetric characteristics and pregnancy outcome. european journal of epidemiology, 8(5), 708-714. khan, m. s. (2015). micronutrients status in children of rural and urban population of bannu region. micronutrients status in children of ru, 11(1), 3-6. kleinbaum, d. g., dietz, k., gail, m., klein, m., & klein, m. (2002). logistic regression. new york: springer-verlag. khaskheli, m. n., baloch, s., sheeba, a., baloch, s., & khaskheli, f. k. (2016). iron deficiency anaemia is still a major killer of pregnant women. pakistan journal of medical sciences, 32(3), 630634 kramer, m. s. (1987). determinants of low birth weight: methodological assessment and meta-analysis. bulletin of the world health organization, 65(5), 663737 lleras-muney a. the relationship between education and adult mortality in the united states. review of economic studies. 2005;72:189–221 lone, f. w., qureshi, r. n., & emmanuel, f. (2004). maternal anaemia and its impact on perinatal outcome in a tertiary care hospital in pakistan. eastern mediterranean health journal, 10(6), 801–807. marmot mg, shipley mj, rose g. inequalities in death—specific explanations of a general pattern? lancet. 1984; 323 (8384):1003– 1006. mazhar, s. (2015). prevalence of anemia and dietary iron intake among female adolescents ( grade 8-12) in lahore, journal of the dow university of health sciences karachi, 9(3), 99–105. journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 135 mehta, k. (2015). prevalence of nutritional anamia among college students and its correlation with their body mass index. international journal of reproduction, contraception, obstetrics and gynecology milman, n. (2011). anemia—still a major health problem in many parts of the world!. annals of hematology, 90(4), 369-377. rizvi, f. (2012). impact of maternal education, and socioeconomic status on maternal nutritional knowledge and practices regarding iron rich foods and iron supplements. ann pak inst med sci, 8(2), 101-5. smith jp. healthy bodies and thick wallets. journal of economic perspectives. 1999;13(2):145–166 nhanes. (2003-2012). the prevalance of anemia and moderate-severe anemia in the us population. pakistan medical research council (1998) national health survey of pakistan. health profile of the people of pakistan, 1990-1994. pakistan medical research council, islamabad. romano ngui, y. a. (2012). association between anaemia, iron deficiency anaemia,neglected parasitic infections and socioeconomic factors in rural children of west malaysia. plos neglected tropical diseases, 6(3). united nations international children’s emergency fund/world health organization, (1999) prevention and control of iron deficiency anaemia in women and children. world health organization, geneva, switzerland. united nations standing committee on nutrition. (2004). fifth report on the world nutrition situation: nutrition for improved development outcomes., (march), 22– 27. van kippersluis h, o'donnell o, van doorslaer e. long run returns to education: does schooling lead to an extended old age? journal of human resources. 2011;46(4):695–721. world health organization. 2007. conclusions and recommendations of the who consultation on prevention and control of iron deficiency in infants and young children in malaria-endemic areas. food nutr bull 28: s621–s627. world health organization. haemoglobin concentrations for the diagnosis of anaemia and assessment of severity. geneva, switzerland, 2011. journal of applied economics and business studies, volume.4, issue 3 (2020) 113-136 https://doi.org/10.34260/jaebs.435 136 world health organization. global nutrition targets 2025: anemia policy brief. (who/nmh/nhd/14.4). geneva: 2014; world health organization. world health statistics 2016: monitoring health for the sdgs, sustainable development goals. world health organization. (2017). nutritional anaemias : tools for effective prevention. warrell, cox, firth, & benz, 2003. american academy of family physician. journal of applied economics and business studies, volume. 4, issue 2 (2020) 205-220 https://doi.org/10.34260/jaebs.4210 205 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x different is better: does difference in age and ethnicity of the directors matter for corporate performance in malaysia? haseeb ur rahman 1*, muhammad zahid,2 and muhammad jehangir 3 1 faculty member and director institute of management sciences, university of science and technology, bannu, khyber pakhtunkhwa, pakistan. 2 associate professor and head of the department, city university of science and information technology peshawar, khyber pakhtunkhwa, pakistan. 3 institute of business and leadership (ibl), abdul wali khan university mardan, khyber pakhtunkhwa, pakistan. abstract the prior diversity-related literature is mostly dominated by the boardroom gender diversity of the top or large companies of the developed countries. consequently, this study investigates the impact of the rarely investigated boardroom diversity-related dimensions like directors’ age and ethnicity on the financial performance of 360 randomly selected non-financial listed companies from a developing country malaysia from 2010 to 2014. the findings revealed that ethnic equality (the presence of directors from all the three major ethnicities of the country) on the board has a significant positive relationship with roa and share price. however, directors’ age has a significant positive association with share price but it has an insignificant effect on roa. the findings of this study provide important insights for the regulators, policymakers, and all other key stakeholders of the developing countries, especially malaysia, where the corporate boards are mostly dominated by men of middle age from chinese ethnicity. keywords age and ethnicity of directors, roa, share price, malaysia jel classification g30, g38, g39 1. introduction agency theory and most of the corporate governance (cg) regulations firmly stand on the notion that managers being agents of the owners (shareholders) are liable to protect their interests. however, managers being opportunistic peruse their own interests instead of the shareholders; hence the corporate board or more specifically directors on the board should be qualified, experienced, independent, vigilant and skilled enough to protect shareholders from the expropriation of management (fama, 1980; fama & jensen, 1983; jensen & meckling, * email: drhaseeb@ustb.edu.pk mailto:drhaseeb@ustb.edu.pk haseeb ur rahman, muhammad zahid & muhammad jehangir 206 1976). in this vein, boardroom diversity among others grabbed substantial attention for strengthening the structure, composition, and roles of the board (rutledge et al., 2016). like other countries, malaysia also aimed to increase boardroom diversity as evidenced by its third cg code malaysian code on corporate governance (mccg) introduced in march 2012 (ismail et al., 2013; malaysian corporate code on governance, 2012; haseeb ur rahman, rehman, et al., 2018; haseeb ur rahman, zahid, et al., 2018). however, among other aspects of the boardroom diversity, only gender aspect could be prioritized and remained a subject of high interest for the regulators, practitioners, academicians, and researchers around the world and in malaysia (baker et al., 2020; haseeb ur rahman et al., 2017a; zahid et al., 2019). the literature paid relatively less attention to other dimensions of boardroom heterogeneity like age, ethnicity, nationality, professional background, and cognition of the directors (baker et al., 2020; ezeanyim, 2020). few studies which investigated the nexus of age and ethnicity of directors with firm performance produced incongruent results. for instance, some studies noted that age and ethnicity of the directors have positive impacts on firm performance by uplifting the quality of decision making (anju, 2020; ezeanyim, 2020; jonson et al., 2020), while others documented that there is no or negative association between them (pandey, 2020). the latter studies reported that heterogenic boards require more time and energy in reaching a consensus as compared to homogenous boards which are quick and good in making unanimous decisions (hambrick & mason, 1984; haseeb ur rahman et al., 2017b). in addition to the communication problem, board members from the diverse background also have their personal or geographical interests which harm coordination, solidarity, consistency, integrity, and motivation of the board (adams & ferreira, 2009; brown, 2016; carter et al., 2010; pandey, 2020). besides, it is also noted that corporations exhibit boardroom heterogeneity for the sake of showing compliance with relevant regulations that do not improve their efficiency or performance (carter et al., 2010; haseeb ur rahman, rehman, et al., 2018; haseeb ur rahman, zahid, et al., 2018). to sum up, most of the prior literature is dominated by boardroom gender diversity leaving a space and motivation for investigating other aspects of the boardroom like directors’ age and ethnicity (baker et al., 2020; zahid et al., 2019). moreover, the studies in the past mostly relied on small or samples of the top companies from the developed countries (anju, 2020; carter et al., 2010; ezeanyim, 2020). subsequently, this study aims to investigate the impact of age and ethnic equality of the directors on firms’ accounting (roa) and market performance (share price) in a stratified random sample of 360 non-financial listed companies from a developing country like malaysia for five years from 2010 to 2014. the study has many contributions. first, the study enriches the previous incongruent boardroom diversity literature mostly focusing on gender diversity in the small or samples of the top companies of the developed countries (anju, 2020; carter et al., 2010; ezeanyim, 2020). second, it contributes to the literature as the age and ethnicity of the directors have rarely been examined in the past. journal of applied economics and business studies, volume. 4, issue 2 (2020) 205-220 https://doi.org/10.34260/jaebs.4210 207 moreover, investigating these is also important especially after the implementation of the minimum retirement age act 2012 on 1st july 2013 in a country that hosts various ethnicities i.e. malay (50%), chinese (22.6%), indigenous bumiputras (11.9%), indians (6.7%) and others (0.7%). third, it contributes to the literature and policy especially after the enactment of mccg 2012 that requires increasing boardroom diversity. this is further pronounced as the previous two codes of the country – mccg 2000 and mccg 2007 not addressed the issue of boardroom diversity and thus the prior literature could not focus on the subject in the regulatory context (bursa malaysia 2012; rahman et al. 2017a; rahman, ibrahim, and che ahmad 2015). fourth, it enriches the literature by using financial performance as an outcome variable that is still considered important for firms’ success, especially in developing countries. therefore, firms and management in these countries are keen to know about the financial returns of any change or reform they introduced or opt for (gul, muhammad, and rashid 2017; rahman, ibrahim, and ahmad 2015). besides, firm financial performance measured by both the accounting and market perspective provides a comprehensive picture as the former is calculated based on facts and figures while the latter reflects firms’ image in the minds of customers, investors, market, and society (haseeb ur rahman et al., 2016). overall, the study updates the literature, regulators, policymakers, and other stakeholders in a multi-ethnic country like malaysia that has recently turned its attention to increasing boardroom heterogeneity as reflected by mccg 2012. the remaining of this paper is organized as the following section reviews prior literature for developing hypotheses while the next sections discuss methodology and findings before reporting the conclusion and recommendations of the study. 2. literature review 2.1 boardroom average age and firm performance like many other issues in management sciences, the previous empirical studies regarding the age of the directors being an important element of the boardroom diversity also lack uniformity. some of these studies favor seniors (ahern & dittmar, 2012; amoll, 2015; demeke, 2016; smith-meyer, 2013) while others support young directors (abdullah & ismail, 2013; darmadi, 2011; grove et al., 2011; hambrick & mason, 1984). supporters of the senior directors claim that young directors commit more mistakes in financial and economic matters due to lack of experience. besides grooming young directors and controlling their risky approach (agarwal et al., 2009; demeke, 2016), they also acquire important information from their already established network and experience. in addition to experience in managing risks and handling organizational problems, they also oppose firms’ investment in vulnerable projects (demeke 2016; francis, hasan, and wu 2012). hence, many studies endorse the significant positive impact of senior directors on firms’ financial performance (ahern & dittmar, 2012; amoll, 2015; anju, 2020; demeke, 2016; ezeanyim, 2020; francis et al., 2012; smith-meyer, 2013). also, they improve firms’ overall efficiency, market value, and haseeb ur rahman, muhammad zahid & muhammad jehangir 208 shareholders’ confidence (demeke, 2016; francis et al., 2012). accordingly, the death (nguyen & nielsen, 2014) or retirement of senior directors decrease share market price (arioglu, 2015). whilst, young directors have negative impacts on firms’ financial performance (ahern & dittmar, 2012; smith-meyer, 2013) and market value (adams & ferreira, 2009; vania & supatmi, 2014). however, in contrast, it is argued that senior directors have low physical stamina and cerebral resilience which affect their cognitive abilities. risk aversion, concerns for financial and career security along with the rigidity to respond regulatory reforms and other challenges posed by the external environment are those issues associated with senior directors which weaken their monitoring abilities (grove et al., 2011). hence, many studies found that the average age of directors has a negative relation with firms’ financial performance (abdullah & ismail, 2013; grove et al., 2011). hambrick and mason, (1984) argued that young directors have a low tendency towards the status quo and willingness to accept challenges that improve their flexibility and adaptability for structural and strategic changes in the organization. many empirical studies found that young directors have a significant positive relation with firms’ financial performance (abdullah & ismail, 2013; darmadi, 2011), market value, and share market price (darmadi, 2011; eversheds report, 2013). interestingly, some studies also found that directors’ age has no significant relation with firms’ financial performance (letting’ et al., 2012; muravyev, 2017; pandey, 2020), share market price (randøy et al., 2006) and firms’ market value (kusumastuti et al., 2007). based on mixed findings and a focus on either accounting or market performance as an outcome variable of prior literature along with a rare empirical investigation of the relationship under consideration after the enactment of mccg 2012 in malaysia necessitates further investigation. the need for re-investigation is also highlighted by the discussion started after the implementation of the minimum retirement age act 2012 on 1st july 2013 in malaysia. therefore, this study establishes the following hypothesis for further investigation. h1: average age of directors has a significant positive impact on firm performance. 2.2. boardroom ethnic equality and firm performance under representing a different culture, ethnic directors have a low level of social interaction and informal relations with their colleagues which strengthen independence and monitoring roles of the board (baker et al., 2020; huo, 2016; nguyen et al., 2015; haseeb ur rahman et al., 2014; zahid et al., 2019). besides improving the transparency of the financial matters (ferreira, 2010; marimuthu & kolandaisamy, 2009), ethnic directors also assist firms in an independent assessment of compensation and nomination of the directors (carter et al., 2010; salloum et al., 2017). they contribute a critical but constructive voice on the board that improves firms’ financial performance (carter et al., 2003; huo, 2016; ramirez, 2003). many empirical studies found that ethnic directors have a significant positive relationship with firms’ journal of applied economics and business studies, volume. 4, issue 2 (2020) 205-220 https://doi.org/10.34260/jaebs.4210 209 financial performance (abdullah & ismail, 2013; anju, 2020; carter et al., 2003; cheong & sinnakkannu, 2014; ezeanyim, 2020; rachagan et al., 2015). shareholders also assign a value to ethnic directors for connecting firms with critical resources and acquiring legitimacy from the external environment (huo, 2016; ntim, 2015). also, they assume that ethnic directors signal firms’ commitment towards equality, transparency, good governance (ferreira, 2010; salloum et al., 2017) and shareholders’ protection (abdullah & ismail, 2013; huo, 2016; ismail et al., 2013; oxelheim & randoy, 2003; salloum et al., 2017) which improve firms’ reputation and political support (amoll 2015; carter et al. 2003; ferreira 2010; huo 2016; ntim 2015; zahid et al. 2019). however, in contrast, it is also argued that the preference of ethnic directors for a particular geographical area and barriers in communication cause conflicts and information asymmetry among the board members (brown, 2016; carter et al., 2010). also, ethnic directors are less effective due to no or low attention from the organization (adams & ferreira, 2009; brown, 2016; pandey, 2020) or showcasing them for the sake of compliance with relevant regulations (brown 2016; zahid et al. 2019). accordingly, many studies found no or significant negative impact of the ethnic directors on firms’ financial performance (brown, 2016; ilona, 2015; pandey, 2020; wang & clift, 2009) including share price (kusumastuti et al., 2007). to sum up, the prior literature is not only scarce but also incongruent. besides, it examined the impact of ethnic directors on either accounting or market performance and that too of the top companies only. furthermore, the introduction of mccg 2012 that required increasing boardroom heterogeneity in a multi-ethnic country like malaysia where the total population of around 30 million accounts for various ethnicities i.e. malay (50%), chinese (22.6%), indigenous bumiputras (11.9%), indians (6.7%) and others (0.7%) also necessitates further investigation of the subject. the importance of this investigation is further magnified by bearing in mind the riot on 13th may 1969 and it's aftershocks that changed the ethnic outlook of the country by implementing new economic policy (nep) and sparking the differences among different ethnicities, especially malay and chinese (abdullah & ismail, 2013; cheong & sinnakkannu, 2014; ismail et al., 2013). accordingly, this study investigates the relationship by establishing the following hypotheses. h 2: ethnic equality on the board has a significant positive impact on firm performance. 2. methods due to potential effects, the study includes age and size (haseeb ur rahman et al., 2017a) of the sample firms as control variables. board size, cg reforms (mccg 2012), and the effects of time (5years) and industry (9 sectors) are other control variables in the study. regarding the sample, there were 960 companies registered in 12 different sectors on bursa malaysia (malaysia's stock exchange) at the end of 2009 (economic planning unit, 2011). by excluding finance, hotels, and mining sectors due to their different governance structure and low https://en.wikipedia.org/wiki/malaysian_new_economic_policy haseeb ur rahman, muhammad zahid & muhammad jehangir 210 representation, the study selected 360 non-financial listed companies through proportionate stratified random sampling from a total of 921 companies registered in nine sectors. following previous studies, data for age and ethnicity of directors on the board collected from annual reports while the information for all other variables of the study extracted from thomson reuters datastream. the following are two econometric models of the study. firm performance𝑖𝑡 = β0 + β1bage𝑖𝑡 + β2 beth𝑖𝑡 + β3bsiz𝑖𝑡 + β4fage𝑖𝑡 + β5fsiz𝑖𝑡 + β6cod𝑖𝑡 + + β7id𝑖𝑡 + β8td𝑖𝑡 + ε𝑖𝑡……………………………. (1) where; firm performance is captured through return on assets (roa) and share price of the ith firm at time t β = beta bage𝑖𝑡 = age of the directors on the board of the ith firm at time t beth𝑖𝑡 = a dummy variable denoting 1 for the presence of directors from all the three major ethnicities (malay, chinese and indians) on the board and 0 otherwise of the ith firm at time t bsize𝑖𝑡 = total number of directors on the board of the ith firm at time t fage𝑖𝑡 = age of the ith firm at time t measured by the number of years since listing fsiz𝑖𝑡 = size of the ith firm at time t measured by the log of the total market capitalization cod𝑖𝑡 = malaysian code on corporate governance 2012 measured as 1 for code and 0 otherwise id𝑖𝑡 = dummy variables for controlling sector-wise effects on ith firm at time t td𝑖𝑡 = dummy variables for controlling time effects of five years on ith firm at time t ε𝑖𝑡 = error term of the ith firm at time t 4. data analyses 4.1. univariate analyses the descriptive statistics reported in table 1 show that sp representing share market price has an average value of 1.958 rm (ringgit malaysia). roa has a mean value of 0.040 and the negative sign of its minimum value explains losses of some of the sample firms. table 1: descriptive statistics minimum maximum mean std. deviation skewness kurtosis sp 0.04 3.445 1.958 0.040 .001 -.073 roa -1.09 6.34 0.055 0.237 .000 -.069 bage 38.83 72.83 56.856 4.827 -.020 .457 beth 0.00 1.00 0.125 0.331 0.270 0.156 fage 1.00 42.00 15.992 7.250 .244 -.302 fsiz 4.07 7.36 5.590 0.589 .561 .148 bsiz 4.00 13.00 7.201 1.699 .663 .290 cod 0.00 1.00 0.600 0.490 journal of applied economics and business studies, volume. 4, issue 2 (2020) 205-220 https://doi.org/10.34260/jaebs.4210 211 regarding the age of the directors, malaysian boards are mostly occupied by middle and old age directors as evidenced by an average of 56.86 years in a range from 38.83 to 72.83 years. in respect to ethnic equality, the statistics show that only 12.5% of the boards host directors from all the three major ethnicities (malay, chinese, and indian) of the country. among control variables, fage has an average value of 16 years while fsize measured by the log of total market capitalization has a mean value of 5.59. similarly, the average size of the board is 7.2 while the cod shows the enactment of the code for the last three years of the study since 2012. given these findings, it is noted that despite a multi-ethnic society, malaysia has a low boardroom gender and ethnic diversity. overall, the boards of most of the firms are dominated by middle-aged males from the chinese and malay ethnicities of the country. to investigate the year-wise statistics for examining the impact of the new code, this study employed comparing the means of average age and ethnicity equality of the directors on the board on yearly basis for 5 years from 2010 to 2014. the statistics for bage in table 2 show that the average age of directors i.e. 55.718 years in 2010 increased to 56.492 in 2011 and 57.124 in 2012. likewise, bage of 57.470 years in 2013 and 57.475 years in 2014 respectively show a slight increase at a decreasing rate. table 2: year-wise mean analyses year 2010 2011 2012 2013 2014 bage 55.718 56.492 57.124 57.470 57.475 beth 0.12 0.12 0.13 0.12 0.13 bage denoting average age of all the directors on the board and beth a dummy for the presence or otherwise of the directors from all the three major ethnicities of the country equality the findings indicate that malaysian firms have started the nomination of some young directors to their boards particularly after the enactment of new code. the year-wise statistics for beth are interesting as their 12% representation in 2010 and 2011 each increased to 13% in 2012. however, it dropped again to 12% in 2013 before increasing to 13% in 2014. the findings indicate that the new code has no substantial contribution to improving compliance with ethnic equality on the malaysian boards. table 3 shows pearson’s correlation of the age and ethnic equality of the directors with accounting (roa) and market measures (share market price) of firms’ performance. the statistics show that beth has a significant positive association with roa. similarly, bsiz and fsiz also have a significant positive correlation with roa. however, bage has an insignificant positive while fage shows an insignificant negative association with roa. moreover, cod has a significant negative relationship with roa. the statistics reported in table 3 also show that bage and beth have significant positive coefficients towards sp. likewise, bsiz, fage, fsiz, and cod also have a significant positive association with sp. haseeb ur rahman, muhammad zahid & muhammad jehangir 212 overall, there is no correlation above 0.8 and thus there is no multicollinearity in both models of the study. table 3: pearson correlation roa sp bage beth bsiz fage fsiz cod roa 1 sp 0.439** 1 bage 0.035 0.303** 1 beth 0.088** 0.128** 0.008 1 bsiz 0.183** 0.280** 0.117** 0.112** 1 fage -0.021 0.272** 0.351** -0.037 0.058* 1 fsiz 0.221** 0.602** 0.297** 0.078** 0.366** 0.309** 1 cod -0.067** 0.084** 0.127** 0.008 -0.026 0.172** 0.051* 1 4.2 multivariate analyses after having no multicollinearity, the data for both the models diagnosed for heteroscedasticity, cross-sectional dependence, and serial correlation. the results of the breusch-pagan/cook-weisberg test for heteroscedasticity test (chi2 (1) = 23.04 and prob. > chi2 = 0.000) reported in table 4 and (chi2 (1) = 1961.81 and prob. > chi2 = 0.000) in table 5 confirm heteroscedasticity in the models. similarly, the significance (pr. = 0.000) of the statistics of 59.556 in table 4 and 89.649 in table 5 evidence cross-sectional dependence in both models. the statistics of wooldridge test for autocorrelation f (1, 319) = 15.084 and prob. > f = 0.000 in table 4 and f (1, 319) = 182.834 and prob. > f = 0.000 in table 5 show the existence of serial correlation. keeping in view the results of all these diagnostics tests, the study employed pooled ordinary least squares (ols) with robust standard errors – drisc. kraal standard errors (dkses) as suggested by hoechle (2007). additionally, the study also employed ols panel corrected standard errors (pcses) that is considered a more consistent and robust estimator in the presence of heteroscedasticity and cross-sectional dependence in the panel data estimation. however, as the estimator is sensitive to serial correlation, thus the study used a one-year lag of roa and sp as predictors. the statistics of f (23, 4) = 267.22 and prob. > chi2 = 0.0000 for dkses and wald chi2 (15) = 1.46000 and prob. > chi2 = 0.0000 for pcses reported in table 4 evidence fitness of model 1. similarly, the findings of f (23, 4) = 23471.34 and prob. > chi2 = 0.0000 for dkses and wald chi2 (15) = 3.51000 and prob. > chi2 = 0.0000 for pcses in table 5 confirm fitness of model 2. 5. discussion table 4 shows the statistical findings for the relationship of age and ethnic equality of the directors with firm performance measured by roa through dkses and ols – pcses estimators. the findings of both the estimators show that an increase in bage has no significant impact on roa. following previous literature, the finding has a plausible journal of applied economics and business studies, volume. 4, issue 2 (2020) 205-220 https://doi.org/10.34260/jaebs.4210 213 explanation that the cognitive ability of the individuals regarding monitoring and decision making gets weaker with an increase in their age that affects firms’ performance. also, the findings could be explained in that senior directors are not only concerned about their financial and career security but also rigid in responding to the regulatory reforms and other challenges posed by the internal and external environment which affect firms’ performance (grove et al., 2011). table 4: boardroom diversity and firm performance (roa) roa dkses xtpcses bage 0.003 (0.002) 0.003 (0.004) beth 0.180** (0.054) 0.097** (0.042) bsiz 0.057** (0.007) 0.040*** (0.010) fage -0.011*** (0.001) -0.007** (0.003) fsize 0.338*** (0.032) 0.190** (0.085) cod -0.119** (0.011) -0.162*** (0.016) lagroa 0.432 (0.148) cons 0.035** (0.005) -1.168* (0.601) industry dummy yes yes year dummy yes yes f(23, 4) 267.22 wald chi2(15) 1.460 prob > chi2 0.000 prob > chi2 0.000 r-squared 0.1056 r-squared 0.274 heteroscedasticity chi2(1) prob > chi2 23.04 0.000 cross sectional dependence pesaran cd pr 59.556 0.0000 autocorrelation f(1, 319) prob > f 15.084 0.0000 standard errors in parentheses, *** p < 0.01, ** p < 0.05,*p < 0.10 the findings are consistent with many previous studies showing no significant relationship between the age of the directors and firms’ performance (letting’ et al., 2012; muravyev, 2017; pandey, 2020). however, the findings are not consistent with (abdullah & ismail, 2013) in malaysia which may probably be due to the difference in methodology or their sample composed of top companies. the estimations of dkses and pcses reported in table 4 also show that ethnic equality – the presence of directors from major three ethnicities of malaysia has a significant positive haseeb ur rahman, muhammad zahid & muhammad jehangir 214 impact on firms’ performance (roa). by consulting the previous studies, the findings have a possible explanation that ethnic directors improve firm performance by augmenting monitoring role of the board and transparency of financial matters as they have no strong social or informal relations with other colleagues mainly due to differences in culture (huo, 2016; nguyen et al., 2015). besides, under their different culture, they also improve the quality of decision making through offering unique perspectives which have a positive impact on firms’ financial performance by countering group thinking (carter et al., 2003; huo, 2016; ramirez, 2003). simply, the findings show similarity with many studies around the world (carter et al., 2003) including uk (nathan, 2016), kenya (amoll, 2015), nigeria (ezeanyim, 2020), india (anju, 2020), australia (jonson et al., 2020) and malaysia (abdullah & ismail, 2013; cheong & sinnakkannu, 2014; marimuthu, 2008; rachagan et al., 2015). however, the findings are different than hassan and marimuthu (2016) in malaysian context which might be due to their different methodology or a small sample of top companies which does not reflect the true picture of the economy. table 5 shows the estimations dkses and ols – pcses for the relationship of age and ethnic equality of the directors with firm market performance measured by share market price. the findings of both the estimators show that the average age of the directors on the board has a significant positive impact on the share market price. by referring to previous studies, the findings have an explanation that senior directors by their experience and ability to oppose managements’ decisions particularly in regard with risky projects enhance shareholders’ confidence as reflected by an increase in share market price (demeke 2016; francis, hasan, and wu 2012). also, the findings could be explained in that shareholders consider senior directors on the board as safe hands for managing risks (arioglu, 2015; nguyen & nielsen, 2014) and improving firms’ overall efficiency (demeke, 2016; francis et al., 2012). the findings are consistent with many previous studies (anju, 2020; demeke, 2016; ezeanyim, 2020; francis et al., 2012; jonson et al., 2020). the dkses and pcses estimations also evidenced that the presence of directors on the board from all three major ethnicities in the country has a significant positive impact on the share price. table 5: boardroom diversity and firm performance (share market price) sp dkses xtpcses bage 0.019** (0.002) 0.012 *** (0.003) beth 0.257*** (0.016) 0.106** (0.030) bsiz 0.035** (0.009) 0.026*** (0.006) fage 0.008** (0.002) 0.005*** (0.001) fsize 0.896*** (0.040) 0.457*** (0.138) cod 0.197*** 0.419*** journal of applied economics and business studies, volume. 4, issue 2 (2020) 205-220 https://doi.org/10.34260/jaebs.4210 215 (0.004) (0.044) lagsmp ------- 0.590*** (0.116) cons 0.201*** (0.003) -4.135*** (0.959) industry dummy yes yes year dummy yes yes f (23, 4) 23471.34 wald chi2(15) 3.510 prob > chi2 0.000 prob > chi2 0.000 r-squared 0.4439 r-squared 0.680 heteroscedasticity chi2(1) prob > chi2 1961.81 0.000 cross-sectional pesaran cd pr 89.649 0.000 autocorrelation f(1, 319) prob > f 182.834 0.000 standard errors in parentheses, *** p < 0.01, ** p < 0.05, *p < 0.10 consulting previous literature, the findings could be explained that ethnic directors signal firms’ positive image particularly in regard with board independence, transparency of financial matters and commitment of firms towards good governance including the protection of shareholders’ interests (abdullah & ismail, 2013; huo, 2016; ismail et al., 2013; oxelheim & randoy, 2003; salloum et al., 2017). these, in turn, enhance shareholders’ confidence as reflected by an increase in share market price (ferreira, 2010; salloum et al., 2017). the shareholders also believe that ethnic directors are helpful in the acquisition of resources and legitimacy for firms (huo, 2016; ntim, 2015) which improve their reputation, political support and worth in the market, among others (amoll, 2015; carter et al., 2003; ferreira, 2010; huo, 2016; ntim, 2015). also, the findings could be explained in that ethnic directors increase share market prices by attracting institutional investors (brammer et al., 2007). the findings of the study endorse many previous studies (anju, 2020; carter et al., 2003, 2010; cook & glass, 2009; ezeanyim, 2020; jonson et al., 2020; ntim, 2015). overall, the findings show that bage has an insignificant (table 4) and significant positive (table 5) association with roa and share price which partially supports h1 of the study. likewise, the findings reveal a significant positive association of beth with roa (table 4) and share price (table 5) that fully support h2 of the study. among control variables, both the estimations show a significant positive impact of the bsiz on roa and share market price. besides an increased diversity, the findings also indicate that shareholders prefer the large size of the board for improving monitoring and the acquisition of resources to improve firm performance. previously, it is found that small boards may suffer regarding expert advice or better conceptualization of problems (jensen & meckling, 1976; pfeffer & salancik, 1978). fage has a significant negative impact on roa that might have a haseeb ur rahman, muhammad zahid & muhammad jehangir 216 rationale that sample firms could not adapt themselves as per the emerging challenges and new developments in the market which facilitate the introduction of substitutes for their products or decreased their market share. the findings are consistent with rahman et al. (2017b) who found that an increase in firms’ age negatively affects their performance. fsize has a significant positive impact on firms’ financial performance (roa) which indicates that large firms have substantial assets the efficient utilization of which improves firms’ financial performance (haseeb ur rahman et al., 2017a). the significant positive association of fage and fsiz with sp has an explanation that shareholders express confidence in old, already established, and large firms due to their experience and potential resilience for the economic and financial crisis (rahman et al. 2017a). as per both the estimations, cod has a significant negative impact on roa. this could be explained in that regulatory reforms complicate the already in practice structure of the firms that affect their efficiency and thus profitability through increased costs, particularly in the short run (haseeb ur rahman et al., 2017a). the significant positive coefficient of cod implies that shareholders believe in the introduction of new code for improving firms’ compliance with good cg practices including the protection of shareholders’ interests. under pcses estimation, no significant relationship between lagroa and roa shows that firms’ previous years’ financial performance has no significant role in the current year. likewise, the significant positive relationship between lagsp and sp indicates that shareholders assign a value to firms’ previous record on the stock market in buying shares. 6. conclusion and recommendations this study aimed to investigate the nexus of rarely focused boardroom diversity-related aspects like age and ethnicity of the directors with firm performance and the impact of regulations on their association if any. the findings explain that mccg 2012 not only addressed the important issue of boardroom diversity but also improved firms’ compliance in this regard. however, the compliance is still not good and the overall statistics are not satisfactory as the malaysian boards are mostly dominated by men of middle age from chinese ethnicity, and only 12.5% of them have ethnic equality-directors from all the three major ethnicities. the insignificant association between directors’ age and roa endorse that senior directors have no significant contribution towards firms’ accounting performance which might be due to their weak physical stamina or monitoring abilities. the significant positive influence of the ethnic diversity on roa explains that firms with boards having directors from all the three major ethnicities of the country have better financial performance than others. the positive findings for the relationship of age and ethnic equality of directors with share market price indicate that shareholders express trust in senior and ethnically diverse boards for increasing firms’ reputational capital and signaling their equality, transparency, experience, and commitment towards good governance, particularly the protection of shareholders’ interests. journal of applied economics and business studies, volume. 4, issue 2 (2020) 205-220 https://doi.org/10.34260/jaebs.4210 217 the study has many contributions and policy implications. first, this study identified the current status, the level of voluntary compliance, and the importance of the boardroom heterogeneity in malaysia. though the government has taken some steps like the introduction of mccg 2012 that recommends an increase in boardroom diversity, thus it is the turn of malaysian firms to ensure their compliance with the new code or voluntarily increase boardroom heterogeneity for reaping its benefits. second, the findings of the study also provide policy insights in that to train, equip and enhance the acquaintance and capabilities of the senior and directors from diverse backgrounds, especially regarding the new technologies and market trends, for the effective discharge of their duties and improving efficiency. third, the regulators and firms may increase boardroom ethnic diversity in anticipation of producing good results regarding their financial performance, positive image, and equality as the stakeholders and more importantly, the customers belong to various ethnic groups of society. finally, they may also encourage the nomination of a few young directors to the boards. overall, the findings indicate that malaysian firms should reflect the composition of the population on their boards as an increase in boardroom diversity could minimize group thinking or the inclination of its decisions towards a particular group or groups of the stakeholders. the study is not free of limitations; hence studies in the future may add value to the subject by considering some other aspects of the boardroom diversity and the qualitative aspect of this investigation as well. references abdullah, s. n., & ismail, k. n. i. k. (2013). gender, ethnic, and age diversity of the boards of large malaysian firms and performance. jurnal pengurusan, 38(2013), 27–40. adams, r. b., & ferreira, d. (2009). women in the boardroom and their impact on governance and performance. journal of financial economics, 94(1), 291–309. agarwal, s., driscoll, j. c., gabaix, x., & laibson, d. (2009). the age of reason. financial decisions over the life-cycle with implications for regulation. brookings papers on economic activity, fall 2009, 51-117. ahern, k. r., & dittmar, a. k. (2012). the changing of the boards: the impact on the firm valuation of mandated female board representation. the quarterly journal of economics, 127(1), 137–197. amoll, d. o. (2015). effect of the board of directors composition on the financial performance of companies listed in the nairobi securities exchange. master thesis. school of business, university of nairobi. anju, v. (2020). a critical review of literature on the impact of workforce diversity (specifically age, gender, and ethnic diversity) on organizational competitiveness. asian journal of management, 11(1), 1–11. arioglu, e. (2015). demographic diversity in the boardroom: : evidence from borsa istanbul. research journal of business and management, 2(1), 18–36. baker, k., pandey, n., kumar, s., & haldar, a. (2020). a bibliometric analysis of board diversity: current status, development, and future research directions. journal of business research, 108, 232–246. brammer, s., millington, a., & pavelin, s. (2007). gender and ethnic diversity among the uk corporate boards. journal compilation, 15(2), 393–403. brown, r. j. (2016). a discourse on diversity: the impact of management team heterogeneity on firm haseeb ur rahman, muhammad zahid & muhammad jehangir 218 performance. undergraduate economic review, 13(1), 1–25. bursa malaysia. (2012). towards boardroom excellence: corporate governance guide. 2nd edition. carter, d. a., d’souza, f., simkins, b. j., & simpson, w. g. (2010). the gender and ethnic diversity of us boards and board committees and firm financial performance. corporate governance: an international review, 18(1), 396–414. carter, d. a., simkins, b. j., & simpson, w. g. (2003). corporate governance, board diversity and firm value. the financial review, 38(1), 33–53. cheong, c. w. h., & sinnakkannu, j. (2014). ethnic diversity and firm financial performance: evidence from malaysia. journal of asia-pacific business, 15(1), 73–100. cook, a., & glass, c. (2009). when markets blink: us stock price responses to the appointment of minority leaders. ethnic and racial studies, 32(7), 1183–1202. darmadi, s. (2011). board diversity and firm performance: the indonesian evidence. corporate ownership and control, 8(1), 1–38. demeke, a. t. (2016). corporate governance mechanisms and firm performance: the case of the ethiopian insurance industry. journal of investment and management, 5(2), 6–16. eversheds report. (2013). the eversheds board report: the effective board. eversheds llp, office: one wood street, london ec2v 7ws. ezeanyim, e. e. e. e.. (2020). diversity management and firm performance in selected breweries in the south east of nigeria. international journal of management and entrepreneurship, 2(1), 1–12. fama, e. (1980). agency problems and theory of the firm. journal of political economy, 88(2), 288– 307. fama, e., & jensen, m. c. (1983). separation of ownership and control. journal of law and economics, 26(1), 1–32. ferreira, d. (2010). board diversity. in corporate governance: a synthesis of theory, research, and practice (pp. 1–228). francis, b., hasan, i., & wu, q. (2012). do corporate boards affect firm performance ? new evidence from the financial crisis. in bank of finland research discussion papers 11. grove, h., patelli, l., victoravich, l. m., & xu, p. t. (2011). corporate governance and performance in the wake of the financial crisis: evidence from us commercial banks. corporate governance: an international review, 19(5), 418–436. gul, s., muhammad, f., & rashid, a. (2017). corporate governance and corporate social responsibility: the case of a small, medium, and large firms. pakistan journal of commerce and social sciences (pjcss), 11(1), 1–34. hambrick, d. c., & mason, p. a. (1984). upper echelons: the organization as a reflection of its top managers. the academy of management review, 9(2), 1–15. hoechle, d. (2007). robust standard errors for panel regressions with cross-sectional dependence. the stata journal, 7(3), 281–312. huo, j. (2016). corporate board diversity: an empirical study of its effects on firm financial performance. master thesis. ilona, d. (2015). directors’ diversity, ownership concentration, and company performance in indonesian listed companies. ph.d. thesis. ismail, abdullah, s. n., & nachum, l. (2013). women on boards of directors of malaysian firms: impact on market and accounting performance. academy of management proceedings. http// dx.doi.org/10.2139/ssrn jensen, m. c., & meckling, w. h. (1976). theory of the firm: managerial behavior, agency costs, and ownership structure. journal of financial economics, 3(4), 305–360. journal of applied economics and business studies, volume. 4, issue 2 (2020) 205-220 https://doi.org/10.34260/jaebs.4210 219 jonson, e. p., mcguire, l., rasel, s., & cooper, b. (2020). older boards are better boards, so beware of diversity targets. journal of management and organization, 26(1), 15–28. kusumastuti, s., supatmi, & sastra, p. (2007). the impact of board diversity on firm value: corporate governance perspectives. journal of accounting and finance, 9(2), 88–98. letting’, n., aosa, e., & machuki, v. (2012). board diversity and performance of companies listed in the nairobi stock exchange. international journal of humanities and social science, 2(11), 172– 182. marimuthu, m. (2008). ethnic diversity on boards of directors and its implications on firm financial performance. the journal of international social research volume, 1(4), 431–445. marimuthu, m., & kolandaisamy, i. (2009). ethnic and gender diversity in boards of directors and their relevance to the financial performance of malaysian companies. journal of sustainable development, 2(3), 139–148. malaysian corporate code on governance, 1 (2012). https://www.sc.com.my/wpcontent/uploads/eng/html/cg/cg2012.pdf (last retrived on 23rd october 2016) muravyev, a. (2017). boards of directors in russian publicly traded companies in 1998–2014: structure, dynamics, and performance effects. economic systems, 41(1), 5–25. https://doi.org/10.1016/j.ecosys.2016.12.001 nathan, m. (2016). ethnic diversity and business performance: which firms? which cities? environment and planning a, 48(12), 2462–2483. nguyen, hagendorff, j., & eshraghi, a. (2015). does ceo cultural heritage affect performance under competitive pressure? available at https://www.newyorkfed.org/medialibrary/media/research/conference/2015/econ_culture/paper s_nguyenhagendorffeshraghi.pdf. 1–67. nguyen, & nielsen, k. m. (2014). what death can tell: are executives paid for their contributions to firm value? management science, 60(12), 2994–3010. ntim, c. g. (2015). board diversity and organizational valuation: unrevealling the effects of ethnicity and gender. journal of management & governance, 19(1), 167–195. oxelheim, l., & randoy, t. (2003). the impact of foreign membership on firm valuation. journal of banking and finance, 27(1), 2369-2392. pandey, d. l. (2020). workforce diversity and performance: an impact assessment. international research journal of marketing & economics, 7(4), 18–27. pfeffer, j., & salancik, g. r. (1978). the external control of organizations: a resource dependence perspective. rachagan, s., marshall, s., poon, w. c., & satkunasingam, e. (2015). board diversity: lessons from malaysia. international journal of corporate governance, 6(2), 1–19. rahman, h.u., ibrahim, m. y., & ahmad, a. c. (2015). the impact of soft regulations on boardroom diversity and shareholders’ confidence in malaysia. 2nd international symposium on technology management and emerging technologies, istmet 2015 proceeding. https://doi.org/10.1109/istmet.2015.7359056 rahman, haseeb ur, ibrahim, m. y., & che-ahmad, a. (2017a). corporate governance reforms and shareholders’ confidence in emerging markets: a case of malaysia. world journal of science, technology and sustainable development, 14(1), 60–74. rahman, haseeb ur, ibrahim, m. y., & che-ahmad, a. (2017b). physical characteristics of the chief executive officer and firm accounting and market-based performance. asian journal of accounting and governance, 8(1), 27–37. rahman, haseeb ur, ibrahim, m. y., & che-ahmad, a. (2017c). physical characteristics of the chief executive officer and firm accounting and market-based performance. asian journal of haseeb ur rahman, muhammad zahid & muhammad jehangir 220 accounting and governance, 8(accepted), upcoming. rahman, haseeb ur, ibrahim, m. y., & che-ahmad, a. (2016). accounting profitability and firm market valuation: a panel data analysis. international conference on leadership and management at kuala lumpur, convention centre, kuala lumpur, malaysia., 1–6. rahman, haseeb ur, ibrahim, m. y., & che ahmad, a. (2015). corporate governance, firm financial performance, and shareholders’ confidence: a proposed analysis of mccg 2012. an international journal global business & management research an international journal, 7(1), 139–147. rahman, haseeb ur, ibrahim, m. y., & zahid, m. (2014). boardroom diversity and firms’ financial performance: a proposed pre and post-analysis of mccg 2012. international symposium on research in innovation and sustainability 2014 (isoris ’14) 15-16 october 2014, malacca, malaysia., 1471–1475. rahman, haseeb ur, rehman, s., & zahid, m. (2018). the impact of boardroom national diversity on firms’ performance and boards’ monitoring in emerging markets: a case of malaysia. city university research journal, 18(1), 1–15. rahman, haseeb ur, zahid, m., & naveed. (2018). does the ‘ business case ’ for academic directors on corporate board stand up ? journal of managerial sciences, xii(3), 190–202. ramirez, s. a. (2003). a flaw in the sarbanes-oxley reform: can diversity in the boardroom quell corporate corruption? st. john’s l. review, 77(2003), 837–866. randøy, t., thomsen, s., & oxelheim, l. (2006). a nordic perspective on corporate board diversity. nordic innovation centre. in agder university college (norway), copenhagen business school (denmark) and lund institute of economic research (sweden). http://www.nordicinnovation.org/global/_publications/reports/2006/the performance effects of board diversity in nordic firms.pdf (last retrieved on 20th august 2016) rutledge, r. w., karim, k. e., & lu, s. (2016). the effects of board independence and ceo duality on firm performance: evidence from the nasdaq-100 index with controls for endogeneity. journal of applied business and economics, 18(2), 49–71. salloum, c., jabbour, g., & mercier-suissa, c. (2017). democracy across gender diversity and ethnicity of middle eastern smes: how does performance differ? journal of small business management, 3(2), 1–13. smith-meyer, a. (2013). gender diversity in the boardroom. available at http://www.femaleboardpool.eu/201301fdp.pdf. vania, h., & supatmi. (2014). the effect of board diversity on the company value of financial institutions in indonesia. international journal of business and management invention, 3(4), 32– 41. wang, & clift, b. (2009). is there a “business case” for board diversity? pacific accounting review, 21(2), 88–103. zahid, m., rehman, h. u., ali, w., khan, m., majed alharthi, qureshi, m. i., & jan, a. (2019). boardroom gender diversity: implications for corporate sustainability disclosures in malaysia. journal of cleaner production, 244, 1–24. https://doi.org/10.1016/j.jclepro.2019.118683 journal of applied economics and business studies, volume 1, issue 1 (2017) 65-76 https://doi.org/10.34260/jaebs.117 65 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online): 2663-693x understanding internal migration in pakistan naila safi1*, zahoor ul haq2 & javed iqbal31 1 bs student, department of economics, abdul wali khan university, mardan. 2 professor, department of economics, abdul wali khan university, mardan. 3 assistant professor, department of economics, abdul wali khan university, mardan abstract this study estimated the effect of socio-economic factors on internal migration for pakistan using the labor force survey data, 2013. this study used a sample of 26013 observations. both bi-variate and multivariate logistic and probit analysis were performed. estimates of logistic and probit regressions show that gender has negative and statistically significant effect on migration. similarly, marriage also has a negative effect on migration indicating as individual gets married, their log-likelihood of migration decreases. income is statistically significant determinant of migration as analysis exhibited that as income increased, the log-likelihood of internal migration increased too. education and employment do not determine migration. this study can be used in planning internal migration as for developing countries like pakistan socioeconomic factors are important determinants of migration. providing employment and income opportunities can significantly help in reducing the migration. besides socioeconomic factors, further investigation is required to better understand the perception of individuals with respect to internal migration which this study could not tackle because of nonexistence of such data in the labour force survey. keywords migration; labour force; socio-economic factors jel classification j15; o15; r2 copyright © 2017 saebr all rights reserved 1. introduction migration is broadly defined as the relocation of residence for a specified duration due to various reasons (hossain, 2001). structural transformation in an economy such as a relatively higher growth of one sector as compared to others creates more employment opportunities in higher growth sectors as compared to the low growth sectors. hence, employment opportunities in higher growing sectors attracts more workers, especially from the low growth sectors, where labourers are laid-off due to low demand. this structural transformation leads to migration as labourers move from one region to another following employment opportunities. structural transformation has also been witnessed in pakistan. 1*corresponding author: nyla.safi@gmail.com naila safi, zahoor ul haq & javed iqbal 66 over the past few decades, services sector growth has surpassed the commodity producing sectors of the country (figure-1). traditionally, the agricultural sector in pakistan used to be the highest contributor to economic growth. the combined growth of both agricultural and industrial sectors has been lower compared to services sectors. resultantly, workers have been migrating from commodity producing sectors, especially agriculture which has demonstrated lower growth than industrial sector, to services sectors. employment opportunities in services sectors primarily exist in urban areas, and therefore workers tend to migrate from rural areas to urban areas for good living. migration is generally of internal and external types. internal migration means that people migrate between the states/villages/districts within a country and international migration means the movement of people from one country to another (usman et al., 2009). internal migration is the old phenomena as observed in many developing countries including pakistan. the trends and nature of migration changes with time and its impact is different on each migrant’s life (hamid, 2010). the reason behind the rural and urban migration is to avail better socio-economic opportunity. imran et al. (2013) explain that one third of the rural population lives below the poverty line which force people to migrate in search for better opportunities. that is why the estimated population in urban areas has been witnessing an increase from 37.9 to 40.54 percent while the population in rural areas is decreasing from 62.1 percent to 59.46 percent. according to neo-classical theory, the decision of migration depends upon the wage differentials and they migrate only in search of the better economic opportunities. todaro (1969) and the harris and todaro (1970) reported that the differences between urban-rural wages/incomes as the primary determinant for migration. lewis (1954) further explain that people migrate from rural agricultural sector to urban industrial sector because agriculture sector has unlimited supply of labour while industrial sector has higher wages. human capital theory states that migration is associated with the costs and returns. figure 1: annual agriculture, industry and services sector’s growth, 2013 to 2018 (percent) 0 1 2 3 4 5 6 7 2013 2014 2015 2016 2017 2018 g ro w th r a te ( % ) agriculture industry agriculture & industry services journal of applied economics and business studies, volume 1, issue 1 (2017) 69-80 https://doi.org/10.34260/jaebs.116 67 according to the theories discussed above and the general behaviour of the economies show that migration take place between different sectors of economy as a result of economic growth and pakistan is no exception. the present study estimates the effect of socioeconomic factors on internal migration in pakistan. studying internal migration helps in urban planning. urbanization is an ongoing phenomenon and investigating internal migration helps in the development of the labour and social policies and infrastructures to meet their residential and living requirements. the migration of workers may also create many social problems and urban issues for example slum areas, urban sprawl, congestion of traffic, urban poverty and therefore it is important to plan for these ahead of time. hence, it is important to understand migration decisions and their implications for urbanization and other socio-economic aspects of life. 2. literature review in recent years, analysis of labour migration in pakistan have got an attention of many researchers. ahmed and sirageldin (1993) analyzed internal migration in pakistan using the investment in human capital paradigm. they concluded that person having college or professional degree are more likely to migrate. probit model has been used in estimating the effect of husband’s age, schooling years of husband, schooling years of wife, occupational groupings, dummy if husband is entrepreneur, dummy representing ownership of the house and land, and number of children. it was concluded that person having college or professional degree are more likely to migrate. the professional and skilled worker have high migration rate. farooq et al. (2005) also investigated the factors of rural-urban migration in faisalabad metropolitan. using the probit model it is concluded that land holding is the man economic opportunity in rural areas of pakistan, landlessness and total land scarcity is positive factor of migration from rural to urban areas. khan et al. (2000) also analyzed the determinants of internal migration using the labour force survey of 1997. they categorized the sample of the survey into economic and non-economic migrants. age, education, type of training (technical and vocational), marital status, location of residence (urban or rural), family type (nuclear or joint) and province of living were the selected variables. memon (2005) also used labour force survey and employed a probit model to estimate the effect of different factors on internal migration in pakistan. the study concludes that income differences between the rural and urban areas are increasing internal migration. the study finds that ownership of land reduces migration as people are more attached to their land and don’t want to migrate. such families usually have women bread earners as well to support their families and avoid migration. in bangladesh, similar nature of the study is conducted by hossain (2001). a micro-level data were collected, and the study concludes that seeking employment opportunities is the primary reason for migration. he concluded that half of the migrant migrate for temporary jobs and one quarter migrate for availing naila safi, zahoor ul haq & javed iqbal 68 permanent jobs. further, it is found that the percentage of migration is high among educated and unemployed. similar researches has been conducted by siddiqi (2004), oda (2005), hamid (2010), haider and kabir (2010), throat et al. (2011), ali et al. (2015), fareed et al. (2016), moses et al. (2017), tripathi and kaur (2017), park and fullerton (1980), ullah (2004), khan et al. (2011), mohammad et al. (1983) and irfan (1986). the main determinants suggested by these previous studies are education level, peace, income, land ownership, profession and gender. 3. research methods this study uses labour force survey (lfs) for the year 2013 consisting of 26,013 observations. this survey provides us information about active and inactive labour force. bi-variate analysis has been carried using chi-squared test statistics. we test the hypothesis that the observed and expected values of the observations are different. the chi-squared test statistics is given in equation 1 is given as follows: 𝜒 = ∑ (𝑥𝑖 − 𝑚𝑖 ) 2 𝑚𝑖 = 𝑘 𝑖=1 ∑ (𝑥𝑖 ) 2 𝑚𝑖 − 𝑛 𝑘 𝑖=1 (1) where 𝜒 represents the chi-squared test statistics, n the number of observations, k mutually exclusive classes, 𝑥𝑖 represent the observed numbers for 𝑖 = 1, 2, 3, … . , 𝑘 and 𝑚𝑖 = 𝑛𝑝𝑖 for all i, where 𝑝𝑖is the probability that an observation occurs in the ith class. the econometric analysis consisted of estimation of the logit and probit models. for the logit model, the 𝐹(𝑋′𝛽) is the cumulative density function of the logistic distribution and the predicted probabilities are limited between the 0 and 1 interval and is given as follows: 𝐹(𝑋′𝛽) = 𝑒 𝑋′𝛽 1+𝑒𝑋 ′𝛽 = exp(𝑋′𝛽) 1+exp(𝑋′𝛽) (2) the matrix of 𝑋′ consists of the exogeneous variables defined in table 1. for probit model case, the 𝐹(𝑋′𝛽) is the cumulative density function of the standard normal distribution (𝜙). 𝐹(𝑋′𝛽) = φ(𝑋′𝛽) = ∫ 𝜙(𝑧)𝑑𝑧 𝑋′𝛽 −∞ (3) the dependent variable used in the analysis is internal migration which is a dummy variable where migrated households are represented as one and zero otherwise. the study uses gender, education, literacy level, income, marital status as the explanatory variables. the detail about the variables used in this study is provided in table 1 below: journal of applied economics and business studies, volume 1, issue 1 (2017) 69-80 https://doi.org/10.34260/jaebs.116 69 table 1: variables used in the analysis variable description gender one if the respondent is male, zero otherwise literacy level one if respondent is literate, zero otherwise income annual income of the respondent in rupees marital status one if never married, two if married, three if widow/ widower, four if divorced employment status one if the respondent is employed, two if unemployed, three if not in labor force graduate one if respondent is graduate and zero otherwise 4. results and discussion the first part presents bi-variate analysis while the second part offer multivariate analysis primarily consisting of econometric analysis. collectively these two parts helps in understanding the internal migration phenomena in the country. 4.1 bivariate analysis this section consists of the cross-tabulation of the socio-economic factors across the migrated and non-migrated individuals indicating variations in migration pattern based on the socio-economic factors. the socio-economic factors include education, gender, literacy level, income, marital status and employment status. we aim to test the null hypothesis that socio-economic factors do not determine internal migration of respondents. data shows that female migrated more than males. table 2 shows that 12.4 percent of the male and 16 percent of the female reported migration. riley and gardner (1993) examined the role of gender in migration decisions in developing countries and found that 45 percent of female and 65 percent of male migrated. however, men migrated mostly out of the country while female mostly migrated within country. chi-squared statistics show that gender is a statistically significant determinant of migration. migration status of respondents by marital status shows that 76.3 percent of the migrated persons are married, and 21.4 percent are not married. marriages lead to more migration as married people may relatively face more financial burden due to larger households and feeling more responsibility. kanwal et al. (2015) studied the socio-economic determinants of rural-urban migration in pakistan and concluded that marriages play a significant role in movement from rural to urban rather than urban to rural. generally, females after marriages move with their spouse from rural to urban and marriages in this case are the core cause for migration in pakistan. hamid (2010) also investigated the role of gender in internal migration. his findings show that marriage is the key factor for migration in the case of female. our analysis also show that marriage has statistically significant association with migration. the variable literacy is binary in nature that is individual are either literate or not. the table shows that 69.3 percent of the migrated are literate. hence, literacy has a direct impact on migration. ahmed and sirageldin (1993) analyzed internal migration in pakistan using the investment in human capital paradigm. their results show that educated people have more opportunities as compared to illiterate and hence they migrate to avail such naila safi, zahoor ul haq & javed iqbal 70 opportunities. however, literacy does not have statistically significant association with migration. literacy can be further classified into different educational level. data shows that as educational level of respondents increases, their migration increases. migration within individuals having post-graduate education is the highest, 17.8 percent, which reduces to 14.9 percent among graduates and 12.6 percent among those having higher secondary school qualification. about 87.5 percent of the illiterate do not migrate. according to the world bank (2017), education is not considered as the only factor for migration. educated people migrate to avail livelihood opportunities that are more secure in nature. chi-squared statistics show that education is a statistically significant determinant of migration. this result is different than literacy, which does not determine migration. hence, it is not literacy rather education level that is important for migration. table 2: bi-variate analysis of the association between socio-economic factors and migration decisions characteristics migrated non-migrated chi-square male 12.4 % 87.6 % 30.27 (0.00) female 16.0% 84.0 % un married 21.4 34.8 236.7 (0.00) married 76.3 63.5 widow 2.0 1.4 divorced 0.3 0.3 illetrate 30.7 31.70 1.20 (0.27) literate 69.30 68.30 no literacy 12.5 87.5 51.27 (0.00) less than ssc 11.7 88.3 ssc 12.8 87.2 hssc 12.6 87.4 graduate 14.9 85.1 post graduate 17.8 82.2 employed 12.8 87.2 2.68 (0.00) un employed 11.5 88.5 not in labour force 21.90 78.1 income group 1 (less than rs. 50,000) 9.7 12.9 186 (0.00) income group 2 (rs. 50,000-2,00000) 60.3 64.9 income group 3 (rs.2,00001-5,00000) 22.4 18.8 income group 4 (rs 5,00001-8,00000) 5.3 2.5 income group 5 (above rs. 8,00000) 2.3 0.9 non-economic reasons 83.30 81.30 0.87 (0.00) economic reasons 16.70 18.70 head of household 57.2 46.2 1032.59 (0.00) spouse 7.9 4.3 son/daughter(unmarried) 13.9 28.9 son/daughter (married) 8.2 11.8 father/mother o.8 0.5 brother/sister 4.0 5.5 other relative 3.2 2.3 servant 3.2 0.4 non relative 1.8 0.1 source: author's own estimations journal of applied economics and business studies, volume 1, issue 1 (2017) 69-80 https://doi.org/10.34260/jaebs.116 71 employment status shows that 12.8 percent of the respondents who are employed have migrated. about 88.5 percent of the unemployed did not migrated. hence, livelihood and employment are important factors in the migration decision. jamil and mohyuddini (2015) analyzed the rural urban migration among educated people. they concluded that trend is changing, and migration is not motivated due to employment rather it is both employment and education opportunities that determine migration. annual income of the respondents is categorized into five groups. data suggest that the proportion of migrated respondents decreases as income level increases. the table shows that 12.5 percent of the respondents are in group-1, 64.3 percent in group-2, 19.3 percent in group-3, 2.8 percent in group-4 and 1.1 percent in group-5. the proportion of respondents not-migrated within an income group decreases as income rises. about 90.1 percent of the respondents did not migrate in group-1, followed by 88 percent in group-2, 85.1 percent in group-3, 76 percent in group-4 and 73 percent in group-5. third, the middle-income groups that is income groups two and three, have the highest proportion of all respondents migrated. collectively groups 2 and 3 account for 83 percent of the migrated respondents (that is 10.6 percent of the total 12.8 percent). only 5.3 and 2.3 percent of the migrated belong to groups 4 and 5, respectively. hence, the middle-income groups look for earning opportunities and migrate to avail these as compared to high income group (that is groups 4 and 5) who already have high annual income and have less motivation to migrate. farah (2001) found that the socio-economic factors like income has statistically significant effect on the decision to migrate. memon (2005) used a probit model to study the effect of socio-economic variables including income on internal migration. he found that urban-rural income differences have increased internal migration among respondents. these effects could have also been picked by the chi-squared statistics, which show a statistically significant association between income and migration. among the migrated respondents, 82.1 percent migrated for non-economic and the remaining for economic reasons. if a migrant identified job transfer, finding a job, education or business as the main reason for migration, then such migration is based on reasons which can be classified as economic reasons and the migrants as economic migrants. other are included in non-economic reasons (khan et al. 2000). there is significant relation between the reason for migration and migration decision. the study conducted by khan et al. (2011) reveals that mostly people migrate from rural to urban area for socio-economic factors. male migrated for economic reason than female. however according to distance such as short and medium in migration, social factors like marriage and migration with household are the main reasons for migration. naila safi, zahoor ul haq & javed iqbal 72 the head of household who migrated were 57.2% where spouse who has not migrated were 78.7%. the son/daughter who were unmarried have not migrated were 93.4% compared to son/daughter who were married and have not migrated were 90.8%. father/mother who have not migrated were 82% and brother/sister who have not migrated were 90.5%. nonrelative who have migrated were 70.2%. servants who decide to migrate were 57.2%. table shows that nonrelatives and servant have migrated mostly. 4.2 multivariate analysis bi-variate analysis does not control for the effect of other variables when association between migration and an individual socio-economic factor is studied. however, it is quite possible that migration could have been influenced by more than one variable at the same time and thus the bi-variate analysis could be misleading. hence, this section presents multivariate analysis to study the effect of a socio-economic variable while controlling for others. internal migration is used as the dependent variable where migrated households are represented as one and zero otherwise. dependent variables with dichotomous response can be estimated through probit and logit estimation techniques. this study uses both the techniques to study the effect of socio-economic factors on internal migration. the socioeconomic variables based on the previous section, includes both qualitative and quantitative variables. the qualitative variables include gender, literacy, marital status, employment status and education level while the quantitative variables include income level. the estimation of both the probit and logit models is provided in table 3 and its explanations are respectively explained in sections 4.2.1 and 4.2.2. 4.2.1 probit estimates many studies including farooq et al. (2005), memon (2005) and akins et al. (1979) estimated the effect of socio-economic factors on internal migration using probit model. table 3 shows the result of probit model with and without controlling for heteroscedasticity. the analysis uses 26,013 observations. the model on overall is statistically significant as indicated by wald-chi statistics. it also shows that majority of the variables are statistically significant and determine internal migration. gender has negative effect on internal migration implying that as compared to men, women migrate less. these results are in-line with kanwal et al. (2015). similarly, unmarried individuals as compared to married have a higher probability of migration. employment and literacy status do not determine internal migration. marr and millered (1987) also found insignificant effect of employment on migration. income has a positive and statistically significant effect on internal migration. hence, an increase in income increases the log-likelihood of internal migration. similar results were also reported by kennan and walker (2011). journal of applied economics and business studies, volume 1, issue 1 (2017) 69-80 https://doi.org/10.34260/jaebs.116 73 4.2.2 logistic estimates kanwal et al. (2015) and hogan and steinnes (1998) used logistic models to estimate the effect of socio-economic factors on internal migration in pakistan. table 3 show results of logistic regression with and without correction for heteroscedasticity. all the models are statistically significant and produced good pseudo r-squared. it is important to mention that primary data, like the one used in this study, usually produce low r-squared. these results are like the results accomplished with probit analysis. the direction of effect and test of the hypotheses (i.e. statistically significance) remained the same. hence, the results are robust to the estimation technique. table 3: probit and logit estimates probit model probit model correcting for heteroscadasticity logit model logistic model correcting for heteroscadasticity parameters co. efficient co. efficient co. efficient co. efficient gender -0.21 (0.00) -0.22 (0.00) -0.41 (0.00) -0.40 (0.00) marital status -0.35 (0.00) -0.29 (0.00) -0.70 (0.00) -0.58 (0.00) employment status 0.10 (0.40) 0.02 (0.77) 0.17 (0.45) 0.04 (0.80) literacy level 0.04 (0.13) -0.007 (0.75) 0.08 (0.11) -0.01 (0.80) grade -0.07 (0.06) -0.05 (0.09) -0.13 (0.07) -0.09 (0.11) income 0.86 (0.00) 0.82 (0.00) 1.50 (0.00) 1.43 (0.00) cons -1.11 (0.00) -1.01 (0.00) -1.84 (0.00) -1.67 (0.00) summary statistics number of observations 26013 wald chi 343.73 pseudo r2 0.025 wald chi 434.43 pseudo r2 0.021 wald chi 338.36 pseudo r2 0.025 wald chi 432.44 pseudo r2 0.02 source: author's own estimation 5. conclusion internal migration is the phenomenon of temporary relocation of residence due to various reasons. structural transformation in an economy such as a relatively higher growth of one sector as compared to another sector creates more employment opportunities in the high growth sectors as compared to the low growth sectors. hence, the employment opportunities in high growing sectors attracts more workers, especially from the low growth sectors, where laborers are laid-off due to low demand. this structural transformation leads to migration as laborers mover from one region to another following employment naila safi, zahoor ul haq & javed iqbal 74 opportunities. this study estimates the effect of socio-economic factors on internal migration in pakistan using labor force survey (lfs) for the year 2012-13. the study uses a sample of 26013 observations. income is the main factor behind the migration. estimates of logistic regression show that gender has negative and statistically significant effect on migration. similarly, marriage has a negative effect on migration, hence as individual gets married, their log-likelihood of migration decreases. income is statistically significant determinant of migration our analysis shows that as income increases the log-likelihood of internal migration increases. education and employment do not determine migration. we conclude that it is important to understand socioeconomic conditions of an individual to understand migration in the country. this study is useful in many ways. it can be used in planning internal migration. in developing countries like pakistan, socio-economic factors are important determinants of migration and providing livelihood opportunities in rural areas can significantly reduce migration to urban areas. providing employment and income opportunities can significantly help in reducing the migration. further investigation is required to better understand the perception of individuals with respect to internal migration, a data that is missing in the existing lfs. references ahmed, a.m. and sirageldin, i. (1993). socio-economic determinants of labour mobility in pakistan. the pakistan development review, pp.139-157. akin, j.s., guilkey, d.k and sickles, r.. (1979). a random coefficient probit model with an application to a study of migration. journal of econometrics, 11(2-3), pp.233-246. ali, h., shafi, m.m., rehman, m.u and jadoon., m.a. (2015). causes and effects of ruralurban migration in rural areas of khyber pakhtunkhwa-pakistan. arts and social sciences journal. 10.4172/2151-6200.1000144. farah, n. (2001). socio-economic & cultural factors affecting migration behavior (a case study of faisalabad [pakistan]. unpublished thesis m.sc., rural sociology, univ. of agri., faisalabad, pakistan. fareed, g., faridi, m.z and ayyoub, m., (2016). socioeconomic determinants of ruralurban migration: evidence from punjab (pakistan). new horizons, 10(2), p.89. farooq, m., mateen, a. and cheema, m.a. (2005). determinants of migration in punjab, pakistan: a case study of faisalabad metropolitan. journal of agriculture and social sciences (pakistan). haider, u and kabir, s. (2010). factors of migration on urban bangladesh: an empirical study of poor migrants in rajshahi city. pakistan journal of social sciences (pjss), 30(2). hamid, s. (2010). rural to urban migration in pakistan: the gender perspective. working papers & research reports, 2010. harris j. and m. todaro (1970) migration, unemployment and development: a two-sector analysis, american economic review, 60, 126-142. journal of applied economics and business studies, volume 1, issue 1 (2017) 69-80 https://doi.org/10.34260/jaebs.116 75 hogan, t.d and steinnes, d.n. (1998). a logistic model of the seasonal migration decision for elderly households in arizona and minnesota. the gerontologist, 38(2), pp.152158. hossain, m.z. (2001). rural-urban migration in bangladesh: a micro-level study. in brazil iussp conference. august (pp. 20-24). imran, f., nawaz, y., asim, m. and hashmi, a.h. (2013). socio-economic determinants of rural migrants in urban setting: a study conducted at city sargodha, pakistan. academic journal of interdisciplinary studies, 2(1), p.71. irfan, m. (1986). migration and development in pakistan: some selected issues. the pakistan development review, 25(4), pp.743-755. jamil, a and mohyuddin, a. (2015). rural urban migration for education. (a case study of district bahawalpur, pakistan). sci int (lahore), 27, pp.4819-24. kanwal, h., naveed, t.a and khan, m.a. (2015). socio-economic determinants of ruralurban migration in pakistan. stud, 4(3) kennan, j and walker, j.r. (2011). the effect of expected income on individual migration decisions. econometrica, 79(1), pp.211-251. khan, a.h., shehnaz, l. and ahmed, a.m. (2000). determinants of internal migration in pakistan: evidence from the labour force survey, 1996-97 [with comments]. the pakistan development review, pp.695-712. khan, h., hassan.t and shamshad. (2011). socio-economic causes of rural to urban migration in india. asia-pacific journal of social science, (3), pp.138-158. lewis w. (1954) economic development with unlimited supplies of labour, manchester school of economics and social studies, 22, 139-91. marr, w.l and millerd, f.w. (1987). migration and the employment status of married women. school of business and economics, wilfrid laurier university. memon, r. (2005). determinants and trends of internal migration in pakistan. internal labour migration in pakistan. institute of developing economies, japan external trade organization, chiba, japan. mohammad, i., demery, l., ghulam mohammad, a., farooqui, f.a., tariq, j., haq, r and khan, g. (1983). migration patterns in pakistan: preliminary results from the plm survey, 1979. moses, l.a.b., guogping, x and john, l.c.l. (2017). august. causes and consequences of rural-urban migration: the case of juba metropolitan, republic of south sudan. in iop conference series: earth and environmental science (vol. 81, no. 1, p. 012130). iop publishing. oda, h. (2005). labour migration from rural pakistan: evidence from village in rain-fed area of punjab. internal labour migration in pakistan. asedp, (72). park, h.y and fullerton, h.h. (1980). rural-urban labour migration: the case of korea. the annals of regional science, 14(1), pp.72-90. riley, n.e. and gardner, r.w. (1993). migration decisions: the role of gender. sidddiqi, w.m. (2004). rural-urban migration: a case study of lahore district. pakistan research repository. university of the punjab, p.1. naila safi, zahoor ul haq & javed iqbal 76 todaro, m. (1969) a model of labour migration and urban unemployment in less developed countries, american economic review, 59, 138-48. thorat, v.a., dhekale, j.s., patil, h.k. and tilekar, s.n. (2011). determinants of ruralurban migration in konkan region of maharashtra. agricultural economics research review, 24, pp.503-509. tripathi, s and kaur, h. (2017). determinants of rural to urban migration in large agglomerations in india: an empirical analysis. ullah, a.a. (2004). bright city lights and slums of dhaka city: determinants of ruralurban migration in bangladesh. migration letters, 1(1), pp.26-41. usman, m.n., naeem, m and khan, z. (2009). socio-economic determinants of migration of people from rural to urban areas of district peshawar. journal of management sciences, 4(2). world bank. (2017). forcibly displaced: toward a development approach supporting refugees, the internally displaced, and their hosts. world bank publications. journal of applied economics and business studies, volume. 6, issue 2 (2022) 55-76 https://doi.org/10.34260/jaebs.624 55 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x good governance: reforming civil service of pakistan abdul wajid rana*1 1 the author is retired federal secretary finance and former member of federal public service commission abstract an effective and proficient civil service is cornerstone for good governance in the country.. civil service reforms not only include recruitment and training but also involves deliberate changes to the structures and processes of public sector organizations with the purpose of good governance in political realm, economic, legislative, judicial and humans’ governance. it is a continuous process with no general starting point, and equally no end. the reform package begins with redefining the role of the state including an objective assessment of current political needs, capability and effectiveness of the state from where must flow institutional and civil service reforms rather than isolated arbitrary decisions. moreover, their recommendations usually covered short-term priorities without addressing key issues faced by the civil services. a gradual holistic restructuring of the existing bureaucratic institutions is recommended which is well integrated and time bound in all tiers of government. specific reforms pertaining to realignment of public service commission, human resource management, civil services hiring procedure, civil officers’ training and performance appraisal system, promotion mechanism, civil servants’ compensation and accountability have been stipulated. keywords civil service, governance jel classification p11, g18 1* a.w.rana@cgiar.org mailto:a.w.rana@cgiar.org abdul wajid rana 56 1. introduction a major challenge faced by developing as well as developed countries is of effective governance. today, governments are facing complex internal and issues ranging from conflicts, financial volatility, social tensions, climate change, infrastructure gaps, negative demographic trends and elite capture of the state. the growing criticism on state institutions’ on their public delivery has resulted into a renewed intertest on reforming the civil service delivery and structure. figure 1: governance indicators 20202 key challenges pakistan has been enduring for more than five decades include: (i) security and regional geo-politics; (ii) political economy factors including civil-military relationship; (iii) weak policy formulation process; (iv) declining quality of public administration system; (v) poorly designed and hastily implemented devolution and fiscal decentralization (2001 and 2010); (vi) dominance of state in the market; (vii) weak economic governance with brief periods of stability; and (viii) entrenchment of vested interests and elite capturing. ensuring political settlements for short-term patronage-based gains, far too long, have undermined effectiveness of the state. today, pakistan ranks in the lowest percentile in most governance indicators (see figure 1). concomitantly, public policy and governance dynamics have undergone major restructuring post-18th constitutional amendment in 2010, transforming pakistan from a layered-cake hierarchical dual federalism to collaborative federalism leading to collaborative and participative governance. article 90 and 129 of the constitution, as amended in 2010, have added new complexities to both public policy formulation and public administration at federal and provincial levels. globalization and advances in communication technology has incited additional challenges. efforts to dismantle the excessive state under the influence of “washington 2 world bank group governance indicators 2020 journal of applied economics and business studies, volume. 6, issue 2 (2022) 55-76 https://doi.org/10.34260/jaebs.624 57 consensus” aiming at removing distortions and fiscal consolidation3 followed by liberalization, privatization, deregulation in 1990s and 2000s and rapid technological advancement required different set of skills, expertise, values, attitude and knowledge in public officials. the ubiquitous use of internet and social media has raised the level of expectations in terms of speed, quality and personalization of public service delivery. with a more educated as well as vocal citizenry savvy in the use of social media, an effervescent and highly vigilant media, vibrant judiciary, active legislative committees and political agenda for change and transformation, public servants are finding themselves under ever keener and intense public scrutiny and expectations. today, there is a broad agreement among policy makers throughout the world that a lean, efficient and professional civil service is a necessary, though not sufficient, requirement for effective governance in any country. the governments around the world are pushing the ideas of “hallow state”4, “lean, efficient and facilitative government”, “paperless government”, “practical government” and decentralization. to coup up with the challenges of 21st century, many countries have reformed their civil service system, such as uk, canada, australia, south africa, brazil, and many countries are on the path of reforming the system to ensure professionalism and efficiency of public administration. some countries have adopted the concept of “incorporation” to foster close public-private sector collaboration and transforming the civil service that promotes competitiveness of the private sector. to the contrary, pakistan remained a straggler despite establishment of national commission for government reforms in 2006 and huge investment of time and resources for securing a pragmatic as well as implementable governance reform plan. in the interregnum, size of the national and sub-national governments in pakistan kept swelling, especially at the lower end of bureaucracy, colonial laws, rules and procedures continued weakening critical areas of governance further, and political polarization and wide-ranging bashing of the bureaucracy triggered impasses and inertia. this paper will take stock of reform efforts since 1947, contemporary approaches for public sector reform, and addressing key elements of civil service reform (csr). the world is now acknowledging that csr remans a continuous process with no particular starting point and similarly no end. csr cannot be performed in isolation, and they should be undertaken with political will and inclusivity for positive and effective results. these reforms should have a focus on curtailing the culture of neo-patrimonialism 3 nadeem-ul-haq, “why civil service reforms do not work”, ssrn electronic journal, february 2007. 4 hallow state where government reduces its role as a direct supplier of public goods by contracting public service delivery out to private organizations abdul wajid rana 58 and kinship politics which hinders modernism. any reform towards bringing an effective change will fail if the institutions of the state broker their authority for the following reasons: (i) power imbalance between the very strong, rigid, and over-centralized bureaucratic institutions and weak democratic institutions; (ii) challenges thrown towards popular sovereignty by non-state actors performing many of state functions; (iii) institutionalization of elaborate system of patronage and clientele making the civil bureaucracy dysfunctional; (iv) bloating federal government despite devolution under 18th constitutional amendment, manpower of the federal government has doubled since 2010 (federal employees census); (v) closed structure of the civil service; (vi) ever expanding employment at low-skills levels with uneven management structure, particularly at sub-national level; (vii) an inappropriate skill-mix and archaic training structure; (viii) weak inter-ministerial or inter-departmental harmonization; (ix) archaic laws, rules and procedures that restrict innovation, knowledge and agile concepts; (x) elite-capturing and endemic rent-seeking making the institutions irrelevant; (xi) ineffective incentive system that rewards the incompetent; (xii) political interference and administrative acquiescence leading to systemic inconsistencies and arbitrariness in decision-making; (xiii) lack of transparency and weak internal accountability; (xiv) increasing public sector wage costs crowding out operational expenditures; (xv) politically oriented bulge in development expenditure squeezing the recurrent budget further; and (xvi) resistance to change and reform. 2. background of reforming the governance and civil services in pakistan 2.1 efforts to reform governance and civil service in pakistan the government has constituted around 40 commissions and committees since 1947 to reorganize and reform the civil service prior to recently constituted task force on institutional reform and austerity and civil service reform, most with flawed composition, and imperfect terms of reference. generally, the proposed reforms with some exception were loosely structured and its primary focus had always been the higher civil service (the central superior services), particularly one cadre, without addressing the issues relating to federal and provincial government structures and lower levels, which journal of applied economics and business studies, volume. 6, issue 2 (2022) 55-76 https://doi.org/10.34260/jaebs.624 59 is 95 percent of the bureaucracy. induction of private sector as well as “expatriate” favourites at disproportionate compensation under the garb of creating a healthy mix of skill-set not only demoralizes the bureaucracy but at times, making the institutions dysfunctional. most commissions and task forces considered civil service reform a technical exercise--where problems are reduced to boxes and then solutions are found to fit into the boxes instead of aligning them to pre-determined politically owned objectives and goals. due to absence of precision in objectives and goals, general direction, a realistic assessment of political as well as cultural contexts, an analysis of existing operating environment and rustic processes and procedures, the effectiveness of these exercises is lost after initial steam. the deliberations during these meetings generally remains perception based or anecdotal rather than an evidence-based analysis. the recommendations are largely motivated by short-term priorities as well as political compromises instead of addressing key issues facing the civil service including compensation that has led to consistent decline in pakistan’s ranking in government effectiveness (see figure 2). figure 2: government effectiveness (wgi) for these reasons, quite often the reports of these commissions/committees were either not made public or shelved without quite often the reports of these commissions/committees were either not made public or shelved without seeking formal approval or implementation5. those implemented, ended up fragmenting the public sector and civil service further rather than bringing any qualitative improvement in governance. the present exercise is no different. aspirations for good governance will not succeed if the requisite conditions for good governance are not there. reforms should not only focus on clearly defined objectives 5 report of national commission for government reforms abdul wajid rana 60 and goals to be accomplished but also on creating conditions necessary to sustain effective governance. it mandates reforming all areas of governance including economic policies which lead to inclusive growth and greater equality, civil service and public administration, corporate governance, regulatory bodies and service delivery, and governance institutions: political, executive, legal and legislative for establishing a transparent and accountable system. similarly, the objective of good governance may not be realized in a political system which is too weak and vulnerable to frequent gridlocks. strong system of checks and balances helps in circumventing state capture by interest groups as there are ample incentives in a country to subvert good governance resulting in efforts at rent-seeking and state capture. strengthening institutions (executive, political, judicial, and markets) could fortify trust in the state, create opportunities for citizen participation, better access to service delivery, and increased investment, foreign and domestic, increase the role of the private sector and alleviate the government’s burden. 2.2 redefining the role of the state and political needs assessment the state has been a dominant player in pakistan since 1960s playing a considerable role ranging from national security and economic development to all areas of service delivery that expanded the public sector disproportionately with weakening institutions germinating the state-capturing. synchronously, the private sector has continuously relied on subsidies, concessions, and rent-seeking. therefore, the starting point for designing reform package is to redefine the role of the state and an objective assessment of current political needs, capability and effectiveness of the state from where must flow institutional and civil service reforms rather than isolated arbitrary decisions. secondly, the task force must recognize that both the governments and the markets are institutions that interact continuously for nibbling their niches and resource allocation, thus, the rules governing the institution are critical, particularly the rules determining how decisions are made within it and how these are aggregated as institutions consist of multiple individuals with differences in preferences and beliefs6. 2.3 building state capacity state must define precise goals and policies that it intends to pursue. building state capacity implies the ability to formulate and implement policy, enforce legislation and deliver services (barkey and parikh, 1991; vom hau, 2012:4, fukuyama, 2013: 4). it assumes the essential capacities that the state requires in development realm are: (a) bureaucratic capacity to manage resources, implement policy and laws; (b) legal capacity 6 e. stiglitz, “markets, states and institutions”, roosevelt institute working paper, june 2017 journal of applied economics and business studies, volume. 6, issue 2 (2022) 55-76 https://doi.org/10.34260/jaebs.624 61 to enforce contracts, property rights and other legal regime effectively and speedily; (c) territorial capacity to employ all elements of national power to defend its territorial boundaries; (d) fiscal capacity to extract tax revenues from society and its diligent management; (e) infrastructural capacity to shape societal behaviour and service delivery; and (f) coercive capacity to deter or repel challenges to internal or external security7. 2.4 inter-dependence of politics, policy, and administration figure 3: policy, politics & administration relationship quite often these commissions/ task forces failed to appreciate the relationship between politics, policy and public administration which is in turn influenced by country’s geo-political position, stage of development, its natural resources endowment, political environment, and human capital. figure 2 vividly demonstrates that administrative and governance priorities are to a large extent dependant on the policy priorities of a country and other endowments including human capital. in the end, political dynamics and politics determine the support for acceptable policies 3. redefining civil service reforms and good governance 3.1 size of the government and organizational structures redefining role of the state leads to reassigning functions between private and public sectors that determines latter’s size, organizational structures (hierarchical, flat or hybrid), decision-making tiers and the skills that are required to perform assigned functions. this necessitates a comprehensive functional review of each level of government as well as ministry/department. functional review may be backed up with reduction in decisionmaking tiers from 5 (section officer to secretary) to 4 (assistant secretary, under secretary, joint secretary, and secretary) and from 4 to 3 in provincial governments which is expected to accelerate decision-making. it must be combined with major 7 badru bukenya and pablo yanguas, “building state capacity for inclusive development: the politics of public sector reform”, working paper no. 25, effective states and inclusive development research centre (esid), the university of manchester, uk. abdul wajid rana 62 restructuring of the support staff introducing common pool sharing concept. 3.2 functional review figure 4: functional review functional review requires litmus test of each function through the prism of katrrd ,i.e. keep the function without change, abolish the function, transfer the function to another public organization, or to an executive department, or to private enterprise, or to nongovernment organization, rationalize the function, reduce certain aspects of the function performance, or decentralize the function. the task force must have clarity as to objectives it intends to accomplish and correspondingly choose one or a combination of the approaches for functional review: (i) pure policy or program reviews reprioritizing policy or programs and eliminating the low-priority program; (ii) pure efficiency reviews focusing on identifying organizational and business process changes for greater efficiency; (iii) upstream program and efficiency review focusing on large-scale restructuring; (iv) mainstream ministry/agency-level functional review focusing on elimination of duplication, consolidation of similar services and merger of similar functions from different ministries/ agencies (v) downstream organizational reviews and business process re-engineering entailing restructure functions and activities within same ministry, automation and journal of applied economics and business studies, volume. 6, issue 2 (2022) 55-76 https://doi.org/10.34260/jaebs.624 63 contracting out some functions8 (manning, et al., 2004). the design of functional review is at figure 49. figure 5: structure of civil service functional reassignment and institutional restructuring through functional review at the federal, sub-national and market levels will provide a fair assessment of manpower as well as skills-set needed which will logically lead to required civil service reform and human resource hiring. however, before reforms in this area are considered, a government needs to choose between centralized or decentralized civil service system. in case of latter, five levels of civil service must be considered to perform assigned functions including inter-governmental, federal, provincial, local or district civil service and agency service system. it will be followed by a decision about choice of civil service structure from amongst (a) merit based open system; (b) merit based closed system; (c) political open system; and (d) political closed system; or a combination of these systems (see figure 5). 3.3 civil service reform landscape civil service reform is clearly not just about training and hiring of officers but also involves voluntary changes to the processes and structure of all the government departments with the sole purpose of good governance, targeting not just the political space but also human, legislative, economic and judicial governance. the literature defines the civil service as “a political institution that promotes joint action among many actors and stakeholders whose effort must be bent toward the goals of public policy…an institution of governance” (mcgregor, 1996). it may be added that 8 nick manning and neil parison, “determining the structure and functions of government program and functional reviews”, the world bank, moscow, october 2004 9 olga lukashenko, vladimir vlajković, and jasna stojanović, towards effective public administration: methodology for functional analysis, 2009 abdul wajid rana 64 the dimensions of collaboration that the civil service had to deal with have multiplied in the 21st century, especially post 18th constitutional amendment. the broad categories of collaborations now include between civil servants and political institutions; (ii) between the civil servants and the lower and higher judiciary; (iii) between private sector and the civil servants; (iv) between major state actors such as media, politicians, academicians, civil society and the interest groups; (v) between provincial and federal governments; and (vi) international community and the federal government. figure 6: civil service reform landscape author’s construct it is essential to understand that csr is “not an end by itself rather means to an end”. as identified by pollitt et al. (2017), the four ends are (i) useful policy design; (ii) enabling efficient government operations; (iii) improving service delivery and quality of service provided by the public organizations and (iv) cutting down public expenditures pollitt et al. (2017) also mentioned some intermediate ends i.e. (i) making bureaucracy completely apolitical; (ii) improving the accountability of the government towards the citizens for all its programs, policies and financial management; (iii) removing al the bureaucratic civil service reform political needs assessme nt objective s of governanc e and csr theoretic al framewor k for reforms political governanc e reform determini ng size and struture of governme nt determinin g skills-set required/ needed in governmen t empowerin g & strengtheni ng recruiting & regulatory bodies merit based recruitme nt restructurin g/ strengtheni ng training institutions/ methods restrucutring/st regnthning training institutions/ methods performanc e evaluation and promotion system technical advisory & support system in ministries compensatio n & incentives system disclosures & accountabili ty system promoting digitalizatio n & egovernance journal of applied economics and business studies, volume. 6, issue 2 (2022) 55-76 https://doi.org/10.34260/jaebs.624 65 restrains so that public officials have opportunities to grow and manage operations themselves. these ends ultimately create conditions for good governance and effective government. figure 6 describes the process for formulating civil service reform in the country 3.4 objectives of public sector reform therefore, fourth generation public sector reforms for building an effective, transparent and accountable state of pakistan must aim at: (i) establishing a professional, politically insulated, fully empowered and constitutionally mandated independent public service commissions at federal and provincial levels with all appointments of chairman and members on the recommendations of a high powered selection board; (ii) creating an inter-governmental, federal, provincial, local/district and agency civil services; (iii) creating a merit-based, politically neutral, well structured, right sized, well trained, high performing and affordable civil service at national and sub-national levels selected through competitive contemporary induction system and managed in an efficient, nondiscretionary, and transparent manner; (iv) performance based annual appraisal system linked with tenured postings; (v) establishing a well-articulated and non-discretionary promotion system linked with achievements at each level; (vi) creating an effective and speedy system of internal accountability; (vii) strong and it based public financial management system and allocation of resources fully aligned with policy priorities; (viii) establishing an equitable, efficient and accountable tax service for improved revenue performance; (ix) efficient and accountable service delivery through public-private partnership or outsourcing with full recovery of user charges; and (x) effective judiciary as well as accountability and transparency in the management of resources to discourage the use of public office for state capture and private gains. 3.5 key principles driving the reform the guiding principles driving the public sector and civil service reform in 21st century, once the objectives are clearly defined, include: (i) a pragmatic and result oriented framework measuring productivity of the organization as well as civil service; (ii) awareness of objectives and administrative structures; (iii) political strategies and engagement with stakeholders; (iv) selecting goal-oriented competencies and skills development; (v) innovations for public service delivery; (vi) professionalization , motivation and moral boosting in civil service; (vii) code of conduct for public sector abdul wajid rana 66 ethics; and (viii) effective and pragmatic anti-corruption strategies10. it is proposed that the federal government’s human capital system be shaped by: collaboration and not authority-driven, agility and not hostage to fixed ideas; outcome-driven, and not process driven, variegated, not monolithic and apolitical not parties. 4. recommendations for approaches to civil service reform and choice decisions before a government embarks upon civil service reforms, a decision is needed regarding the approach it intends to pursue for the reforms. it can choose amongst: (a) a business-oriented approach to government or a quality and performance-oriented approach; (b) emphasis on functional responsiveness and public service delivery; (c) retreat of government institutions in favor of commercial market enterprises (deregulation, privatization and marketisation). a comparison between tradition and contemporary approach and paradigm shift in various roles (see table 1). table 1: approaches to civil service reform intended purpose traditional contemporary level of change incremental extensive and comprehensive service career permanency and life-long employment -continuing employment -contracts culture maintaining administrative state management for business efficiency and markets civil service system collective and coherent • centralized • closed and bounded disaggregated units • decentralized • open and permeable relationship with politicians separate and formally independent responsive and vulnerable role of civil servant follower and implementer source of expertise and institutional memory recruitment of civil servants aptitude and skills service value orientation and integrity of character roel of senior management administrative managers visionaries/technocrats/experts appointment and promotion to senior management bias for managerial skills and annual performance reports competitive process/ insulated from politics; major consideration are: competence encompassing managerial and technical skills, and potentials; management style subservience hierarchy/ participatory and consultative 10 joan nwasike and dunstan maina (eds.), key principles for public sector reforms, the commonwealth, 2018 journal of applied economics and business studies, volume. 6, issue 2 (2022) 55-76 https://doi.org/10.34260/jaebs.624 67 authority operating perspective of the government regulation assistance and public service civil service relationship with other branches of government an adjunct of other branches autonomy from the political will of the regime relationship with citizens administration citizen charter/ service first notwithstanding above, public sector reforms have to be weaved around six key factors: (a) pace, coherence and sequencing of the reforms; (b) consensus leadership and political support to inspire and sustain the reforms and to mitigate resistance; (c) communication strategy to communicate reforms to a larger audience; (d) design adaptation requiring ability of the change management teams at all levels to adapt and real time adjustment to deal with emerging challenges; (e) legal/strategic framework as anchors for providing legal cover to reforms, wherever required; and (f) linking the reform framework to deliverables and performance. prudently crafted reforms having strong political support and are enforced by competent and motivated leadership and staff are likely to succeed. compromises and weaknesses in any of these domains will never have desired impact. incidentally, most public sector and civil service reforms till-date suffered from compromises and weaknesses in all these domains. suggestions on some of key elements of csr are discussed below. 4.1 all-pakistan service civil service of pakistan was constituted following an agreement between the governor general and the governors of the provinces of the provinces and civil service of pakistan (composition and cadre) rules, 1954 were notified. article 179 of the 1956 constitution recognized an all-pakistan service, a service common to the federation and the provinces. article 240 of 1973 constitution reiterated the provisions for an allpakistan service and 1954 rule have been protected under article 268. so, quite often the critique on all-pakistan service is misplaced. 4.2 federal secretary’s appointment recently, the task force recommended selection of a federal secretary through a high-powered committee for a tenure of three years. nevertheless, addition of three ministers in the committee for selection of a federal secretary will politicize the bureaucracy further. appointing the secretary on probation for six months initially will leave him at the mercy of the minister-in-charge and will lead to political compromises which will lead to politicizing the bureaucracy more than ever. nowhere in the world in parliamentary democracies, a permanent secretary or a federal secretary is appointed on abdul wajid rana 68 probation and there is no such provision in the civil servant act or rules. 4.3 revamping the public service commissions recruitment into civil service pre-supposes an apparatus manned by personnel of high intellect, quality, integrity, and leadership. while public service commissions retain constitutional protection and mandate in all saarc11 countries with complete independence, it has been left to the mercy of legislature in pakistan since 1973. its autonomy has been further diluted by down-grading it to an attached department of the establishment division. all commissions (federal and provincial) in pakistan are relegated to centers of political patronage. it is important that all the commissions must be either provided constitutional protection and mandate like other saarc countries or restore its statutory independence as held by the supreme court of pakistan in its judgement of march 16, 2015 “we, however, note that the federal public service commission is an independent statutory body which also derives validity from article 242 of the constitution”. secondly, the number of members may be increased to 15 because of ever increasing load. thirdly, the law must stipulate qualifications for appointment of members and chairman rather than leaving it to the discretion of the government. fourthly, the government may notify revised fpsc act, fpsc rules and fpsc conduct of business rules vetted by the ministry of law in 2015 to accomplish its autonomy. fifthly, commissions also requires improvement in technology, staff’s quality and its examination system. 4.4 human resource management (hrm) and plan human resource plan and management system is missing at all levels of the government. staff allocation is not aligned to assigned mandate or goals and programs reflected in the rules of business or medium-term budgetary framework. absence of key elements of hrm, such as, job description or performance indicators creates inefficiency and opaqueness in personnel management. absence of hr planning results in overstaffing and difficulty in promotion or transfer. purpose specific cells or wings sanctioned continue infinitely irrespective of their requirement resulting into adjustment of favourites against higher-scale positions and wasteful expenditure. hrm plan must lead to rationalize staffing of each ministry to determine its optimal size corresponding to assigned functions, preparing job description for each staff position, and develop performance indicators and contracts for desk officer to senior management to determine skills needs and effective performance evaluation. once it is done, the government must 11 south asian association of regional cooperation journal of applied economics and business studies, volume. 6, issue 2 (2022) 55-76 https://doi.org/10.34260/jaebs.624 69 segregate all positions at federal level into: (i) competitive hiring position; (ii) senior executive service positions; (iii) contract hiring positions; and (d) support staff positions. cross-merger of such positions or manpower must be discouraged to avoid tensions. the government must also determine cadre strength of each service and transparent sharing of position in bps-17 to 22 and notify rules for induction of officers of other services into the secretariat service. it also needs to design and develop career development path for officers in bps-17 to 21 to groom the talent and each ministry must notify delegation of authority. 4.5 merit based hiring in civil service induction into civil service is not just about filling current vacancies, as is generally perceived, but making a continuous, long-term investment in attracting a persistent flow of highly qualified and skilled men and women capable of accomplishing the assigned mission now and in the future. it is proposed that for the competitive examination: (a) all aspirants must qualify civil service aptitude test-i and ii; (b) qualified candidates (1:20 ratio for each vacancy) may undertake written examination (i) four compulsory subjects and (ii) cluster-based examination in at least four optional subjects; (c) divide the current occupations into four clusters: (i) administrative services-pakistan administrative service, police services of pakistan, and secretariat service; (ii) pakistan foreign service and commerce and trade group; (iii) financial management services-inland revenue service, pakistan custom service, pakistan audit and accounts service; and (iv) general management services-information group, military and land cantonments, postal service group, railways group; (d) psychometry test of qualified candidates; and (e) viva voce. induction of armed forces officers as well as provincial management services against their quota into specified federal services must be governed through competitive process including written examination and viva voce. for others, the commission and the government must review and refurbish the induction system as it is quite discretionary now. gradually, the whole system of examination may be reformed to align with modern practices in developed countries, such as, united kingdom, australia, canada, usa, etc. 4.6 training initial and adaptational training is of paramount importance in a career system. there is a need to redefine the philosophy, format, and duration of various trainings as well as restructuring and strengthening training institutions in pakistan. training must be geared towards not only improving knowledge, skills and abilities but also attitudinal changes, inculcating positive values and engendering an excellent work culture. the development of expertise and specialization, so critical to deal with 21st century challenges, can be abdul wajid rana 70 made possible through advanced study in specified fields required in the government. mode of trainings may be changed from general lecturing by retired civil servants to domain related case studies, syndicate discussions, focused group policy papers, off-thetray exercises, whole-of-government and inter-agency coordination, e-governance, negotiation skills, conflict resolution, public speech, and international best practices. 4.7 performance appraisal system performance appraisal system in pakistan remains a tool of control rather than innovation and development. annual performance evaluation report (per) because of its implications for promotion, career prospects and premature retirement has significant consequences. however, in pakistan, it has no no quantifiable targets and objective standards as garvin (1993) puts it that “if you cannot measure it, you cannot manage it. secondly, postings and transfers as well as salary progression are not linked to it, therefore, it is of least consequences except when it comes to promotion of a civil servant. the system continues to be susceptible to political and bureaucratic manipulations as observed by justice dorab patel, chairman, 1991 service reform commission that “acrs have long ceased to give a true picture of the performance of a civil servant because reporting officers did not dare to make adverse entries in the acrs of their subordinates for fear of harassment. therefore, all attempts to improve the method of promotion will have little effect until our values change…. the most important qualities required in a civil service are integrity and character, and as the assessment of these qualities by another officer can only be subjective..., acr system has broken down because officers are reluctant to pass adverse remarks against their subordinates.” a fair, efficient, and effective performance appraisal system is the lynchpin of civil service reform. three performance measurement models are in vogue: (a) balanced scorecard model developed by kaplan (2010) and isoreita (2008), establishing link between financial measures with outcome measures; (b) smart model of cross and lynch (1992)12 linking performance measures at each level of organizational hierarchy; and (c) performance prism model of neely et al (2001)13 incorporating the idea of inclusiveness of stakeholders in performance management. various techniques have been developed over time including ranking method, graphic rating scale, critical incidence, narrative essay, management by objective, human resource accounting, 360 degree and 12 k. f. cross (1992). for good measure. cma magazine, vol. 66 no. 3, pp.20-24 13 neely, a., adams, c., & crowe, p. (2001). the performance prism in practice. measuring business excellence, 5(2), 6-12. journal of applied economics and business studies, volume. 6, issue 2 (2022) 55-76 https://doi.org/10.34260/jaebs.624 71 720 degrees to assess performance. it is important that the government may create a reliable and comprehensive system for work performance; develop key indicators for measuring performance at all levels, groups and classification of public service; and objective assessment against those measurable performance indicators. the system must encompass a determination of annual work targets aligned with medium-term budgetary framework, a mid-year review of work targets, an open process of performance appraisal for appraising personnel functions including training, placement, promotion and salary progression as well as work planning, supervision, motivation, counselling, and development. professional skills of officers at senior management level must be assessed to three functional areas: leadership, professional skills, domain areas specialization, policy formulation, and programs implementation. 4.8 promotion civil service promotion is increasingly hinging on patronage, political dimension, and regional and gender considerations. such political biases are more pronounced in countries with diverse ethnicity and political polarization, like us. promotion to each next higher scale be linked with: per record upto the year immediately preceding the dpc or csb, required length of service and a minimum of service in each of the lower scales/rank, successful completion of mandatory training and ranking achieved as well as degree course in in one of the prescribed specialized fields, updated filing of income tax return and declaration of assets duly reconciled, variety of experience for promotion to bps-19 and above, compliance to rotation, pending inquiries and end outcome thereof, functional ability, leadership, and potential for higher management, and accomplishments. services may be dispensed with after two consecutive supersession and civil servant must be provided grounds for supersession with necessary legal amendment. 4.9 compensation the objectives of equitable compensation are: attracting high quality talent, retain and reward talent, inducing performance and accountability, and total cost is fiscally sustainable. there are different theories to determine compensation for the executives: (i) cognitive theory based on employees’ perception of their working environment and affects motivation; (ii) equity theory which deals individuals’ perception of the ratio of inputs (work performance) and outcomes (reward) in relation to people who perform the same job; (iii) human capital theory which argues that accumulation of skills and knowledge gained through education and experience is a key element of compensation; (iv) managerial power theory, i.e. perceptions created that managers act in their own abdul wajid rana 72 interests at the expense of the interest of shareholders; (v) structural theory in which compensation is a direct function of the number of organizational levels below top management; and (vi) contingency/institutional theory where external and internal factors influence compensation decisions. a combination of cognitive, human capital, equity, structural and institutional theories determine the compensation. today, major reform challenges the government is facing is to enhance capacity of ministries within reasonable personnel cost. it is important that the government may move out of “one-size-fits-all” notion introduced in 1973 and develop separate wage scales by profession, by individual sector and ministries, skills and qualification, and performance and index it with cost of living. perks may be gradually monetized. however, it will require a fiscal framework for financial sustainability combined with moving from pay-as-you-go pension system to contributory pension. thus, mere compensation reform will yield the desired result. it must be combined with reforming the archaic pension system in pakistan. 4.10 accountability figure 7: accountability “too much centralization undermines the flexibility needed to serve government’s greatly varied purposes. too much decentralization undermines its embrace of core principles,” as referred to above. therefore, a system of accountability focused on results is the theme for civil/public service in 21st century (see figure 6). there is a need to establish standing discipline authority/ committees for expeditious disposal of inquiries under efficiency and discipline rules against civil servants and revamping the services tribunal to strengthen internal accountability. it is also proposed to decentralize the system of internal accountability by delegating powers to the administrative secretaries for disciplinary action against official’s up to desk accountab ility political & pubic results financial legislativejudicial media managerial journal of applied economics and business studies, volume. 6, issue 2 (2022) 55-76 https://doi.org/10.34260/jaebs.624 73 (section) officers and deputy secretaries. it is equally important to establish statutory whistle blower system with full legal protection to the whistle blower to make the system more transparent and corruption free. 5. conclusion there are two options available for civil service/public service reforms: first, a radical restructuring of the existing bureaucratic order. secondly, a piece-meal but holistic reform of the existing bureaucratic institutions with a clear time-frame. i tend to favour the later approach as the country has experienced appalling results of “big-bang approach” to restructuring of these institutions in 2001 and 2010. however, the gradual approach must be well integrated, and time bound in all tiers of the government: (federal, provincial, and local) rather open-ended exercise with no impacting result. specific reforms needed include functional rationalization, human resource induction, training and management, pay reform, and public financial management. likewise, reforming the internal and external accountability including the anti-corruption institutions and disciplinary regime may also be part of this reform the current constitutional scheme infers decentralization, co-production, collaborative and participative governance. statutory autonomy pre-supposes autonomous institutions and agencies and pressure free decision-making. it must be known that political resistance to rationality, restraint, embeddedness and autonomy leads to patrimonialism, corruption, oligarchy, and capture which weakens state’s capacity. acknowledging that the political leadership can only deliver on its promises through public servants, it is in the interest of the government as well as interest of the nation to have a professional and capable civil service that continues to be strengthened and developed to deal with the challenges of 21st century and deliver services to the people. for this to happen, the government need to cultivate an environment where every civil servant must: (i) realize that they will not and cannot let the public down; (ii) acknowledge that they cannot betray the trust of the public; (iii) learn from the world, innovate and implement the best; (iv) keep grooming the next generation to meet the higher expectations of the public; and (v) fair and equitable compensation package. the key values underpinning the civil service for producing such civil servants must include: leadership, integrity, innovation, ability for strategic thinking and planning, capacity to formulate public policies to respond to emerging challenges and their implementation, competence to take decisions and continual learning with constitutional security of tenure and job. recognizing what michel foucault said that “power is everywhere” and “comes from everywhere,” so in this sense is neither an agency nor a structure. instead, it is a kind of ‘metapower’ or ‘regime of truth’ that pervades society, and which is in constant flux and negotiation.” abdul wajid rana 74 references bezes, philippe, jeannot, gilles. “the development and current features of the french civil service system”, van der meer frits, civil service systems in western europe, cheltenham, edward elgar, p. 185-215, 2011. biesen, guy van-biesen, (2006)., “overview of civil service selection procedures in eu context”, european personnel selection office (epso), belgium, march 2006 bukenya, badru, and yanguas, “building state capacity for inclusive development: the politics of public sector reform”, working paper no. 25, effective states and inclusive development research centre (esid), the university of manchester, uk. burns, john p., “civil service reform in china”, oecd journal on budgeting – volume 7 – no. 1 – issn 1608-7143, 2007 cho,prof. chang-hyun, “professional civil service of korean government”, republic of korea civil service commission http://www.csc.go.kr/ 11/2003. congressional budget office, “comparing the compensation of federal and privatesector employees, 2011 to 2015, www.cbo.gov/publication/52637 cross, k.f., (1992). for good measure. cma magazine, vol. 66 no. 3, pp.20-24. devine, donald j., “reforming the federal bureaucracy: challenge and opportunity”, the heritage foundation, dec. 10, 2018. edmans, alex, gabaix, xavier, and jenter, dirk., “executive compensation: a modern primer”, ecgi working paper series in finance, http://ssrn.com/abstract_id=2576707 farazmand, ali (ed.), administrative reform in developing nations, 2002 garvin, david a., “building a learning organization”, harvard business review, july– august 1993 issue. government of pakistan, national commission of government reforms report, 2008 greckhamer, thomas, cross-cultural differences in compensation level and inequality across occupations: a set-theoretic analysis, february 2011, organization studies 32(1):85-115, doi:10.1177/0170840610380806 grimshaw, damian, marino, stefania, and rubery, jill, “public sector pay and procurement in the uk”, ewerc, manchester business school, university of manchester, november 2012. haddon, catherine, “reforming the civil service”, institute of government, july 2012, https://www.instituteforgovernment.org.uk/sites/default/files/publications/cmps %20in%20template%20final.pdf hanif, ayesha, jabeen, nasira, and jadoon, zafar iqbal, “performance management in public sector: a case of civil service in pakistan”, south asian studies vol. 31, no. 1, january – june 2016, pp.99 – 116 international monetary fund, (june 2016). “case studies on managing government compensation and employment—institutions, policies, and reform challenges”, http://ssrn.com/abstract_id=2576707 https://www.researchgate.net/scientific-contributions/79726119_thomas_greckhamer?_sg%5b0%5d=h0bz9rqqjnw9ywdcgn4wfddfzjroglmc2h2taecf9-memxxyhppopigatvioymraamn9ty0.ynph9wvdxxftu-hajfs4kztop2bymnaciu-3lwkgcu4cempdrewlxaqea_9y2v2ff2p481lc7t-copa-tcfxpg&_sg%5b1%5d=h61ehdsqbokbratyrglq5qp5fpf9skxpusjq6pjmsj4pqvgzgs2tyumnfvab5v0co-dtyjpr2yzazzuz.cgombmzrv5d6p1mqvz3jvf71u2wzjpveswurqln2einjcm13xno_tgm3zsjw6rkwisg2wmxguvb1bksvxfw0ag https://www.instituteforgovernment.org.uk/sites/default/files/publications/cmps%20in%20template%20final.pdf https://www.instituteforgovernment.org.uk/sites/default/files/publications/cmps%20in%20template%20final.pdf journal of applied economics and business studies, volume. 6, issue 2 (2022) 55-76 https://doi.org/10.34260/jaebs.624 75 http://www.imf.org/external/pp/ppindex.aspx išoraitė, margarita, “the balanced scorecard method: from theory to practice”, intellectual economics 2008, no. 1(3), p. 18–28. kaplan, robert s., “conceptual foundations of the balanced scorecard”, harvard business school, working paper 10-074, 2010 lee, eun-suk, and kim, seongsu, “best practices and performance-based hr system in korea”, seoul journal of business, volume 12, number 1 (june 2006) lukashenko, olga, vlajković, vladimir, and stojanović, jasna, towards effective public administration: methodology for functional analysis, 2009. manning, nick, and parison, neil, “determining the structure and functions of government program and functional reviews”, the world bank, moscow, october 2004. muhyiddin, “corruption and development: a cross country analysis”, journal of development planning (edition: number 03/xii april-june 2007). neely, a., adams, c., & crowe, p., (2001). the performance prism in practice. measuring business excellence, 5(2), 6-12. nwasike, joan, and maina, dunstan (eds.), key principles for public sector reforms, the commonwealth, 2018 o’reilly, dermot, and reed, mike, “leaderism’: an evolution of managerialism in uk public service reform”, public administration vol. 88, no. 4, 2010 (960–978), blackwell publishing ltd, 9600 garsington road, oxford ox4 2dq, uk and 350 main street, malden, ma 02148, usa. pollitt, christopher, bouckaert, geert, public management reform, fourth edition, 2017 poocharoen, ora-orn, and mai, chiant, “meritocracy in asia pacific status, issues, and challenges”, article in review of public personnel administration · june 2013, https://www.researchgate.net/publication/258183551 rana, shan-aman, “discretion in a bureaucracy-evidence from pakistan”, policy brief 37303, international growth center, june 2019. stiglitz, joseph e., “markets, states and institutions”, roosevelt institute working paper, june 2017. sulemane, jose´ a., and kayizzi-mugerwa, steve, reforming africa’s institutions ownership, incentives, and capabilities, united nations university wider, 2003 therkildsen, ole, public sector reforms and the development of productive capacities in ldcs, background paper no. 1, unctad, march 2008, https://www.diis.dk/files/media/publications/import/extra/ldcr2009_therkildsen_e n.pdf ul-haq, nadeem, (2007). “why civil service reforms do not work”, ssrn electronic journal, february 2007. ulleberg, inger, “incentive structure as a capacity development strategy in public service delivery”, unesco, 2009 us national academy of public administration, “no time to wait: building a public http://www.imf.org/external/pp/ppindex.aspx https://www.diis.dk/files/media/publications/import/extra/ldcr2009_therkildsen_en.pdf https://www.diis.dk/files/media/publications/import/extra/ldcr2009_therkildsen_en.pdf abdul wajid rana 76 service for the 21st century”, july 2017. wild, leni, chambers, victoria, king, maia, and harris, dan, “common constraints and incentive problems in service delivery”, working paper 351, overseas development institute, www.odi.org.uk/ wilkins, john, “strategic leadership of public sector”, commonwealth governance handbook, 2013-14. world bank, governance indicators 2020. world bank, “poland: reforming government pay setting practices”, june 11, 2013. http://www.odi.org.uk/ journal of applied economics and business studies, volume. 6, issue 3 (2022) 67-78 https://doi.org/10.34260/jaebs.634 67 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x does investor attention matter? fresh evidences from wavelet approach muhammad asif khan1, muhammad akbar2, besma hkiri3 & noman khan4 1 department of management sciences, university of swabi, swabi, pakistan 2 birmingham city business school, birmingham city university, birmingham, uk 3 department of finance and economics, college of business, university of jeddah, jeddah, saudi arabia 4 department of management sciences, comsats university islamabad, attock campus, attock, pakistan abstract the covid-19 pandemic drastically damaged business activities that not only affected conventional financial markets but also upset islamic securities. given the severity of the recent pandemic, this study looked at the returns of the investor attention index, islamic bonds, and stock indexes in the occurrence of the ads business condition index. bivariate and multivariate wavelet analysis was employed on the daily data from january 2, 2020, to july 27, 2020. the study results indicate that before april 2020, there was a negative coherence between the investor attention index and the ads, islamic bonds, and stock returns. after that date, however, there is a positive relationship between the attention index and islamic bonds. in addition, the relationship between investors’ attention and the ads index shows both short-term and long-term correlations, but the long-term correlations are clearer. it has implications for household investors by empirically revealing the significance of google trends for islamic capital markets in the pre and during pandemic situation. our results show that the way investors use the google search engine is a key factor in how prices respond to new information. keywords investor attention index, google search volume, ads business condition index, islamic stocks and bonds, wavelet analysis jel classification d80, c91, d19 1. introduction the global covid-19 pandemic traumatized the global financial and economic conditions, led by that of the us, after an exponential increase in both the number of patients and deaths since january 2020 onward. the covid-19 epidemic shattered the us economy muhammad asif khan, muhammad akbar, besma hkiri & noman khan 68 massively (thorbecke, 2020); stock market volatility reached an all-time high, surpassing previous highs in december 2008, october 1987, and the 1929 market crash (baker, bloom, davis, & kost, 2020). the virus spread sporadically in china from mid-january 2020. however, the world, including the us, gave it due attention in march 2020, after who officially declared it as a global pandemic. the late response deterred the us household’s attention to covid-19’s social and economic impacts. us households realized the severity of the situation after the government announced the covid-19 emergency policy to shut down business activities. the anxiety induced by the depressed situation caused a more severe impact on the us financial market in march 2020 than in march 2019. risk-averse household investors paid more attention to seeking shelter by investing in alternative investments (conlon, corbet, & mcgee, 2020). extant literature suggests that islamic financial institutions and islamic stocks are more resilient and perform better in adverse market conditions (alkhazali et al., 2014; farooq & zaheer, 2015). during the coming covid-19 crisis, it is important to look into the link between the investors' attention index and islamic stocks and bonds. the capacity attention theory explains that individuals use more mental effort when the difficulty level of the situation is high; the higher the complexity or seriousness, the higher attention is paid (kehneman, 1973). using google search volume as a measure of investor attention, smales (2020a) established that the us stock returns are negative due to high investor attention towards the covid-19 pandemic. however, smales (2020b) reports that higher investors’ attention also led to higher stock returns in a few sectors. it implies that investors’ attention may have a negative or positive influence on stock returns depending on their understanding of the relationship between business conditions and the characteristics of the financial asset. moreover, the relationship may be strong if the shift in investors’ attention is collective over time (ballinari, audrino, & sigrist, 2022). using google search volume, akarsu and suer (2022) investigated investor attention effects on stock returns. they found that investors pay more attention to stock returns in both developed and developing countries. chen, tang, yao, and zhou (2021) used principal component analysis and partial least square methods to construct investor attention to examine stock returns. they reported that the investor attention index's predictive capability is mostly due to the reversal of transient price pressure and the greater forecasting ability for riskier stocks. dong, wu, fang, gozgor, and yan (2022) developed a theoretical model to explain how investor attention influences equity returns. they found a negative (positive) link between the fraction of continuous attention and cross-section stock returns. nevertheless, none of the previous studies have focused on the time-frequency coherence of investor attention, particularly in the islamic capital market. so, journal of applied economics and business studies, volume. 6, issue 3 (2022) 67-78 https://doi.org/10.34260/jaebs.634 69 the study looked at the relationship between investors' attention, the aruoba-diebold-scotti (ads) business conditions index (a proxy for business conditions), and the returns on islamic stocks and islamic bond indexes. this research accords with the empirical literature in three unique manners. first, it employs the capacity attention theory introduced by kehneman (1973) in the framework of the covid-19 epidemic. it postulates that mental energy is concentrated in attention, which is dependent upon suitable resources to fulfill situational attentional demand for a given task. there is a direct proportionality between the level of mental effort and attention and the level of task difficulty. second, as islamic stocks and bonds are considered safe havens and are expected to perform better in uncertain and vulnerable times (razak et al., 2016), we explore the coherency among the investor attention index, the ads index, and islamic bond and stock indices. lastly, as far as the researchers know, we are the first to look at how the investor attention index, the ads business conditions index, and the islamic bond and stock indices in the us are linked and how they change over time. the rest of the article goes like this: section 2 talks about the data and empirical approach; section 3 talks about the main conclusions; and section 4 gives the conclusions. 2. research methods 2.1 data description this research used daily data from january 2, 2020, to july 27, 2020 for the covid-19 induced investors’ attention index, the ads business condition index, and islamic stock and bond indices. we obtained the covid-19-related keyword series from google trends, ads business condition index values, s&p500 sharia stock values, and islamic bond index values from the datastream database. the ads business condition index is meant to be a proxy for real business conditions. it has seasonally adjusted economic indicators like initial jobless claims, industrial production, payroll employment, real personal income minus transfer income, quarterly real gdp, and real manufacturing and trade sales. the intensity of google search reflects the household’s interest in events and markets. to avoid potential bias due to an increasing number of users, the series is categorized from 1– 100 by dividing the total number of searches on all total keywords searched during the period (adachi, masuda, & takeda, 2017; khan et al., 2019). we make an investors' attention index by using principle component analysis on the 23 keywords related to covid-19 that google trends for u.s. households suggests in the "related search query" section. muhammad asif khan, muhammad akbar, besma hkiri & noman khan 70 2.2 empirical approach (wavelet analysis) 2.1.1. bivariate wavelet analysis we employ different wavelet tools including bivariate and multivariate wavelet analysis. the technique of cross wavelet has the capability to initially decompose and then rearrange x(t) function (rua & nunes, 2009) as: 𝑥(𝑡) = 1 𝐶𝛹 ∫ ∞ 0 [∫ ∞ −0 𝑤𝑥 (𝑢, 𝑠)𝛹𝑢,𝑠(𝑡)𝑑𝑢] 𝑑𝑠 𝑠2 , 𝑠 > 0 (1) wavelet coherence is a useful method for calculating the coefficients of local correlation among series in a temporal frequency domain. wavelet coherence is measured by the absolute smoothed cross wavelet value, which is standardized by the sum of each series' smoothed individual wavelet power spectrum: 𝑅2(𝑢, 𝑠) = |𝑆(𝑠−1 𝑊𝑥𝑦(𝑢,𝑠))| 2 𝑆(𝑠−1|𝑊𝑥(𝑢,𝑠)| 2(𝑠−1|𝑊𝑦(𝑢,𝑠)| 2) (2) where, s represents the smoothing parameter; the r^2 (u,s) is alike to coefficient of correlation which encounters the ensuing dissimilarity〖 0≤r〗^2 (u,s)≤1. if the squared wavelet coherence value is close to zero, it means that the correlation between the two series is weak, while a value of one or close to one means that the correlation is perfect or strong (torrence & webster, 1999). 2.1.2. multiple wavelet analysis next, we employ partial and multiple wavelet coherence (hereafter, pwc and mwc respectively) techniques. both these techniques allow including control variables (see, liu et al., 2007; veleda et al., 2012). pwc analysis detects wavelet coherence between two time series after eradicating the power of a third time series. mihanović et al. (2009) considers pwc similar to simple correlation that is mathematically given as: (3) the mwc permits the inclusion of additional series as conditioning factor that is overlooked in bivariate wavelet methods. mwc also enables to analyze the joint effect of two series on the third one. mwc is like multiple correlations, and it looks at how the effects of multiple explanatory variables on an explained variable are:     , ),(1),(1 ),(),(),( ),,( 2 12 2 2 2 121 21 2 xxrxyr xyrxyrxyr xxyrp −− − =  journal of applied economics and business studies, volume. 6, issue 3 (2022) 67-78 https://doi.org/10.34260/jaebs.634 71 𝑅𝑀2(𝑦, 𝑥2𝑥1) = r2(y,x1)+r 2(y,x2)−2re[r(y,x1).r(y,x2 ) ∗.r(x1,x2) ∗] (1−r(x1,x2) (4) 3. results and discussion figure 1. investors’ attention index vs ads business condition index figure 2. investors’ attention index vs islamic stock index figure 3. investors’ attention index vs islamic bond index figure 4. ads business condition index vs islamic stock index muhammad asif khan, muhammad akbar, besma hkiri & noman khan 72 figure 5. ads business condition index vs islamic stock index figure 6. islamic stock index vs islamic bond index 3.1 wavelet coherence figure 1 depicts continuous cross wavelet coherence between the investor attention index and the ads index. it shows coherency in the short term (2-4 bands of scale), especially over two sub-periods (investment horizons). strong coherency is more noticeable during the subperiod ranging from 4/21/2020 to the middle of may 2020, whereas small islands of orange color are scattered during the sub-period from 5/19/2020 to the end of the period. it suggests a mitigated lead-lag coherence between the two indices. over the first sub-period, the two indices are in an anti-phase (negative relationship), indicating a bi-directional causal relationship where the ads business condition index is the leading factor. it implies that if the business condition index is showing a lower value than average, household investors will pay more attention to the covid-19 outbreak (epidemic conditions). moreover, the lead-lag direction suggests that household investors focused more attention on the covid-19 outbreaks when closely linked with the real business conditions of the economy. during the second sub-period, the covid-19 attention index was conspicuously driving the real-time business index. after disrupting the real business conditions in the us market, household investors also realized the seriousness of covid-19 outbreaks and responded accordingly. our outcomes are in line with the capacity attention theory of kehneman (1973). figure 2 reports the coherency between the investors’ attention index and the islamic stock index for our sample period. it shows a relatively high coherency across various time horizons between the two variables. moderate coherency from 0.7 to 0.8 is specifically viewed in the short run, corresponding to a 4 to 8 scale of band. the left and up-direction arrows (quadrant iii) indicate that the islamic stock index leads the investors’ attention index in a negative journal of applied economics and business studies, volume. 6, issue 3 (2022) 67-78 https://doi.org/10.34260/jaebs.634 73 relationship. it implies that us households started focusing on the covid-19 outbreaks, depicting a link with the islamic stock market while, at the same time, the us businesses and stock markets were already disfigured. remarkably, the relationship (in the long run) is reversed from the start of march 2020 to the middle of may 2020, when investors’ attention leads the islamic stock index in a negative relationship. this finding is in line with the capacity attention theory, which states a situation attains more mental attention at the level of its severity. in contrast to the empirical literature, which shows that investor attention has a positive effect on stock returns (bank, larch, and peter, 2011), our results point to a negative relationship. figure 3 depicts the wavelet coherence plot of the investors’ attention index and the islamic bond index. there is a non-coherency zone shown in the short and medium run, indicating no causality between the investors’ attention index and the islamic bond index. looking at the left side and upwards arrows (quadrant iii) in the long run frequency (16-32 band of scale), we find a negative relationship between the two indexes where the islamic bond index is the leading factor. it implies that if there is an increase in investors’ attention index, co-movements decrease with the islamic bond returns over the sub-period 1/29/2020– 5/19/2020. the fear and uncertainty caused by the covid-19 outbreak has a negative impact on investment decisions. over the sub-period 3/24/2020-5/19/2020, the plot depicts a positive relationship where the investors’ attention index is leading. it implies that the islamic bond index, considered as a safe haven asset, is linked to the investors’ attention index in the long run over our sample period. since this is a positive relationship, it means that more investors will switch to the islamic bond market when bad things happen or when things are uncertain. the ads business condition index and the islamic stock index co-movement are depicted in figure 4. the coherency between these indexes, especially localized in the short run, is moderately significant over the sub-period 2/25/2020 to 3/24/2020. while the covariation changes from an in-phase (positive) to an anti-phase (negative) relationship over the same period, the ads business condition index always leads the islamic stock index. the ads business condition index has a negative impact on islamic stock returns. in the long run, it is clearly shown that the ads business condition index and the islamic stock index are in-phase (positive). our findings are consistent with razak et al. (2016), who observed that the islamic stock market performs better during vulnerable economic conditions. figure 5 shows that the ads business condition index leads the islamic bond market index in the short run almost over the full sample period. however, the relationship between the indexes remains difficult to interpret, especially over the sub-period 1/29/2020 to 2/25/2020. over this period, business activity tends to go up and down, implying swings in the ads muhammad asif khan, muhammad akbar, besma hkiri & noman khan 74 business condition index between leading and lagging the islamic bond index in the short run. further, it also suggests a negative relationship during january and february months when us market conditions were stable as the covid-19 outbreak was not yet in full swing. by the end of april, covid-19 epidemic conditions will become more serious in the us and the ads business condition index has a positive relationship with the islamic bond market. figure 6 shows the islamic bond index leads the islamic stock index over the medium-term as well as the short-term horizons, implying that investment in the islamic bond index may be the optimal substitute for investors during eras of uncertainty and anxiety (flight-to-quality). in the long run, however, the islamic stock index has a negative effect on the islamic bond index. this is in line with what park, fang, and ha (2019) found. figure 7. pwc for bond index vs. investors’ attention index| ads index figure 8. pwc for islamic stock index vs. investors’ attention index | ads index figure 9. mwc for bond index| investors’ attention index & ads index figure 10. mwc for islamic stock index| investors’ attention index & ads index 3.2 multiple and partial wavelet outcomes figure 7 visualizes the partial wavelet coherency between the investors’ attention index journal of applied economics and business studies, volume. 6, issue 3 (2022) 67-78 https://doi.org/10.34260/jaebs.634 75 and the islamic bond index, incorporating the ads business condition index as a control variable. the highest level of coherency between the islamic bond index and investors’ attention index is observed in the long term horizon from 1/29/2020 to 5/19/2020, whereas there is no coherency in the short-run over the sample period (except at the end of march 2020). it ranges from 0.7 to 0.8, indicating a strong relationship. with the ads business condition index, the two indices have only a long-run correlation. one way to explain this is that there is a negative relationship between the ads business condition index and the investors' attention index, which makes short-term correlation go away. figure 8 shows same for the islamic stock index and the investors’ attention index. these findings suggest that the investor attention index has more pronounced implications in long run, particularly when the majority of investors have homogenous beliefs directly linked with market uncertainty and complexity. it means that investors will put in more time and effort to understand the market situation when they are more worried about it. we combine the effects of the investor attention index and the ads index to show the islamic bond index movement over time and across scale bands in figure 9. the joint effect of these variables on the islamic bond index is significant and is observed across all scale bands and over the sample period. in the short and medium term, small islands of orange color are scattered over the sample period, indicating the combined impact on movement of the islamic bond index. more interestingly, the joint effect on the islamic bond index is strongest over the long term (ranges from 0.9 to 1) and is exactly localized over the sub-period of 1/29/2020 to 5/19/2020. the combined synchronized effects for the islamic stock index are depicted in figure 10, which shows a similar pattern to those reported for the islamic bond index in figure 9 over the long run. however, the combined impact is less pronounced in short and medium-runs and especially localized over the sub-period from 1/29/2020 to 3/24/2020. it implies that during an epidemic outbreak, investors prefer passive strategies and respond to long-term return expectations. when epidemics directly or indirectly affect business conditions and in the end financial markets investors become more careful about their investments and pay more attention to them. 4. conclusion we examined the time-varying coherence of the investor attention index and the ads index with islamic stock and bond indices' returns from january 1, 2020, to july 27, 2020. our empirical findings supported the capacity theory of attention as we reported higher investors’ attention in the critical covid-19 pandemic period from mid-march to mid-may 2020. interestingly, the investors’ attention is seconded by ads business conditions, which intensify the household investors’ anxiety about financial markets. it seems to say that muhammad asif khan, muhammad akbar, besma hkiri & noman khan 76 investors were worried about the financial markets because businesses were closing because of the government's covid-19 lockdown policy. our findings also exhibit a negative coherence between islamic stock returns and the investor attention index, in line with the information discovery theory of attention. furthermore, we reported a positive long-term relationship between business conditions and islamic stock returns, confirming the notion that islamic stocks perform better in times of vulnerable business conditions. in line with the extant literature, our findings support the notion that islamic bonds provide a safe haven for investors in times of uncertainty and anxiety, and we report that the ads business conditions index leads the islamic bond index in the long run. our findings imply increased investors’ attention towards business conditions and (islamic) capital markets during the covid-19 pandemic crisis. based on what we've seen in the real world, policymakers and regulators should take steps and make policies to make it easier to get market information and make market shocks less complicated. references adachi, y., masuda, m., & takeda, f. (2017). google search intensity and its relationship to the returns and liquidity of japanese startup stocks. pacific-basin finance journal, 46, 243-257. akarsu, s., & süer, ö. (2022). how investor attention affects stock returns? some international evidence. borsa istanbul review, 22(3), 616-626. al-khazali, o., lean, h., and samet, a. 2014. “do islamic stock indexes outperform conventional stock indexes? a stochastic dominance approach”, pacific-basin finance journal, 28, 2946. alnemer, h. a., hkiri, b., & khan, m. a. (2021). time-varying nexus between investor sentiment and cryptocurrency market: new insights from a wavelet coherence framework. journal of risk and financial management, 14(6), 275. andrei, d., & hasler, m. (2014). investor attention and stock market volatility. review of financial studies, 28, 33–72. aruoba, s.b., f.x. diebold, and c. scotti (2009), “real time measurement of business conditions,” journal of business and economic statistics, 27, 417–427. baker, s. r., bloom, n., davis, s. j., kost, k., sammon, m., & viratyosin, t. (2020). the unprecedented stock market reaction to covid-19. the review of asset pricing studies, 1-17. ballinari, d., audrino, f., & sigrist, f. (2022). when does attention matter? the effect of investor attention on stock market volatility around news releases. international review of financial analysis, 82, 102185. journal of applied economics and business studies, volume. 6, issue 3 (2022) 67-78 https://doi.org/10.34260/jaebs.634 77 bank, m., larch, m., & peter, g. (2011). google search volume and its influence on liquidity and returns of german stocks. financial markets and portfolio management, 25(3), 239. barber, b. m., & odean, t. (2008). all the glitters: the effect of attention and news on the buying behaviour of individual and institutional investors. review of financial studies, 21, 785– 818. brunnermeier, m., g. gorton, and a. krishnamurthy (2011), “risk topography,” nber macroeconomics annual 2011, 26(1), 149–176. chen, j., tang, g., yao, j., & zhou, g. (2022). investor attention and stock returns. journal of financial and quantitative analysis, 57(2), 455-484. conlon, t., corbet, s., & mcgee, r. j. (2020). are cryptocurrencies a safe haven for equity markets? an international perspective from the covid-19 pandemic. research in international business and finance, 101248. dong, d., wu, k., fang, j., gozgor, g., & yan, c. (2022). investor attention factors and stock returns: evidence from china. journal of international financial markets, institutions and money, 77, 101499. farooq, moazzam, and sajjad zaheer. "are islamic banks more resilient during financial panics?." pacific economic review 20, no. 1 (2015): 101-124. hkiri, b., hammoudeh, s., aloui, c. and shahbaz, m., 2018. the interconnections between us financial cds spreads and control variables: new evidence using partial and multivariate wavelet coherences. international review of economics & finance, 57, pp.237-257. kahneman, d. (1973). attention and effort (vol. 1063). citeseer. khan, m. a., ahmad, r., azmi, a., & akbar, m. (2019). a new sentiment index for the islamic stock market. investment analysts journal, 48(2), 146-172. khan, m. a., hernandez, j. a., & shahzad, s. j. h. (2020). time and frequency relationship between household investors’ sentiment index and us industry stock returns. finance research letters, 36, 101318. liu, y., san liang, x., & weisberg, r. h. (2007). rectification of the bias in the wavelet power spectrum. journal of atmospheric and oceanic technology, 24(12), 2093-2102. mihanović, h., orlić, m., & pasarić, z. (2009). diurnal thermocline oscillations driven by tidal flow around an island in the middle adriatic. journal of marine systems, 78, s157-s168. odean, t. (1999). do investors trade too much? american economic review, 89, 1279–1298. park, k., fang, z., & ha, y. h. (2019). stock and bond returns correlation in korea: local versus global risk during crisis periods. journal of asian economics, 65, 101136. razak, a. r., ismail, n., & aridi, n. a. (2016). is islamic stock market no different than conventional stock market? an evidence from malaysia. international business muhammad asif khan, muhammad akbar, besma hkiri & noman khan 78 management, 10(17), 3914-3920. rua, a., & nunes, l. c. (2009). international comovement of stock market returns: a wavelet analysis. journal of empirical finance, 16(4), 632-639. sharif, a., aloui, c., & yarovaya, l. (2020). covid-19 pandemic, oil prices, stock market, geopolitical risk and policy uncertainty nexus in the us economy: fresh evidence from the wavelet-based approach. international review of financial analysis, 101496. smales, l. a. (2020a). investor attention and global market returns during the covid-19 crisis. available at ssrn 3623915. smales, l. a. (2020b). investor attention and the response of us stock sectors to the covid-19 crisis. available at ssrn 3625487. sousa, r., aguiar-conraria, l., & soares, m. j. (2014). carbon financial markets: a time— frequency analysis of co2 prices. physica a: statistical mechanics and its applications, 414, 118–127. tantaopas, p., padungsaksawasdi, c., & treepongkaruna, s. (2016). attention effect via internet search intensity in asia-pacific stock markets. pacific-basin finance journal, 38, 107–124. thorbecke, w. (2020). the impact of the covid-19 pandemic on the us economy: evidence from the stock market. torrence, c. and compo, g.p., 1998. a practical guide to wavelet analysis. bulletin of the american meteorological society, 79(1), pp.61-78. torrence, c., & webster, p. j. (1999). interdecadal changes in the enso–monsoon system. journal of climate, 12(8), 2679-2690. uluceviz, e. and yılmaz, k., 2018. measuring real-financial connectedness in the us economy (no. 1812). working paper. veleda, d., montagne, r., & araujo, m. (2012). cross-wavelet bias corrected by normalizing scales. journal of atmospheric and oceanic technology, 29(9), 1401-1408. vlastakis, n., & markellos, r. n. (2012). information demand and stock market volatility. journal of banking & finance, 36, 1808–1821 journal of applied economics and business studies, volume. 6, issue 3 (2022) 45-66 https://doi.org/10.34260/jaebs.633 45 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x remittances, human capital nexus: exploring the role of exchange rate in asia iqra noor1, babar hussain2 & umbreen zahra3*1 1 ms scholar, international institute of islamic economics, international islamic university islamabad, pakistan 2 assistant professor, international institute of islamic economics, international islamic university islamabad, pakistan 3 phd scholar, international institute of islamic economics, international islamic university islamabad, pakistan abstract this paper analyses the impact of remittances on human capital through the interaction of exchange rate in asia. there is considerable debate regarding the relative contribution of international migrant’s remittances to sustainable economic development. remittances are considered as temporary income. if exchange rate remains stable, then remittances will help in curtailing income hurdles of households as well as for government and helps to increase investment in education. investment in education increases skill level of labor force and create an environment for long term growth. we use panel data of 54 asian countries for the period of 1980-2020 and employ hausman test and percentile analysis. the paper concludes that remittances have a positive and significant impact on human capital. however, the substituting relationship of remittances and exchange rate is observed in stimulating human capital in asian countries. the developing countries having low growth, when exchange rate is overvalued (appreciate), it is not only pushing the economy further toward slower growth due to decrease in foreign competitiveness, increase in the current account deficit and increase unemployment but also cause to decrease main source of earning remittances growth in these countries. therefore, this study keywords remittances, exchange rate, human capital, fixed effects, random effects, hausman test jel classification f24, o24, j24 1 *amberen2010@gmail.com iqra noor, babar hussain & umbreen zahra 46 suggests that the government of developing country implements such policies that helps in the stabilization of their currencies (manage floating exchange rate). 1. introduction human capital is getting much importance around the world in the recent era of globalization. human capital is considered as a most important productive asset even at a level of technological advancement (adenutsi, 2010). schultz (1961) describes that the main difference in the income level of developed and developing countries depends upon the quality of population which is directly related to the stock of human capital. investment in education is considered as a way for a long term economic growth (melton, 1965; romer, 1990). to get the long term competitive advantage in emerging economies human capital plays an important role as compared to financial and physical capital. human capital is an invisible asset that represents the capabilities, competencies, and commitment of people in particular areas (mozumder & islam 2017). most developing and emerging economies have not enough finance to invest in the education sector. that’s why developing countries face low level of development, higher unemployment, and low wage rate due to higher population growth. as a result, young professionals and labors may have developed an irresistible desire to move abroad in search of jobs (adenutsi, 2010). remittances earned by migration included employee’s compensation, migrant transfers, and worker remittances (sghaier, 2021). remittances are substantially considered as a financial injection for developing countries as it serves as a source to overcome the liquidity constraints for consumption and investment objectives in recipient countries (nguyen & nguyen, 2015). lucas and stark (1985) also exposed two different motives of sending remittances back to home country such as self-interest and pure altruism motive. in self-interest motive people may spend their remittances to increase assets in home town by investing for homes, land and cattle’s. migrants sometimes invest in assets through remittances to increase their prestige and enhance their living standards. in pure altruism motive migrants spend money on their family members back in their home country. according to becker (1974) when migrants considered pure altruism motive they derive positive utility by spending on consumption and overall wellbeing of family in the home country. in general, the available empirical evidence also shows mixed results. some studies show positive impact of remittance on education, some show negative and few show no impact of remittances on education (akanbi, 2017; awusi, 2016; azam & raza, 2016; azizi, 2018; journal of applied economics and business studies, volume. 6, issue 3 (2022) 45-66 https://doi.org/10.34260/jaebs.633 47 dietz et al., 2015; mozumder & islam, 2017; naeem & arzu, 2017; pilařová & kandakov, 2017). the controversy of literature indicates that there are some missing linkages between remittances and human capital and there are no direct linkages between the said variables. therefore, this paper tries to find some other inconspicuous factors which play an intermediary role in explaining their relationship as one of the key determinants of remittances is the exchange rate. the exchange rate brings fluctuation in the number of remittances which is directed to home countries (olubiyi, 2015; pant & budha, 2016; rahman et al., 2019). lin (2011) explains that the remittances growth decrease with the appreciation of the real exchange rate. if migrant living’s county currency appreciates due to any reason as compared to home country currency. the amount of remittances upsurge in the home country and households of the home country become rich and now they have more money after consumption expenditure then they invest in human capital (yang, 2008). this study tries to analyze the impact of remittances on human capital by exploring the role of exchange rate. we are motivated to do this in-depth analysis because human capital is considered as most crucial productive assets (adenutsi, 2010) and remittances considered as lifeline for people as it became largest rising source of capital inflow as compared to fdi and official development assistance (oda) in developing countries (world bank, 2019). this study significantly helps to understand that, how developing countries respond in case of transitory changes in the economic conditions. outcome of this study significantly contribute that, with stable exchange rate, international remittances inflow will curtail income hurdles of households and government and helps both of them to increase investment in education. investment in education will increase skill level of labor force and create environment for long term growth, which further pushes the country to sustainable development. the remaining part of the study is organized in the following section. section ii reviews the previous literature section iii is devoted to data and empirical specification. section iv discusses empirical results and section v concludes the study. 2. literature review there is a huge empirical literature that explains the impact of remittances on human capital. there exist positive, negative, and insignificant impacts of remittances on human capital. the positive linkages between remittances and exchange rate are supported through the direct channel at macro level data (adenutsi, 2010; amakom & iheoma, 2014; azizi, 2018; mohamed & sivarajasingham, 2019; mozumder & islam, 2017; ngoma & ismail, 2013). adenutsi (2010) finds that in the low-income region of ssa, migrant remittances positively and significantly affect human capital development through long run enrollment in secondary school. iqra noor, babar hussain & umbreen zahra 48 moreover, mozumder and islam (2017) explore that remittances positively impact secondary school enrollment rate and average years of schooling. another study by azizi (2018) suggests that remittances help in improving the rate of school enrollment, school completion, and enrollment rate in private schools. amakom and iheoma (2014) confirm that the remittances inflow increases the income level in the country that cushions the literacy rate. recently, the study of mohamed and sivarajasingham (2019) conclude that the remittances have long term equilibrium relationship with human capital in sri lanka through ardl bound test. ngoma and ismail (2013) suggest that on average, an increase in remittances increases the years of schooling for both secondary and higher levels. as remittances lessen financial constraints and produce spillover impact in the remittances recipient countries for further schooling opportunities. sometimes remittances positively affect human capital through an indirect channel, which are some intermediate variables. dzeha et al., (2018) highlight that when there is increase in the inflow of remittances it helps out in fulfilling technological expenses as a result that it spurs innovative processes, skill acquisition and training that increase human development in terms of education. mohammed (2021) analyzes ssa countries from the period of 2004-2018 and he finds that remittances play an important role in human capital development especially in those countries which have well-developed institutions. azam and raza (2016) depict that immigrant remittances positively impact human capital development and suggest that the economic governance system strengthens the relationship between immigrant remittances and human capital. borja (2020) finds that the secondary education enrollment ratio increase through remittances in those countries where the rate of corruption is low. according to ziesemer (2012) remittances have a stronger impact on education through savings as the amount of high savings helps the government to increase public expenditures on education. the studies at micro-level demonstrate that remittances are effective for children’s education. salas (2014) expounds that the probability of sending children to private schools increases with overseas remittances by controlling parents’ absenteeism. brempong and asiedu (2015) discover that remittances increase the likelihood for households to enroll their children in primary school and secondary school. hines and simpson (2019) observe that more educational expenses are inclined toward those households who earn remittances from abroad for kenya in 2009. the previous literature supports the negative as well as insignificant impact of remittance on human capital such as tsaurai and ngcobo (2018), akanbi (2017) and dietz et al., (2015). tsaurai and ngcobo (2018) reveal that the remittances decrease investment in human capital because the amount of remittances utilizes in speculative activities of stock market instead of journal of applied economics and business studies, volume. 6, issue 3 (2022) 45-66 https://doi.org/10.34260/jaebs.633 49 investment in human capital. akanbi (2017) suggest that migration in term of remittances negatively affect human capital due to low skilled emigrants for 19 ssa countries between the time span of 1990-2013. dietz et al., (2015) find that the remittances negatively affect school attendance of children due to successful migration and work abroad even with a low level of education. awusi (2016) find that the remittances have insignificant impact on education of developed countries because in developed countries it is a legal right of people to attain education up to secondary level. naeem and arzu (2017) argue that remittances have insignificant effect on human development. they justify that other factors are important for human development instead of remittances these factors include fdi, foreign aid, and official development assistance. moreover, pilařová and kandakov (2017) detect that remittances insignificantly bear on the attendance of high school youth aged 16 to 20 years. people intend to move abroad for a gain of high return jobs. earlier studies support exchange rate depreciation and appreciation both have different impacts on remittances growth. olubiyi (2015) discover that depreciation of real exchange rate caused to decrease in remittances inflow in nigeria. real exchange rate depreciation shows that there are unfavorable economic conditions in back home country. therefore, remittances inflow decrease. exchange rate depreciation reduces remittances inflow in the country because value of cash held in form of assets decreases with the depreciation of exchange rate and migrants decide to cut down the amount of their remittances (jijin et al., 2021; omon, 2021). but some time depreciation increases the remittances inflow in the country because altruistic reason motivates migrants to remit more money back with currency depreciation for welfare purposes (kuncoro, 2020; rahman et al., 2019). opposite to the depreciation effect, appreciation of exchange rate has different impact on human capital. for example, lin (2011) postulated that exchange rate appreciates in tonga it decreases the growth of remittances in term of tonga domestic currency. sometimes exchange rate cannot control magnitude of remittances flow. literature shows the linkages between exchange rate and human capital like as guo et al., (2019) enunciate that with an exchange rate appreciates in china as compared to the u.s dollar, then the cost of foreign education will decrease and more students will decide to go abroad to study. moreover, jeanneney and hua (2011) conclude that the with an appreciation of the currency workers increase their education level as education cost decrease and the benefit from education increase with high expected returns of education. but gylfason et al., (1999) give counter argument, if appreciation in exchange rate due to any reason which cause to slowdown economic activities like as lower income level and lower investment in human iqra noor, babar hussain & umbreen zahra 50 capital. however, some studies advocate that exchange rate depreciation opposing to appreciation effect, decrease human capital such as ogege (2019) infer that a rise in the exchange rate (depreciation) affects negatively on education index. kaur and sirohi (2013) suggest that the indian students who are planning for foreign education or currently studying abroad are pinch by rupee depreciation. javid and ahmad (2014) discover that exchange rate devaluation increase the educational cost and decreases the amount of budget for students forcing the student to search for an alternative to full fills their educational expense. emmanuel (2017) also supports above stated relationship and finds that one-year previous lagged exchange rate increase or decrease government expenditure on human development (health and education) also increase or decrease in nigeria during the period from 1986 to 2015. few studied in literature analyzed the combined impact of remittances and exchange rate on human capital. yang (2005) concludes that positive exchange rate shock in migrant destination country leads to an appreciation of migrant’s currency against the domestic currency. migrants send the greater amount of remittances that is associated with greater investment in child’s schooling, less child labor, and more educational expenditure in households of the origin country. moreover, yang (2008) also gives similar results that the exchange rate favorable shocks help in the appreciation of migrant’s currency. as a result, it raises the amount of households’ overseas remittances and increases their expenditures in nonconsumption areas like educational expenditures. so, a longer stay of children in school will decrease the child labor force and increase capital intensive entrepreneurship. this study intended to fill the gap by novel findings of this research to check the influence of remittances variation on human capital at the macro level of panel of 54 countries through the channel of the exchange rate. as this question is not discussed at broader aspect earlier and only one study which is time series and based on philippines data by yang (2008). 3. methodology and data on the theoretical side, two theories link the impact of remittance on human capital through the exchange rate. friedman (1957) gives permanent income hypothesis which postulates that people face random variation in their income level on annual basis. this variation may be due to promotion, lottery, transfer payment and remittances that receive from migrants for a short span of time. the pattern of consumption does not change with the change of remittances receipts because the average propensity to consume is the function of permanent income. therefore, the extra amount of remittances would be saved or invested e.g. physical capital or human capital. journal of applied economics and business studies, volume. 6, issue 3 (2022) 45-66 https://doi.org/10.34260/jaebs.633 51 the second theory is given by mundell (1963) and fleming (1962) which is known as the mundell fleming condition. this condition explains that the government has two options either to choose credibility or flexibility. calvo and reinhart (2002) explain that in the credibility fixed exchange rate exist and announcements about the fluctuation of exchange rate are made in advance. such announcements eliminate the volatility of the exchange rate as a result it attracts international inflow of capital like in the form of remittances. on the other hand, flexibility link with the flexible exchange rate. according to flexibility, when there is a flexible exchange rate no clear announcements are made about the change in the exchange rate. therefore, exchange rate flexibility may reduce the level of international inflow of capital like remittances. in order to examine the impact of the remittances on human capital, we follow the model posited by ngoma and ismail (2013) with little modifications the model has taken the following functional form: 𝐻𝑖𝑡 = 𝛽0 + 𝛽1𝑅𝑖𝑡 + 𝛽2𝐸𝑅𝑖𝑡 + 𝛽3(𝑅𝑖𝑡 ∗ 𝐸𝑅𝑖𝑡 ) + 𝑋𝑖𝑡 𝛽 ′ + 𝜆𝑖 + 𝜀𝑖𝑡……………. (1) i and t denote the number of countries and number of observations in the panel over time. where, 𝐻𝑖𝑡 represents the human capital (human capital index (hci) based on years of schooling and returns to education). where, β0 shows intercept term, rit represents the worker’s remittances as a percentage of gdp, erit is the real effective exchange rate index, xit vector represents control variables that potentially have an effect on human capital, λi is the unobserved country specific fixed effect variables such as institutional and geographical factors and εit is the time varying error term. β1 denotes, the elasticity of the human capital according to worker’s remittances, β2 interpret, the elasticity of the human capital according to real effective exchange rate and β3 represents the elasticity of the human capital according to worker’s remittances through the index of the real effective exchange rate. it shows either exchange rate compliment or substitute remittances impact on human capital. if we take differential of eq (l) with respect to the remittances, it takes the following form: 𝜕𝐻𝑖𝑡 𝜕𝑅𝑖𝑡 = 𝛽1 + 𝛽3𝐸𝑅𝑖𝑡 if (β3>0) then it will complement the relationship and impact of remittances on human capital strengthening by exchange rate. if (β3<0) then it will substitute the relationship and impact of remittances on human capital weakened by exchange rate. as this paper uses panel data, pooled ols, fixed effect (fe) and random effect (re) iqra noor, babar hussain & umbreen zahra 52 models employ to analyze the impact of remittances on human capital. hausman test is use to decide among fe and re, which one is most appropriate for this study. according to the hausman test, if error term and regressors are not correlated then fixed/random effects techniques both are appropriate for study. but test suggests that if the error term is correlated with any regressors, then the random effects technique gives inconsistent results and the fixed effects technique is appropriate for study. 3.1 data, variables definition and data sources this study uses the data set of 54 asian countries for the period of 41 years from 19802020 table 3.1: variable description category variables definition expected sign source dependent variable human capital index of human capital per person, based on years of schooling and returns to education for country “i” at time “t” n/a pwt 10.0 independent variables remittances worker’s remittances, (% of gdp) for country “i” at time “t” (+/-) wdi exchange rate real effective exchange rate index (cpibased) for country “i” at time “t” (+/-) brugel database control variables openness trade as (% of gdp) for country “i” at time “t” (+) wdi population (annual %) of population growth for country “i” at time “t” (-) wdi economic development measured through gdp per capita (constant 2015 us$) for country “i” at time “t” (+) wdi inflation inflation measured through consumer price index by (annual %) for country “i” at time “t” (+) wdi foreign direct investment net inflows of foreign direct investment as (% of gdp) for country “i” at time “t” (+) wdi savings measure by gross domestic savings as percentage of gdp (+) wdi democratic development polity2 for country “i” at time “t” ranges from -10 (full autocracy) to +10 (full democratic) (+) polity v source: compiled by researchers from wdi, pwt, brugel database, polity v our main focus is on asian region and only 54 countries are selected on the basis of availability of data during the considered period. in addition, data of many variables was collected from world development indicators (2021) and the data of human capital was taken from pwt. journal of applied economics and business studies, volume. 6, issue 3 (2022) 45-66 https://doi.org/10.34260/jaebs.633 53 4. preliminary findings, results and discussion descriptive statistics include mean, minimum, maximum, and standard deviation of each variable. in table 4.1 descriptive statistics give the overview of the data. some variables have negative sign in their minimum values like as population growth rate which has minimum value of -4.533 which observed in syrian country in the year 2014. it was due to civil war that started in 2011. the war has brought devastation to syria. there has been a huge loss in terms of human deaths and refugee crisis (khan & khan, 2017). maximum value of population growth is observe 17.512 for qatar in year 2007. at that time the country has the highest share of immigrants in the world because approximately 87% of the population in the state of qatar were foreigners (seshan, 2012). further, serious efforts were made related to health status to accomplish the mdgs (al-thani et al., 2014). inflation rate which has minimum and maximum value of -18.109 and 3373.76 respectively. in 2004 deflation was experienced with the value of -18.109 in bhutan. while, hyperinflation with the value of 3373.76 was observed in armenia in 1994 due to dissolution of the soviet union (shifflett, 2021). lowest fdi value -37.173 experienced in mongolia in the year 2016. inflow of fdi into mongolia decreased drastically owing to many wrong steps taken by mongolia’s political leadership. such as piece of legislation related to “strategic entities foreign investment law” passed in 2012 (orji et al., 2018). huge fdi inflow gained in hong kong in 2015 with figure of 58.519. this was became possible due to corporate restructuring there (unctad, 2016). gross domestic savings in timor-leste was -136.873 in 2002 due there independence (freitas, 2013). higher volume of savings gained in turkmenistan in 2010 with value of 87.827. table 4.1: variable description variable obs mean std. dev. min max hc 1,509 2.264 0.061 1.035 4.352 remitt 1,434 4.862 7.156 0 44.126 reer 1,819 117.215 68.9530 0.38 964.6 popg 2,210 2.060 1.753 -4.533 17.512 gdppc 1,884 10264.78 15661.57 13.898 111657.4 to 1,779 94.864 66.769 0.021 442.62 inf 1,707 14.294 108.183 -18.109 3373.76 fdi 1,879 3.391 5.813 -37.173 58.519 gds 1,703 23.285 23.985 -136.873 87.827 polity2 1,563 -1.287 6.822 -10 10 source: own estimation then to explore the expected relationship between variables, the correlation matrix is presented in table 4.2 presented the relationship between the dependent variable (human iqra noor, babar hussain & umbreen zahra 54 capital) and explanatory variables personal remittances, gdp per capita, trade openness, fdi and savings have a positive relationship with human capital. while real effective exchange rate population and inflation have a negative relationship with human capital. table 4.2: correlation matrix var hc remitt reer pop gdp pc to inf fdi gds polity 2 hc 1 remitt 0.097 1 rerr -0.11 -0.01 1 pop -0.19 -0.009 0.07 1 gdppc 0.49 -0.283 0.062 0.371 1 to 0.336 0.07 -0.031 0.133 0.29 1 inf -0.18 -0.048 -0.244 -0.08 -0.146 -0.15 1 fdi 0.316 0.08 2e-04 0.371 0.172 0.576 -0.08 1 gsd 0.086 -0.676 0.037 0.216 0.515 0.155 -0.08 0.052 1 polity 2 0.086 0.013 -0.124 -0.28 -0.13 -0.17 0.011 -0.13 -0.2 1 pooled regression and fixed effect model results report in table 4.3 and table 4.4. both have the same effects but the magnitude of all coefficients is different. worker’s remittances have a statistically significant and positive impact on human capital at the 1% in all columns of table 4.3 and table 4.4. estimated results of remittances concluded that remittances play an important role in enhancing human capital development in asia. the significance of remittances shows that an increase in remittances loosens income constraints, especially in developing countries. as emigrants get better income opportunities in host countries, it prompts them to send a handsome amount of remittances back for increasing investment in human capital and enhancing welfare by keeping altruistic motives in mind. these findings are consistent with the studies of (adenutsi, 2010; azam & raza, 2016; huay et al., 2019; mohamed & sivarajasingham, 2019). the coefficient of the real effective exchange rate is significant and has an inverse impact on human capital. when the exchange rate appreciates in an economy due to any reason, it causes to slowdown economic activities, income level and lowers the investment in human capital (gylfason et al., 1999). journal of applied economics and business studies, volume. 6, issue 3 (2022) 45-66 https://doi.org/10.34260/jaebs.633 55 table 4.3: pooled regression regarding the interaction of remittances and exchange rate the value of the coefficient of interaction in both tables’ column 4 exhibited that when exchange rate appreciates it decreases the remittances growth in terms of domestic currency. government and people face a decrease in income earned through migrant’s remittances as a result they decrease to investing in human capital. it means that exchange rate appreciation substitutes (weaker) the remittance’s impact on human capital (lin, 2011; nekoei, 2013; yang, 2008). variables (1) baseline (2) (remitt + reer) (3) (remitt + reer + control) (4) (remitt + er + control + interaction) constant 2.337*** (0.000) 2.425*** (0.000) 1.888*** (0.000) 1.823*** (0.000) remitt 0.990*** (0.000) 0.104*** (0.000) 0.852*** (0.000) 0.243*** (0.000) reer -0.678*** (0.000) -0.110*** (0.000) -0.669** (0.044) remitt* reer -0.139*** (0.012) pop -0.822*** (0.000) -0.810*** (0.000) gdppc 0.484*** (0.000) 0.492*** (0.000) to 0.195*** (0.000) 0.203*** (0.000) inf -0.282*** (0.000) -0.275*** (0.000) fdi 0.966*** (0.000) 0.937*** (0.000) gds 0.391*** (0.000) 0.380*** (0.000) polity2 0.849*** (0.000) 0.784*** (0.000) observations 1,047 1,008 782 782 countries 36 34 29 29 wald-test 20.84*** (0.000) 35.36*** (0.000) 804.04*** (0.000) 820.26*** (0.000) r-square 0.0204 0.0349 0.5375 0.5425 note: dependent variable is human capital index, remitt is remittances inflow, reer is index of real effective exchange rate, pop is population growth, gdppc is per capita gdp, to is trade openness, inf is inflation, fdi is foreign direct investment, gds is gross domestic savings, polity2 is used for democracy. in parentheses p-value is given, *, **, *** suggest level of significance respectively at 10%, 5% and 1% level. iqra noor, babar hussain & umbreen zahra 56 table 4.4: fixed effects regression variables (1) baseline (2) (remitt + reer) (3) (remitt + reer + control) (4) (remitt + er + control + interaction) constant 2.281*** (0.000) 2.362*** (0.000) 1.803*** (0.000) 1.743*** (0.000) remitt 0.100*** (0.000) 0.106*** (0.000) 0.766*** (0.000) 0.242*** (0.000) reer -0.667*** (0.000) -0.109*** (0.000) -0.649** (0.045) remitt*reer -0.146*** (0.007) pop -0.744*** (0.000) -0.734*** (0.000) gdppc 0.611*** (0.000) 0.615*** (0.000) to 0.233*** (0.000) 0.241*** (0.000) inf -0.242*** (0.000) -0.237*** (0.000) fdi 0.914*** (0.000) 0.885*** (0.000) gds 0.405*** (0.000) 0.391*** (0.000) polity2 0.817*** (0.000) 0.748*** (0.000) observation 1,047 1,008 782 782 countries 36 34 29 29 f-stat (p-value) 21.02*** (0.000) 17.58*** (0.000) 99.21*** (0.000) 90.78*** (0.000) r-square 0.0204 0.0349 0.5455 0.5499 hausman test 66.31*** (0.000) 52.36*** (0.000) note: see table (4.3) for level of significance and variable description. hausman test suggest that fixed effect model is appropriate for this study. regarding the group of control variables, population annual percentage growth has a negative and statistically significant impact on human capital at the 1 percent level. with an increase in population size burden on the available resources of the finance sector increases. it became difficult for families and the government to maintain existing education levels. as a result, population growth decreases human capital investment (hassan et al., 2013; mohammed, 2021; rosenzweig, 1988). journal of applied economics and business studies, volume. 6, issue 3 (2022) 45-66 https://doi.org/10.34260/jaebs.633 57 table 4.5: interactive effect of remittances and exchange rate gdp per capita is used as a proxy to measure the economic development level for countries. the coefficient value of gdp per capita affects human capital positively and significantly at a 1 percent level. when economic development expands, it increases income per capita and wealth. capacities of people and countries increase to build skills, training, and investment in human capital (cohen & soto, 2007; mozumder & islam, 2017; tsaurai & ngcobo, 2018). the variable of trade as a percentage of gdp is used to measure trade openness. results reveal that the coefficient of trade openness has a positive and highly significant impact on variables pols fem rem constant 1.823*** (0.000) 1.743*** (0.000) 1.8227* (0.000) remitt 0.243*** (0.000) 0.242*** (0.000) 0.243*** (0.000) reer -0.669** (0.044) -0.649** (0.045) -0.669** (0.044) remitt * reer -0.139*** (0.012) -0.146*** (0.007) -0.139*** (0.012) pop -0.810*** (0.000) -0.734*** (0.000) -0.810*** (0.000) gdppc 0.492*** (0.000) 0.615*** (0.000) 0.492*** (0.000) to 0.203*** (0.000) 0.241*** (0.000) 0.203*** (0.000) inf -0.275*** (0.000) -0.237*** (0.000) -0.275*** (0.000) fdi 0.937*** (0.000) 0.885*** (0.000) 0.937*** (0.000) gds 0.380*** (0.000) 0.391*** (0.000) 0 .380*** (0.000) polity2 0.784*** (0.000) 0.748*** (0.000) 0.784*** (0.000) observations 782 782 782 countries 29 29 29 wald-test, f-stat (p-value) 820.26*** (0.000) 90.78*** (0.000) 820.26*** (0.000) r-square 0.5425 0.5499 0.5425 hausman test 52.36*** (0.000) note: see table (4.3) for the level of significance and variable description. pols is pooled regression, fem and rem is (fixed/random) effect models. hausman test suggest that fixed effect model is appropriate for this study. iqra noor, babar hussain & umbreen zahra 58 years of schooling at a 1% level. the main reason for low human capital investment is financial barriers. these financial barriers are decreased by financial development through an increase in the level of trade openness. it boosts education expenditures by the government and investment in human capital (chinn & ito, 2002; mozumder & islam, 2017; ozcan, 2018). the consumer price index is used to measure inflation’s impact on human capital and it effect on years of schooling adversely. because the general level rise in prices affect negatively general public and government budget and erodes the purchasing power of money. with the rising levels of prices, it became difficult for people to maintain the existing level of education, because they have to spend their resources fulfilling their basic necessities instead of investing in education (kamalu & ibrahim, 2021; sahoo & sethi, 2020). foreign direct inflows positively and significantly affect human capital at a 1% level. when foreign investment increase in country, it affects human capital directly and indirectly. through direct means, foreign companies provide scholarship opportunities to their employees to upgrade their technical know-how and skills. through indirect means fdi provides employment opportunities in the host countries, it pushes financial development and income level up that support investment in human capital (blomström & kokko, 2002; michie, 2001; tsaurai & ngcobo, 2018). proxy of gross domestic savings is used here to check savings impact on average year schooling and returns to education. savings exert a positive, and significant effect at a 1% level on human capital. higher savings rate at the aggregate level significantly contributes to improving per capita income. likewise, when gross domestic savings increase, it endorses long-run productive investments such as education, and health because, the financial sector of any country is able to invest in the productive sector and to fulfill private sector credit demands due to higher domestic savings (leff, 1969; mozumder & islam, 2017). polity2 is the composite measure used as a proxy of democratic development. polity2 affects positively, and significantly at 1% level to human capital. a government that is selected through a democratic way, spends more on education (klomp & de haan, 2013; mozumder & islam, 2017). regression analysis of pols, fem, and rem displayed that remittances have a direct and positive impact on hc and confirmed the permanent income hypothesis and altruistic model of sending remittances. the exchange rate has a direct but negative impact on hc. indirect impact represents exchange rate weaker the remittances effect on human capital. to find the overall impact of variables through percentile analysis table 4.5 is build. journal of applied economics and business studies, volume. 6, issue 3 (2022) 45-66 https://doi.org/10.34260/jaebs.633 59 table 4.6 is developed by using table 4.5. 25th, 50th, and 75th percentile of exchange rate and remittances are considered for estimation in the table given below. table 4.6: impact of remittances at different levels of exchange rate results of pols, fem and rem in table 4.6 indicate that the remittances effect positive the human capital at all percentile of real effective exchange rate. conclusion this paper uses panel data set to analyze the impact of remittances, exchange rate on human capital of 54 asian countries from 1980 to 2020. techniques of pooled and fixed effect models were applied for panel data analysis. hausman test was utilized to decide among (fixed and random) effect models. at last, percentile analysis was carried out to check the overall impact of interaction term. our pooled and fixed effects estimations reveal that the direct effects of remittances on human capital are positive and significant. it indicates that remittances are fostering human capital in asia. as an increase in remittances loosens income constraints, especially in developing countries and pave the way for investment in human capital in term of education. according to the direct effects of exchange rate on human capital, there exist negative relationship between remittances and human capital. if a country exposed to any external or internal shock that leads to increase in the exchange rate (appreciation) in the home country. which slowed down the overall economic activities in the economy by decreasing foreign exports, and manufacturing sector production. as a result, it pushed down the growth rate and investment in human capital. the indirect effects of remittances through exchange rate exerted a negative effect and explain that exchange rate substitutes the remittance’s ability to increase human capital. when the exchange rate increases its lead to decrease remittances growth as the amount of remittances decreases in terms of domestic currency. government and people face a decrease in income earned through migrant’s remittances as a result they decrease to invest in human capital. reer pols fem rem p25= 93.55 0.011*** (0.000) 0.010*** (0.000) 0.011*** (0.000) p50= 105.19 0.976*** (0.000) 0.089*** (0.000) 0.976*** (0.000) p75= 122.63 0.727*** (0.000) 0.632*** (0.000) 0.727*** (0.000) notes: pols is pooled regression, fem and rem is (fixed/random) effect models. p25, p50, and p75 are the 25th percentile, 50th percentile and 75th percentile coefficient values of reer respectively. *, **, *** are 10, 5 and 1 % significance level respectively. iqra noor, babar hussain & umbreen zahra 60 references adenutsi, d. e. (2010). long-run macroeconomic impact of international migrant remittances on human development in low-income countries: a panel analysis of sub-saharan africa. journal of international economic studies(24), 113-132. akanbi, o. a. (2017). impact of migration on economic growth and human development: case of sub-saharan african countries. international journal of social economics, 44(5), 683-695. al-thani, m. h., sadoun, e., al-thani, a. a., khalifa, s. a., sayegh, s., & badawi, a. (2014). change in the structures, dynamics and disease-related mortality rates of the population of qatari nationals: 2007–2011. journal of epidemiology and global health, 4(4), 277-287. amakom, u., & iheoma, c. g. (2014). impact of migrant remittances on health and education outcomes in sub-saharan africa. iosr journal of humanities and social science, 19(8), 33-44. awusi, e. o. (2016). the impact of remittances on human capital formation in developing and developed countries: a comparative panel data approach. azam, m., & raza, s. a. (2016). do workers' remittances boost human capital development? the pakistan development review, 55(2), 26. azizi, s. (2018). the impacts of workers' remittances on human capital and labor supply in developing countries. economic modelling, 75, 377-396. becker, g. s. (1974). a theory of social interactions. journal of political economy, 82(6), 30. blomström, m., & kokko, a. (2002). fdi and human capital: a research agenda. borja, k. (2020). remittances, corruption, and human development in latin america. studies in comparative international development, 55(3), 305-327. calvo, g. a., & reinhart, c. m. (2002). fear of floating. the quarterly journal of economics, 117(2), 379-408. chinn, m. d., & ito, h. (2002). capital account liberalization, institutions and financial development: cross country evidence: national bureau of economic research cambridge, mass., usa. cohen, d., & soto, m. (2007). growth and human capital: good data, good results. journal of economic growth, 12(1), 51-76. dietz, b., gatskova, k., & ivlevs, a. (2015). emigration, remittances and the education of children staying behind: evidence from tajikistan. dzeha, g. c. o., turkson, f. e., agbloyor, e. k., & abor, j. y. (2018). total factor productivity growth and human development: the role of remittances in africa. african journal of economic and sustainable development, 7(1), 47-72. emmanuel, o. d. (2017). exchange rate deregulation and human capital. department of economics, faculty of social sciences, kogi state university. journal of applied economics and business studies, volume. 6, issue 3 (2022) 45-66 https://doi.org/10.34260/jaebs.633 61 fleming, j. m. (1962). domestic financial policies under fixed and under floating exchange rates. staff papers, 9(3), 369-380. freitas, t.(2013). country risk analysis: a macroeconomic risk (presented at development future conference, 22-23 nov 2013, uts). friedman, m. (1957). the permanent income hypothesis a theory of the consumption function (pp. 20-37): princeton university press. guo, m., zhang, y., & ye, j. (2019). does a foreign degree pay? the return to foreign education in china. review of development economics, 23(1), 415-434. gyimah brempong, k., & asiedu, e. (2015). remittances and investment in education: evidence from ghana. the journal of international trade & economic development, 24(2), 173-200. gylfason, t., herbertsson, t. t., & zoega, g. (1999). a mixed blessing: natural resources and economic growth. macroeconomic dynamics, 3(2), 204-225. hassan, m. u., mehmood, h., & hassan, m. s. (2013). consequences of worker s remittances on human capital: an in-depth investigation for a case of pakistan. middle-east journal of scentific research, 14(3), 443-452. hines, a. l., & simpson, n. b. (2019). migration, remittances and human capital investment in kenya. economic notes: review of banking, finance and monetary economics, 48(3), e12142. huay, c. s., winterton, j., bani, y., & matemilola, b. t. (2019). do remittances promote human development? empirical evidence from developing countries. international journal of social economics, 46(10), 1173-1185. javid, m., & ahmad, m. (2014). indian falling rupee impacting foreign education. available at ssrn 2401294. jeanneney, s. g., & hua, p. (2011). how does real exchange rate influence labour productivity in china? china economic review, 22(4), 628-645. jijin, p., mishra, a. k., & nithin, m. (2021). macroeconomic determinants of remittances to india. economic change and restructuring, 1-20. kamalu, k., & ibrahim, w. h. b. w. (2021). foreign remittances, good governance and human development in sub-saharan africa: evidence from fmols and dols. the journal of management theory and practice (jmtp), 38-48. kaur, n., & sirohi, r. (2013). effect of rupee depreciation on common man. international journal of scientific and research publications, 3(10), 1-14. khan, h. u., & khan, w. (2017). syria: history, the civil war and peace prospects. journal of political studies, 24(2). klomp, j., & de haan, j. (2013). political regime and human capital: a cross-country analysis. social indicators research, 111(1), 45-73. kuncoro, h. (2020). the role of exchange rate in remittance inflows: evidence from indonesia. economics bulletin, 40(2), 1508-1521. leff, n. h. (1969). dependency rates and savings rates. the american economic review, 59(5), 886-896. iqra noor, babar hussain & umbreen zahra 62 lin, m. h. h. (2011). determinants of remittances: evidence from tonga: international monetary fund. lucas, r. e., & stark, o. (1985). motivations to remit: evidence from botswana. journal of political economy, 93(5), 901-918. melton, r. (1965). schultz's theory of" human capital". the southwestern social science quarterly, 46(3), 264-272. michie, j. (2001). the impact of foreign direct investment on human capital enhancement in developing countries. report for the organization for economic co-operation and development (oecd). mohamed, a. a., & sivarajasingham, s. (2019). dynamic relationship between workers’ remittances and human capital formation in sri lanka. 6(7), 11. mohammed, u. (2021). remittances, institutions and human development in sub-saharan africa. journal of economics and development. mozumder, l., & islam, m. a. (2017). effects of remittances on human capital development: an empirical analysis. bangladesh journal of agricultural economics, 36(454-2017-036), 18-11. mundell, r. a. (1963). capital mobility and stabilization policy under fixed and flexible exchange rates. canadian journal of economics and political science/revue canadienne de economiques et science politique, 29(4), 475-485. naeem, m. z., & arzu, s. (2017). the role of remittances on human development: evidence from developing countries. bulletin of business and economics (bbe), 6(2), 74-91. nekoei, a. (2013). immigrants' labor supply and exchange rate volatility. american economic journal: applied economics, 5(4), 144-164. ngoma, a. l., & ismail, n. w. (2013). do migrant remittances promote human capital formation? evidence from 89 developing countries. migration and development, 2(1), 106-116. nguyen, c. v., & nguyen, h. q. (2015). do internal and international remittances matter to health, education and labor of children and adolescents? the case of vietnam. children and youth services review, 58(c), 28-34. ogege, s. (2019). analysis of the impact of inflation, interest rate and exchange rate on economic development. international journal of commerce and finance, 5(1), 121132. olubiyi, e. (2015). does exchange rate affect remittances in nigeria? the review of finance and banking, 7(1), 31-45. omon, i. j. (2021). determinants of migrants’ remittances in the west africa monetary zones (wamz). journal of research in humanities and social science, 9(1), 8795. orji, a., damdinsuren, a., ramgoolam, g., perez, m., & dzvukutu, t. fdi and international investments: experiences in some selected countries. development initiatives and south-south cooperation, 1. journal of applied economics and business studies, volume. 6, issue 3 (2022) 45-66 https://doi.org/10.34260/jaebs.633 63 ozcan, b. (2018). the impact of financial development on human capital: evidence from emerging market economies. international journal of economics and financial issues, 8(1), 258. pant, b., & budha, b. b. (2016). remittances and exchange rate linkages: experiences of nepal. nepal rastra bank working paper, 33. pilařová, t., & kandakov, a. (2017). the impact of remittances on school attendance: the evidence from the republic of moldova. international journal of educational development, 55, 11-16. rahman, m. r., selim, m. m. i., haque, m. a. u., smrity, d. y., & islam, m. s. (2019). examining the effect of exchange rate and overseas employment on remittances: an evidence from bangladesh. theoretical economics letters, 9(7), 2593-2605. romer, p. m. (1990). endogenous technological change. journal of political economy, 98(5), s71-s102. rosenzweig, m. r. (1988). human capital, population growth, and economic development: beyond correlations. journal of policy modeling, 10(1), 83-111. sahoo, m., & sethi, n. (2020). does remittance inflow promote human development in subsaharan africa? an empirical insight. global economy journal, 20(04), 1-23. salas, v. b. (2014). international remittances and human capital formation. world development, 59(c), 224-237. schultz, t. w. (1961). investment in human capital. the american economic review, 51(1), 1-17. sghaier, i. m. (2021). remittances and economic growth in mena countries: the role of financial development. econ. altern(1), 43-59. seshan, g. (2012). migrants in qatar: a socio-economic profile. journal of arabian studies, 2(2), 157-171. shifflett, z. (2021). the ends of four modern hyperinflations (doctoral dissertation). tsaurai, k., & ngcobo, l. (2018). the role of liquidity in the remittances-human capital development nexus in emerging economies. acta universitatis danubius. œconomica, 14(7), 101-119. world bank, (2019, april). migration and remitances. recent developments and outlook. migration and development brief, 31 unctad, g. (2016). world investment report: investor nationality: policy challenges. geneva: unctad, 1-232. yang, d. (2005). international migration, human capital, and entrepreneurship: evidence from philippine migrants' exchange rate shocks (vol. 3578): world bank publications. yang, d. (2008). international migration, remittances and household investment: evidence from philippine migrants’ exchange rate shocks. the economic journal, 118(528), 591-630. ziesemer, t. h. (2012). worker remittances, migration, accumulation and growth in poor developing countries: survey and analysis of direct and indirect effects. economic iqra noor, babar hussain & umbreen zahra 64 modelling, 29(2), 103-118. appendix table a1: list of countries included into sample status of countries countries status of countries countries status of countries countries li afghanistan umi jordan hi qatar umi armenia umi kazakhstan lmi philippine umi azerbaijan lmi kiribati lmi samoa hi bahrain hi korea, rep hi saudi arabia lmi bangladesh hi kuwait hi singapore lmi bhutan lmi kyrgyzstan lmi solomon island hi brunei darussalam lmi lao democratic republic lmi sri lanka lmi cambodia umi lebanon li syria li dpr korea hi macao sar lmi tajikistan umi fiji umi malaysia umi thailand umi georgia umi maldives lmi timor-leste hi hong kong lmi mongolia umi turkey lmi india lmi myanmar umi turkmenistan lmi indonesia lmi nepal hi uae lmi iran hi oman lmi uzbekistan umi iraq lmi pakistan lmi vanuatu hi israel lmi png lmi viet nam hi japan umi prc li yemen total = 18 total = 18 total = 18 * hi (high income), umi (upper middle income), lmi (lower middle income), li (low income) source: compiled by the researcher from (world development indicator, 2021). journal of applied economics and business studies, volume. 6, issue 3 (2022) 45-66 https://doi.org/10.34260/jaebs.633 65 table a2: compact prior literature on remittances effects on human capital macro studies author(s) country, period, technique dependent variable explanatory variables findings aslam and sivarajasingham (2019) sri lanka, f 1975 2017, ardl, dicky-fuller, ecm hdi remittances, education expenditures, health expenditures remittances have positive significant impact in long-run and short-run adenutsi (2010) sub saharan african countries, 1987-2007, fixed effect hd remittances, trade openness, inflation, education expenditures remittances have positive significant long run impact azam and raza (2016) (lower, lower middle, upper middle and high) income countries, 1996-2013, fixed effects secondary school enrollment remittances, gdp per capita, fdi, governance migrant remittances have positive significant impact, remittances impact became stronger through economic governance ngoma and ismail (2013) developing countries, 19702010, gmm average years of secondary and tertiary schooling remittances, gdp per capita, pop growth, remittances have a positive impact azizi (2018) developing countries, 19902015, iv approach education remittances, education expenditures, labor force participation rate, mortality rate remittances raise, school completion, school enrollment rate and private school enrollment mozumdar and islam (2013) global perspective,199602010, gls educational attainment, gross attendance and enrollment rate remittances, gdp per capita, gross domestic savings, polity2, inflation, trade openness, education expenditures remittances increase average years schooling and secondary enrollment tsaurai and ngcobo (2018) emerging economies, 19952014 , pooled ols, fixed effects, random effects hdi remittances, stock market, liquidity, inflation, trade, gdp per capita, openness, fdi, remittances have positive direct impact, stock market liquidity weaken remittances impact iqra noor, babar hussain & umbreen zahra 66 naeem and arzu (2017) developing countries, 2014, least square, arch hdi remittances, gender inequality index, trade as a gdp, government effectiveness remittances have positive but insignificant impact mohammed (2021) sub saharan african countries, 2004-2018, gmm hdi remittances, inflation, pop growth, democracy, economic freedom, gdp growth remittances have a positive impact, while institutions substitute the remittances impact on hd. hassan et al. (2013) pakistan, 19812011, ardl secondary school enrollment remittances, pop growth, per capita income, fdi, remittances have an adverse impact source: own compilation note: gls (generalized least square), ols (ordinary least square), iv (instrumental variable), (gmm (generalized method of the moment), ardl (auto regressive distributed lag bound testing), hdi (human development index), fdi (foreign direct investment), pop (population) journal of applied economics and business studies, volume. 6, issue 2 (2022) 77-94 https://doi.org/10.34260/jaebs.625 77 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x renewable and non-renewable energy consumption and economic growth in pakistan: a disaggregated analysis shazia farhat durrani1, inayatullah jan2 & sidra pervez3 1 institute of development studies (ids), the university of agriculture peshawar, pakistan 2 institute of development studies (ids), the university of agriculture peshawar, pakistan 3 department of business administration, iqra university, islamabad, pakistan abstract this study addresses the research question of how renewable and non-renewable energy consumption (ec) affects economic growth (gdp) in pakistan over a period of 1972-2015. the study extends the basic production function having labor and capital as the mainstream variables by adding major energy sources of pakistan. the results of toda-yamamoto granger causality test confirm that no causality exists between gdp and labor force and bidirectional causality exists between gdp and gross capital formation. moreover, the results confirm a unidirectional relationship (growth hypothesis) between hydroelectricity consumption (hec) and nuclear energy consumption (nec) and a bidirectional relationship (feedback hypothesis) between fossil fuel consumption (ffc) and gdp, respectively. the findings suggest for an efficient utilization of existing energy resources along with diversification and expansion of the renewable energy resources. the study recommends for the government policy to avoid energy conservation as it can hamper gdp growth in pakistan. keywords energy consumption, economic growth, casual relationship, disaggregated analysis, government policy jel classification q13, q42, q48 1. introduction the mainstream economic theory of production considers labor and capital as the main factors of production. nevertheless, the neo-classical aggregate production function complements energy as an additional and necessary input factor in the production model (shahbaz et al., 2014; chiou-wei et al., 2016). in a broader sense, energy is a vital input for shazia farhat durrani, inayatullah jan & sidra pervez 78 all production processes (azam et al., 2015). it is required for domestic, agricultural, industrial, and transportation purposes (kahouli, 2017). thus, a secure, adequate, and accessible energy supply is important for socioeconomic development of a country (jan et al., 2017; rafindadi & ozturk, 2017; durrani et al., 2021; li et al., 2022; wang et al., 2022). pakistan is an energy deficient country (jan, 2012; javed et al., 2016; jan & lohano, 2021). over the last two decades, pakistan is trapped in the worst crisis of energy (javid et al., 2013; jan et al., 2017) which has severely affected economic growth (gdp) in the country (jan & akram, 2018). fossil fuels, renewable energy (hydroelectricity), and nuclear energy are the major sources constituting pakistan’s total energy mix. in pakistan, energy consumption (ec) and gdp growth are highly correlated (jan & akram, 2018). figure 1 illustrates the historical trend in total ec and gdp in pakistan. the figure shows that increase in gdp is accompanied by increased ec. during 1972-2015, pakistan’s fossil fuels consumption has increased from 7.039 to 68.870 million tons of oil equivalent (mtoe) whereas hydroelectricity consumption (hec) has increased from 4 to 34.6 terawatt-hour (twh). until 1999, there was no considerable change in nuclear energy consumption (nec). however, after 1999 the consumption of nuclear energy started to grow and reached to 4.7 twh by 2015 (bps, 2017). figure 1: disaggregated ec and gdp growth in pakistan, 1972-2015 (bps, 2016; world bank, 2017) 0 50 100 150 200 250 0 10 20 30 40 50 60 70 80 1 9 7 2 1 9 7 3 1 9 7 4 1 9 7 5 1 9 7 6 1 9 7 7 1 9 7 8 1 9 7 9 1 9 8 0 1 9 8 1 1 9 8 2 1 9 8 3 1 9 8 4 1 9 8 5 1 9 8 6 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 1 9 9 7 1 9 9 8 1 9 9 9 2 0 0 0 2 0 0 1 2 0 0 2 2 0 0 3 2 0 0 4 2 0 0 5 2 0 0 6 2 0 0 7 2 0 0 8 2 0 0 9 2 0 1 0 2 0 1 1 2 0 1 2 2 0 1 3 2 0 1 4 2 0 1 5 b il li o n $ / m t o e t w h fossil fuel consumption (mtoe) hydroelectricity consumption (terawatt-hours) nuclear consumption(terawatt-hours) gdp at market prices (in billions,constant 2010 us$) journal of applied economics and business studies, volume. 6, issue 2 (2022) 77-94 https://doi.org/10.34260/jaebs.625 79 this study is motivated by the fact that ec is an important determinant of gdp growth. in countries like pakistan, high rates of ec are highly correlated with high gdp growth (jan & akram, 2018). considering this scenario, this paper attempts to address the research question that what is the causal link between energy consumption and economic growth in pakistan? a number of studies have examined the ec-gdp nexus (jan et al., 2020; durrani et al., 2021; husaini & lean, 2022; oryani et al., 2022; wang et al., 2022; zhang et al., 2022). however, this study is different from other studies in several aspects. firstly, we determine the ec-gdp relationship by taking labor and capital as additional variables, and thereby, evade the problem of specification error that could possibly arise by the omission of relevant variables from the model. secondly, we carry out a disaggregated analysis which gives us source specific results regarding different types of energy sources in pakistan. to our knowledge, none or very few studies investigate the energy-growth relationship by employing fossil fuel consumption (ffc), hec, and nec altogether in a single study. thirdly, we used toda-yamamoto (t-y) causality test for detecting the direction of causal relationship. the t-y causality approach allows us to test for cointegration even if the variables are integrated of order i(0) or i(1) or the combination of both orders i.e. i(0) and i(1). the approach can also be used disregarding either the variables are cointegrated or not. fourthly and most importantly, the previous studies have explored the ec and gdp relationship without considering structural break in the analysis. we use break point test to determine the break point while estimating the energy-growth nexus. the combination of these different methodological approaches, different data period, and country specific outcomes and inferences make the paper novel and an original contribution to literature. this paper is organized in various sections in the following manner. after introduction, we present a literature review. in section three, we provide details of the methodology used during this research. in this section, we provide relevant information on the data, variables, and econometric technique used in this study. section four is related to the empirical results of unit roots test and t-y granger causality test. in the last section, we provide conclusion and policy implications. 2. literature review literature on energy-growth nexus provides evidence of mixed and conflicting results (yuan et al., 2008; jan et al., 2020; oryani et al., 2022). the conflicting nature of results is because of the heterogeneity of data sets and temporal and methodological variations in various studies. based on the causality between ec and gdp, four types of hypotheses have shazia farhat durrani, inayatullah jan & sidra pervez 80 been identified in literature (apergis et al., 2010; ikhide & adjasi, 2015; marques et al., 2016; thao & chon, 2016; zaidi & ferhi, 2019; durrani et al., 2021; filippidis et al., 2021; husaini & lean, 2022; zhang et al., 2022). in the first case, the results exhibit a causality that runs from gdp to ec (kraft & kraft, 1978; al-iriani, 2006; li et al., 2011; ouedraogo, 2013; dudzevičiūtė & šimelytė, 2017; furouka, 2017). this kind of hypothesis is referred to as conservation hypothesis. the conservation hypothesis confirms that gdp is the major driver of ec (oryani et al., 2022). this hypothesis recommends for energy conservation policies having little or no effects on gdp. in the second case, ec leads to gdp, and is referred to as growth hypothesis (siddiqui, 2004; kakar & khilji, 2011; arouri et al., 2014; mutascu, 2016; gozgor et al., 2018; husaini & lean, 2022). in this case, energy conservation policies are not recommended as they may negatively affect gdp. the growth hypothesis suggests for increased energy production and consumption which flourishes gdp. in the third case, a bidirectional relationship is asserted between ec and gdp. this type of relationship is summed up in the feedback hypothesis (omri & chaibi, 2014; alper & oguz, 2016; kahia et al., 2016; rodríguez-caballero & ventosa-santaulària, 2016; tiba & omri, 2017; durrani et al., 2021;). according to this hypothesis, ec and gdp are complementary to each other and a change in one causes a change in the other. this is the reason why feedback hypothesis emphasizes on energy exploration and efficiency policies. the fourth case, the neutrality hypothesis, confirms no causality between ec and gdp (zhang & cheng, 2009; jalil & feridun, 2014; yildirim et al., 2014).the neutrality hypothesis calls for energy efficiency policies. a recap of literature on causality between ec and gdp is provided in table 1. table 1: literature on ec-gdp nexus study period country methods relationship omri & chaibi (2014) 19902011 developed and developing countries dsems and gmm ec↔gdp for pakistan pin (2014) 1982-2011 oecd countries ardl bounds test, vecm granger causality mixed results ahmed & azam (2015) 30 years, varying 119 countries granger causality test mixed results alper & oguz (2016) 1990–2009 new eu member countries asymmetric causality test and ardl bounds test ecgdp (czech republic) destek (2016) 1971-2011 newly asymmetric ec− gdp https://www.sciencedirect.com/science/article/pii/s1364032116001787#! https://www.sciencedirect.com/science/article/pii/s1364032116001787#! https://www.sciencedirect.com/science/article/pii/s0960148116303512#! journal of applied economics and business studies, volume. 6, issue 2 (2022) 77-94 https://doi.org/10.34260/jaebs.625 81 industrialized countries causality approach (brazil and malaysia) luke (2016) 1990-2011 sub-sahara african countries panel cointegration, granger causality test ecgdp marques et al., (2016) 1965-2013 global ardl bound test and t-y causality test ec↔gdp nadeem & munir (2016) 1972-2014 pakistan ardl bound test, granger causality test mixed results narayan (2016) 1984-2010 135 countries panel data predictive regression model ec− gdp tang et al., (2016) 1971-2011 vietnam cointegration and granger causality ec→gdp thao & chon (2016) 1990-2012 oecd countries stochastic distance function ec→gdp carmona et al., (2017) 1980-2013 oil exporting countries cointegration and granger causality ec↔gdp koçak & şarkgüneşi (2017) 1990–2012 black sea and balkan countries panel cointegration, heterogeneous panel causality approach ec−gdp (turkey) gozgor et al., 2018 1990-2013 oecd countries panel ardl and pqr ec→gdp jan et al., 2020 1972–2015 pakistan ardl bound test ec→gdp durrani et al., 2021 1972-2015 pakistan t-y causality test ec↔gdp husaini & lean, 2022 1980-2017 asia threshold estimation ec→gdp oryani et al., 2022 1976-2017 iran ardl, cointegration test, causality ecgdp note: ec→gdp shows the direction of causality running from ec to gdp. eggdp signifies the direction of causality running from gdp to ec. ec↔gdp symbolizes bidirectional causality between gdp and ec. ec−gdp means no causality between ec and gdp. ardl means autoregressive distributed lag, vecm means vector error correction model, t-y means toda-yamamoto causality test, dsems means dynamic simultaneousequation models and gmm means generalized method of moments. https://www.sciencedirect.com/science/article/pii/s030142151630550x#! https://www.sciencedirect.com/science/article/pii/s030142151630550x#! shazia farhat durrani, inayatullah jan & sidra pervez 82 3. methodology 3.1.data sources and variables in this study, we employed the basic production function containing only two variables, i.e. capital and labor and extended this basic function by adding ffc, hec, and nec in it. annual time series data on gdp, gross capital formation (k), labor (l), ffc, hec and nec for pakistan for the period 1972-2015 has been used in this study. data on gross domestic product (gdp) and gross capital formation was obtained from the world development indicators (wdi) database (wb, 2017). data on labor force was taken from various issues of the economic survey of pakistan (published by the ministry of finance, government of pakistan). data on fossil fuels consumption, hec, and nec was retrieved from british petroleum’s (bps) statistical review of world energy 2016 (bps, 2017). we use gdp in constant us$ 2010 as a dependent variable (li et al., 2011). explanatory variables include labor, gross capital formation, ffc, hec, and nec. labor is measured in million whereas gross capital formation is measured in constant us$ 2010. fossil fuels consumption is measured in million tons of oil equivalent (mtoe) whereas hec and nec are measured in terawatt-hour (twh) (jan et al., 2020; durrani et al., 2021). all the variables were measured in natural logarithms. e-views v.10 was used for data analysis. 3.2.model specification we examined the direction of causality between pakistan’s major sources of ec (at disaggregated level) and gdp using the following basic model: ( , , , , ) t t t t t t y f l k ffc hec nec= (1) where y denotes gdp, l denotes labor, k denotes gross capital formation, ffc means fossil fuel consumption, hec means hydroelectricity consumption, and nec means nuclear energy consumption. all of the study variables are converted into log form. the econometric model to be estimated is: 1 2 3 4 5t o t t t t t t lny lnl lnk lnffc lnhec lnnec      += + + + + + (2) where 𝛽0 = intercept, 𝛽1 to 𝛽5= coefficients that are interpreted as elasticity in logarithmic models, 𝜀𝑡 = error term in time t. 3.3.unit root tests it is essential to check time series data for the unit root. in the presence of a unit root, the model will generate spurious, biased and meaningless results (gujarati & porter, 2009). we journal of applied economics and business studies, volume. 6, issue 2 (2022) 77-94 https://doi.org/10.34260/jaebs.625 83 used the augmented dickey-fuller (adf) (chiou-wei, 2008) and break point (bp) (lee, 2006) unit root tests to avoid unit root problem. we conducted unit root tests both with (eq. 3) and without allowing for a time trend (eq. 4) (oh & lee, 2004). where, ∆yt means 1st differenced value of variable to be tested in time t; α means intercept; βt means time trend; yt−1means the first lag of variable; δ means parameter to be estimated; p means number of lags; and εt means error term in time t. our null and alternative hypotheses are: h0: δ = 0 (depicting nonstationarity) and ha: δ < 0 (depicting stationarity). the null hypothesis of unit root tests is compared with the significance levels of 1%, 5%, and 10%. in order to reject the null hypothesis, the probability value of adf or bp statistics should be less than the specified significance level. besides, if the probability of trend is found significant at 1%, 5%, and 10% level, then the results of model with intercept and trend are accepted (i.e. eq. 3). nonetheless, if the trend is found insignificant, then the decision about stationarity of a variable is made on the basis of model with intercept only (i.e. eq. 4). 3.4.optimum lag selection in this study, the akaike’s information criterion (aic) has been used for selecting optimum number of lags for the model. the aic is significant over other criteria if the number of observations is small (liew & khimm, 2004). for 60 observations or below, aic is a more reliable and accurate criterion. 3.5.toda-yamamoto causality test the direction of causality was determined by using toda-yamamoto causality test (toda & yamamoto, 1995; leiva & liu, 2018). t-y test is preferred for determining causality because this test can be applied without considering the integration order of the selected variables. it means that we can use this test if all variables are integrated at levels or at difference or both. this test can also be applied regardless of the presence or absence of cointegration (soytas & sari, 2009). the following general form of the equations has to be estimated: 1 1 11 p t t t t ti y y y     − −=  = + + + + (3) 1 1 11 p t t t ti y y y    − −=  = + + + (4) 1 1 t h d k d t i t i j t j i j y y x    + + − − = = = + + +  (5) shazia farhat durrani, inayatullah jan & sidra pervez 84 where, d is the maximum order of integration of the variables; h and k are optimum lags of y and x, and εt is the error term. 4. analysis and discussion 4.1.unit root tests the results of adf and bp unit root tests for gdp (y), l, k, ffc, hec, and nec along with their order of integration are illustrated in table 2. the table confirms that the unit root tests produce mixed results about the variables being i(0) and i(1). integration order of each variable is decided following adf test that does not consider structural break in the data series and a break point unit root test that considers a single structural break when testing for unit root. table 2: unit root tests results variable adf decision bp decision level ι lny -1.858222 -0.2666 (1992) lnl -0.696935 -3.862556 (1996) lnk -3.382422** i(0) -4.532775 (2004) i(0) lnffc -3.268340** i(0) -4.407659 (1978) lnhec -2.267476 -3.713505(2003) lnnec -1.403973 -4.306537**(1999) ι & γ lny -1.020664 -1.013616 (2009) lnl -1.681005 -0.719640 (1996) lnk -2.316333 -5.130789** (1991) lnffc 0.396322 -2.726607 (2005) lnhec -1.966210 -4.169948(1986) lnnec -2.900759 -5.606595**(2002) first difference ι lny -4.359346*** -5.162705** (1992) lnl -6.962707*** -7.16885*** (2010) lnk -5.752976 -4.84811*** (1993) lnffc -4.214846*** -5.82469*** (2004) lnhec 6.966780*** -7.29408***(1988) lnnec -6.905983*** -8.06017***(1999) ι & γ lny -4.781955*** i(1) -5.161639 (2003) i(1) lnl -6.909658*** i(1) -8.75798*** (1996) i(1) 1 1 t h d k d t i t i j t j i j x x y    + + − − = = = + + +  (6) journal of applied economics and business studies, volume. 6, issue 2 (2022) 77-94 https://doi.org/10.34260/jaebs.625 85 lnk -5.952431*** -6.34264*** (2005) lnffc -4.988945*** -5.519503** (2003) i(1) lnhec -7.432628*** i(1) -7.57556***(1979) i(1) lnnec -6.881003*** i(1) -10.2694***(1999) i(1) note: ***, **, * represents significance at 1%, 5%, and 10% respectively. the years in the parentheses indicates break year. ι shows intercept and ι & γ show intercept and trend. table 2 shows that the two unit root tests provide mixed and somewhat contradicting results. both tests concur that gdp, labor, hec and nec are stationary at first difference and are integrated of order i(1). however, the results are contradicting for ffc. for ffc, the results of adf test indicate stationarity at level, i.e. i(0), whereas the results of the bp test show startionarity and first difference, i.e. i(1). for labor, both adf and bp unit root tests are showing the same order, i.e. i(0). comparing the results of two unit root tests, the results of bp test are preferred due to incorporation of structural break in it, and hence, are used to decide integration order of variables. according to bp unit root test, it is concluded that except for labor, all other variables failed to reject the null hypothesis at level. hence, the unit roots results indicate that the dependent variable and all explanatory variable except labor are stationary at first difference and are i(1). 4.2.optimum lag selection we use var lag order selection criteria to select the appropriate number of lags (razzaqi et al., 2011). we used different lag order selection criteria to decide the lag length (zhang & cheng, 2009). the results of var lag order selection criteria for var model are provided in table 3. the table confirms that the number of lags selected by aic is two. in our study, three among five criteria are selecting one as optimal lag. however, we select two lag as optimum because the model run with two lags perform better and pass all the diagnostic tests. in contrast, model with one lag exhibits the problem of serial correlation and dynamic instability. the auto-regressive (ar) root graph and other relevant tests applications confirm that the model is dynamically stable at two lags (see figure 2) and is free from non-normality, serial correlation, and heteroskedasticity issues. shazia farhat durrani, inayatullah jan & sidra pervez 86 figure 2: inverse roots of auto-regressive characteristic polynomial table 3: var lag order selection criteria lag logl lr a fpeb aicc scd hqe 0 148.5196 na 4.00e-06 -6.757055 -6.339111 -6.604863 1 230.7338 36.3552* 8.84e-08 -10.57238 -9.987257* -10.35931* 2 235.5195 7.470385 8.57e-08* -10.61071* -9.858407 -10.33676 3 238.3091 4.082382 9.21e-08 -10.55166 -9.632187 -10.21684 * indicates optimum lags selected by the criterion (at 5% level). asequential modified lr test statistic (lr), bfinal prediction error (fpe), cakaike information criterion (aic), dschwarz information criterion (sc), and ehannan-quinn information criterion (hq) 4.3.results of t-y granger causality test the t-y granger causality test was carried out by using modified wald (mwald) test to investigate the direction of causality between ec and gdp (alper & oguz, 2016). in our model, the maximum integration order is i(1) and maximum lag length is 2 lags. table 4 shows results of the t-y granger causality test. the results confirm that in panel a, all -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 journal of applied economics and business studies, volume. 6, issue 2 (2022) 77-94 https://doi.org/10.34260/jaebs.625 87 variables except labor reject the null of non-granger causality at 10%, 1%, and 5% levels respectively. these results suggest that labor does not cause gdp. however, the remaining variables such as gross capital formation, ffc, hec, and nec are significantly granger causing gdp of pakistan. the results are in consensus with those of gozgor et al., (2018) who found that both renewable and non-renewable energy consumption are positively associated with economic growth in 29 oecd countries. in panel b, gross capital formation, ffc, and nec are significantly granger causing labor at 5% level and the rest of the variables fail to reject the non-causality null in case of gdp and labor, and hec and labor. in case of panel c, however, the results are opposite to panel b. in panel c, the gdp and hec are granger causing the gross capital formation at 5% and 1% level respectively. contrary to that, labor, ffc, and nec fail to cause the gross capital formation, as none of them reject the null of non-causality at any prescribed significance level. in case of panel d, with ffc as dependent variable, only gdp is significantly causing the ffc at 10% level. all other variables including labor, gross capital formation, hec, and nec are failed to reject the null of non-causality. in panel e, no other variable other than nec is significantly granger causing the hec. in penal f, where nec is taken as a dependent variable in mwald test, only ffc was causing nec. the remaining four variables failed to reject the null of non-causality. similar results were reported by husaini & lean (2022) and oryani et al., (2022). table 4: results of t-y granger causality test dependent variable excluded variables chi-square probability panel a lny lnl 2.287107 0.3187 lnk 4.808387* 0.0903 lnffc 5.721649* 0.0572 lnhec 10.28256*** 0.0059 lnnec 6.577388** 0.0373 panel b lnl lny 0.229032 0.8918 lnk 6.280045** 0.0433 lnffc 8.934992** 0.0115 lnhec 3.461423 0.1772 lnnec 8.002591** 0.0183 panel c lnk lny 7.021431** 0.0299 lnl 1.738061 0.4194 lnffc 3.949095 0.1388 lnhec 11.32432*** 0.0035 lnnec 3.507478 0.1731 panel d lnffc lny 5.678164* 0.0585 shazia farhat durrani, inayatullah jan & sidra pervez 88 lnl 0.604023 0.7393 lnk 2.020159 0.3642 lnhec 0.754212 0.6858 lnnec 1.918292 0.3832 panel e lnhec lny 0.108051 0.9474 lnl 3.515819 0.1724 lnk 0.556998 0.7569 lnffc 2.811650 0.2452 lnnec 6.832299** 0.0328 panel f lnnec lny 0.781551 0.6765 lnl 3.379927 0.1845 lnk 2.911404 0.2332 lnffc 6.330418** 0.0422 lnhec 4.471844 0.1069 note: ***, **, * represents respective significance at 1%, 5%, and 10% levels. we provide an overview of the above results and the associated direction of causality in table 5. the table shows that no causality exists between gdp and labor force. however, a bidirectional causal relationship between gdp and gross capital formation occurs. the results further validate a growth hypothesis for hec and gdp and nec and gdp. the findings are in consensus with those of aqeel & butt (2001) and wolde-rufael (2004) who confirmed growth hypothesis between electricity use and gdp growth for pakistan and shanghai, respectively. likewise, the study by omri & chaibi (2014) supported growth hypothesis between nec and gdp for belgium and spain. in case of ffc and gdp, a feedback hypothesis is confirmed for pakistan. similar findings were reported by bildirici & bakirtas (2014) who found bidirectional causality between coal consumption and gdp for china. table 5: direction of causality between variables no granger causality unidirectional granger causality bidirectional granger causality lny – lnl lnhec → lny lny ↔ lnk lnhec − lnl lnnec → lny lny ↔ lnffc lnffc − lnk lnk → lnl lnnec − lnk lnffc → lnl lnhec − lnffc lnnec → lnl lnhec → lnk lnffc → lnnec lnnec → lnhec note: −, →, ↔ indicate no, unidirectional, and bidirectional granger causality journal of applied economics and business studies, volume. 6, issue 2 (2022) 77-94 https://doi.org/10.34260/jaebs.625 89 at1%, 5%, and 10% levels, respectively. 5. conclusion this study probes causality between major sources of ec and gdp of pakistan over the period 1972 to 2015. adf and bp unit root tests were used for testing stationarity of the data series. only gross capital formation was found stationary at level. the rest of all variables became stationary after differencing and were integrated of order i(1). given the mixed integration order, t-y granger causality test was employed for investigating the direction of causality among the variables of interest. results of the granger causality tests confirm the existence of feedback hypothesis for gdp and ffc. besides, the findings also confirm the growth hypothesis for hydroelectricity and nec and gdp. based on the findings of this study, it is recommended that government should focus on expanding the supply of energy along with the diversification of energy mix. the focus of the government policy should be on increasing the renewable energy sources in the total energy mix. this will promote economic growth as well as global environmental sustainability which are hampered by carbon emission from non-renewable energy consumption. 6. acknowledgements the authors thank the anonymous reviewers for their constructive remarks which helped in improving the manuscript. the authors confirm that the study was not financed by any organization. references ahmed, m. and m. azam (2015). ‘causal nexus between energy consumption and economic growth for high, middle and low income countries using frequency domain analysis’, renewable and sustainable energy reviews, 60, pp. 653-678. al-iriani, m.a. (2006). ‘energy-gdp relationship revisited: an example from gcc countries using panel causality’, energy policy, 34, pp. 3342-3350. alper, a. and o. oguz (2016). ‘the role of renewable energy consumption in economic growth: evidence from asymmetric causality’, renewable and sustainable energy reviews, 60, pp. 953-959. apergis, n., j. payne, k. menyah and y. rufael (2010). ‘on the causal dynamics between emissions, nuclear energy, renewable energy, and economic growth’, ecological economics, 69(1), pp. 2255-2260. aqeel, a. and m.s. butt (2001). ‘the relationship between energy consumption and economic growth in pakistan’, asia-pacific development journal, 8(2), pp. 101110. arouri, m., g.s. uddin, p. kyophilavong, f. teulon and a.k. tiwari (2014). ‘energy utilization and economic growth in france: evidence from asymmetric causality test’. ipag working paper series, 102. shazia farhat durrani, inayatullah jan & sidra pervez 90 azam, m., a.q. khan, b. bakhtyar and c. emirullah (2015). ‘the causal relationship between energy consumption and economic growth in the asean-5 countries’, renewable and sustainable energy reviews, 47, pp. 732-745. bildirici, m.e. and t. bakirtas (2014). ‘the relationship among oil, natural gas and coal consumption and economic growth in bricts (brazil, russian, india, china, turkey and south africa) countries’, energy, 65, pp. 134-144 british petroleum statistics. (2016). ‘bp statistical review of world energy’. https://www.bp.com/en/global/corporate/energy-economics/statistica-review-ofworld-energy/downloads.html [accessed on 28 june 2019]. carmona, m., e. congregado, j. feria and j. iglesias (2017). ‘the energy-growth nexus reconsidered: persistence and causality’, renewable and sustainable energy reviews, 71, pp. 342-347. chiou-wei, s.z., c. ching-fu and z. zhu (2008). ‘economic growth and energy consumption revisited evidence from linear and nonlinear granger causality’, energy economics, 30, pp. 3063-3076. chiou-wei, s-z., z. zhu, s-h. chen and s-p. hsueh (2016). ‘controlling for relevant variables: energy consumption and economic growth nexus revisited in an egarch-m (exponential garch-in-mean) model’, energy, 109, pp. 391-399. destek, m.a. (2016). ‘renewable energy consumption and economic growth in newly industrialized countries: evidence from asymmetric causality test’, renewable energy, 95, pp. 478-484. dudzevičiūtė, g. and a. šimelytė (2017). ‘export, energy consumption and economic growth inter-linkages: the case of lithuania’, scientific annals economics and business, 64(3), pp. 395-410. durrani, s.f., i. jan and m., ahmad (2021). ‘do primary energy consumption and economic growth drive each other in pakistan? implications for energy policy’, biophysical economics and sustainability, 6(8). https://doi.org/10.1007/s41247021-00090-x. filippidis, m., p. tzouvanas and i. chatziantoniou (2021). ‘energy poverty through the lens of the energy-environmental kuznets curve hypothesis’, energy economics, 100 (105328). https://doi.org/10.1016/j. eneco.2021.105328. furouka, f. (2017). ‘renewable electricity consumption and economic development: new findings from the baltic countries’, renewable and sustainable energy reviews, 71, pp. 450-463. gozgor, g., c.k.m. lau and z. lu (2018). ‘energy consumption and economic growth: new evidence from the oecd countries’, energy, 153, pp. 27-34. gujarati, d. and d. porter (2009). basic econometrics. 5th ed. usa: mcgraw-hill irwin. husaini, d.h. and h.h. lean 2022. ‘renewable and non-renewable electricity-growth nexus in asia: the role of private power plants and oil price threshold effect’, resources policy, 78(102850). https://doi.org/10.1016/j.resourpol.2022.102850. journal of applied economics and business studies, volume. 6, issue 2 (2022) 77-94 https://doi.org/10.34260/jaebs.625 91 ikhide, e. and c. adjasi (2015). ‘the causal relationship between renewable and nonrenewable energy consumption and economic growth: the case study of nigeria’. paper presented at the economic society of south africa at uct: south africa. jalil, a. and m. feridun (2014). ‘energy-driven economic growth: energy consumptioneconomic growth nexus revisited for china’, emerging markets finance and trade, 50(5), pp. 159-168. jan, i. and h.d. lohano (2021). ‘uptake of energy efficient cookstoves in pakistan’, renewable and sustainable energy reviews, 137(110466). https://doi.org/10.1016/j.rser.2020.110466. jan, i., s.f. durrani and h. khan (2020). ‘does renewable energy efficiently spur economic growth? evidence from pakistan’, environment, development and sustainability, 23, pp. 373-387. jan, i. and w. akram (2018). ‘willingness of rural communities to adopt biogas systems in pakistan: critical factors and policy implications’, renewable and sustainable energy reviews, 81, pp. 3178-3185. jan, i. (2012). ‘what makes people adopt improved cookstoves? empirical evidence from rural northwest pakistan’, renewable and sustainable energy reviews, 16(5), pp. 3200-3205. jan, i., s. ullah, w. akram, n.p. khan, s.m. asim, z. mahmood, m.n. ahmad and s.s. ahmad (2017). ‘adoption of improved cookstoves in pakistan: a logit analysis’, biomass and bioenergy, 103, pp. 55-62. javed, m.s., r. raza, i. hassan, r. saeed, n. shaheen, j. iqbal and s.f. shaukat (2016). ‘the energy crisis in pakistan: a possible solution via biomass-based waste’, journal of renewable and sustainable energy, 8 (043102). javid, a.y., m. javid and z.a. awan (2013). ‘electricity consumption and economic growth: evidence from pakistan’, economics and business letters, 2(1), pp. 2132. kahia, m., m.s. ben-aissa and l. charfeddine (2016). ‘impact of renewable and nonrenewable energy consumption on economic growth: new evidence from the mena net oil exporting countries (noecs)’, energy, 116(1), pp. 102-115. kahouli, b. (2017). ‘the short and long run causality relationship among economic growth, energy consumption and financial development: evidence from south mediterranean countries (smcs)’, energy economics, 68, pp. 19-30. kakar, z.h. and b.a. khilji (2011). ‘energy consumption and economic growth in pakistan’, journal of international academic research, 11(1), pp. 33-36. kraft, j. and a. kraft (1978). ‘relationships between energy and gnp’, the journal of energy and development, 3, pp. 401-403. koçak, e. and a. şarkgüneşi (2017). ‘the renewable energy and economic growth nexus in black sea and balkan countries’, energy policy, 100, pp. 51-57. lee, c.c. (2006). ‘the causality relationship between energy consumption and gdp in g11 countries revised’, energy policy, 34, pp. 1086-1093. shazia farhat durrani, inayatullah jan & sidra pervez 92 leiva, b. and z. liu (2019). energy and economic growth in the usa two decades later: replication and reanalysis’, energy economics, 82, pp. 89-99. li, c., t. lin, y. chen, y. yan, z. xu (2022). ‘nonlinear impacts of renewable energy consumption on economic growth and environmental pollution across china’, journal of cleaner production, 368(133183). https://doi.org/10.1016/j.jclepro.2022.133183. li, y., feng, n., and fang, n. (2008). ‘relationship between energy consumption and economic growth: empirical study based on data on hebei province from 1980 to 2008’, system engineering procedia, 1, pp. 117-123. liew, v., and khim, s. (2004). ‘which lag length selection criteria should we employ? economics bulletin, 3(33), pp. 1-9. luke, n. (2016). ‘examining the renewable energy consumption-economic growth nexus in sub-saharan african countries’, msc honors program theses. http://scholarworks.uni.edu/hpt/241[accessed on 15 june 2019]. marques, m.l., a.j. fuinhas and c.a. marques (2016). ‘on the global energy consumption and economic growth nexus: a long time span analysis’, international energy journal, 16, pp. 143-150. mutascu, m. (2016). ‘a bootstrap panel granger causality analysis of energy consumption and economic growth in the g7 countries’, renewable and sustainable energy reviews, 63, pp. 166-171. nadeem, s. and k. munir (2016). ‘energy consumption and economic growth in pakistan: a sectoral analysis. https://mpra.ub.uni-muenchen.de/74569/ [accessed on 15 january 2018]. narayan, s. (2016). ‘predictability within the energy consumption-economic growth nexus: some evidence from income and regional groups’, economic modelling, 54, pp. 515-521. oh, w. and k. lee (2004). ‘causal relationship between energy consumption and gdp revisited: the case of korea 1970–1999’, energy economics, 26, pp. 51-59. omri, a. and a. chaibi (2014). ‘nuclear energy, renewable energy, and economic growth in developed and developing countries: a modeling analysis from simultaneousequation models’. http://www.ipag.fr/fr/accueil/la-recherche/publicationswp.html [accessed on 02 august 2022]. oryani, b., h. kamyab, a. mоridiаn, z. azizi, s. rezania, s. chelliapan (2022). ‘does structural change boost the energy demand in a fossil fuel-driven economy? new evidence from iran’, energy, 254(c). 124391. https://doi.org/10.1016/j.energy.2022.124391. ouedraogo, n.s. (2013). ‘energy consumption and economic growth: evidence from the economic community of west african states’, energy economics,36, pp. 637-647. pin, h.l. (2014). ‘renewable energy consumption and economic growth in nine oecd countries: bounds test approach and causality analysis’, the scientific world journal of applied economics and business studies, volume. 6, issue 2 (2022) 77-94 https://doi.org/10.34260/jaebs.625 93 journal, 2014, pp. 1-6. rafindadi, a.a. and i. ozturk (2017). ‘impacts of renewable energy consumption on the german economic growth: evidence from combined cointegration test’, renewable and sustainable energy reviews, 75, pp. 1130–1141. razzaqi, s., f. bilquees and s. sherbazz (2011). ‘dynamic relationship between energy and economic growth: evidence from d8 countries’, pakistan development review, 50(4), pp. 437-58. rodríguez-caballero, c.v. and d. ventosa-santaulària (2016). ‘energy-growth long-term relationship under structural breaks: evidence from canada, 17 latin american economies and the usa’, energy economics, 61, pp. 121-134. shahbaz, m., m. arouri and f. teulon (2014). ‘shortand long-run relationships between natural gas consumption and economic growth: evidence from pakistan’, economic modelling, 41, pp. 219-226. siddiqui, r. (2004). ‘energy and economic growth in pakistan’, pakistan development review, 43(2), pp. 175-200. soytas, u. and r. sari (2009). ‘energy consumption, economic growth, and carbon emissions: challenges faced by an eu candidate member’, ecological economics, 68, pp. 1667-1675. tang, c.f., b.w. tan and i. ozturk (2016). ‘energy consumption and economic growth in vietnam’, renewable and sustainable energy reviews, 54, pp. 1506-1514. thao, n.t.n. and v.l. chon (2016). nonrenewable, renewable energy consumption and economic performance in oecd countries: a stochastic distance function approach’. http://veam.org/wp-content/uploads/2016/08/70.-nguyen-thi-nganthao.pdf [accessed on 7 july 2022]. tiba, s. and a. omri (2017). ‘literature survey on the relationships between energy, environment and economic growth’, renewable and sustainable energy reviews, 69, pp. 1129-1146. toda, h.y. and t. yamamoto (1995). ‘statistical inference in vector autoregressions with possibly integrated processes’, journal of econometrics, 66 (1-2), pp. 225-250. wang, q., z. dong, r. li and l. wang (2022). ‘renewable energy and economic growth: new insight from country risks’, energy. 238(c). 122018. https://doi.org/10.1016/j.energy.2021.122018. wolde-rufael, y. (2004). ‘disaggregated industrial energy consumption and gdp: the case of shanghai, 1952-1999’, energy economics, 26, pp. 69-75. world bank. (2017). world development indicators database. https://databank.worldbank.org/data/reports.aspx/source=2&country=pak [accessed on 28 december 2017] yildirim, e., d. sukruoglu and a. aslan (2014). ‘energy consumption and economic growth in the next 11 countries: the bootstrapped autoregressive metric causality approach’, energy economics, 44, pp. 14-21. yuan, j.h., j.g. kang, c.h. zhao and z.g. hu (2008). ‘energy consumption and economic shazia farhat durrani, inayatullah jan & sidra pervez 94 growth: evidence from china at both aggregated and disaggregated levels’, energy economics, 30(6), pp. 3077-3094. zaidi, s. and s. ferhi (2019). ‘causal relationships between energy consumption, economic growth and co2 emission in sub-saharan: evidence from dynamic simultaneous-equations models’, modern economy, 10, 2157-2173. zhang, q., s.a.r. shah and l. yang (2022). ‘an appreciated response of disaggregated energies consumption towards the sustainable growth: a debate on g-10 economies’, energy, 254(a). 124377. https://doi.org/10.1016/j.energy.2022.124377. zhang, x.p. and x.m. cheng (2009). ‘energy consumption, carbon emissions, and economic growth in china’, ecological economics, 68(10), pp. 2706-2712. journal of applied economics and business studies, volume. 6, issue 1 (2022) 25-52 https://doi.org/10.34260/jaebs.612 25 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x person-organization fit as determinant of constructive deviance behavior: the mediating role of psychological ownership and the moderating role of collectivist orientation. a conditional process analysis kamil hussain1,saima aftab2 & mahwash ghafoor chaudhry3 123 department of management sciences, university of wah, wah cantt, pakistan abstract deviant behaviors of employees have been into research considerations for long however, research on constructive deviance behaviors, keeping in view its non-contingency, still needs consideration. this study focuses on determining how person-organization fit (p-o fit) predicts constructive deviance behavior (cdb) among employees. to analyze the mechanism involved, psychological ownership (po) is studied as a mediating mechanism between the direct relationship of personorganization fit and constructive deviance behaviors. moreover, collectivist orientation (co) is taken as a moderating variable on the relationship of psychological ownership and constructive deviance behaviors. a total of 561 managerial employees from informatics industry were included in the sample. findings show that person-organization fit, psychological ownership and collectivist orientation were significantly related to constructive deviance behaviors. psychological ownership mediates the link between person origination fit and construct deviance behaviors, while collectivist orientations moderate the relationship between psychological ownership and constructive deviance behaviors. at the low, moderate, and high levels of collectivist orientation (moderator), conditional indirect effects of person-organization fit on constructive deviance behaviors through psychological ownership (mediator) were significant. future research studies may examine other antecedents (such as building coalition and networking) of constructive deviance behavior through other mediating mechanism (such as keywords personorganization fit, psychological ownership, constructive deviance behaviours, collectivistic orientation jel classification d21, d22, d23 kamil huusain, saima aftab &mahwash hussain 26 psychological empowerment), while the consequences (such as creative performance and innovative performance) of constructive deviance behaviors are also needed to be explored to create awareness about the positive side of deviance behaviors. 1. introduction workplace deviance is contagious for the organizations garg et al. (2022). relationship of workplace deviance behaviors and organizational level outcomes is center of debate for researchers (garg et al., 2022). workplace deviance can be explained as an employee’s voluntary behavior to negate and violate the organizational norms and values and hence potentially risking the very survival of the organization (robinson & bennett, 1995; tekmen&kaptangil, 2022). behaviors are considered deviant in case of a disregard for organizational and societal rules and regulations (mishra et al., 2021). counterproductive work behaviors at individual, group and organizational levels, are synonymously been used by scholars to address deviance behaviors in organizations (dalal, 2005; lau et al., 2003; uggerslev et al., 2012), causing low morals and high stress among employees leading to increased levels of employee turnover (o'leary-kelly et al., 1996). however, recent empirical research contests that deviant behavior can be constructive as well as destructive (martinko et al., 2002). past studies focused more on negative side of deviance behaviors i.e., counterproductive behaviors and more research studies are needed to explore their positive side i.e., constructive deviance behaviors, as it holds consequences and implications for both individual as well as organizations. constructive deviance behaviors (cdb) can be defined as “voluntary action that violates significant organizational rules while also contributing to the organization's and its members' well-being” (galperin, 2003). employees engaged in these behaviors have significant voluntariness and intention to do well (yildiz et al., 2015). for instance, increase in innovation (hussain & rehman, 2020) and creativity (galperin, 2012; gino &wiltermuth, 2014; krau, 2008; pascale &sternin, 2005) are reported as positive consequences of constructive deviance behaviors. past literature has also discussed some antecedents of cdb such as person–organization fit (p-o fit) (rahman et al., 2022; ucar et al., 2021). p-o fit can be described as an integration and understanding that exists between the employee and the organization in order to make each of them fulfill the respective needs (boon and biron (2016). the p-o fit explains the logic behind the needssupply fit that is the congruence of the individual’s demand and supply by the organization in response to that demand, an integral part of employment agreement (ghielen et al., journal of applied economics and business studies, volume. 6, issue 1 (2022) 25-52 https://doi.org/10.34260/jaebs.612 27 2021; grobler &holtzhausen, 2018). once an individual realizes a fit between him and his organization a feeling of psychological ownership (po) develops (mehmood et al., 2021). the mediating effect of psychological ownership between the direct relationship of constructive deviance behavior and person organization fit has not been extensively researched. po can be defined as a psychological compatibility that exists between individual and the organization they work in (chawla, 2020; koburtay&alzoubi, 2021). individual, cognitive and behavioral tendencies are the key to behavioral outcomes on the job (rafferty &minbashian, 2019). research found that collectivist orientation (co) is an important factor that impacts an employee’s behaviors to greater extent (zulfiqar et al., 2019). co can be defined as views about one’s self, of how his goals and the goals of the group he works in are interdependent (hagemann et al., 2020). one’s perception about the relationship with other group memebers and group cohesiveness plays a key role in driving human behaviors at work (hagemann et al., 2020; hofstede, 2001). the culture of the organizations determines how its people behave (shahzad et al., 2013). software housing industry is the fastest developing around the world (chureemas, 2021). innovation and creativity is the blood and bones of software houses (haque &oino, 2019). previous research has focused mainly on the negative side of deviance behaviors (yıldız&alpkan, 2015), while the focus of this study is to explore the positivity attached with cdb. this area of research is still untapped (galperin, 2012). this study is taking po fit as a predictor of cdb, with a moderating impat of co on the direct relationship of po and cdb. in light of the potential and significance the present area of research holds for the academia and in view of the valuable contribution the present the study will make to the in the existing body of knowledge, the present study attempts to study the constructive deviance behaviors. the relationships between the variables of the current study have been taken on the basis of empirical support/explicit theoretical support or implicit theoretical support in the existing literature. hence, ample literature support has been presented in table 1 below, to support the logic of the relationships between variables. table 1: gap analysis iv mediator moderator dv empirical support explicit theoretical support implicit theoretical support p-o fit cdb elçi et al. (2008) and liu et al. (2011) p-o fit po cdb jawad et al. (2013); cable and judge (1996); liu et al. (2011); kamil huusain, saima aftab &mahwash hussain 28 sharkawi et al. (2013) po co cdb chung and moon (2011) the table above shows that unique relationships studied in this research is novel and needed investigation. these gaps were also suggested to be studied by chung and moon (2011) and yildiz et al. (2015). 1.1 problem statement according to the world economic forum report (2015), pakistan ranks 111th out of 139 countries in terms of creativity, according to the martin prosperity institute's global creativity index (2015), and 99th out of 132 countries according to the world intellectual property organization's (wipo) global innovation index (2021), indicating that there are issues in pakistani organizations that do not allow or give employees the independence to be creative. we are living in such fascinating times that sincere, smart, and honest management of organizations must be willing to take risky judgments and actions that may contradict corporate conventions and policies. management must reevaluate and re-construct their approach to management, which may include making difficult decisions. constructive deviance behaviors are one of those courageous and unpleasant moves that can benefit the organization. second, there has been very little research in pakistan on the topic of constructive deviance behaviours. so, the study's purpose was to raise awareness in pakistan about deviance behaviours, namely that not all deviance behaviours are destructive; some are constructive and can be useful in fostering creativity and innovation in pakistani organizations, and to close the gap in the area of constructive deviance behaviours and their relationship with the aforementioned independent variable, mediator and moderator. 1.2 contextual analysis software development is without a doubt one of the most crucial strategic industries for any country's economic progress. software, according to (al-jaghoub, 2004), can work as a catalyst for any country's economic progress. acs and mueller (2008) found that software enterprises make a beneficial contribution to both knowledge and traditional economies. india's software sector is a great illustration of software development's good contribution to practically all of the country's industrial sectors (nasscom, 2006). according to pakistan software export board (pseb), pakistan's it and ites export remittances totaled $1.231 billion in 2019, which is significantly less than india's $137 billion in software exports in the same year, accounting for only 5% journal of applied economics and business studies, volume. 6, issue 1 (2022) 25-52 https://doi.org/10.34260/jaebs.612 29 of the country's total exports, according to pakistan bureau of statistics report (2019). despite the fact that pakistan's it sector is expanding, the country still lags behind its regional counterparts in terms of it. as a result, pakistan cannot be regarded a powerhouse in the field of information technology based on its proportion of worldwide it sales. on the pakistan software export board, there are currently 4464 it companies listed, and approximately 20,000 it graduates enter the market each year. freelancers account for a significant share of pakistan's total it exports. pakistan is the third-largest source of freelance programmers, trailing only the united states and india. so, while there are indicators of significant progress in the country's it sector, statistics show that nine out of ten startups fail, while those who survive generate novel and creative goods and services. as a result, only those companies that offer unique and creative products and services succeed, and software development companies are no exception. in fact, compared to other industries, software development organizations are required to come out with innovative goods more frequently. various academics have suggested focusing on constructive deviance behaviours as a solution (robbins & galperin, 2010; vadera et al., 2013) 2. review of literature 2.1 relationship between p-o fit and cdb previous studies consider p-o fit as an important organizational factor and report different micro, meso and macro level impacts. it reportedly has a positive relationship with various job attitudes i.e., job satisfaction , organizational citizenship behavior, organizational commitment, organizational performance (bretz jr & judge, 1994; elçi et al., 2008; liu et al., 2010; o'reilly iii et al., 1991) turnover intentions, job satisfaction. (arthur jr et al., 2006; oh et al., 2014), and job performance (hoffman &woehr, 2006). however, a few past studies found that a low degree of p-o fit can negatively result in form of dissatisfaction, alienation, increased turnover intentions and counter-productive workplace behaviors (jawad et al., 2013; liu et al., 2010; sharkawi et al., 2013). though existing research work investigating the relationship between p-o fit with cdb is limited but it still holds potential of further exploration considering the prediction made by the past studies. according to cable and judge (1996), p-o fit positively influences various job attitudes. the concept of right person for the right job is the critical philosophy of p-o fit (yildiz et al., 2015). according to the person environment theory, individuals are always in search of those organizations whose environment is compatible with their personality traits i.e., one’s abilities, values, needs and wants as it gives them opportunities to make best possible use of these personality kamil huusain, saima aftab &mahwash hussain 30 traits (kristof-brown &guay, 2011). p-o fit can be conceptualized as value congruence creating compatibility between an individual and the organization; he/she works in (kristof, 1996). boon and biron (2016) described p-o fit as integration existing between an employee and the organization and measures the extent of integration in terms of similarity that exist between them enabling them to fulfill their respective needs. uggerslev et al. (2012), suggest that an individual’s intentions to join an organization are influenced by value congruence existing among individual and organization. unlike the p-o fit focuses on general supply of the demands made by an individual, the needssupply approach explains the supply of individual demands in context of specific jobs i.e., person-job fit (kristof-brown &guay, 2011; kristof, 1996). cdb is defined as the deliberate attempt on behalf of organizational employees to defy norms and values aimed at the wellbeing of members and organization as a whole (galperin (2003); spreitzer and sonenshein (2004)). according to warren (2003) individual expressing cdb hold on strictly to the hyper norms while going against values of the reference group. constructive deviance being a broad phenomenon encompasses different kinds of deviance behaviors warren (2003) such as tempered radicalism (meyerson & scully, 1995) , whistle-blowing (near & miceli, 1985), counter-role behaviors (staw&boettger, 1990), exercising voice (van dyne &lepine, 1998), principled organizational dissent (graham, 1986), creative or functional disobedience (brief et al., 2001), difference types of pro-social behaviors (o'reilly & chatman, 1986; puffer, 1987) and some types of ocb (van dyne et al., 1994). other than these, creative performance (amabile et al., 1996), extra-role behaviors (vandyne et al., 1995), pro-social rule breaking (morrison, 2006) and issue selling (dutton & ashford, 1993) have been discussed in literature. creative performance, pro-social rule breaking, issue selling, and extra-role behaviors were also included under the rubric of cdb by (vadera et al., 2013). cdb is considered to be a strong prosocial behavior focusing on doing well for others suffering in organization (galperin (2003); spreitzer and sonenshein (2004)). according to yildiz et al. (2015) a relationship of p-o fit and cdb is possible through various mediating factors. the positive relationship exiting between perception, behaviors and attitudes can be explained in context of the social exchange theory by blau (1964). theory of social exchange has formed basis of various research works in the discipline of social sciences (suárez‐mendoza and zoghbi‐manrique‐de‐lara (2007). the reciprocal nature of the theory explains how negative or positive perception leads to the resultant behavior (positive or negative) through some associated positive or journal of applied economics and business studies, volume. 6, issue 1 (2022) 25-52 https://doi.org/10.34260/jaebs.612 31 negative attitude (greenberg & scott, 1996). considering the social exchange theory explaining reference to the relationship between perception, attitude, and behavior it can be concluded that perception of fit exiting between individual and organization greatly impacts the attitude pertaining to the psychological learning. this influences cdb. considering the above discussion, the relationship between p-o fit and cdb can be hypothesized as follows. h1: person-organization fit has direct positive relationship with constructive deviance behaviors. 2.2 relationship between p-o fit and psychological ownership the concept of po was presented by pierce et al. (2001). it can be described as a feeling experienced by individuals giving them a sense of passion and ownership over some target. according to (reb & connolly, 2007) an individual experiencing po develops mental state leading to possessive attitude towards things over which he may or may not have any right or ownership (park et al., 2013). this feeling of possession can be extended to any object or idea (baer & brown, 2012) or job (wang et al., 2018) eventually resulting in feelings of territorial responses (kirk et al., 2018). though the feeling of po for any object/idea begins with a possessive feeling but afterwards it develops into a positive attitude leading to strong emotions where individuals takes full control and responsibility of the object / idea as he starts visualizing the it as extension of himself (van dyne & pierce, 2004). the existence of two psychological states i.e., cognitive state and affective state, adds to the complexity of this phenomenon. under the cognitive state an individual experiences feeling of ownership over an object or idea whereas in affective state individual thinks he has control over the idea or object which conforms to his beliefs, ideas, and thoughts (pierce et al, 2003). various scholars have linked idea of p-o fit with recruitment process (han et al. (2015). recruiters choose candidates based on their demand-abilities, and applicants choose organizations based on their need-supply (cable & judge, 1997; kristof-brown, 2000). organizations are always on the lookout for people whose personality qualities align with the organization's standards and ideals (morley, 2007) (o'reilly iii et al., 1991). similarly, applicants exert efforts to select the organizations which are compatible with their personalities (schneider, 2001). this concept not only facilitates individuals in making decision about joining an organization but it also helps organizations in predicting whether an individual is interested in joining the organization or not (cable &derue, 2002). previous research studies on the p-o fit find that it has a positive role in reducing turnover intention, increasing job satisfaction and kamil huusain, saima aftab &mahwash hussain 32 individual’s commitment to the organization (ambrose et al., 2008; cable & judge, 1996; kristof‐brown et al., 2005) improves organizational performance (ambrose et al., 2008), enhances extra-role behaviors (lauver& kristof-brown, 2001) and employee trust (sekiguchi, 2007). according to (wagner et al., 2003) the sense of ownership and belonging that an individual experiences is dependent on the level of compatibility that exist between an individual and organizations. in view of above arguments theorizing the relationship between p-o fit and po, following hypothesis is proposed: h2: person-organization fit has a direct positive relationship with psychological ownership. 2.3 relationship between po and cdb po has been the focus empirical research by many researchers (liu et al., 2012; mayhew et al., 2007; ozler et al., 2008; pierce et al., 2003; pierce et al., 2004; sieger et al., 2011; sieger et al., 2013). po has extensively been investigated in relation with deviance behaviors (avey et al., 2009) such as risk-taking propensity (pierce et al., 2001) extra-role behaviors (o’driscoll et al., 2006) and organizational citizenship behavior (mayhew et al., 2007). yet it still lacks rigorous research supporting this relationship based on empirical findings. po being intensity of possessiveness due to an individual’s psychological attachment and having cognitive and affective elements, results in constructive deviance behavior (pierce et al. (2003). previous research studies (mayhew et al., 2007; van dyne & pierce, 2004) have also validated and supported the positive relationship between po and its behavioural outcomes, such as job satisfaction and organisational commitment (mayhew et al., 2007; o'driscoll et al., 2006; van dyne & pierce, 2004). (dalal, 2005; lepine et al., 2002). an individual’s ability of po for evaluative judgment due to affective state enables them to engage in activities and behaviors for the organizational benefit (pierce et al. (2003). however, avey et al. (2009) made valuable contribution to the research on po by exploring the negative effect of po on deviant behaviors. hence, it can be hypothesized as: h3: psychological ownership has a direct positive relationship with constructive deviance behaviors. 2.4 po as a mediating factor between the relationship of p-o fit and cdb the phenomenon of p-o fit can be further extended with its two different characteristics, supplemental and complementary fit, when considering the importance of the right person for the right job (kristof, 1996). according to muchinsky and journal of applied economics and business studies, volume. 6, issue 1 (2022) 25-52 https://doi.org/10.34260/jaebs.612 33 monahan (1987), supplementary fit occurs when an individual's personality traits are similar to those of others in an organizational environment, whereas complementary fit occurs when an individual's personality traits create or contribute to the organizational environment by providing something valuable that was previously lacking. according to needs-supplies view, a perfect match exists between an employee and its organization, if the organization can meet the expectation of its employees. on the other hand, according to demand-abilities perspective a perfect match exits if an employee’s skills and abilities fulfill needs and requirements of the organization (caplan, 1987). blau's (1964) social exchange theory, which suggests that people's perceptions, attitudes, and behaviours are all linked, provides a solid theoretical foundation for proposing po as a mediator in the relationship between p-o fit and cdb. according to cable and judge (1996), there is a favorable association between p-o fit and employee attitudes. a direct association between p-o fit and cdb is not achievable, according to yildiz et al. (2015), and a major mediating component plays a role. when a p-o fit is created, higher quantities of po are produced, which leads to higher levels of cdb (ambrose et al., 2008). (pierce et al., 2003). as a result, based on the above discussion, the following hypothesis can be proposed. h4: psychological ownership mediates the relationship between person-organization fit and constructive deviance. 2.5 role of co as a moderator on direct relationship of po and cdb co happens to be one of the key dimensions of hofstede (1980) study placing greater emphasis on the goals and well-being of the groups. individualism versus collectivism dimension is one of the most used dimensions in cross cultural studies among all the rest with salient features of collectivistic/individualistic orientation commonly found in eastern as well as western countries. but there exist a tendency of collectivism in asian countries unlike the inclination towards individualism in western countries (jun & lee, 2007). the concept of co has been the center of attention and research for many scholars in past who investigated this phenomenon in context of the effect of cultural orientation on employee behavior and attitude (jiang et al., 2020; ramamoorthy& flood, 2002, 2004). one of the studies found that the collectivistic orientation impacts the cognitive and behavioral tendencies of an individual which in turn affects their behavioral outcomes. co is based on the concept on mutual commitment focusing on the establishment of collective goals and desire to getting along well with others in the group (li and chen, 2015). collectivists possess characteristics of being interdependent and in contract to the individualistic, collectivistic prefer the collective over personal benefits hofstede (2001). one of the kamil huusain, saima aftab &mahwash hussain 34 reason could be the fact that approximately 60 % of individuals, consider themselves as a member of the group, give more importance to goals of the group over their personal ones (allocentric) while in individualistic culture contains approximately 60% selfcentered individuals (idiocentric) (triandis& suh, 2002). the concept of co has extensively been explored in connection with people’s motivation, feelings and conduct (hofstede, 2001). at psychological level of analysis the concept of collectivist can be defined as individual’s keenness and dedication towards the welfare of groups and its members, norms and objectives (dierdorff et al. (2011). many scholars investigating the possibility of impact of cultural orientation on resources owned by a person argue that collectivist people being committed to the group and organizational welfare prefer group goals over their personal goals (jang et al., 2018) which leads to development of close and strong interpersonal relationship among members of the group (kim et al., 1994) and do not prefer to separate themselves from their group members by competing with them (kitayama et al., 1995). therefore, it can be suggested that co and cdb are negatively associated as it provokes individual’s defiance of organizational norms and values. yet some past studies report that some types of deviance behaviors such as extra role behavior (cho &faerman, 2010), organizational citizenship behaviors (organ et al., 2005) and taking charge (love & dustin, 2014) has a positive relationship with co. considering previous literature, the present research attempts to study the effect of co as a moderator on the relationship of po and cdb. hence, it is hypothesized as follows: h5: co moderates the direct positive relationship of po and cdb h6: co negatively moderates the mediational role of po between p-o fit and cdb. 3. research methodology 3.1 research design the current study aims to empirically test proposed relationships on basis of testing quantitative data gathered through amos software. the nature of the current research is cross-sectional as it involves collecting data on the hypothesized relationships at a specific point in time. 3.2 sampling and data collection individuals working in software houses of rawalpindi and islamabad region of pakistan as software engineers, graphics and animation designers, web page developers and others in the same field form population for the present study. the reason for selecting these employees is the fact that due to increasing demands of flexibility, journal of applied economics and business studies, volume. 6, issue 1 (2022) 25-52 https://doi.org/10.34260/jaebs.612 35 innovative change, and creative advancements these employees are most tempted to violate organizational norms and values for the sake of exercising innovative ideas. this stress prompts them to break organizational rules and regulations for the sake of organizational benefit and thus results in expressing cdb. existing research also suggests on focusing the employees in information sector as they are more likely to exhibit different types of cdb (yildiz et al. (2015) . a representative sample (sekaran & bougie, 2016) from the target population working in software companies in twin cities (islamabad and rawalpindi), namely awamimarkaz software technology park, ksl software technology park, evacuee complex software technology park and rose center software technology park, are selected using the judgment sampling technique. the reason of using judgment sampling is that it allows researchers to go directly to their target population of interest and increases the relevance of the sample to the population of interest. as the employees of software houses are most relevant to the population of interest and fit the particular criteria of creativity and innovation which requires the employees to be constructively deviant of the norms and values of the organization for the wellbeing of the organization. a total of 561 full and valid questionnaires were returned from a total of 600 disseminated, resulting in a 93 percent response rate from the respondents. 3.3 measurement instruments questionnaire consisting of a total 26 items is used to collect data. the questionnaire is designed on a 7-point likert scale. a total of 9 items measuring cdb are adapted from the study of galperin (2012). sample item includes “bent a rule to satisfy a customer’s needs”. p-o fit is measured using items adapted from the research work of valentine et al. (2002). sample item includes “i feel that my personal values are good fit with this organization”. a seven items scale is used to measure po adapted from the study of van dyne and pierce (2004). sample item includes “this is my organization”, while a seven item scale of co has been measured by scale adapted from the study of robert and wasti (2002). sample item includes “employees are taken care of like members of a family”. the measures in this study were tested to ensure reliability and validity since the adapted items taken from previous studies were designed on 5-point likert scale. 3.4 analysis technique instrument validation on basis of the convergent and discriminant validity is done using factor analysis whereas the techniques of structure equation modeling is used to measure the association between the latent variables (babin et al., 2008). structured kamil huusain, saima aftab &mahwash hussain 36 equation modeling is a second generation multivariate analysis as it has the capability of combining the multiple regression with factor analysis (hair et al. (2012). 4. results 4.1 demographic statistics a representative sample (sekaran & bougie, 2016) from the target population working in software companies in twin cities, are selected using the judgment sampling technique. a total of 561 full and valid questionnaires were returned from a total of 600 disseminated, resulting in a 93 percent response rate from the respondents.percentages calculated on basis of the qualification of respondents indicate that most of the respondents (52%) were bachelor degree holders, 32% were having a master degree and remaining respondents (16 %) of the participants were ms qualified. journal of applied economics and business studies, volume. 6, issue 1 (2022) 25-52 https://doi.org/10.34260/jaebs.612 37 4.2 measurement model figure 2 below explicates the measurement model showing correlationamongpo, po fit, cdb and co. figure 1: measurement model kamil huusain, saima aftab &mahwash hussain 38 table 2: confirmatory factor analysis journal of applied economics and business studies, volume. 6, issue 1 (2022) 25-52 https://doi.org/10.34260/jaebs.612 39 4.3 discriminant and convergent validity discriminant validity assesses if a construct has a strong correlation with measurements that are predicted to be different (churchill, 1999). it is the measurement of how far the model's measurements differ from one another. all discriminant validity difficulties were overcome by removing components that were more linked with other measures than their own.similarly, "convergent validity" is "the extent to which several ways of measuring a variable provide the same results” (s. w. o'leary-kelly &vokurka, 1998). all the values of ave were greater than acceptable threshold level of 0.5, ensuring convergent validity of the scale.reliability of constructs was assessed using cronbach’s alpha statistics. results show that ca values lie within the range between 0.756 and 0.925 which were very much in the acceptable range to establish the reliability of constructs (nunnally & bernstein, 1994). 4.4 correlation table 3shows the correlation values indicating that there is a significant relationship among all the constructs at a 0.01 level of significance. the researcher rules out the possibility of multi-collinearity which needs to be resolved if correlation values are greater than 0.80 (hair et al. (2006), which is not the case for this study table3: correlation variables 1 2 3 4 constructive deviance behaviours 1.00 person-organization fit 0.283** 1.00 psychological ownership 0.612** 0.322** 1.00 collectivist orientation 0.522** 0.482** 0.362** 1.00 **p<0.01 4.5 model fitness according hair et al. (2006) various indices of fit such as gfi, agfi, cfi and rmsea that determines the model fits the data.table 5 provides the goodness of fit indices for the measurement model, with a gfi of 0.95, agfi of 0.92, cfi of 0.91, and rmsea of 0.076, all of which are within the acceptable range (hair et al, 2006). kamil huusain, saima aftab &mahwash hussain 40 table 4: goodness of fit indices 4.6 mediation analysis the present study investigates meditational effect of po on relationship existing between the p-o fit and cdb. figure 3 below presents path model for the meditational effect. figure 2: mediation path model table 5: goodness of fit indices for mediation path model journal of applied economics and business studies, volume. 6, issue 1 (2022) 25-52 https://doi.org/10.34260/jaebs.612 41 as indicated in the table above, the model fit indices for the mediation model indicates some issues which need to be resolved through of the modification indices. 4.7 structural model after modification indices model modification indices suggest the links to change in the model while checking for improvement in chi square value. starting with the largest sensible modification and correlating the error terms of same constructs with highest correlational value an improved model was developed. the improvements made in model by introducing modification indices are evident in figure 4 below, resulting in improved values of the model fit indices of the resultant model. kamil huusain, saima aftab &mahwash hussain 42 figure 3: structural model the value of cmin/df increased from 0 to 2.61 which is within prescribed range of 1 to 3 (hair et al. (2006). similarly, the modification indices improved other values such as cfi=0.977, gfi=0.957, agfi=928, and rmsea=0.046 which also lie within the acceptable range. table 6: goodness of fit indices after modification indices 4.8 standardized parameter estimates (mediation results) the results of the mediation of po on relationship of p–o fit and cdb can be seen in table 7. table7:standardized direct / indirect effect the results in table 7 indicate significant correlation existing among po, cdb and p-o fit. p-o fit significantly effects the po which in turn significantly impacts the cdb. p-o fit also bears a significant direct relationship with cdb. it also has significant indirect relationship with cdb through po as a mediator. since both the direct and indirect relationship was significant it means that po partially mediates the relationship between p-o fit and cdb. preacher et al. (2007) believes that significance of indirect standardized path is ample to establish mediation effect in a relationship and since they haven’t focused on partial mediation so they ignore the need to focus on the direct path. the results of indirect regression effect indicate that the indirect effect of p-o fit on journal of applied economics and business studies, volume. 6, issue 1 (2022) 25-52 https://doi.org/10.34260/jaebs.612 43 cdb has increased from 0.11 to 0.175. this shows the significance of the mediational impact of po. 4.9 moderated mediation analysis results the present study uses process macro developed by preacher et al. (2007) for conducting conditional process analysis. table 8 below provides the results for moderated mediation. results show that co moderates the meditational impact of po on association existing between p-o fit and cdb. table 8:moderated mediation the findings of the conditional process analysis indicate that all the tvalues are greater than 2 and all the p-values are less than 0.5. this means that p-o fit, po, and co positively and significantly effect cdb. the values of interaction term i.e., int_1 (po×co) having beta values of -0.091 with p-value of 0.001 and t-value of -3.274. the results in table 8 above indicate that po and cdb are negatively moderated by the co. the beta values reduced from 0.085 to 0.059 and to 0.033 with the increasing values of co (from 3.5 to 4.7 and to 6.0 respectively). 4.10 hypothesis summary the following table provides a summary of the accepted or rejected proposed hypothesis. table 9: summary of hypotheses kamil huusain, saima aftab &mahwash hussain 44 5. discussion and conclusion the current study intended to investigate the role of po as a mediator between p-o fit and cdb. furthermore, the current study evaluates the moderation effect of po on the relationship existing between p-o fit and cdb. social exchange theory by blau (1964) explains and establishes the association among perception, attitude and behaviors. this provides a sound foundation for current study by suggesting that perception of individual such as p-o fit can result in cdb. past research also supports the positive relationship of p-o fit with other variables such as organizational citizenship behavior, satisfaction with job as well as the organization, organizational performance and commitment (bretz jr & judge, 1994; elçi et al., 2008; liu et al., 2010; o'reilly iii et al., 1991). it also highlights the relationship of p-o fit and turnover, dissatisfaction, and counterproductive work behaviors. the results of the present study indicate positive relationship of p-o fit, po and cdb. the meditational role of po between the p-o fit and cdb proposed by the present study is also proved in the light of the findings. the results indicate that when a p-o fit is established between individual and its organization, it creates feeling of po, and this induces an individual to engage in cdb. despite the fact, that limited literature exists on relationship of p-o fit with cdb, yet some of the past studies have reported evidences suggesting the positive relationship of p-o fit with cdb (chung and moon journal of applied economics and business studies, volume. 6, issue 1 (2022) 25-52 https://doi.org/10.34260/jaebs.612 45 (2011) and other behaviors such as extra role behaviors and organizational citizenship behaviors (avey et al., 2009; van dyne & pierce, 2004). results similar to the present study have been reported by yildiz et al. (2015) which shows that relationship between p-o fit and cdb is mediated by po and the individual experiencing high po behave in a responsible manner and engage in cdb. this relationship between p-o fit and cdb is also supported by the theory of stewardship, according to which an individual experiencing high po are like stewards for the whole organization and are dedicated to organizational wellbeing (davis et al., 1997). the finding of the present study indicate that a significant relationship exists among co, po and cdb. results show that the impact of po on cdb is reduced by co, as it negatively moderates the relationship, conforming to similar results reported by study of chung and moon (2011). co prevents employees from setting themselves apart from the their team / groups and abstain from breaking organizational norms and values (kitayama et al., 1995). this negative moderation effect of co explains as to why the individuals having collectivistic orientation refrain from constructive deviance behaviors and do not go against the organizational norms and values. according to the conditional process analysis, co moderates the mediating relationship of po and cdb. it means the mediation impact of po on cdb is weekend or reduced by co. results suggest that individual sense of po is based on good fit among individuals and organization (pierce et al., 2001) which materializes in the form of cdb (chung & moon, 2011) as the collectivistic individuals prefer to engage in behaviors of strictly following norms, values, polices and ensuring organizational wellbeing and refrain from action of violating them (van dyne &pierce, 2004). similar finding were made by (chung & moon, 2011), who found that co negatively moderates between po and cdb. the phenomenon of constructive deviance has gained considerable attention of the academia, researchers as well as the practitioners in recent times (neall & tuckey, 2014; nielsen & einarsen, 2012). limited research has been done by scholars on the investigating and exploring positive side of cdb as compared to the negative side of this phenomenon. this has come as a surprise for the scholars and practitioners because until now it has not been expected that this negative concept could have a positive side also. the present study adds value to the existing research by exploring the untapped research area i.e., the positive side of the cdb. investigating the benefits associated this variable in relation p-o fit, po, co, cdb as well as the dynamic interaction of these constructs with one another provides a way forward for the future researchers to explore kamil huusain, saima aftab &mahwash hussain 46 cdb in relation to other work behaviours and organizational and individual level outcomes. journal of applied economics and business studies, volume. 6, issue 1 (2022) 25-52 https://doi.org/10.34260/jaebs.612 47 references acs, z. j., & mueller, p. (2008). employment effects of business dynamics: mice, gazelles and elephants. small business economics, 30(1), 85-100. al-jaghoub, s. (2004). building a knowledge-based economy: using icts for development and the role of the national state: a case study of jordan the university of manchester]. ambrose, m. l., arnaud, a., & schminke, m. (2008). individual moral development and ethical climate: the influence of person–organization fit on job attitudes. journal of business ethics, 77(3), 323-333. avey, j. b., avolio, b. j., crossley, c. d., & luthans, f. (2009). psychological ownership: theoretical extensions, measurement and relation to work outcomes. journal of organizational behavior: the international journal of industrial, occupational and organizational psychology and behavior, 30(2), 173-191. babin, b. j., hair, j. f., & boles, j. s. (2008). publishing research in marketing journals using structural equation modeling. journal of marketing theory and practice, 16(4), 279-286. baer, m., & brown, g. (2012). blind in one eye: how psychological ownership of ideas affects the types of suggestions people adopt. organizational behavior and human decision processes, 118(1), 60-71. blau, p. m. (1964). social exchange theory. retrieved september, 3(2007), 62. boon, c., & biron, m. (2016). temporal issues in person–organization fit, person–job fit and turnover: the role of leader–member exchange. human relations, 69(12), 2177-2200. bretz jr, r. d., & judge, t. a. (1994). the role of human resource systems in job applicant decision processes. journal of management, 20(3), 531-551. cable, d. m., & derue, d. s. (2002). the convergent and discriminant validity of subjective fit perceptions. journal of applied psychology, 87(5), 875. cable, d. m., & judge, t. a. (1996). person–organization fit, job choice decisions, and organizational entry. organizational behavior and human decision processes, 67(3), 294-311. caplan, r. d. (1987). person-environment fit theory and organizations: commensurate dimensions, time perspectives, and mechanisms. journal of vocational behavior, 31(3), 248-267. chawla, p. (2020). impact of employer branding on employee engagement in business process outsourcing (bpo) sector in india: mediating effect of person–organization fit. industrial and commercial training. cho, t., & faerman, s. r. (2010). an integrative model of empowerment and individuals' inrole and extra-role performance in the korean public sector: moderating effects of organizational individualism and collectivism. international public management journal, 13(2), 130-154. chung, y. w., & moon, h. k. (2011). the moderating effects of collectivistic orientation on psychological ownership and constructive deviant behavior. international journal of business and management, 6(12), 65. chureemas, r. (2021). assessing the application of intellectual property law for software development in relation to international law and the industry response. abac journal, 41(2), 259-270. kamil huusain, saima aftab &mahwash hussain 48 dalal, r. s. (2005). a meta-analysis of the relationship between organizational citizenship behavior and counterproductive work behavior. journal of applied psychology, 90(6), 1241. davis, j. h., schoorman, f. d., & donaldson, l. (1997). toward a stewardship theory of management. academy of management review, 22(1), 20-47. dierdorff, e. c., bell, s. t., & belohlav, j. a. (2011). the power of “we”: effects of psychological collectivism on team performance over time. journal of applied psychology, 96(2), 247. elçi, m., alpkan, l., & çekmecelioğlu, h. (2008). the influence of person organization fit on the employee’s perception of organizational performance. 4th international strategic management conference, sarajevo, bosnia-herzegovnia, galperin, b. l. (2003). can workplace deviance be constructive? in misbehaviour and dysfunctional attitudes in organizations (pp. 154-170). springer. galperin, b. l. (2012). exploring the nomological network of workplace deviance: developing and validating a measure of constructive deviance. journal of applied social psychology, 42(12), 2988-3025. garg, n., kumari, s., & punia, b. (2022). resolving stress of university teacher: exploring role of workplace spirituality and constructive workplace deviance. south asian journal of business studies. ghielen, s. t. s., de cooman, r., & sels, l. (2021). the interacting content and process of the employer brand: person-organization fit and employer brand clarity. european journal of work and organizational psychology, 30(2), 292-304. gino, f., & wiltermuth, s. s. (2014). evil genius? how dishonesty can lead to greater creativity. psychological science, 25(4), 973-981. greenberg, j., & scott, k. s. (1996). why do workers bite the hands that feed them? employee theft as a social exchange process. grobler, a., & holtzhausen, m. m. (2018). supervisory trust to be earned: the role of ethical leadership mediated by person-organisational fit. south african journal of economic and management sciences, 21(1), 1-11. hagemann, v., ontrup, g., & kluge, a. (2020). collective orientation and its implications for coordination and team performance in interdependent work contexts. team performance management: an international journal. hair, j. f., black, w. c., babin, b. j., anderson, r. e., & tatham, r. l. (2006). multivariate data analysis 6th edition. in: new jersey: prentice hall. hair, j. f., ringle, c. m., & sarstedt, m. (2012). partial least squares: the better approach to structural equation modeling? long range planning, 45(5-6), 312-319. han, t.-s., chiang, h.-h., mcconville, d., & chiang, c.-l. (2015). a longitudinal investigation of person–organization fit, person–job fit, and contextual performance: the mediating role of psychological ownership. human performance, 28(5), 425-439. haque, a. u., & oino, i. (2019). managerial challenges for software houses related to work, worker and workplace: stress reduction and sustenance of human capital. polish journal of management studies, 19. journal of applied economics and business studies, volume. 6, issue 1 (2022) 25-52 https://doi.org/10.34260/jaebs.612 49 hofstede, g. (1980). culture and organizations. international studies of management & organization, 10(4), 15-41. hofstede, g. (2001). culture's recent consequences: using dimension scores in theory and research. international journal of cross cultural management, 1(1), 11-17. hussain, k., & rehman, h. (2020). antecedents of constructive deviance behaviors and the mediating role of psychological empowerment: conditional process analysis. abasyn university journal of social sciences, 13(1). jang, s., shen, w., allen, t. d., & zhang, h. (2018). societal individualism–collectivism and uncertainty avoidance as cultural moderators of relationships between job resources and strain. journal of organizational behavior, 39(4), 507-524. jawad, m., tabassum, t. m., raja, s., & abraiz, a. (2013). study on work place behaviour: role of person-organization fit, person-job fit & empowerment, evidence from pakistan. journal of business and management sciences, 1(4), 47-54. jun, j. w., & lee, h. s. (2007). cultural differences in brand designs and tagline appeals. international marketing review. kim, u. e., triandis, h. c., kâğitçibaşi, ç. e., choi, s.-c. e., & yoon, g. e. (1994). individualism and collectivism: theory, method, and applications. sage publications, inc. kirk, c. p., peck, j., & swain, s. d. (2018). property lines in the mind: consumers’ psychological ownership and their territorial responses. journal of consumer research, 45(1), 148-168. kitayama, s., markus, h. r., & lieberman, c. (1995). the collective construction of self esteem. in everyday conceptions of emotion (pp. 523-550). springer. koburtay, t., & alzoubi, a. (2021). the linkages between person-organization spirituality fit and workers’ psychological well-being. the journal of social psychology, 161(1), 103118. krau, e. (2008). work, creativity, inventions and society. man and work, 16, 46-54. kristof‐brown, a. l., zimmerman, r. d., & johnson, e. c. (2005). consequences of individuals'fit at work: a meta‐analysis of person–job, person–organization, person–group, and person–supervisor fit. personnel psychology, 58(2), 281-342. lau, v. c., au, w. t., & ho, j. m. (2003). a qualitative and quantitative review of antecedents of counterproductive behavior in organizations. journal of business and psychology, 18(1), 73-99. lauver, k. j., & kristof-brown, a. (2001). distinguishing between employees' perceptions of person–job and person–organization fit. journal of vocational behavior, 59(3), 454470. liu, b., liu, j., & hu, j. (2010). person-organization fit, job satisfaction, and turnover intention: an empirical study in the chinese public sector. social behavior and personality: an international journal, 38(5), 615-625. liu, d., chen, x.-p., & yao, x. (2011). from autonomy to creativity: a multilevel investigation of the mediating role of harmonious passion. journal of applied psychology, 96(2), 294. liu, j., wang, h., hui, c., & lee, c. (2012). psychological ownership: how having control kamil huusain, saima aftab &mahwash hussain 50 matters. journal of management studies, 49(5), 869-895. love, m. s., & dustin, s. l. (2014). an investigation of coworker relationships and psychological collectivism on employee propensity to take charge. the international journal of human resource management, 25(9), 1208-1226. mayhew, m. g., ashkanasy, n. m., bramble, t., & gardner, j. (2007). a study of the antecedents and consequences of psychological ownership in organizational settings. the journal of social psychology, 147(5), 477-500. mehmood, s. a., malik, m. a. r., akhtar, m. s., faraz, n. a., & memon, m. a. (2021). organizational justice, psychological ownership and organizational embeddedness: a conservation of resources perspective. international journal of manpower. mishra, m., ghosh, k., & sharma, d. (2021). unethical pro-organizational behavior: a systematic review and future research agenda. journal of business ethics, 1-25. neall, a. m., & tuckey, m. r. (2014). a methodological review of research on the antecedents and consequences of workplace harassment. journal of occupational and organizational psychology, 87(2), 225-257. nielsen, m. b., & einarsen, s. (2012). outcomes of exposure to workplace bullying: a metaanalytic review. work & stress, 26(4), 309-332. nunnally, j. c., & bernstein, i. h. (1994). psychological theory. in: new york: mcgraw-hill. o'leary-kelly, a. m., griffin, r. w., & glew, d. j. (1996). organization-motivated aggression: a research framework. academy of management review, 21(1), 225-253. o'reilly iii, c. a., chatman, j., & caldwell, d. f. (1991). people and organizational culture: a profile comparison approach to assessing person-organization fit. academy of management journal, 34(3), 487-516. o’driscoll, m. p., pierce, j. l., & coghlan, a.-m. (2006). the psychology of ownership: work environment structure, organizational commitment, and citizenship behaviors. group & organization management, 31(3), 388-416. organ, d. w., podsakoff, p. m., & mackenzie, s. b. (2005). organizational citizenship behavior: its nature, antecedents, and consequences. sage publications. ozler, h., yilmaz, a., & ozler, d. (2008). psychological ownership: an empirical study on its antecedents and impacts upon organizational behaviors. problems and perspectives in management(6, iss. 3), 38-47. park, c. w., eisingerich, a. b., & park, j. w. (2013). from brand aversion or indifference to brand attachment: authors' response to commentaries to park, eisingerich, and park's brand attachment–aversion model. journal of consumer psychology, 23(2), 269-274. pascale, r. t., & sternin, j. (2005). your company’s secret change agents. harvard business review, 83(5), 72-81. pierce, j. l., kostova, t., & dirks, k. t. (2001). toward a theory of psychological ownership in organizations. academy of management review, 26(2), 298-310. pierce, j. l., kostova, t., & dirks, k. t. (2003). the state of psychological ownership: integrating and extending a century of research. review of general psychology, 7(1), 84-107. pierce, j. l., o'driscoll, m. p., & coghlan, a.-m. (2004). work environment structure and journal of applied economics and business studies, volume. 6, issue 1 (2022) 25-52 https://doi.org/10.34260/jaebs.612 51 psychological ownership: the mediating effects of control. the journal of social psychology, 144(5), 507-534. preacher, k. j., rucker, d. d., & hayes, a. f. (2007). addressing moderated mediation hypotheses: theory, methods, and prescriptions. multivariate behavioral research, 42(1), 185-227. rafferty, a. e., & minbashian, a. (2019). cognitive beliefs and positive emotions about change: relationships with employee change readiness and change-supportive behaviors. human relations, 72(10), 1623-1650. rahman, m. f. w., kistyanto, a., & surjanti, j. (2022). does cyberloafing and person‐ organization fit affect employee performance? the mediating role of innovative work behavior. global business and organizational excellence, 41(5), 44-64. reb, j., & connolly, t. (2007). possession, feelings of ownership, and the endowment effect. judgment and decision making, 2(2), 107. robbins, d. l., & galperin, b. l. (2010). constructive deviance: striving toward organizational change in healthcare. journal of management and marketing research, 5, 1. robert, c., & wasti, s. a. (2002). organizational individualism and collectivism: theoretical development and an empirical test of a measure. journal of management, 28(4), 544566. robinson, s. l., & bennett, r. j. (1995). a typology of deviant workplace behaviors: a multidimensional scaling study. academy of management journal, 38(2), 555-572. schneider, s. l. (2001). in search of realistic optimism: meaning, knowledge, and warm fuzziness. american psychologist, 56(3), 250. sekiguchi, t. (2007). a contingency perspective of the importance of pj fit and po fit in employee selection. journal of managerial psychology. shahzad, f., iqbal, z., & gulzar, m. (2013). impact of organizational culture on employees job performance: an empirical study of software houses in pakistan. journal of business studies quarterly, 5(2), 56. sharkawi, s., rahim, a. r. a., & azuradahalan, n. (2013). relationship between person organization fit, psychological contract violation on counterproductive work behaviour. international journal of business and social science, 4(4). sieger, p., bernhard, f., & frey, u. (2011). affective commitment and job satisfaction among non-family employees: investigating the roles of justice perceptions and psychological ownership. journal of family business strategy, 2(2), 78-89. sieger, p., zellweger, t., & aquino, k. (2013). turning agents into psychological principals: aligning interests of non‐owners through psychological ownership. journal of management studies, 50(3), 361-388. spreitzer, g. m., & sonenshein, s. (2004). toward the construct definition of positive deviance. american behavioral scientist, 47(6), 828-847. suárez‐mendoza, m. j., & zoghbi‐manrique‐de‐lara, p. (2007). the impact of work alienation on organizational citizenship behavior in the canary islands. international journal of organizational analysis. tekmen, e. e., & kaptangil, k. (2022). the determinants of constructive deviant behaviour kamil huusain, saima aftab &mahwash hussain 52 of frontline tourism employees: an exploration with perceived supervisory support and intrinsic motivation. journal of tourism, sustainability and well-being, 10(1), 5874. ucar, a. c., alpkan, l., & elci, m. (2021). the effects of person–organization fit and turnover intention on employees’ creative behavior: the mediating role of psychological ownership. sage open, 11(4), 21582440211066924. uggerslev, k. l., fassina, n. e., & kraichy, d. (2012). recruiting through the stages: a meta‐ analytic test of predictors of applicant attraction at different stages of the recruiting process. personnel psychology, 65(3), 597-660. vadera, a. k., pratt, m. g., & mishra, p. (2013). constructive deviance in organizations: integrating and moving forward. journal of management, 39(5), 1221-1276. valentine, s., godkin, l., & lucero, m. (2002). ethical context, organizational commitment, and person-organization fit. journal of business ethics, 41(4), 349-360. van dyne, l., & pierce, j. l. (2004). psychological ownership and feelings of possession: three field studies predicting employee attitudes and organizational citizenship behavior. journal of organizational behavior: the international journal of industrial, occupational and organizational psychology and behavior, 25(4), 439-459. wagner, s. h., parker, c. p., & christiansen, n. d. (2003). employees that think and act like owners: effects of ownership beliefs and behaviors on organizational effectiveness. personnel psychology, 56(4), 847-871. wang, h. j., demerouti, e., blanc, p. l., & lu, c. q. (2018). crafting a job in ‘tough times’: when being proactive is positively related to work attachment. journal of occupational and organizational psychology, 91(3), 569-590. warren, d. e. (2003). constructive and destructive deviance tn organizations. academy of management review, 28(4), 622-632. yildiz, b., alpkan, l., ates, h., & sezen, b. (2015). determinants of constructive deviance: the mediator role of psychological ownership. international business research, 8(4), 107. zulfiqar, s., sadaf, r., popp, j., vveinhardt, j., & máté, d. (2019). an examination of corporate social responsibility and employee behavior: the case of pakistan. sustainability, 11(13), 3515. journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 111 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x agriculture for sustainable development: empirical evidence on carbon footprint in agriculture production, expansion, and trade adnan habib1, saima sarwar2*, uzair ahson3 and alvina sabah idrees4 1department of economics, government college university, lahore, pakistan. 2department of economics, government college university, lahore, pakistan. 3department of economics, government college university, lahore, pakistan. 4department of economics, government college university, lahore, pakistan. abstract the international food production is reason of almost one-third of the total humans caused greenhouse gas (ghg) emissions. the sustainable development goals (sdgs) include agriculture sector growth and management of climate change. carbon footprint is an indicator of ghg intensity, created from different economic actions. it represents more than fifty percent of the total ecological footprint and used as managing tool for estimating environmental pollution. the current study investigates the relationship between agriculture sector specific indicators and carbon footprint in fifty-six countries by using panel econometrics. the outcomes of the study provide strong evidence on the presence of environmental kuznets curve (ekc). the analysis also shows that there exists negative association between carbon footprint and agricultural development. however agriculture sector expansion by employing environment friendly methods and technologies decreases carbon footprint in selected countries. furthermore, the relationship between carbon footprint and agricultural exports is found positive. it implies that agricultural exports encourage the carbon footprint growth by stimulating the production and transport of agricultural commodities. finally, there is positive relationship between carbon footprint and high scale of agricultural production, which supports the concept of production-based emission. these findings emphasizes that too much use of fertilizers in agriculture sector fosters the carbon footprint growth and damages natural environment of countries. keywords carbon footprint, agriculture, determinants, agricultural development, agricultural production, agricultural exports jel classification q00, q01, q17 * saimasarwar@gcu.edu.pk mailto:saimasarwar@gcu.edu.pk adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 112 1. introduction serious efforts are required in current decade to achieve sdgs, as deadline is approaching in year 2030. sustainable development relies on the three domains: social, economic, and environment, which perform the important part in decision making for the policymakers (dernbach and mintz, 2011). sustainable development through environmental sustainability requires that use of resources must be less than refill rate, even though population is raising carbon emissions, but overall level should remain low (khan, 1995). green growth, environmentally sustainable economic growth is considered a main tool to accomplish sustainable development. economic growth with help of ecological and environmental sustainability is regarded as key goal of development policies across world. to attain green growth, control of co2 emissions is necessary, which is only possible with innovations in methods of supply and use of green technologies to help cleaner production (wiebe and yamano, 2016). the second and thirteenth sdgs state that the world needs urgent work on climate change and agriculture sector growth. improvements and rearrangements are required in food production, distribution, and consumption to achieve sustainable economic growth. current picture of world climate advocates that zero net ghg emissions are required to prevent the world from being adversely affect by climate change in future. international climate policy formulated in the twenty-first congress of the parties (cop21) of the unfccc in paris aims to control the rise in the worldwide temperature to sufficiently below two degrees celsius above preindustrial levels and taking efforts to limit the temperature increase to one and half degree celsius above preindustrial levels (unfccc, 2015). policy emphasis the part of different sectors in climate change mitigation, particularly focusing on agriculture sector. agriculture has been always main part of debate on carbon emission, its sources and effects, and carbon sink. in the past, agriculture production was household-based with short distance transportation, little energy consumption and little use of fuel (amate and de molina, 2013). according to krey et al. (2012) agriculture consumes less electricity and heat, transportation, and industry. currently, agriculture sector is a major contributor of climate change as it produces a large quantity of ghg emissions. almost one-third of the total ghg emissions originates from the agriculture sector (gilbert, 2012). in last decade, ghg emissions has increased due to large scale production in agriculture (calvin et al., 2016). mbow et al., 2021 report that the global food system produces twenty-one percent to thirty-seven percent annual ghg emissions. according to international atomic energy agency (iaea) agricultural activities generates almost journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 113 thirty percent of total ghg emissions, which are reason of climate change and global warming. the food system is main reason of co2 emissions due to change in land-use, which is result of clearing land for pastures and crop production. fourteen percent of annual human caused co2 emissions are due to change in net land-use (le quéré et al., 2018) and ten percent of these are from agriculture (mbow et al., 2021). application of urea and lime in agriculture also originate co2 emissions. high demand of food by world population, the greater demand for meat and dairy products, and the intensification of agricultural methods has increased rate of ghg emissions. the international food system including fertilizer production, storage of food and its packing is reason for thirty-three percent of total human caused ghg emissions (gilbert, 2012). these emissions comprise of three gases i.e., n2o, ch4, and co2. co2 emissions in agriculture sector has increased due to more use of energy in the sector. these include emissions from either tractor and other machines fuel or from fertilizers manufacturing and their transportation (vermeulen et al., 2012). these emissions are measured as transport and energy emissions in the accounting framework of intergovernmental panel on climate change (ipcc). these emissions can be reduced by decarbonization of energy generation sources. agriculture sector consumes around eleven percent of the total land of earth for yield production and employs seventy percent of the entire water surface (fao, 2003; fao, 2011). current scenario of global warming and climate changes has a great impact on the sustainability of agricultural systems. therefore, springmann et al. (2018) describe the need to reduce agricultural emissions to maintain world environment. poore and nemecek (2018) investigate efforts on reducing environmental effects of food through consumers and producers. use of carbon footprint and its estimations in literature has been increased in last decade. the carbon footprint determines the carbon amount being produced by economic activity. the carbon footprint is a demand for biologically productive space, and it is major component of the ecological footprint (gfn, 2018). it is used as managing tool for evaluating the environmental impact of different nations and main indicator of pollution in the environmental economics, as it is significant factor to improve the understanding of environmental degradation. ecological footprint estimates a nation’s usage of grazing land, cropland, forests, and fishing grounds to supply resources and to absorb co2 from the burning of fossil fuels (gfn, 2018). notably, the carbon footprint describes greater than fifty percent of all ecological footprint in several countries. moreover, the carbon footprint is generally accepted indicator of ghg adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 114 concentration, resulting from different economic actions. scientists and policymakers treat it as a control indicator due to its increasing importance. the paper examines the agriculture specific determinants of carbon footprint, specifically aiming on the economic growth and development, agricultural development, agricultural trade (in terms of agricultural exports) and agricultural production. next section of paper provides a systematic review of past literature on carbon footprint especially with respect to agriculture and agri-related subsectors. third section is development of a conceptual framework to examine agriculture-specific determinants of carbon footprint. fourth section describes estimation technique including a detailed account of indicators and methodological framework. fifth section provides estimates of econometric models. it comprises of discussions on findings and determinants of carbon footprint. last section concludes analysis of the current study. 2. literature review current literature on agriculture sector emissions focuses on estimations or measurement of emissions (castesana et al., 2018; rehman et al., 2019). different works exist on the macro level estimation methods of ghg emissions. there are two methods to estimate ghg emissions: the production-based and consumption-based approach (mózner, 2013). the consumption-based method declares emissions are emitted-emissions, which occur during the supply-chain of commodities utilized within a country, regardless of their production tertiary. therefore, countries are liable for emissions caused somewhere else due to its consumption, while the production-based method is based on domestic emission due to inventories (peters and hertwich, 2008; mózner, 2013). the association between consumption-based emissions and gdp is higher than the gdp’s association with territorial emissions (ipcc, 2014). so, carbon footprint and consumption expenditure strongly correlate. the ipcc (2014) identifies various income and non-income features as main cause of carbon emissions in the recent years, like consumption expenditures, production methods, transport infrastructure, waste management, energy systems, population growth and trends in demographic structure (urbanization). per capita emission of carbon footprint also depends on some other nonincome factors like geography, diet, and lifestyle (gaia, 2012; corsten et al., 2013). empirical investigations to assess agriculture related determinants of carbon footprint at the nations level are rare in the past works, particularly from global point of view. only a few studies explore the factors of agriculture sector emissions. kastratović (2019) examines the consequences of foreign direct investment on agricultural journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 115 emissions. he only considers co2 emissions but does not estimate long run and short run links. due to globalization, foreign direct investment (fdi) inflows and trade openness has increased. these are also major reasons of gas emissions for the host nations (pao and tsai, 2011; naranpanawa, 2011; ren et al., 2014; le et al., 2016). international trade is also a reason of the difference in carbon emissions (peters and hertwich, 2008; bows and barrett, 2010). agriculture is a main cause of nitrogen dioxide emissions to the environment (audet et al., 2017). fertilizer and cattle are the key determinants of co2 emissions in agriculture sector (luo et al., 2017). mineral fertilizers, agriculture waste burning, and manure management are the main causes of ammonia for argentina. different studies report their findings related to growth of population in poor and rich countries, which causes to increase ghg emissions. according to jorgenson and clark (2010) population elasticity of ghg emissions for developed countries is 1.65 percent and for developing countries it is 1.27 percent. poumanyvong and kaneko (2010) estimate the value of population elasticities of ghg emissions as 1.75 for low-income, 1.23 for middleincome, and 1.12 for high-income nations. furthermore, increase in labor productivity decreases intensity of gas emissions. most of past studies only measures gas emissions or don’t provide theoretical background for the empirical examination. robaina-alves and moutinho (2014), luo et al. (2017) and castesana et al. (2018) analyze the determinants of co2 emissions from agriculture sector. robaina-alves and moutinho (2014) decompose agriculture sector emissions into factors like economic growth, human capital accumulation, energy structure, energy intensity, emissions components, and labor productivity. they find that nitrogen use per cultivated area is main determinant of gas emissions. ben jebli and ben youssef (2019) find that agriculture sector value addition, and per capita increase in renewable combustible material and consumption of waste decrease co2 emissions in the long run. they provide long run and short run factors of agricultural emissions at world level instead of individual country. studies on carbon footprint also report product-level data and certain geographical area. according to muthu (2014) among three major chinese crops rice is producing highest percentage of carbon footprint, wheat is second in the list and maize is third. fodders of agricultural farms produce more than two-third of the carbon footprint. the work of jacobsen et al. (2014) on livestock shows that one kg of pig’s meat produces carbon footprint equivalent to almost 5.7 kg co2. according to verge et al. (2008) and desjardins et al. (2014) beef produces largest amount of carbon footprint among meat adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 116 production of different animals. among the dairy products, milk powders generate highest carbon footprint in dairy production, butter is second and cheese is third in the list. past literature also provides cross-country differences in carbon footprint. in addition to this, within a country agriculture seems to produce varied amount of carbon footprint. carbon footprint of countries depends on country size, role of agriculture in country, agricultural production, population, and use of technology (balogh, 2019). china is main producers of the worldwide carbon footprint and contributor of climate change. carbon footprint of china’s crop output are more than eight percent of the country’s overall emissions and sixty-six percent of the carbon footprint are of the agrochemical origin. energy consumption and irrigation produces twenty-two percent on average, while machinery management and plastic film contributes below ten percent of the overall carbon footprint in country’s crop output (muthu, 2014). agriculture sector performs an important part in the economy by providing food and nutrition, alongside further environmental, social, and economic effects (li et al., 2016), but this area of economy is also a major cause of co2 emissions causing climate change and global warming (oenema et al., 2001; tubiello et al., 2013; calvin et al., 2016; agovino et al., 2019). technological advancement can reduce co2, ammonium cation and nitrogen dioxide emissions from the agriculture sector (cole et al., 1997). strategies to lessen the gas emissions from agriculture sector can be formed by focusing on demand side factors (franks and hadingham, 2012) rather than supply side factors like land use, crop production decisions, soil fertility and forests (de pinto et al., 2016). sebri and abid (2012), chen et al. (2017), waheed et al. (2018) and paul et al. (2018) work on the demand side economic factors of agriculture sector ghg emissions. ben jebli and ben youssef (2017) find bidirectional causality between co2 emissions and agriculture sector in the long run and short run. ghg emissions from agricultural output per unit have been decreased globally (bennetzen et al., 2016). the fast growth in technology and economic integration in last decade produced high emissions from agriculture sector (wri, 2015). agriculture sector emissions varies country to country (dace and blumberga, 2016). li et al. (2016) work on the importance of long run and short run factors of agriculture sector gas emissions. giannadaki et al. (2018) find that agriculture sector ammonia gas emissions increase particulate matter air pollution, have significant harmful effects on human being health and increases rate of mortality. fifty percent decrease in agriculture sector emissions can save more than two hundred thousand lives per year in the selected fifty-nine countries and can provide economic benefits of billions of us dollars. the mortality journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 117 rate can be decreased eighteen percent with an annual economic benefit of billions of us dollars. decrease in agriculture sector emissions creates social and financial benefits. hence, an analysis of the agriculture-specific determinants of emissions across different income level nations is useful. the topic of the current study is so crucial that it must be focused because reduction in agriculture sector emissions gives significant social and economic benefits. identification of agriculture-specific factors of gas emissions is important for policy making in this area. 3. theoretical framework consumption side analysis reveals that high income, developed and populated countries consumes processed food products, more meat and demand great amount of food products that cause greater carbon footprint. there exists a positive association between co2 emissions per person and gdp per person (ang, 2007). the ekc hypothesis has been proposed by grossman and krueger (1995) for examining the relationship between ecological pollution and economic growth like kuznets (1955)1 inverted u-hypothesis for income inequality and economic development. grossman and krueger (1995) state that pollution increases and environmental quality decreases at early stage of economic growth, but after achieving certain level of per capita income, environment quality improves with economic growth. therefore, per capita pollution emissions are the inverted u-shaped function of income per person (stern, 2004; ozturk and acaravci, 2010; al-mulali, saboori, and ozturk, 2015; hussain and dey, 2021). economic development and industrial growth increase the use of natural resources, which damages environment, but after development environmental quality increases due to usage of cleaner technologies in the post-industrial stage (munasinghe, 1999). the present study examines ekc hypothesis on carbon footprint. presence of inverted u-shaped long run link between countries’ carbon footprint and economic growth (gdp per capita) suggests: carbon footptint = f(gdp per capita, gdp per capita2) (1) balogh and jambor (2017) finds that agricultural development by engaging environment friendly technologies reduces agricultural co2. the relationship between carbon footprint and agricultural development implies: carbon footptint = f(agricultural value added) (2) 1 kuznets (1955) suggests that inequality increases in the country at the early stage of income growth, but it moves back towards greater equality as economic growth endures. adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 118 trade of processed meat products and animal feed has increased due to globalization (kearney, 2010). it has reduced burden on environment (balogh and jambor, 2017) and decreased carbon footprint of countries through technological improvements. agriculture sector trade effects carbon footprints by increasing food production and transport use. so, the relationship between carbon footprint and agricultural exports declares: carbon footptint = f(agricultural exports) (3) environmental degradation increases with higher scale of agricultural production (foley et al., 2011; baccini et al., 2012; grace et al., 2014; henders et al., 2015). so, carbon footprint rises with increase in agricultural production, which depends on agricultural machinery and use of fertilizers. effect of tractors and fertilizers use on carbon footprint propose: carbon footptint = f(fertilizers, tractors) (4) consumption habits of urban and rural populations have varied effect on the natural resources use. more urbanization in the countries also effect carbon emissions and spatial development of cities cause global warming through ghg emissions (sethi, 2017). so, countries with greater percentage of rural population have less carbon footprint as compared to a country that has more urban population. therefore, carbon footprint also depends on percentage of rural population in the country. carbon footptint = f(rural population) (5) figure 1 provides framework to find agriculture-specific determinants of carbon footprint. figure 1: carbon footprint, gdp per capita and agriculture sector gdp per capita agriculture sector development agriculture sector exports agriculture sector production source: authors’ developed theoretical framework based on literature review. journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 119 4. data and methods this section describes the path of empirical estimation of discussed phenomena. it includes details of data and methods employed in the analysis. 4.1. data and variable selection data consists of seven indicators for fifty-six countries selected from barilla center for food and nutrition list for reference years 1980 to 2017. the list of sample countries with their respective sustainable agriculture index is provided in table a (appendix). food and agriculture organization (fao) of the united nations, world development indicators (wdi) of world bank, and global footprint network (gfn) are the sources of data. selection of indicators is based on framework developed in last section. detail description of the selected variables is mentioned in table b (appendix). 4.2 econometric specifications this study employed panel data models in examining the association between the carbon footprint and its determinants. panel data provides more observations with the help of pooling time series information across cross-sections and permits for informative data, high efficiency, high variability, more degrees of freedom and lesser collinearity among indicators (gujarati, 2004). these advantages are possible if and only if individual heterogeneity does not exist in the data (baltagi, 2008). this empirical study is conducted by using the following econometric techniques (also mentioned in table 1). table 1: methodological framework to study the relationship between the carbon footprint and its determinants step inquiry test/ method objective 1 cross-sectional dependence (cd) cd test by pesaran (2015) to unveil cross-country dependence between indicators. 2 slope homogeneity slope homogeneity test by pesaran and yamagata (2008) to reveal the heterogeneity between the cross-sections. 3 unit root problem cadf and cips tests by pesaran (2007) to verify order of integration of variables. 4 long run cointegration westerlund (2007) cointegration tests to confirm presence of cointegration between variables. 5 estimation of long run relationship dcce mg by chudik and pesaran (2015) to decide long run link between the variables. 6 granger causality dumitrescu and hurlin (2012) causality test to study the direction of relationship and causal links between the variables. source: authors’ developed framework. adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 120 4.2.1 cd test firstly, cd test is required to examine cross-country dependence as cd analysis is the basic step in the study of panel data. cross-correlation between error terms of econometric model and existence of non-zero error covariance between them indicates cd. presence of cd is crucial problem in panel unit root and cointegration tests. these tests are irrelevant if cd assumption does not hold due to correlations among individual cross-sections (chang, 2002). second-generation tests like cross-sectionally augmented im, pesaran, shin (cips) and tests of westerlund cointegration assume heterogeneity and dependence among cross-sections. different factors such as common shocks or model misspecification may cause cd dependence (cerrato, 2002). if cd exists among units but not considered, results can be significantly biased (breusch and pagan, 1980; pesaran, 2004). cd is defined as error term of panel individuals is related in the econometric model. shock of economy effects individuals and influence other units in the panel. this has been described in following equation: y(i,t) = α(i) + β(i)x(i,t) + u(i,t) (6) where cov(u(i,t), u(j,t)) ≠ 0 (7) in all equations of paper, the subscripts ‘i’ and ‘j’ show the cross-sections i.e., i = 1, 2, …, n and j = 1, 2, …, n and the subscript ‘t’ shows the time i.e., t = 1, 2, …, t. the lagrange multiplier cdlm test by pesaran (2015) is applied to exam cd. this cdlm test is effective when n is greater than t. null hypothesis of cdlm test considers weak cd across countries against the alternative hypothesis in which the strong cd is present. statistics of cd test is: cd = √ 2t n(n−1) ∑ ∑ ρ̂(i,j) n j=i+1 n−1 i=1 ~n(0,1)i, j (8) here ρ̂(i,j)is the sample approximation of the pairwise correlation of errors. ρ̂(i,j) = ∑ û(i,t) t t=1 û(j,t) √(∑ û(i,t) 2t t=1 )√(∑ û(j,t) 2t t=1 ) (9) here û(i,t) 2 is estimate of u(i,t) in equation 6. journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 121 4.2.2 slope homogeneity test the second step is application of slope homogeneity test (pesaran and yamagata, 2008) to reveal the existence of slope homogeneity among units. this test does not allow cd (atasoy, 2017). tests statistics can be expressed as: ∆̃sh= √n √2k −2 ( 1 n s̃ − k) (10) ∆̃ash= √n √ 2k(t−k−1) t+1 −2 ( 1 n s̃ − k) (11) here ∆̃sh is delta tilde and ∆̃ash is the adjusted delta tilde. 4.2.3 unit root tests the third step employs panel unit root tests to confirm integration order of variables. literature reports two kinds of panel unit root tests. the first type is described as firstgeneration unit root tests; this type does not allow for cd and may offer deceptive results (dogan and seker, 2016). the second type is termed as second-generation unit root tests; which considers cd (e.g., phillips and sul 2003; bai and ng 2004; smith et al. 2004; moon and perron 2004; pesaran, 2007). khan et al. (2020) recommend use of second generation non-parametric and parametric tests to avoid from being wrong. the current study employed cross-sectionally augmented dicky-fuller (cadf) and cross-sectionally augmented im pesaran shin (cips) tests by pesaran (2007). cips test also encounters cd heterogeneity and provides reliable results. panel unit root test of pesaran (2007) permits n t ⟶ σ to obtain a positive real value. the equations of cadf test can be written as: ∆y(i,t) = α(i) + β(i)y(i,t−1) + γ(i)y̅(i,t−1) + δ(i)∆y̅(i,t−1) + e(i,t) (12) the t-ratio of ols estimates β(i) in equation 12, defined by t(i)(n, t), referred as a cadf statistics for i, and the average of its t-ratios is cips(n, t) = n−1 ∑ t(i)(n, t) n i=1 (13) this equation provides the panel unit root test statistic. cips(n, t) is a crosssectionally augmented edition of the test statistic suggested by im, pesaran and shin (2003) and is described as a cips statistic. adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 122 4.2.4 westerlund panel cointegration tests the fourth step entails the analysis of the long run relationship among carbon footprint and its determinants. the panel cointegration tests given by westerlund’s (2007) have been used in the current study. these tests check presence of cointegration by deciding whether the error correction exists for individual panel members and for panel. these tests show high power and improved size accuracy as compared to the methodology developed by pedroni (2004), which are error-based tests. these tests provide robust and reliable results and managing cross-sectional dependency of the residual terms (kapetanios et al., 2011). the panels tests and the group-mean tests use null hypothesis of no cointegration to decide existence of cointegration. westerlund (2007) suggests four panel cointegration test statistics i.e., ga, gt, pa, and pt. these are normally distributed and based on error correction mechanism (ecm). the logic here is to check the lack of cointegration by deciding whether error correction occurs for individual panel members and for whole panel. the westerlund cointegration equations can be written as: without intercept: y(i,t) = β(i)∆y(i,t−1) + γ(i)∆x(i,t−1) + δ(i)(y(i,t) − θ(i)x(i,t−1)) + u(i,t) (14) with intercept: ∆y(i,t) = α(i) + β(i)∆y(i,t−1) + γ(i)∆x(i,t−1) + δ(i)(y(i,t−1) − θ(i)x(i,t−1)) + u(i,t) (15) with intercept and trend: ∆y(i,t) = α(i) + β(i)∆y(i,t−1) + γ(i)∆x(i,t−1) + ω(i)t + δ(i)(y(i,t−1) − θ(i)x(i,t)) + u(i,t) (16) here δi provides an estimation of the error-correction speed toward the long run equilibrium. the two statistics gt and pt are calculated with the standard errors of the parameters of error correction, calculated in normal way. the other two tests, ga and pa are based on two standard errors developed by newey and west (1994). these statistics are adjusted standard errors for autocorrelations and heteroskedasticity. the test statistics can be described as: gt = 1 n ∑ δ(i) s.e(δ(i)) n i=1 (17) journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 123 ga = 1 n ∑ tδ(i) δ(i)(1) n i=1 (18 pt = δi s.e(δ(i)) (19) pa = tδ(i) (20) 4.2.5 long run estimation final step is evaluation of long run relationship among carbon footprint and agriculture sector variables, in the presence of cointegration. for this purpose, this study estimates different models with chudik’s and pesaran’s (2015) dynamic commoncorrelated effects mean group (dcce mg) method. it is extension of cce method developed by pesaran (2006) to heterogeneous panel models along with lagged dependent regressor and/or weakly exogenous regressors. this approach involves two conditions: first, enough lags of cross-section means should be involved in individual equations of the panels; and second, the number of cross-section means should be equal to or greater than the number of unobserved common factors. dcce mg estimation method requires sufficiently large time series length of the panel. this method also deals cd and heterogenous slope coefficient. the cd is shaped using cross-sectional means of the dependent variable and regressors to account for the unobserved common factors. estimations are also robust to structural breaks in dcce mg method. in this approach, set of unobservable common factor is considered as a common dynamic process, which provides useful interpretations. different models of carbon footprint need to be assessed to ensure long run relationship between variables. dcce mg specification for estimated models of carbon footprint are: ln_carbonfp(i,t) = α0(i) + α1(i)ln_carbonfp(i,t−1) + α2(i)ln_gdppc(i,t) + α3(i) ln_gdppc(i,t) 2 + α4(i)agriva(i,t) + ∑(d(i)z(i,s)) + εr1(i,t) (21) ln_carbonfp(i,t) = β0(i) + β1(i)ln_carbonfp(i,t−1) + β2(i)ln_gdppc(i,t) + β3(i) ln_gdppc(i,t) 2 + β4(i)agriva(i,t) + β5(i)ln_agriexp(i,t) + ∑(d(i)z(i,s)) + εr2(i,t) (22) ln_carbonfp(i,t) = γ0(i) + γ1(i)ln_carbonfp(i,t−1) + γ2(i)ln_gdppc(i,t) + γ3(i) ln_gdppc(i,t) 2 + γ4(i)agriva(i,t) + γ5(i)ln_agriexp(i,t) + γ6(i)ln_fer(i,t) + γ7(i) ln_trac(i,t) + ∑(d(i)z(i,s)) + εr3(i,t) (23) adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 124 ln_carbonfp(i,t) = δ0(i) + δ1(i)ln_carbonfp(i,t−1) + δ2(i)ln_gdppc(i,t) + δ3(i) ln_gdppc(i,t) 2 + δ4(i)agriva(i,t) + δ5(i) ln_trac(i,t) + δ6(i)ruralpop(i,t) + ∑(d(i)z(i,s)) + εr4(i,t) (24) ln_carbonfp(i,t) = λ0(i) + λ1(i)ln_carbonfp(i,t−1) + λ2(i)ln_gdppc(i,t) + λ3(i) ln_gdppc(i,t) 2 + λ4(i)agriva(i,t) + λ5(i)ln_agriexp(i,t) + λ6(i)ln_fer(i,t) + λ7(i) ln_trac(i,t) + δ8(i)ruralpop(i,t) + ∑(d(i)z(i,s)) + εr5(i,t) (25) here, natural log of carbon footprint ln_carbonfp(i,t) is dependent variable, ln_carbonfp(i,t−1) is lag of the dependent, ln_gdppc(i,t) is natural log of gdp per capita, ln_gdppc(i,t) 2 is square term of natural log of gdp per capita, agriva(i,t) is agricultural value added, ln_agriexp(i,t) is natural log of agricultural exports, ln_fer(i,t) is natural log of fertilizers quantity used in agriculture, ln_trac(i,t) is natural log of tractors employed in agriculture and ruralpop(i,t) is rural population percentage in total population. furthermore z(i,s) is a (1 x k+1) vector including the cross-sectional averages at time ‘s’ and the sum is over s = t, t-1, ..., t-p (p are the lags of the crosssectional means). while εr1(i,t), εr2(i,t), εr3(i,t), εr4(i,t), and εr5(i,t) are the error terms in dcce mg equations. 4.2.6 dumitrescu-hurlin (d-h) causality test application of dumitrescu-hurlin causality test (2012) ensures the direction and causal association between indicators after the estimation of long run relationship. the null hypothesis of the d-h causality says that there is no causal association between indicators against the alternative hypothesis, which reflects that there is causal link between indicators. the model can be written as: y(i,t) = βi + ∑ δ(i) j y(i,t−j) p j=1 + ∑ θi j t(i,t−j) p j=1 (26) here, j donates the lag length, while δ (j) j represents the autoregressive parameters. 5. results and findings analysis begins with averages and standard deviations estimates of indicators (reported in table 2). selected countries have on average 1.36e+08 global hectors carbon footprint, 17353.33 us dollars (constant 2010) gdp per capita, agriculture value added equals to 10.58 percent of gdp, 9.1e+12 us dollars agricultural exports, 503.13 tractors per hundred square kilometer of (arable) land, 145.43 kg fertilizers per hectare of arable land, and 39.53 percent rural population out of total population. journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 125 table 2: descriptive statistics of indicators variable all countries mean standard deviation carbon footprint (global hectors) 1.36e+08 3.80e+08 gdp per capita (constant 2010 us$) 17353.33 19259.24 agriculture value added (% of gdp) 10.58 12.23 agricultural exports (1000 us$) 9.1e+09 1.7e+10 tractors (per 1002 km of (arable) land) 503.13 929.70 fertilizers (kg per hectare of arable land) 145.43 152.46 rural population (% of total population) 39.53 23.23 source: authors’ calculations using the dataset from faoun, wdi and gfn. due to globalization and liberalization, world has developed a replica of small town. actions in one economy impact nearby countries. hence, it is necessary to address the cd in the analysis of panel data. so, application of lagrange multiplier test of error cd checks cd in carbon footprint model. cdlm statistics values of the lagrange multiplier test by pesaran (2015) displays that there is cd in carbon footprint model. p-values of cdlm test statistics reject the null hypothesis of not any cross-sectional independence at one percent significance level. in other words, there exists dependency across nations. table 3 displays the results of cd tests. table 3: lagrange multiplier tests of error cd variables cd statistics p-value ln_carbonfp(i,t) 241.811 0.00*** ln_gdp(i,t) 241.747 0.00*** ln_gdp2(i,t) 241.277 0.00*** agriva(i,t) 231.824 0.00*** ln_agriexp(i,t) 241.692 0.00*** ln_fer(i,t) 218.005 0.00*** ln_trac(i,t) 194.359 0.00*** ruralpop(i,t) 239.339 0.00*** overall model 37.496 0.00*** note: h0: errors are weakly cross-sectional dependent. *** indicates significant at 1% level. source: authors’ calculations using the dataset from faoun, wdi and gfn. figure 2 reports kernel density estimates and graph of residuals of carbon footprint regression model. density graph confirms that residuals do not follow normal distribution. adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 126 figure 2: kernel density estimates and graph source: graph is based on authors’ estimates of carbon footprint model. table 4 reports results of homogeneity pesaran and yamagata (2008) test and confirms that there is heterogeneity in the data. it suggests that the panel data constants are heterogeneous, and the slopes differ across cross-sections. it also suggests that relationships of economic variables can be distinct in different countries. so, assumption of homogeneity in model estimation can provide ambiguous and biased results (alam et al. 2018). table 4: slope homogeneity test by pesaran and yamagata test-statistics value p-value ∆̃sh 46.371 0.00*** ∆̃ash (adjusted) 53.081 0.00*** note: *** indicates significant at 1% level. source: authors’ calculations using the dataset from faoun, wdi and gfn. the presence of dependence and panel data heterogeneity requires the secondgeneration unit root tests and the panel cointegration method of dcce mg for the estimations. order of integration has been confirmed to avoid spurious regression estimations (ulucak and bilgill, 2018). therefore, cadf and cips have been used. table 5 shows the cadf test results. stated statistics of unit roots are for logged values except agriculture value added and rural population. journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 127 table 5: cadf unit root tests variables level first difference order intercept intercept and trend intercept intercept and trend ln_carbonfp(i,t) -1.994** -2.048 i(0) ln_gdp(i,t) -2.131*** -2.619*** i(0) ln_gdp2(i,t) -2.113*** -2.559** i(0) agriva(i,t) -2.475*** -2.813*** i(0) ln_agriexp(i,t) -2.204*** -2.652*** i(0) ln_fer(i,t) -1.824 -2.423 -4.693*** -4.830*** i(1) ln_trac(i,t) -2.214*** -2.700*** i(0) ruralpop(i,t) -2.359*** -2.665*** i(0) note: h0: all series are non-stationary. *** and ** imply significant at 1%, and 5% level respectively. source: authors’ calculations using the dataset from faoun, wdi and gfn. result confirms that carbon footprint, gdp per capita, gdp per capita square, agriculture value added, agriculture exports, tractors and rural population indicators are level stationary as p-values of test statistics reject null hypothesis of unit root for all indicators. so, order of integration for all these indicators is zero i.e. i(0). only fertilizers variable is first difference stationary as p-values of statistics accept null hypothesis of unit root at level and reject at first difference. so, order of integration for fertilizers indicator is one i.e. i(1). table 6 displays the results of cips tests at level and at first difference. stated unit root statistics are for logged values except agriculture value added and rural population. result confirms that carbon footprint, agriculture value added, agriculture exports, fertilizers, tractors, and rural population indicators are level stationary as p-values of test statistics reject null hypothesis of unit root for all indicators. so, order of integration for all these indicators is zero i.e. i(0). indicators gdp per capita and gdp per capita square are first difference stationary as p-values of test statistics accept null hypothesis of unit root at level and reject at first difference. so, order of integration for gdp per capita and gdp per capita square is one i.e. i(1). table 6: cips panel unit root tests variables level first difference order intercept intercept and trend intercept intercept and trend ln_carbonfp(i,t) -2.337*** -2.363 i(0) ln_gdp(i,t) -1.638 -2.287 -4.264*** -4.612*** i(1) ln_gdp2(i,t) -1.634 -2.260 -4.161*** -4.547*** i(1) agriva(i,t) -2.545*** -2.769*** i(0) ln_agriexp(i,t) -2.570*** -2.991*** i(0) ln_fer(i,t) -2.239*** -2.871*** i(0) ln_trac(i,t) -2.456*** -2.789*** i(0) ruralpop(i,t) -1.561*** -2.354 i(0) note: h0: homogeneous non-stationary. *** indicates significant at 1% level. source: authors’ calculations using the dataset from faoun, wdi and gfn. adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 128 here, results of cips are important as compared to the estimates of cadf, because it provides more consistent results in case of cd and panel data slope heterogeneity. results of cips tests provides mixed order for integration for indicators and requires application of westerlund’s (2007) panel cointegration tests to ensure long run association between indicators. statistics of cointegration tests are shown in table 7 for carbon footprint models. statistics gt and pt provide evidence of cointegration between carbon footprint and its determinants for all models of westerlund i.e., without intercept, with intercept, and with intercept and trend. so, confirmation of cointegration suggests that there exists long run association between carbon footprint, gdp per capita, gdp per capita square, agriculture value added, agriculture exports and fertilizers. table 7: westerlund’s panel cointegration tests specification statistic statistics z-value p-value without intercept gt -2.623 -3.123 0.001*** ga -8.021 3.609 1.000 pt -18.216 -3.385 0.000*** pa -7.106 0.871 0.808 with intercept gt -2.872 -1.776 0.038** ga -8.576 5.809 1.000 pt -19.552 -1.660 0.049** pa -7.756 3.196 0.999 with intercept and trend gt -3.343 -2.666 0.004*** ga -9.474 7.947 1.000 pt -22.133 -1.291 0.098* pa -9.613 5.033 1.000 note: h0: no integration. ***, ** and * mean significant at 1%, 5% and 10% level. source: author’s calculations using the dataset from faoun, wdi and gfn. table 8 shows the correlation between carbon footprint indicator and other variables. correlation estimates confirm that carbon footprint has positive association with indicators of gdp per capita, agriculture exports, tractors, and fertilizers; and negative association with indicators of gdp per capita square, agriculture value added and rural population. table 8: correlation between ln_carbonfp(i,t) variable and other indicators variable measure ln_gdp(i,t) 0.5620 ln_gdp2(i,t) -0.5461 agriva(i,t) -0.5377 ln_agriexp(i,t) 0.7720 ln_fer(i,t) 0.4468 ln_trac(i,t) 0.5672 ruralpop(i,t) -0.5090 source: authors’ calculations using the dataset from faoun, wdi and gfn. journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 129 ratification of cointegration between variables allow estimation of long run relationship with the help of dcce mg. table 9 reports heterogeneous parameter estimates by using dcce mg approach. results of estimations include estimated coefficients, their p-values and post estimation tests of dcce mg models (model i to v). estimation of different models with dcce mg technique confirms long run relationship between variables. coefficient of gdp per capita is statistically significant and its positive sign in estimated models shows that carbon footprint increases with increase in gdp per capita. coefficient of gdp per capita square is also statistically significant, but its sign is negative in all models. this implies that carbon footprint increases with increase in countries’ income in the developing phase of economic growth (gdp per capita), but later it begins to decrease in the developed period, confirming the ekc hypothesis (grossman and krueger, 1995). coefficient of agriculture value added is also significant and negative, which suggests inverse relationship between carbon footprint and agriculture value added. this inverse relationship indicates agricultural development decreases carbon footprint accordance with balogh and jambor (2017) and balogh (2019). improved technologies due to agriculture growth provide grounds for negative relationship between carbon footprint and agriculture development. therefore, it helps to lower resource usage and environmental pollution through environment-friendly technologies and methods (munasinghe, 1999). agricultural exports coefficient is also significant, but it is positive, which means agricultural exports brings increase in carbon footprint. therefore, results are in line with ang (2009), chebbi et al. (2011), balogh and jambor (2017) and balogh (2019) declare that agricultural export quantity burdens environment by increasing carbon footprint in exporting country. agriculture sector production variable tractor does not affect carbon footprint, but coefficient of fertilizers is statistically significant and positive. so, results of models only confirm the fertilizers as significant determinant of carbon footprint from the selected agricultural sector inputs. therefore, agricultural production heavily based on fertilizer increases carbon footprint, which certifies production-based emission approach (foley et al., 2011; grace et al., 2014; henders et al., 2015; balogh, 2019). moreover, the carbon footprint does not depend on the share of rural population or urban population in the total population of country, which is contrary to the results of sethi (2017) and balogh (2019). adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 130 table 9 also reports results of f tests and cd tests by pesaran (2015) for all estimated models. f test estimates the overall significance of econometric model. reported f-statistics and their respective p-values for all estimated models reject the null hypothesis ‘estimated model is statistically insignificant’ and accept the alternative hypothesis ‘estimated model is statistically significant’. cd-statistics of all estimated models and their respective p-values do not reject the null hypothesis of weak cd against the strong cross-sectional dependency in the variables. cd test results guarantee that there is a presence of weak cross-sectional dependency in the data, which will not persist in the long run. additionally, table 9 also provides the estimate of turning point of gdp per capita of sample countries, where carbon footprint is at a maximum level. turning point estimate of gdp per capita is approximately us$ 39940 for selected countries. carbon footprint declines when the value of gdp per capita becomes us$ 39940 or greater. therefore, the relationship between carbon footprint and gdp per capita follows the pattern of inverted u-shaped curve. table 9: heterogeneous parameter estimates by using dcce mg variables functional forms (dependent variable: ln_carbonfp(i,t)) i ii iii iv v ln_gdp(i,t) 17.50 (0.000)*** 13.45 (0.002)*** 15.29 (0.013)** 8.831 (0.028)** 17.82 (0.029)** ln_gdp2(i,t) -0.774 (0.001)*** -0.614 (0.008)*** -0.777 (0.025)** -0.428 (0.072)* -0.959 (0.041)** agriva(i,t) -0.027 (0.023)** -0.035 (0.001)*** -0.346 (0.006)*** -0.020 (0.025)** -0.038 (0.001)*** ln_agriexp(i,t) 0.053 (0.012)** 0.052 (0.050)** 0.055 (0.036)** ln_fer(i,t) 0.046 (0.024)*** 0.041 (0.028)** ln_trac(i,t) 2.265 (0.563) 9.848 (0.363) 5.512 (0.454) ruralpop(i,t) -0.006 (0.766) 0.004 (0.882) constant -0.093 (0.022)** -0.097 (0.019)** -0.141 (0.058)* -0.052 (0.516) -0.088 (0.370) l.ln_carbonfp(i,t) 0.023 (0.008)*** 0.025 (0.008)*** 0.079 (0.038)** 0.224 (0.000)*** 0.066 (0.079)* post estimation tests fstatistics 7.500 (0.000)*** 3.220 (0.000)*** 1.330 (0.000)*** 5.120 (0.000)*** 1.990 (0.000)*** cdstatistics (pesaran, 2015) 1.170 (0.240) -0.250 (0.804) -0.570 (0.572) -1.560 (0.118) -1.310 (0.121) inflection point on ekc journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 131 turning point2 39940$ 𝜏3 = exp ( −ω̂1 2ω̂2 ) ω̂1= 8.90 ω̂2= -0.42 note: p-values of test statistics of dcce mg estimators are in the parenthesis. ***, ** and * suggest significant at 1%, 5% and 10% level. source: authors’ calculations using the dataset from faoun, wdi and gfn. lastly, the findings of dumitrescu-hurlin panel causality tests given in the table 10 confirms that gdp per capita, gdp per capita square, agriculture value added, agricultural exports and fertilizers have acausal relationship with the carbon footprint in selected economies. table contains statistics and p-values of dumitrescu-hurlin causality tests. table 10: dumitrescu-hurlin panel causality tests null hypothesis w-bar z-bar p-value remarks ln_gdp(i,t) does not cause ln_carbonfp(i,t) 2.88 9.98 0.00*** it does grangercause ln_gdp2(i,t) does not cause ln_carbonfp(i,t) 2.80 9.51 0.00*** it does grangercause agriva(i,t) does not cause ln_carbonfp(i,t) 2.84 9.76 0.00*** it does grangercause ln_agriexp(i,t) does not cause ln_carbonfp(i,t) 2.22 6.45 0.00*** it does grangercause ln_fer(i,t) does not cause ln_carbonfp(i,t) 3.47 13.09 0.00*** it does grangercause note: statistics and p-values of dumitrescu-hurlin causality test results are reported. *** suggests significant at 1% level. source: authors’ calculations using the dataset from faoun, wdi and gfn. 6. conclusion macroeconomic analysis of agriculture sector inputs and outputs in the study contributes to the existing literature by evaluating agriculture-specific determinants of carbon footprint. the study investigates the relationship between agriculture sector indicators and carbon footprint by using panel econometrics. this study finds that carbon footprint depends on economic development of countries and confirm ekc hypothesis. therefore, early stage of economic growth and development upsurge carbon footprint and damages environment, then later, after a turning point, carbon footprint decrease and environmental quality increases. 2 turning point is based on the dcce mg specification: ln_carbonfp(i,t) = ω0(i) + ω1(i)ln_carbonfp(i,t−1) + ω2(i)ln_gdppc(i,t) + ω3(i) ln_gdppc(i,t) 2 + ω4(i)agriva(i,t) + ω5(i)ln_agriexp(i,t) + ω6(i)ln_fer(i,t) + ∑(d(i)z(i,s)) + 𝑒(i,t) 3 estimation formula by stern (2004). adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 132 the study finds that there is a long run relationship between agriculture value added and carbon footprint. rise in value addition of agricultural sector in country reduces carbon footprint. therefore, agriculture sector expansion by employing environment friendly methods and technologies decreases carbon footprint in countries. furthermore, the positive relationship between carbon footprint and agricultural exports proves that there exists a long run relationship between agriculture sector exports and carbon footprint. it implies that agricultural exports encourage the carbon footprint growth by stimulating the production and transport of agricultural commodities. finally, the current study also assures that there exists a long run relationship between high scale of agricultural production and carbon footprint. carbon footprint has positive relationship with high scale of agricultural production, which supports the concept of productionbased emission. this relationship underlines that too much use of fertilizers in agriculture sector fosters the carbon footprint growth and damages natural environment of countries. the conclusion of the current study has some strong implications for policy formulation by suggesting that in agricultural production processes, innovative techniques should be introduced that could ensure less environmental damage and bring considerable decrease in carbon footprint. departments related to agriculture sector must promote research and development culture for innovative solutions to the current issues in production practices. furthermore, trade sector must devise policies to incentivize green agricultural exports. finally, such production techniques must be adopted which are environment friendly and may contribute to reducing carbon footprint i.e. the use of organic fertilizers may provide eco-friendly agriculture. references agovino, m., casaccia, m., ciommi, m., ferrara, m., & marchesano, k. (2019). agriculture, climate change and sustainability: the case of eu-28. ecological indicators, 105, 525543. al-mulali, u., saboori, b., & ozturk, i. (2015). investigating the environmental kuznets curve hypothesis in vietnam. energy policy, 76, 123-131. ang, j. b. (2007). co2 emissions, energy consumption, and output in france. energy policy, 35(10), 4772-4778. audet, j., wallin, m. b., kyllmar, k., andersson, s., & bishop, k. (2017). nitrous oxide emissions from streams in a swedish agricultural catchment. agriculture, ecosystems & environment, 236, 295-303. baccini, a., goetz, s.j., walker, w.s., laporte, n.t., sun, m., sulla-menashe, d., hackler, j., beck, p.s.a., dubayah, r., friedl, m.a., samanta, s. and houghton, r.a. (2012). estimated carbon dioxide emissions from tropical deforestation improved by carbondensity maps. nature climate change, 2, 182–185. journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 133 bai, j., & ng, s. (2004). a panic attack on unit roots and cointegration. econometrica, 72(4), 1127-1177. balogh, j. m. (2019). agriculture-specific determinants of carbon footprint. studies in agricultural economics, 121(3), 166-170. balogh, j.m. and jámbor, a. (2017). determinants of co2 emission: a global evidence. international journal of energy economics and policy, 7 (5), 217–226. baltagi, b. h. (2008). econometric analysis of panel data (vol. 4). chichester: john wiley & sons. ben jebli, m., & ben youssef, s. (2017). renewable energy consumption and agriculture: evidence for cointegration and granger causality for tunisian economy. international journal of sustainable development & world ecology, 24(2), 149-158. ben jebli, m., & ben youssef, s. (2019). combustible renewables and waste consumption, agriculture, co2 emissions and economic growth in brazil. carbon management, 10(3), 309-321. bennetzen, e. h., smith, p., & porter, j. r. (2016). agricultural production and greenhouse gas emissions from world regions—the major trends over 40 years. global environmental change, 37, 43-55. bennetzen, e. h., smith, p., & porter, j. r. (2016). decoupling of greenhouse gas emissions from global agricultural production: 1970–2050. global change biology, 22(2), 763781. bows a. and barrett, j. (2010). cumulative emission scenarios using a consumption-based approach: a glimmer of hope? carbon management, 1, 161–175. breusch, t. s., & pagan, a. r. (1980). the lagrange multiplier test and its applications to model specification in econometrics. the review of economic studies, 47(1), 239-253. calvin, k. v., beach, r., gurgel, a., labriet, m., & rodriguez, a. m. l. (2016). agriculture, forestry, and other land-use emissions in latin america. energy economics, 56, 615624. castesana, p. s., dawidowski, l. e., finster, l., gómez, d. r., & taboada, m. a. (2018). ammonia emissions from the agriculture sector in argentina; 2000–2012. atmospheric environment, 178, 293-304. cerrato, m., & sarantis, n. (2002). the cross-sectional dependence puzzle. london guildhall university. chang, h. j. (2002). kicking away the ladder: development strategy in historical perspective. anthem press. change, i. p. o. c. (2014). climate change: impact, adaptation, vulnerability. chebbi, h., olarreaga, m. and zitouna, h. (2011). trade openness and co2 emissions in tunisia, middle east development journal, 3 (1), 29–53. chen, s., & liu, c. c. (2017). from demand response to transactive energy: state of the art. journal of modern power systems and clean energy, 5(1), 10-19. adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 134 chudik, a., & pesaran, m. h. (2015). common correlated effects estimation of heterogeneous dynamic panel data models with weakly exogenous regressors. journal of econometrics, 188(2), 393-420. cole, m. a., rayner, a. j., & bates, j. m. (1997). the environmental kuznets curve: an empirical analysis. environment and development economics, 2(4), 401-416. corsten m., worrell, e., rouw, m. and van duin, a. (2013). the potential contribution of sustainable waste management to energy use and greenhouse gas emission reduction in the netherlands. resources, conservation and recycling, 77, 13–21. dace, e., & blumberga, d. (2016). how do 28 european union member states perform in agricultural greenhouse gas emissions? it depends on what we look at: application of the multi-criteria analysis. ecological indicators, 71, 352-358. de pinto, v., reina, s., gupta, a., messina, a., & mahalakshmi, r. (2016). role of cysteines in mammalian vdac isoforms' function. biochimica et biophysica acta (bba)bioenergetics, 1857(8), 1219-1227. dernbach, j. c., & mintz, j. a. (2011). environmental laws and sustainability: an introduction. sustainability, 3(3), 531-540. desjardins, r. l., worth, d. e., vergé, x. p. c., vanderzaag, a., janzen, h., kroebel, r., & dyer, j. a. (2014). carbon footprint of agricultural products-a measure of the impact of agricultural production on climate change. in agriculture and agro-food canada [online]. dogan, e., & seker, f. (2016). the influence of real output, renewable and non-renewable energy, trade and financial development on carbon emissions in the top renewable energy countries. renewable and sustainable energy reviews, 60, 1074-1085. dumitrescu, e. i., & hurlin, c. (2012). testing for granger non-causality in heterogeneous panels. economic modelling, 29(4), 1450-1460. fao (2003). world agriculture: towards 2015/2030. an fao perspective. earthscan publications ltd., london, uk. fao (2011). the state of the world’s land and water resources for food and agriculture (solaw) – managing systems at risk. food and agriculture organization of the united nations, rome and earthscan, london, uk. fao (2017). carbon footprint on banana food supply chain. world banana forum good practices collection. foley, j.a., ramankutty, n. and brauman, k.a. (2011). solutions for a cultivated planet. nature, 478, 337–342. franks, j. r., & hadingham, b. (2012). reducing greenhouse gas emissions from agriculture: avoiding trivial solutions to a global problem. land use policy, 29(4), 727-736. gaia (2012). on the road to zero waste: successes and lessons from around the world. gaia — global alliance for incinerator alternatives, quezon city, philippines. giannadaki, d., giannakis, e., pozzer, a., & lelieveld, j. (2018). estimating health and economic benefits of reductions in air pollution from agriculture. science of the total environment, 622, 1304-1316. journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 135 gilbert, n. (2012). one-third of our greenhouse gas emissions come from agriculture. gilbert, o. (2012). the ecology of urban habitats. springer science & business media. global footprint network (2018). climate change. goyette-desjardins, g., auger, j. p., xu, j., segura, m., & gottschalk, m. (2014). streptococcus suis, an important pig pathogen and emerging zoonotic agent—an update on the worldwide distribution based on serotyping and sequence typing. emerging microbes & infections, 3(1), 1-20. grace, j., mitchard, e. and gloor, e. (2014). perturbations in the carbon budget of the tropics. global change biology, 20 (10), 3238–3255. grossman, g. m., & krueger, a. b. (1995). economic growth and the environment. the quarterly journal of economics, 110(2), 353-377. gujarati, d. n., bernier, b., & bernier, b. (2004). econométrie (pp. 17-5). brussels: de boeck. henders, s., persson, u.m. and kastner, t. (2015). trading forests: land-use change and carbon emissions embodied in production and exports of forest-risk commodities. environmental research letters, 10 (12), 1748–9326. im, k. s., pesaran, m. h., & shin, y. (2003). testing for unit roots in heterogeneous panels. journal of econometrics, 115(1), 53-74. infante-amate, j., & de molina, m. g. (2013). the socio-ecological transition on a crop scale: the case of olive orchards in southern spain (1750–2000). human ecology, 41(6), 961969. jacobsen, r., vandermeulen, v., van huylenbroeck, g., & gellynck, x. (2014). the carbon footprint of pigmeat in flanders. in assessment of carbon footprint in different industrial sectors, volume 1 (pp. 167-189). springer, singapore. jorgenson, a. k., clark, b., & kentor, j. (2010). militarization and the environment: a panel study of carbon dioxide emissions and the ecological footprints of nations, 1970– 2000. global environmental politics, 10(1), 7-29. kapetanios, g., pesaran, m. h., & yamagata, t. (2011). panels with non-stationary multifactor error structures. journal of econometrics, 160(2), 326-348. kastratović, r. (2019). impact of foreign direct investment on greenhouse gas emissions in agriculture of developing countries. australian journal of agricultural and resource economics, 63(3), 620-642. kearney, j. (2010). food consumption trends and drivers. philosophical transactions of the royal society b: biological sciences, 365(1554), 2793-2807. khan, m. a. (1995). sustainable development: the key concepts, issues and implications. keynote paper given at the international sustainable development research conference, 27–29 march 1995, manchester, uk. sustainable development, 3(2), 63-69. khan, s. a. r., yu, z., belhadi, a., & mardani, a. (2020). investigating the effects of renewable energy on international trade and environmental quality. journal of environmental management, 272, 111089. adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 136 krey, v., o'neill, b. c., van ruijven, b., chaturvedi, v., daioglou, v., eom, j., ... & ren, x. (2012). urban and rural energy use and carbon dioxide emissions in asia. energy economics, 34, s272-s283. kuznets, s. (1955). economic growth and income inequality. the american economic review, 45(1), 1-28. le quéré, c., andrew, r. m., canadell, j. g., sitch, s., korsbakken, j. i., peters, g. p., ... & zaehle, s. (2016). global carbon budget 2016. earth system science data (online), 8(2). le quéré, c., andrew, r. m., friedlingstein, p., sitch, s., hauck, j., pongratz, j., ... & zheng, b. (2018). global carbon budget 2018. earth system science data, 10(4), 2141-2194. levin, a., lin, c. f., & chu, c. s. j. (2002). unit root tests in panel data: asymptotic and finitesample properties. journal of econometrics, 108(1), 1-24. li, t., baležentis, t., makutėnienė, d., streimikiene, d., & kriščiukaitienė, i. (2016). energyrelated co2 emission in european union agriculture: driving forces and possibilities for reduction. applied energy, 180, 682-694. luo, j. h., wu, j., & lin, w. (2017). thinet: a filter level pruning method for deep neural network compression. in proceedings of the ieee international conference on computer vision (pp. 5058-5066). mbow, c., halle, m., el fadel, r., & thiaw, i. (2021). land resources opportunities for a growing prosperity in the sahel. current opinion in environmental sustainability, 48, 85-92. moon, h. r., & perron, b. (2004). testing for a unit root in panels with dynamic factors. journal of econometrics, 122(1), 81-126. moutinho, v., robaina-alves, m., & mota, j. (2014). carbon dioxide emissions intensity of portuguese industry and energy sectors: a convergence analysis and econometric approach. renewable and sustainable energy reviews, 40, 438-449. mózner, z. v. (2013). a consumption-based approach to carbon emission accounting–sectoral differences and environmental benefits. journal of cleaner production, 42, 83-95. munasinghe, m. (1999). is environmental degradation an inevitable consequence of economic growth: tunneling through the environmental kuznets curve. ecological economics, 29(1), 89-109. muthu, s. s. (ed.). (2014). assessment of carbon footprint in different industrial sectors, volume 2. springer science & business. naranpanawa, a. (2011). does trade openness promote carbon emissions? empirical evidence from sri lanka. the empirical economics letters, 10(10), 973-986. network, g. f. (2018). ecological footprint. newey, w. k., & west, k. d. (1994). automatic lag selection in covariance matrix estimation. the review of economic studies, 61(4), 631-653. oenema, a., brug, j., & lechner, l. (2001). web-based tailored nutrition education: results of a randomized controlled trial. health education research, 16(6), 647-660. journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 137 ozturk, i., & acaravci, a. (2010). co2 emissions, energy consumption and economic growth in turkey. renewable and sustainable energy reviews, 14(9), 3220-3225. pao, h. t., & tsai, c. m. (2011). modeling and forecasting the co2 emissions, energy consumption, and economic growth in brazil. energy, 36(5), 2450-2458. paul, b., khan, m. a., paul, s., shankarganesh, k., & chakravorty, s. (2018). termites and indian agriculture. in termites and sustainable management (pp. 51-96). springer, cham. pedroni, p. (2004). panel cointegration: asymptotic and finite sample properties of pooled time series tests with an application to the ppp hypothesis. econometric theory, 20(3), 597625. pesaran, m. h. (2006). estimation and inference in large heterogeneous panels with a multifactor error structure. econometrica, 74(4), 967-1012. pesaran, m. h. (2007). a simple panel unit root test in the presence of cross‐section dependence. journal of applied econometrics, 22(2), 265-312. pesaran, m. h. (2015). testing weak cross-sectional dependence in large panels. econometric reviews, 34(6-10), 1089-1117. pesaran, m. h., & yamagata, t. (2008). testing slope homogeneity in large panels. journal of econometrics, 142(1), 50-93. pesaran, m. h., schuermann, t., & weiner, s. m. (2004). modeling regional interdependencies using a global error-correcting macroeconometric model. journal of business & economic statistics, 22(2), 129-162. peters, g. p., & hertwich, e. g. (2008). co2 embodied in international trade with implications for global climate policy. phillips, p. c., & sul, d. (2003). dynamic panel estimation and homogeneity testing under cross section dependence. the econometrics journal, 6(1), 217-259. poore, j., & nemecek, t. (2018). reducing food’s environmental impacts through producers and consumers. science, 360(6392), 987-992. poumanyvong, p., & kaneko, s. (2010). does urbanization lead to less energy use and lower co2 emissions? a cross-country analysis. ecological economics, 70(2), 434-444. rehman, a., rauf, a., ahmad, m., chandio, a. a., & deyuan, z. (2019). the effect of carbon dioxide emission and the consumption of electrical energy, fossil fuel energy, and renewable energy, on economic performance: evidence from pakistan. environmental science and pollution research, 26(21), 21760-21773. ren, s., yuan, b., ma, x., & chen, x. (2014). international trade, fdi (foreign direct investment) and embodied co2 emissions: a case study of chinas industrial sectors. china economic review, 28, 123-134. robaina-alves, m., & moutinho, v. (2014). decomposition of energy-related ghg emissions in agriculture over 1995–2008 for european countries. applied energy, 114, 949-957. sebri, m., & abid, m. (2012). energy use for economic growth: a trivariate analysis from tunisian agriculture sector. energy policy, 48, 711-716. adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 138 smith, m. t., edwards, r. r., robinson, r. c., & dworkin, r. h. (2004). suicidal ideation, plans, and attempts in chronic pain patients: factors associated with increased risk. pain, 111(1-2), 201-208. springmann, m., clark, m., mason-d’croz, d., wiebe, k., bodirsky, b. l., lassaletta, l., ... & willett, w. (2018). options for keeping the food system within environmental limits. nature, 562(7728), 519-525. stern, d. i. (2004). the rise and fall of the environmental kuznets curve. world development, 32(8), 1419-1439. tubiello, f. n., salvatore, m., rossi, s., ferrara, a., fitton, n., & smith, p. (2013). the faostat database of greenhouse gas emissions from agriculture. environmental research letters, 8(1), 015009. united nations framework convention on climate change (2015). indc submission. vergé, x. p. c., dyer, j. a., desjardins, r. l., & worth, d. (2008). greenhouse gas emissions from the canadian beef industry. agricultural systems, 98(2), 126-134. vergé, x. p. c., dyer, j. a., desjardins, r. l., & worth, d. (2008). greenhouse gas emissions from the canadian beef industry. agricultural systems, 98(2), 126-134. vergé, x. p., maxime, d., dyer, j. a., desjardins, r. l., arcand, y., & vanderzaag, a. (2013). carbon footprint of canadian dairy products: calculations and issues. journal of dairy science, 96(9), 6091-6104. vermeulen, s. j., aggarwal, p. k., ainslie, a., angelone, c., campbell, b. m., challinor, a. j., ... & wollenberg, e. (2012). options for support to agriculture and food security under climate change. environmental science & policy, 15(1), 136-144. waheed, r., chang, d., sarwar, s., & chen, w. (2018). forest, agriculture, renewable energy, and co2 emission. journal of cleaner production, 172, 4231-4238. westerlund, j., & edgerton, d. l. (2007). a panel bootstrap cointegration test. economics letters, 97(3), 185-190. westerlund, j., & edgerton, d. l. (2007). new improved tests for cointegration with structural breaks. journal of time series analysis, 28(2), 188-224. wiebe, k. s., & yamano, n. (2016). estimating co2 emissions embodied in final demand and trade using the oecd icio 2015: methodology and results. appendix journal of applied economics and business studies, volume. 5, issue 4 (2021) 111-140 https://doi.org/10.34260/jaebs.546 139 table a: list of sample countries sustainable agriculture index (for high-income countries) 1. austria 79.9 2. denmark 79.6 3. israel 78.3 4. germany 78.0 5. poland 78.0 6. netherlands 77.1 7. belgium 74.6 8. czech republic 74.5 9. australia 73.4 10. japan 73.4 11. south korea 73.4 12. canada 73.0 13. sweden 72.7 14. france 71.0 15. italy 70.2 16. portugal 69.7 17. estonia 69.6 18. united states 68.6 19. argentina 66.9 20. spain 66.6 21. lithuania 66.5 22. croatia 66.3 22. greece 65.1 24. slovenia 63.0 25. united kingdom 61.5 26. united arab emirates 56.9 27. slovakia 54.6 28. latvia 53.7 29. luxembourg 53.6 30. saudi arabia 52.4 sustainable agriculture index (for middle-income countries) 31. colombia 76.5 32. cote d'ivoire 73.9 33. zambia 72.7 34. cameroon 72.2 35. tunisia 70.1 34. mexico 61.3 37. turkey 68.3 38. romania 68.0 39. kenya 66.6 40. nigeria 66.6 31. india 65.5 42. lebanon 65.1 43. brazil 64.2 44. jordan 64.0 45. indonesia 61.1 46. china 60.7 47. ghana 57.4 48. russian federation 53.9 49. south africa 52.4 sustainable agriculture index (for low-income countries) 50. rwanda 71.0 51. tanzania 70.5 52. zimbabwe 70.5 53. uganda 68.9 54. mozambique 68.4 55. burkina faso 67.5 56. ethiopia 66.6 note: index values are ranked: zero to hundred, while hundred is maximum sustainability value and best progress to meet societal, economic, and environmental main indicators of performance. source: economist intelligence unit, food sustainability index 2018 table b: indicators selected for the estimation adnan habib, saima sarwar, uzair ahson and alvina sabah idrees 140 sr. no. indicator (unit) abbreviation and description 1 carbon footprint (global hectors) carbonfp: it evaluates co2 emissions related with use of fossil fuel. in the ecological footprint4 estimates, emission quantities are estimated into biologically productive areas required for absorbing this co2 emission. the carbon footprint is considered in the ecological footprint because it is a rival user of bio productive space. increasing co2 concentrations in the atmosphere represent an increase of ecological debt. it is generally quantified in global hectares. 2 gdp per capita (constant 2010 us$) gdppc: gdp (gross domestic product) per capita is the gdp divided by midyear value of population. it is total of values added by all producers in the country and taxes on product, excluding subsidies not included in the value of products. 3 agriculture value added (percentage of gdp) ln_agriva: agriculture value added is measured as percentage of gdp of country. value added of agriculture is the net output value of sector after totaling all final outputs and deducting intermediate inputs. 4 agricultural exports (1000 us$) ln_agriexp: agricultural exports indicator measures value of agriculture sector exports in thousand us dollars. 5 tractors (per 100 sq. km of arable land) trac: tractors are the indicator of machinery employed in agricultural sector. this indicator measures tractors per 100 square km of arable land. 6 fertilizers (kg per hectare of arable land) fer: fertilizer indicator measures utilization of fertilizers in kilograms per hectare of arable land of agriculture sector. 7 rural population (percentage of total population) ruralpop: rural population is percentage of total population who lives in rural parts of country. note: current study uses natural log of indicators of carbon footprints, gdp per capita, agricultural exports, tractors, and fertilizers in the estimations. source: faoun, wdi and gfn. 4 it is an estimate of area consists of water and biologically productive land, which a population, individual, or action requires to create the natural resources it utilizes and to take in the waste it creates, employing resource management practices and exi sting technology. it is estimated in global hectares. nation's footprint include sea and land from all other countries. it normally mentions ecological footprint of consumption. journal of applied economics and business studies, volume. 6, issue 3 (2022) 79-100 https://doi.org/10.34260/jaebs.635 79 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x analyzing the role of workplace vigilantism on job outcomes: a study of healthcare professionals of peshawar muhammad naveed iqbal1*1, anjum ihsan2 & shahid jan kakakhel3 1 phd scholar, islamia college university, peshawar 2 assistant professor, islamia college university, peshawar 3 associate professor, islamia college university, peshawar abstract the overall objective of the study was to develop and test a model to find out the relationship between workplace vigilantism, turnover intention, work mindfulness, and change oriented organizational citizenship behavior (ocb). in this connection, the research aim was to focus on health sector nurses of three hospitals of district peshawar, khyber pakhtunkhwa, the population which is equal to 1800. by using the simplified formula for proportions, taro yamane and the technique of uma sekaran (2012) were utilized to calculate the sample size as well as which is 317. however, 350 questionnaires were distributed due to a possible non-response rate. to avoid any biases, the probability-stratified sampling technique was used. each hospital was considered as a separate stratum due to the inter-group heterogeneity and from each stratum, respondents were selected subject to their availability. below is the brief information representing the name of hospitals and the number of nurses’ staff located in the district of peshawar. the result of all three hypotheses indicated a statistically significant and positive association between workplace vigilantism, turnover intention, work mindfulness, and change oriented organizational citizenship behavior (ocb). the study was limited to health sector nurses with only three hospitals in peshawar. in the future, research can have a moderating effect on perceived organizational support to strengthen the vigilante to work for the ultimate benefit of the organization. keywords workplace vigilantism, perceived organizational support, work mindfulness, turnover intention, change oriented ocb jel classification c91, d23 1 naveed.hr@hotmail.com muhammad naveed iqbal, anjum ihsan & shahid jan kakakhel 80 1. introduction the existence of vigilantes has always been in limelight in the workplace in different settings. contemporary literature in the information sciences brings on record the spread of what has been referred to as “vigilantism,” “online vigilantism,” or “diligilism,” in the online global world where social media is used as a tool by certain individuals or groups to look and admonish individuals, organizations for alleged violation of rules and lawlessness (graso. m et.al, 2020). according to a recent survey by decelles and aquino (2017), a population of 2258 u.s employees inquired about workplace vigilantes’, among them, 18% reported currently working with workplace vigilantes, and 42 reported having worked with vigilantes. they also asserted that the vigilant could summon the vision of citizens who are frustrated by the lawlessness happening around them and take it upon themselves to dispense justice in the untamed land where there is a weak government and corrupt authorities. moreover, due to the rise of the modern state as well as its highly developed form of a system of criminal justice, there can be thinking that vigilantes are now obsolete but there are many instances when the vigilantes being private citizens not only uncover but also fight the crimes without having any authority to do so. besides, vigilantes are different from whistleblowers and peer reporters who might also intend to change the misbehavior of their colleagues but unlike the vigilantes, they depend on the organizational reporting systems or its authorities to help them in achieving this aim, and vigilantes themselves act, take a self-appointed role within a community as jury, judge and provider of justice. furthermore, while the fields like criminology, sociology, and political science have since long recognized that vigilantes play a role in society, however, little research has been conducted in the field of management that can explain why vigilantes emerge in organizations. also, scholars in criminology and sociology have theorized social conditions which produce vigilantes and the outcomes of their activities on individuals and communities, yet the suggested antecedents of the vigilante emergence prevail at the societal level – historical, political or economic conditions or ineffectiveness of the judiciary and law enforcement. these antecedents are too far to explain adequately what leads the employees to become vigilantes in organizations. such antecedents are also more likely to remain constant across different organizations which mean that organizations, as well as the employees, are exposed to the same external factors existing at the country, state, or local community level. therefore, research is warranted to develop theories in the management field to explain the antecedents of vigilante behavior and its consequences or outcomes and different factors influencing workplace vigilantism. kamran (2021) highlighted that vigilante justice can be seen in the past in many countries under unsettled conditions in situations when informally organized groups have been found journal of applied economics and business studies, volume. 6, issue 3 (2022) 79-100 https://doi.org/10.34260/jaebs.635 81 in an attempt to supplement or replace the legal procedure or fill the void in the instances where there has been corruption in the institutional justice. when it happens then the victims very often become the oppressors and after this, no one can restrain them. as regards, pakistani society, apparently frustrations and anger just need a wily demagogue to channel them. it happens to be the anger against the system which in the past has motivated people to be involved in violence. he also viewed that the vigilante is mainly concerned with the common welfare and uses force only if all the other means fail. this implies that in pakistan the lack of normative order in terms of the unobservance of legal procedures sparks the emergence of vigilantes in society in general and in different sectors or organizations in particular. this calls for the role of academia to investigate the antecedents and consequences of vigilantism and the different factors influencing this phenomenon. there is negligible research on vigilante behavior in the pakistani context. the health services sector in pakistan is already fragile and criticized for not delivering services as per international standards (kurji et al., 2016). owing to covid-19, the flaws in the health system are further highlighted. the system due to poor control mechanisms and ineffective management failed to deliver. in this context, the role of the workplace vigilante seems to be of great relevance as a vacuum is created providing ample space for the vigilantes to report and punish individuals/employees for violation of organizational rules and policies. in this view, the study aims to examine such violations in the context of the health sector of khyber pakhtunkhwa, pakistan, to strengthen the healthcare system of pakistan. the overall objective of the study is to develop and test an integrated model to analyze the role of workplace vigilantism on job outcomes such as turnover intention, work mindfulness, and change-oriented organizational citizenship behavior (ocb). the significance of the study is a broader statement that emphasizes the shreds of evidence that clarify how and why there was a need to perform the current research study. it is a justification of the importance of the study and how the results and outcomes of the study will contribute to the literature, knowledge, theory, society, new scholars, researchers, and policymakers to take benefit from the study. to provide benefits to society, the current quantitative study was employed. 2. literature review a vigilante is described as "a member of a volunteered committee organized to repress and punish crime rapidly (as when judicial processes are judged inadequate); a self-appointed doer of justice" (bjorgo& mares, 2019). frequently feeling that formal institutions are inadequate, vigilantes take matters of law, justice, ethics, crime, and punishment into their own hands — a decision rather than having the legal obligation to do so. this is because muhammad naveed iqbal, anjum ihsan & shahid jan kakakhel 82 vigilantes choose to take issues into their own hands rather than comply with the law when it is required of them to do so. according to saucier & webster (2013), psychological theory suggests that vigilantism may be a more general phenomenon, or perhaps even a trait-like predisposition that influences people's interactions with their environment, leads them to believe that their views are morally superior and more accurate than those of others, and motivates them to behave in ways that cast them into the role of the self-appointed doer of justice (johnston, 1996). according to bateson (2021) if being a vigilante is a personality feature, then it may be more common than the eccentric and extreme characters seen in fiction, as well as the reallife examples of people proclaiming themselves to be crime fighters without the formal license to do so, would imply. there may be more vigilantes present in society and organizations than is commonly believed, but this conclusion is highly dependent on how one defines a vigilante and what behaviors qualify as vigilante acts (trotter, 2017). organizations are systems in which many people interact and frequently disagree as well as impose ethical standards and conduct, which indicates that there may emerge vigilantes among workers who feel that the company is not adequately addressing employees' alleged misbehavior in the workplace. we explore the prevalence of what we will refer to as the vigilante syndrome in the labor force of the united states as a first step toward establishing irrefutably that workplace vigilantes do exist (bateson, 2021). 2.1 workplace vigilantism and employee turnover intention according to halbesleben et al., (2014). there are benefits associated with individual mindfulness for employees. differences in individuals and their environments might affect how they react to situations involving earning additional resources or avoiding loss. the person may regain energy after the event (halbesleben et al., 2014), which may affect the resource-related effects of awareness. gunasekara and zheng (2019) found that workplace vigilantism may impair the connection between individual mindfulness and job engagement. in addition, zhang et al. (2018) found that assistance from organizations has the potential to positively influence the connection between individual mindfulness and work engagement. there is a potential that the work states employees impact their reactions to workplace stress. collective vigilantism, which is defined as "law enforcement conducted without legal authority by a group of self-appointed persons," is one of the work states having a significant association with individual mindfulness. are distinguished by their attentiveness and focus on present events, as well as their experiential and nonjudgmental processing of team experiences (yu &zellmer-bruhn, 2018). individual mindfulness indicates cognition at the individual journal of applied economics and business studies, volume. 6, issue 3 (2022) 79-100 https://doi.org/10.34260/jaebs.635 83 level, while team mindfulness demonstrates organizationally shared views among team members. team mindfulness is the practice of paying attention as a group to the current events and stimuli occurring inside and outside the team. as a consequence of a decrease in taskrelated distractions in teams with high mindfulness, the association between individual mindfulness and recovery level is strengthened (widiatmika&darma, 2018). as a consequence of this decreased distraction from task activities, team members with a high level of mindfulness can discern more subtlety within the team (morrison et al., 2014). this prevents resource-intensive processes from developing in response to irrelevant inputs, such as the display of conflict (slagter et al., 2011). in addition, attentive teams have a less judgmental response to stress (good et al., 2016), which decreases the chance of unfavorable reactions (glomb et al., 2011). in this kind of team atmosphere, it is unlikely that the dispute would affect the team members (wise et al., 2022). 2.2 workplace vigilantism and employee turnover intention the majority of academics hypothesize that employee burnout is at the heart of their choice to quit their employment (e.g., cordes& dougherty, 1993; cotton & tuttle, 1986; spector, 1986). the turnover intention was connected to burnout and workplace vigilantism from political, economic, and psychological changes (mccabe et al., 2008), organizationspecific factors associated with decreasing work autonomy (spector, 1986), and individual events contributing to psychological stress (cordes& dougherty, 1993) this greater urge to resign is directly tied to burnout associated with emotional tiredness, depersonalization, and personal success (maslach, 1978; maslach& jackson, 1986). depersonalization is the effect of dehumanizing attitudes held by businesses toward their employees (maslach& jackson, 1981). it is connected to treating other people in an impersonal manner, being cruel toward other people, and experiencing emotional exhaustion. every single one of these facets of burnout may be traced back to workplace vigilantism in some way. alternately, personal success, the third component of burnout, has an influence on an individual's mental state as a result of the individual's dissatisfaction with their professional successes and the injustices that exist inside an organization. as a result of the fact that the three components of burnout occur in the context of emotional or psychological exhaustion and have an effect on the perspectives that workers hold regarding their jobs, burnout has the potential to lessen people's desire to continue working in their respective businesses (chen et al., 2022). or you can coerce him into seeking retribution. whenever employees are fully drained, they are more likely to have low levels of solidarity with coworkers, jealousy, and a diminished sense of gratitude. this is because emotionally drained individuals are more likely to have negative thoughts about themselves. as a direct consequence of this, they have a greater propensity to either desert their organizations or turn become vigilantes. muhammad naveed iqbal, anjum ihsan & shahid jan kakakhel 84 2.3 vigilante changed oriented behavior van dyne et al. (1995) define change-oriented behaviors by combining organizational citizenship behavior with proactive activity. this limits change-oriented behaviors theoretically. van dyne et al. (1995) created a conceptual framework to differentiate between affiliative and challenging employee activities. this conceptual framework was implemented in the sphere of organizational citizenship. podsakoff et al., (2009) provided further support for this viewpoint when they reviewed empirical data on citizenship. according to their results, "one important difference between ocb dimensions is whether they are affiliative or demanding in nature" (podsakoff et al., 2009). although participating in change-oriented behaviors such as voicing problematic beliefs may be considered an activity, the degree to which these difficulties are seen as troublesome differs widely from individual to individual (burris, 2012). the term refers to everything else that highlights change over difficulties (choi, 2007, mcallister et al., 2007, van dyne et al., 2008). positive change-oriented constructs include leadership (morrison & phelps, 1999), having a voice (burris, 2012, van dyne &lepine, 1998), and engaging in change-oriented citizenship action (bettencourt, 2004, choi, 2007, 'seppala et al., 2012, van dyne et al., 2008). parker and his colleagues emphasized proactive employee initiatives to achieve change (e.g., bindl and parker, 2010, parker et al., 2010). parker and collins (2010) categorized proactive activity as work, strategic, and personenvironment fit. the study focuses on proactive work behaviors that try to effect internal organization transformation through vigilantism to restrict the study's scope. even though all of these are important, the study chose to concentrate on proactive work behaviors since they enable us to do so. the study observed that some research on proactive conduct alludes to the presence of positive actor intentions, but does not make this argument directly (grant & ashford, 2008). given that both constructive and destructive actions might be deemed proactive, this may provide a difficulty (spitzmuller& van dyne, 2013). for example, behavior that is counterproductive in the workplace, severe supervision, theft, bullying, and a vast array of other detrimental activities are proactive yet have a negative goal. to clarify any ambiguities and advance our conceptualizations, we thus do not use the term proactive. unlike reactions to organizationally-led change programs, which were the focus of previous studies, we concentrate special emphasis on change-oriented activities undertaken by employees (choi, 2011, oreg et al., 2011, rafferty et al., 2013). in light of the previously described conceptual obstacles, we put a premium on proactive, change-oriented staff actions that contribute to the growth of the firm. there is no such linked find in the literature between vigilantism and employee changed oriented behavior. as the vigilante faced many challenges journal of applied economics and business studies, volume. 6, issue 3 (2022) 79-100 https://doi.org/10.34260/jaebs.635 85 during the process of punishing wrongdoers it led to different changed behaviors. this specific study aimed to find out that link. a conceptual framework based on literature figure: 1 a conceptual framework based on literature (based on the literature and conceptual model, three hypotheses have been established) h1: there is a positive association between workplace vigilantism and work mindfulness h2: there is a positive association between workplace vigilantism and change oriented ocb. h3: there is a negative association between workplace vigilantism and turnover intention. 3. research methods the onion model from the study of saunders et al (2007) was used to organize the current research study. this research model provides a detailed, step-by-step procedure for conducting a research study. the philosophical concept of the study is based on a positivist approach. the approach of the study is deductive as three hypotheses have been established from the literature and theory. the methodological choice of the study is to conduct a monomethod quantitative study. accordingly, a survey questionnaire strategy was adopted to conduct primary data. according to the situation and study demand, a cross-sectional time horizon was implemented to conduct primary data in one visit. workplace vigilantism vigilante turnover intention vigilante work mindfulness intention vigilante change oriented ocb intention muhammad naveed iqbal, anjum ihsan & shahid jan kakakhel 86 figure: 2 3.1 population and sample as this research aims to focus on the health sector of khyber pakhtunkhwa, the population for the study was the nursing staff of the public sector hospitals located in district peshawar which is equal to 1800. by using the simplified formula for proportions, taro yamane and the technique of uma sekaran (2012) were utilized to calculate the sample size as well as which is 317. however, 350 questionnaires were distributed due to a possible non-response rate. to avoid any biases, the probability-stratified sampling technique was used. each hospital was considered as a separate stratum due to the inter-group heterogeneity and from each stratum, respondents were selected subject to their availability. below is the brief information representing the name of hospitals and the number of nurses’ staff located in the district of peshawar. table 1: target population sr. # medical training institutes (mti) population of nurses 1 lady reading hospital mti peshawar 800 2 hayatabad medical complex mti peshawar 500 3 khyber teaching hospital mti peshawar 500 1800 journal of applied economics and business studies, volume. 6, issue 3 (2022) 79-100 https://doi.org/10.34260/jaebs.635 87 3.2 data analysis researchers have utilized structural equation modeling's multivariate statistical technique for the current research study to estimate and evaluate causal linkages (sem). using the phrase "path analysis," a geneticist may examine the influence of one or more independent factors on a route map and see how they interact with one another. structural equation modeling (pls-sem) may be useful when trying to predict a large number of dependent variables from a big number of independent variables (abdi, 2007). 3.3 assessment of measurement model in structural equation modeling, two models are used to analyze the data. the measurement model and the structural model. the measurement model is established when all the criterion gets achieved including validity and reliability (hair et al, 2021). the first stage of the measurement model is to establish convergent validity. convergent validity is about the item loadings of each variable, average variance extract (ave), analysis of discriminant validity, cross-loading of items, the test of fornell and larcker, and hetrotraitmonotrait (htmt). internal consistency is the composite reliability of the scales. 3.4 assessment of reflective model a reflective measurement is the first step of the measurement model. all of the constructs of the study are reflective. according to hair et al. (2019), three main assessment criteria are required at the beginning. in this section, three criteria are necessary to establish such as internal consistency, convergent validity (indicator reliability also referred to as outer loading and average variance extracted ave), and finally discriminant validity. muhammad naveed iqbal, anjum ihsan & shahid jan kakakhel 88 figure 3: reflective measurement model 3.5 cronbach’s alpha, ave, and composite reliability in sem-pls, cronbach’s alpha technique is used for finding internal consistency. due to the deficiencies in cronbach alpha, composite reliability is used as a more reliable way to measure internal constancy (hair et al., 2021). 3.6 item loadings the degree of consistency of the items related to each measurable variable (hair et al., 2019). the threshold for item loading set by the experts is the value should be greater than 0.708 while the value of ave should be greater than 0.5 (ramayah et al., 2018). 3.7 indicator reliability / items loading regarding the threshold of outer loading, few studies set the range but the outer loading should be greater than 0.7, and the loading value equal to and greater than 0.708 is more reliable. journal of applied economics and business studies, volume. 6, issue 3 (2022) 79-100 https://doi.org/10.34260/jaebs.635 89 table 2: item loadings of the construct change oriented ocb turnover intention work mindfulness workplace vigilantism coocb1 0.763 coocb2 0.83 coocb3 0.652 coocb4 0.688 ti1 0.726 ti2 0.839 ti3 0.846 ti4 0.838 ti5 0.532 wm3 0.846 wm4 0.859 wm5 0.841 wv1 0.71 wv2 0.849 wv3 0.844 wv4 0.792 some of the items of a variable that did not achieve the criteria were dropped one by one starting with the items having a low score. 3.8 convergent validity the convergent validity which is also known as ave is the outcome after taking the calculation of the outer loadings of each variable (ave) (hair et al., 2019). the rule of thumb for ave should be above 0.05. table 3 represents cronbach’s alpha, composite reliability, and average variance extracted (ave). table 3: measurement of convergent validity cronbach's alpha rho_a composite reliability average variance extracted (ave) change oriented ocb 0.717 0.74 0.825 0.542 turnover intention 0.811 0.851 0.873 0.586 work mindfulness 0.806 0.807 0.886 0.721 workplace vigilantism 0.812 0.82 0.877 0.641 in line with the above, cronbach’s alpha coefficient of all variables was found above the threshold like change oriented org citizenship behavior = 0.717, turnover intention = 0.811, work mindfulness = 0.806, workplace vigilantism = 0.812. moreover, all the values of cr and ave were also established as per the criteria set by experts (hair et al., 2021). muhammad naveed iqbal, anjum ihsan & shahid jan kakakhel 90 3.9 discriminant validity the third stage of the measurement model is the establishment of discriminant validity. discriminant validity is the degree to which items of each variable are strictly dissimilar from the other items of the variable (hair et al., 2019). table 4: cross loadings change oriented ocb turnover intention work mindfulness workplace vigilantism coocb1 0.763 0.45 0.461 0.325 coocb2 0.83 0.436 0.439 0.415 coocb3 0.652 0.402 0.335 0.284 coocb4 0.688 0.379 0.419 0.305 ti1 0.463 0.726 0.736 0.387 ti2 0.436 0.839 0.465 0.322 ti3 0.422 0.846 0.418 0.299 ti4 0.476 0.838 0.401 0.336 ti5 0.347 0.532 0.236 0.141 wm3 0.44 0.449 0.846 0.36 wm4 0.486 0.531 0.859 0.355 wm5 0.507 0.622 0.841 0.34 wv1 0.366 0.224 0.27 0.71 wv2 0.397 0.328 0.315 0.849 wv3 0.348 0.354 0.357 0.844 wv4 0.359 0.389 0.373 0.792 table 4 is representing cross-loadings. in cross-loading, the difference between items loading across its variable is not less than 0.1. it shows that none of the item loads is below 0.1 hence discriminant validity is not an issue for the present research (ramayah et al., 2018). the second criterion for evaluating discriminant validity (dv) is the use of fornell-larcker (1981). below mentioned table 5 represents fornell larcker's criteria. table 5: fornell larcker criteria change oriented ocb turnover intention work mindfulness workplace vigilantism change oriented ocb 0.736 turnover intention 0.564 0.766 work mindfulness 0.562 0.627 0.849 workplace 0.458 0.41 0.414 0.801 journal of applied economics and business studies, volume. 6, issue 3 (2022) 79-100 https://doi.org/10.34260/jaebs.635 91 vigilantism the fornell-larcker criterion is the square root of the items for each variable, and these items' values should be higher than all the other bivariate correlations (ramayah et al., 2018). due to certain issues in cross-loadings and fornell larcker criteria, another concept of detecting discriminant validity introduced by henseler (2015) is htmt which is the ratio of within-trait and between-trait correlations. according to the criteria of henseler et al., (2015), the higher value of htmt which is 0.90 leads to an absence of discriminant validity. table 6: htmt change oriented ocb turnover intention work mindfulnes s workplace vigilantism change oriented ocb turnover intention 0.735 work mindfulness 0.741 0.729 workplace vigilantism 0.595 0.47 0.508 before moving towards the structural model, make sure that all the criteria of the measurement model have been established and none of them are left. in the second phase, the researcher is more interested to establish all the criteria of the structural model (hair et al., 2019).in the first step of a structural model, the examination of the issues related to collinearity is measured. in the second step, the examination of the coefficient of determination (r2) is established, while in the third step, the examination of the effect size of the model constructs (f2) is established, and finally to fourth step to forecast the predictive relevance (q2) of the structural model is established. muhammad naveed iqbal, anjum ihsan & shahid jan kakakhel 92 figure 4: evaluation of structural model path coefficients and direct relationship evaluation of the path coefficient of the structural model is the second most important criterion which represents the flow of path between the variables and their relationships (ramayah et al., 2018). table 7 is the showing path coefficient of the constructs. table 7: path coefficient original sample (o) standard deviation (stdev) t statistics (|o/stdev|) p values workplace vigilantism -> change-oriented ocb 0.458 0.043 10.566 0.00 workplace vigilantism -> turnover intention 0.41 0.046 8.909 0.00 workplace vigilantism -> work mindfulness 0.414 0.047 8.895 0.00 the result indicates that all of the relationships between variables are statistically significant. the result shows that result of direct hypotheses is positive and significant for the current journal of applied economics and business studies, volume. 6, issue 3 (2022) 79-100 https://doi.org/10.34260/jaebs.635 93 study. workplace vigilantism (β=0.458, t=10.566, p=<.05) indicates a positive and significant relationship with change-oriented ocb. accordingly, workplace vigilantism (β=0.410, t=8.909, p=<.05) shows a significant and positive relationship with turnover intention. similarly, workplace vigilantism (β=0.414, t=8.895, p=<.05) indicates a positive and significant relationship with work mindfulness. 3.10 coefficient of determination r2 in the structural model, the researcher should evaluate the predictive power of the model through the coefficient of determination r2. table 8: result of r square r square r square adjusted change oriented ocb 0.209 0.207 turnover intention 0.168 0.166 work mindfulness 0.171 0.169 table 8 shows the values of the coefficient of determination (r2). the r2 value of change oriented org citizenship behavior is 0.209 while the value of adjusted r2 is 0.207 which proposed that workplace vigilantism and workplace fatigue both explain a 20% change in change oriented organization citizenship behavior. the result showed a moderated predictive measures r2 value. accordingly, the r2 value of turnover intention is 0.168 and the adjusted r2 value is 0.166 showing moderated effect finally, the r2 value of work mindfulness is 0.171 and its adjusted r2 value is 0.169 showing the moderated effect of workplace vigilantism on work mindfulness. 4. conclusion and recommendations the overall objective of the study was to develop and test a model to find out the relationship between workplace vigilantism, turnover intention, work mindfulness, and change oriented organizational citizenship behavior (ocb). in this connection, the research aim was to focus on health sector nurses of three hospitals of district peshawar, khyber pakhtunkhwa, the population which is equal to 1800. by using the simplified formula for proportions, taro yamane and the technique of uma sekaran (2012) were utilized to calculate the sample size as well as which is 317. however, 350 questionnaires were distributed due to a possible non-response rate. to avoid any biases, the probability-stratified sampling technique was used. each hospital was considered as a separate stratum due to the inter-group heterogeneity and from each stratum, respondents were selected subject to their availability. below is the brief information representing the name of hospitals and the number of nurses’ muhammad naveed iqbal, anjum ihsan & shahid jan kakakhel 94 staff located in the district of peshawar. the result of the first hypothesis indicated a statistically significant and positive association between workplace vigilantism and anger with proactive behavior.a prior study by podsakoff et al., (2009) stated that vigilante in the workplace may not be appreciated by colleagues and peers which ultimately affects organizational behaviors like work mindfulness. the current study has tested the relationship between workplace vigilantism and work mindfulness and found a strong association between them. the current result supports the hypothesis that workplace vigilantism has a strong relationship with work mindfulness. accordingly, the current study has tested the association between workplace vigilantism and change-oriented ocb. the result shows a significant correlation between them. the current study is in line with the prior study results that constructive efforts in workplace vigilantism are to identify and implement changes concerning work methods, policies, and procedures to improve the situation and performance’ (bettencourt, 2004). hence, a vigilante who challenges the existing flow of activities and wants to transform the organization in light of the laid down rules and regulations but is not acknowledged and appreciated will be negatively affected. (bettencourt, 2004). the result of the hypothesis shows a statistically significant association between workplace vigilantism and turnover intention. the result of the study confirmed the claim of (podsakoff et al., (2009) that vigilante behaviors are strongly associated with an increase in state depletion and subsequent turnover intentions. recommendations the statistical pieces of evidence of the current research study provide practical implications in the way of offering directions to the health sector employees of khyber pakhtunkhwa to support the one who acts as a vigilante in the workplace as his role is to identify uncivil and criminal activities just for the sake of the betterment of the organization. similarly, the behavior of change-oriented ocb is also a constructive effort performed by individuals to identify and implement changes concerning work methods, policies, and procedures to improve the situation and performance. this research study emphasizes the organization to encourage the behavior of change-oriented ocb among employees to maintain and enhance the social, motivational, and psychological environment for task performance. finally, the reward mechanism will boost the role of vigilantes to reduce the incidents of uncivil and immoral activities in the workplace. future direction journal of applied economics and business studies, volume. 6, issue 3 (2022) 79-100 https://doi.org/10.34260/jaebs.635 95 future direction is a continuous process to bring more improvement in a specified area. in the present research study, there is one independent variable while three are the dependent variables. in the future, research can have a moderating effect on perceived organizational support to strengthen the vigilante to work for the ultimate benefit of the organization. references ahmed, a., & villegas, p. (2016). as mexican frustrations rise, so does number of lynchings. the new york times, a1-l. amitaietzioni. (1961). complex organizations: a sociological reader. holt, rinehart, and winston. ashforth, b. e., & humphrey, r. h. (1997). the ubiquity and potency of labeling in organizations. organization science, 8(1), 43-58. bateson, r. (2021). the politics of vigilantism. comparative political studies, 54(6), 923955. burke, p. j., & tully, j. c. (1977). the measurement of role identity. social forces, 55(4), 881-897. burrows, b. (1976). discharge behavior of redox term galvanic cells. journal of the electrochemical society, 123(2), 154. caldwell, d. f., chatman, j. a., &o'reilly, c. a. (1990). building organizational commitment: a multi-firm study. journal of occupational psychology, 63(3), 245261. cardinal, l. b., sitkin, s. b., & long, c. p. (2004). balancing and rebalancing in the creation and evolution of organizational control. organization science, 15(4), 411431. chen, f. x., graso, m., aquino, k., lin, l., cheng, j. t., decelles, k., &vadera, a. k. (2022). the vigilante identity and organizations. organizational behavior and human decision processes, 170, 104136. choi, m. (2011). employees' attitudes toward organizational change: a literature review. human resource management, 50(4), 479-500. colquitt, j. a., noe, r. a., & jackson, c. l. (2002). justice in teams: antecedents and consequences of procedural justice climate. personnel psychology, 55(1), 83-109. cordes, c. l., & dougherty, t. w. (1993). a review and an integration of research on job burnout. academy of management review, 18(4), 621-656. cordes, c. l., & dougherty, t. w. (1993). a review and an integration of research on job muhammad naveed iqbal, anjum ihsan & shahid jan kakakhel 96 burnout. academy of management review, 18(4), 621-656. cordes, c. l., & dougherty, t. w. (1993). a review and an integration of research on job burnout. academy of management review, 18(4), 621-656. cotton, j. l., & tuttle, j. m. (1986). employee turnover: a meta-analysis and review with implications for research. academy of management review, 11(1), 55-70. decelles, k. a., & aquino, k. (2020). dark knights: when and why an employee becomes a workplace vigilante. academy of management review, 45(3), 528-548. dimaggio, p. j., & powell, w. w. (1991). the iron cage revisited. durkheim, e. (1949). the division of labor in society. translated from the french by george simpson. free press. erikson, e. h. (1962). reality and actuality an address. journal of the american psychoanalytic association, 10(3), 451-474. ferris, d. l., brown, d. j., berry, j. w., &lian, h. (2008). the development and validation of the workplace ostracism scale. journal of applied psychology, 93(6), 1348. gelʹfand, i. m., minlos, r. a., & shapiro, z. y. (2018). representations of the rotation and lorentz groups and their applications. courier dover publications. gelfand, m. j., nishii, l. h., & raver, j. l. (2006). on the nature and importance of cultural tightness-looseness. journal of applied psychology, 91(6), 1225. glomb, k., kringelbach, m. l., deco, g., hagmann, p., pearson, j., &atasoy, s. (2021). functional harmonics reveal multi-dimensional basis functions underlying cortical organization. cell reports, 36(8), 109554. goldstein, d. m. (2003). " in our own hands": lynching, justice, and the law in bolivia. american ethnologist, 30(1), 22-43. gonzález candía, j., garcíacoliñanco, l., caro, c. l., & romero hernández, n. (2014). strategy and culture of innovation, resources management, and generation of ideas: practices for managing innovation in business. pensamiento&gestión, (36), 109135. grant, a. m., & ashford, s. j. (2008). the dynamics of proactivity at work. research in organizational behavior, 28, 3-34. halbesleben, j. r. (2021). individual-level outcomes of employee engagement: conservation of resources framework. in a research agenda for employee engagement in a changing world of work. edward elgar publishing. journal of applied economics and business studies, volume. 6, issue 3 (2022) 79-100 https://doi.org/10.34260/jaebs.635 97 hernandez-espallardo, m., osorio-tinoco, f., & rodriguez-orejuela, a. (2018). improving firm performance through inter-organizational collaborative innovations: the key mediating role of the employee’s job-related attitudes. management decision. higazi, a. (2016). farmer-pastoralist conflicts on the jos plateau, central nigeria: security responses of local vigilantes and the nigerian state. conflict, security & development, 16(4), 365-385. hing, b. o. (2001). vigilante racism: the de-americanization of immigrant america. mich. j. race & l., 7, 441. huggins, m. k. (ed.). (1991). vigilantism and the state in modern latin america: essays on extralegal violence (pp. 1-18). new york: praeger. jacobs, d., carmichael, j. t., & kent, s. l. (2005). vigilantism, current racial threat, and death sentences. american sociological review, 70(4), 656-677. jacobs, d., carmichael, j. t., & kent, s. l. (2005). vigilantism, current racial threat, and death sentences. american sociological review, 70(4), 656-677. jacobs, d., carmichael, j. t., & kent, s. l. (2005). vigilantism, current racial threat, and death sentences. american sociological review, 70(4), 656-677. jensen, m. c., &meckling, w. h. (1976). theory of the firm: managerial behavior, agency costs, and ownership structure. journal of financial economics, 3(4), 305-360. kirsch, t. g., &grätz, t. (eds.). (2010). domesticating vigilantism in africa. boydell& brewer. knudsen, h. k., ducharme, l. j., & roman, p. m. (2008). clinical supervision, emotional exhaustion, and turnover intention: a study of substance abuse treatment counselors in the clinical trials network of the national institute on drug abuse. journal of substance abuse treatment, 35(4), 387-395. loughry, m. l., & tosi, h. l. (2008). performance implications of peer monitoring. organization science, 19(6), 876-890. maslach, c. (1978). the client's role in staff burn‐out. journal of social issues, 34(4), 111124. maslach, c. (1986). stress, burnout, and workaholism. maslach, c., & jackson, s. e. (1981). the measurement of experienced burnout. journal of organizational behavior, 2(2), 99-113. mazzei, j. (2009). death squads or self-defense forces?: how paramilitary groups emerge muhammad naveed iqbal, anjum ihsan & shahid jan kakakhel 98 and challenge democracy in latin america. univ of north carolina press. mccabe, b. c., feiock, r. c., clingermayer, j. c., & stream, c. (2008). turnover among city managers: the role of political and economic change. public administration review, 68(2), 380-386. merton, r. k. (1949). on sociological theories of the middle range [1949]. morrison, e. w., & milliken, f. j. (2000). organizational silence: a barrier to change and development in a pluralistic world. academy of management review, 25(4), 706725. naumann, s. e., & bennett, n. (2000). a case for procedural justice climate: development and test of a multilevel model. academy of management journal, 43(5), 881-889. near, j. p., &miceli, m. p. (1985). organizational dissidence: the case of whistle-blowing. journal of business ethics, 4(1), 1-16. neroulias, n. (2011). seattle’s superhero vigilante has his day in court. reuters. nieves, j., quintana, a., & osorio, j. (2016). organizational knowledge, dynamic capabilities, and innovation in the hotel industry. tourism and hospitality research, 16(2), 158-171. oreg, s., vakola, m., &armenakis, a. (2011). change recipients’ reactions to organizational change: a 60-year review of quantitative studies. the journal of applied behavioral science, 47(4), 461-524. osorio, j., schubiger, l. i., & weintraub, m. (2021). legacies of resistance: mobilization against organized crime in mexico. comparative political studies, 54(9), 15651596. ouchi, w. g. (1977). the relationship between organizational structure and organizational control. administrative science quarterly, 95-113. parker, s. k., & collins, c. g. (2010). taking stock: integrating and differentiating multiple proactive behaviors. journal of management, 36(3), 633-662. parsons, t., &shils, e. a. (1951). values, motives, and systems of action. toward a general theory of action, 33, 247-275. pratten, d. (2008). the politics of protection: perspectives on vigilantism in nigeria. africa, 78(1), 1-15. rafferty, a. e., jimmieson, n. l., &armenakis, a. a. (2013). change readiness: a multilevel review. journal of management, 39(1), 110-135. journal of applied economics and business studies, volume. 6, issue 3 (2022) 79-100 https://doi.org/10.34260/jaebs.635 99 rosenbaum, h. j., &sederberg, p. c. (1974). vigilantism: an analysis of establishment violence. comparative politics, 6(4), 541-570. sampson, r. j., raudenbush, s. w., & earls, f. (1997). neighborhoods and violent crime: a multilevel study of collective efficacy. science, 277(5328), 918-924. skarlicki, d. p., & folger, r. (1997). retaliation in the workplace: the roles of distributive, procedural, and interactional justice. journal of applied psychology, 82(3), 434. slagter, m., &wilderom, c. (2018). team coaching and effective team leadership. in handbuchschlüsselkonzepteim coaching (pp. 593-602). springer, berlin, heidelberg. spector, p. e. (1986). perceived control by employees: a meta-analysis of studies concerning autonomy and participation at work. human relations, 39(11), 10051016. spitzmuller, m., & van dyne, l. (2013). proactive and reactive helping: contrasting the positive consequences of different forms of helping. journal of organizational behavior, 34(4), 560-580. trevino, l. k., & victor, b. (1992). peer reporting of unethical behavior: a social context perspective. academy of management journal, 35(1), 38-64. trevino, l. k., & victor, b. (1992). peer reporting of unethical behavior: a social context perspective. academy of management journal, 35(1), 38-64. tripp, t. m., bies, r. j., & aquino, k. (2007). a vigilante model of justice: revenge, reconciliation, forgiveness, and avoidance. social justice research, 20(1), 10-34. trottier, d. (2017). digital vigilantism as the weaponization of visibility. philosophy & technology, 30(1), 55-72. turner, r. h. (1978). the role and the person. american journal of sociology, 84(1), 1-23. vaisey, s. (2009). motivation and justification: a dual-process model of culture in action. american journal of sociology, 114(6), 1675-1715. victor, b., & cullen, j. b. (1988). the organizational bases of ethical work climates. administrative science quarterly, 101-125. wharton, a. s. (1999). the psychosocial consequences of emotional labor. the annals of the american academy of political and social science, 561(1), 158-176. widiatmika, p. h., &darma, g. s. (2018). good corporate governance, job motivation, and organization culture impact company financial performance. jurnalmanajemenbisnis, 15(3), 82-99. muhammad naveed iqbal, anjum ihsan & shahid jan kakakhel 100 wise, s., yeganegi, s., &laplume, a. o. (2022). startup team ethnic diversity and investment capital raised. journal of business venturing insights, 17, e00314. wrong, d. (1994). the problem of order. simon and schuster. yandell, b. s. (2017). practical data analysis for designed experiments. routledge. yu, l., &zellmer-bruhn, m. (2018). introducing team mindfulness and considering its safeguard role against conflict transformation and social undermining. academy of management journal, 61(1), 324-347. zheng, j., he, c., qin, y., lin, g., park, w. d., sun, m., ... & liu, s. (2019). co‐expression analysis aids in the identification of genes in the cuticular wax pathway in maize. the plant journal, 97(3), 530-542. zhu, l., martens, j. p., & aquino, k. (2012). third-party responses to justice failure: an identity-based meaning maintenance model. organizational psychology review, 2(2), 129-151. zizumbo-colunga, d. (2017). community, authorities, and support for vigilantism: experimental evidence. political behavior, 39(4), 989-1015. journal of applied economics and business studies, volume.4, issue 4 (2020) 55-74 https://doi.org/10.34260/jaebs.443 55 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x analysis of natural resources and environment, politico-economic conditions and their influences on tourist behavioural intentions in hunza: mediating effect of tourist satisfaction faqeer muhammad1*, kifayat ullah2 & rehmat karim3 1 department of economics, karakoram international university, university road gilgit. 2 department of economics, karakoram international university, university road gilgit. 3 assistant professor, karakoram international university, gilgit-baltistan. abstract this study aims to explore the influence of natural resources and environment (nre), politico-economic conditions (pec) on tourist behavioral intension (tbi) in hunza, pakistan. the study further investigates the mediating role of tourist satisfaction (ts) on the given variables. partial least square structural equation modeling (plssem) technique has been applied to conceptualize the research frame and to test the proposed hypotheses. primary data was collected by using convenient sampling technique for analysis from 220 tourists who visited tourism nucleus sites of hunza. the finding of the study reveals that natural resources and environment, politico-and economic conditions have a significant positive impact on tourist’s behavioral intensions. moreover, tourist’s satisfaction partially mediates the positive relationships among natural resources and environment, political & economic conditions and tourist’s behavioral intensions. the findings of the study extend the understanding that presence of natural resources along with healthy environment and stable political & economic conditions of a destination are the key determinants for sustainable tourism development. keywords natural resources and environment (nre), tourist satisfaction (ts), tourists behavioural intensions (tbi), mediation jel classification p28, z3, l83 1. introduction tourism plays an important role in the economy of a country and it has become one of the global businesses and fastest emergent economic segments therefore, various * faqeer@kiu.edu.pk faqeer muhammad, kifayat ullah & rehmat karim 56 economies consider it as an optimal tool for local development. similarly, in recent decades, tourism has been considered as an integral area that has an influence on the economic development. therefore, to promote tourism, it is essential to developed destinations according to the state of art infrastructure. in addition, investments in clean environment, heritage and culture are also required for tourist satisfaction. ariya, wishitemi and sitati (2017), philemon (2015), lai, hitchcock, lu and liu (2018) and gnanapala (2015) have shown that safety and security are the major concerns for the tourist satisfaction. however, factors which influence the perception of tourists is word of mouth (lai et al., 2018). similarly, the other factors of tourist satisfaction and attractions are natural beauty and wildlife resource (ariya et al, 2017). gnanapala (2015) highlighted that disrespectful attitude of custom officials at airports and beach boys have negative effect on tourist satisfaction. infrastructure development, online services and local transport are other services which paly vital role in selection of the destination (rehman, 2012). however according to alim, ray and hossain (2016) role of transportation is not significant in the selection of the destination whereas alim et al., (2016) found in their study that food and beverages are the major determinants in selecting the destination. according to a recent study conducted by gallup pakistan (2019, p. 7), “tourism could be a potential game changer that could revitalize the struggling economy of the country”. keeping this in view, pakistan has tremendous potential in all kinds of tourism, for example religious, expedition and culture tourism. for foreign tourists, museum sites are more popular and it attracts 50% more tourists as compared to cultural sites (gallup pakistan, 2019). likewise, the opening of kartarpur corridor will have great potential of religious tourism in pakistan. among the other tourist destinations, hunza, located in northern part of pakistan has unique and diverse seasons, culture and tradition which attracts many domestic and international tourists. in addition, the recent inflow of domestic tourists to the area is due to improvement in political stability, and improved law and order situation. therefore, this study aimed to explore the natural environment, political and economic conditions on tourist satisfaction and tourist behaviour intention. furthermore, this study also investigates the mediating effect of the tourist satisfaction. the current research is the first hand research in the study area using pls-sem technique due to which it has unique significances for the stakeholders associated with tourism sector. the study also aims to provide policy recommendations to the policy makers. journal of applied economics and business studies, volume.4, issue 4 (2020) 55-74 https://doi.org/10.34260/jaebs.443 57 2. literature review according to ariya, wishitemi and sitati (2017) the most important attributes to tourist’s attraction are wildlife resources, safety and security. in addition, the other key factor, which determines the destination attractiveness is the quality of the infrastructure. the attributes which displeasure tourists are various fees e.g. high park fees and inflated facilities. improving the accessibility and enhancing the quality of the roads for better choice for the tourists. li et al., (2017) used the “second order structural equation model” to analyze the tourist’s perception whereas in this study, the authors discussed the various dimensions of the crowding and congestions in the study area. the outcome of the research showed the negative effect on attractiveness of the destination due to perception of crowding and its direct effect on tourists’ satisfaction. neuts and nijkamp (2012) also emphasized on the high-density destination crowding and recommended to improve the infrastructure of the destinations which increases the positive tourists’ perception about the destination. they also highlighted that a positive influence of attractions, ancillary services, amenities and accommodation on memorable travel experience. alim, ray and hossain (2016) explored the significant factors which attract the tourist in selection of the destination. the findings have shown that the major factors which paly role in selection of destination is food and beverage while transportation has minor role in selection of the destination. philemon (2015) findings showed that international tourists have better perception in some areas i.e. landscape, culture and landscape, however the results have shown the concerns of tourists on tour guiding quality, safety and security. the authors suggested that the policy makers should improve the weak areas and further improve the developed areas of the tourism. in addition, they emphasized on the promotion of positive image of the destination in the national media to counter the negative propaganda. banki et al., (2014) explored the moderating role of the affective image (ai) of destination in examining the association between tourists’ behavioral intention (tbi) and tourist satisfaction (ts). for empirical analysis, the authors conducted their study in the mountain tourism destination by using sem modeling. the results revealed an insignificant relationship between ai and tbi. in addition, the study showed that ai has a moderating effect in explaining the relationship between tbi and ts. canny (2013) conducted a study to explore the various dimensions of tourists’ service quality on satisfaction of the tourists. he also examined the tourist satisfaction on future intentions of the tourists. the results showed a positive and influential influence of service quality on perception of the tourists. similar relation has been observed between tourists’ intention and tourist satisfaction. faqeer muhammad, kifayat ullah & rehmat karim 58 bertan and altintaş (2013) used the one-way variance analysis to assess the effects of various demographic characteristics on their perception about the destination. the outcomes of the study revealed the significant difference between demographic characteristics and perception of the destination. the study suggested that service quality, the employee should be trained and educated in various services, facilities and designed of rooms to enhance the tourism business. rahman (2012) carried out the study about tourist perception of bangladesh as a tourist destination by using sem. the author also examined the most influential features, which are essential for tourists. the results of the sem has shown that in choosing destination brand the influential factors are; “brand image, internet adoption and customers satisfaction” rahman (2012, p. 86). in addition, the important factors of needs and satisfaction of tourist in selection of destination are reputation of visiting place, online services and the system of local transport. similarly, to attract more domestic and international tourists there is need to enhance the infrastructure, which is required for tourism development (dragicevic, stankov & аrsenovic, 2011). wong and yeh (2009) investigated the relationship among tourist risk perception, hesitation and knowledge using structural equation modelling. the results showed that tourist risk is influencing the tourist hesitation however; the knowledge of tourist can moderate the association between tourist risk perception and hesitation. in addition, “hesitating tourists represent fish that have not yet been caught by the nets of tourism managers” (wong & yeh, 2009, p. 18). 3. theoretical framework and research methodology 3.1 conceptual model and hypotheses development assaf and josiassen (2012) used tourism performance theory to determine the drivers of tourism destination image and perception. moreover, authors analyzed the determinants of tourism performance of destination by measuring various factors; economics conditions, environmental aspects, infrastructure, natural and cultural resources. the current research used tourism performance theory to conceptualize the model. the current research model consists of three exogenous latent constructs i.e. natural resources and environment (nre) with eight observed variables, politicoeconomic conditions (pec) with five items and tourists satisfaction (ts) with eight indicators. the model also includes one endogenous latent construct, named tourists behavioral intensions (tbi) with five indicators. conceptual model of the study and the hypothesized relationships among dependent and independent latent constructs are given in the schematic diagram as under: journal of applied economics and business studies, volume.4, issue 4 (2020) 55-74 https://doi.org/10.34260/jaebs.443 59 figure 1: conceptual model (schematic diagram) from the above schematic diagram, we can draw following four hypotheses: h1: natural resources and environment (nre) have a significant positive effect on tourists behavioral intentions (tbi) h2: tourists satisfaction (ts) will mediate the positive relationship between natural resources and environment (nre) and tourists’ behavioral intentions (tbi) h3: politico-economic conditions (pec) have a significant positive effect on tourists behavioral intentions (tbi) h4: tourists satisfaction (ts) will mediate the positive relationship between politicoeconomic conditions (pec) and tourists’ behavioral intentions (tbi) 3.2 research methodology the study proposed “partial least squares structural equation modelling” (plssem) technique to explore the relationships between natural resources & politicoeconomic and environmental conditions, tourists’ satisfaction and tourist’s behavioral intensions in hunza, pakistan. pls-sem is a multivariate statistical method that evaluates both measurement and structural model in order to find the relationships between study constructs and their observed indicators and among the latent constructs all together (hair, hult, ringle, & sarstedt, 2013). in the tourist behavioral intension model, tourist satisfaction (ts) was introduced as a mediator variable to examine its indirect role on the direct relationship between natural resources and environment (nre), politico-economic conditions (pec) and tourist’s behavioral intensions h3 nre pec ts tbi h1 h4 h2 faqeer muhammad, kifayat ullah & rehmat karim 60 (tbi). the study identified influencing factors of natural resources and environment, politico-economic conditions, tourist’s satisfaction and tourist behavioral intensions after an extensive literature review of some previous studies like lai et al. (2018), tukamushaba, xiao & ladkin (2016), banki et al. (2014), wong and yeh (2009), beerli and martın (2004), baker and crompton (2000) etc. finally, we developed eight items for natural resources & environment (nre), five items for political & economic conditions (pec), eight items for tourist’s satisfaction and five items for tourist’s behavioral intensions (tbi). 3.2 latent constructs (list of influencing factors) the study identified influencing factors for tourist satisfaction (ts), natural resources and environment (nre), political and economic conditions (pec) and tourists behavioral intentions (tbi) after an extensive literature review. finally, we developed eight items for tourist satisfaction (ts), eight items for natural resources and environment (nre) and five items for political and economic conditions (pec) and tourists behavioral intentions (tbi) each. details are given in table 1 as under. table 1: list of factors items codes tourist satisfaction (ts) ts1 i am satisfied from the quality of hotels and accommodation ts2 vacation met all expectations ts3 perform services right at the first time ts4 well established on line transaction ts5 offering multiple choices in travel services ts6 ease of access to destination ts7 i am satisfied from the food available in restaurants ts8 health services and facilities are available natural resources and environment (nre) nre1 weather is good in hunza valley nre2 there are protected nature reserves, lakes and mountains. nre3 variety and uniqueness of flora and fauna nre4 hunza valley is naturally attractive and beautiful nre5 the environment is neat and clean nre6 there is overcrowding in the area nre7 there is no air and noise pollution nre8 there is traffic congestion in the area political and economic conditions (pec) pec1 political stability in the area journal of applied economics and business studies, volume.4, issue 4 (2020) 55-74 https://doi.org/10.34260/jaebs.443 61 pec2 economic development is more in hunza pec3 the crime rate is low in hunza pec4 no fear of terrorist attacks pec5 prices of the goods and services are affordable in hunza tourists behavioral intensions (tbi) tbi1 if i had to decide again, i would choose this destination tbi2 i will recommend this destination to friends and family tbi3 i will speak highly of this destination to friends and relatives tbi4 i intend to holiday in this destination within the next year tbi5 consider hunza as your choice to visit in the future 3.3 data collection before going to data collection, a pilot study was conducted to test the reliability of the questionnaire items. after clear understanding and testing of questionnaire for reliability, we executed the final survey in the study area. in order to collect micro data from the respondents we developed questionnaire, which consisted of two sections. section one of the questionnaire was designed to receive the data regarding respondent’s demographic variables like gender, age education and marital status etc. section two consisted items related to the influencing factors of natural resources & environment, political and economic conditions, tourists’ satisfaction and tourists’ behavioral intentions. the statements in this section were designed in to five-point likert-scale and respondent’s responses were recorded from “strongly disagree = 1 to strongly agree = 5 (strongly disagree = 1, disagree = 2, neutral = 3, agree = 4, and strongly agree = 5)”. the target population for this study consisted all those national tourists who were present in the restaurants, guesthouses, tourist huts, and tourist spots in the famous tourist destinations of hunza and gojal valleys (aliabad, altit, karimabad and gulmit) in gilgit baltistan, pakistan in the month of july 2019 at different times. using convenience-sampling technique only those tourists were approached who were willing to participate in the survey. respondents were informed about the objectives of the research and confidentiality of the data before the questionnaires’ distribution. a total of 220 questionnaires were returned, and all were deemed fit for further analysis. this sample size not only fulfills the widely used minimum sample size criteria (10 responses per indicator) but also fulfil the preferred sample size criteria (20 responses per indicators). the current research used spss for descriptive analysis and to evaluate the demographic profile of the respondents. similarly, smart pls-sem is used for investigation of the research model. to ensure reliability and validity, we investigated faqeer muhammad, kifayat ullah & rehmat karim 62 reflective measurement model first and after then we examined structural model (hair et al., 2017). for quality assurance of outer model, we used pls algorithm method. 4. results following table describes descriptive statistical analysis i.e. mean, standard deviation, skewness and kurtosis values. results reported in table 2 showed satisfactory results regarding the normality of the data distribution. table 2. mean, standard deviation, kurtosis and skewness values items mean standard deviation excess kurtosis skewness ts1 2.709 1.306 -1.042 0.392 ts2 2.736 1.222 -0.954 0.381 ts3 2.759 1.308 -1.122 0.233 ts4 2.864 1.272 -1.034 0.218 ts5 2.745 1.331 -1.020 0.336 ts6 2.918 1.339 -1.144 0.185 ts7 2.773 1.349 -1.109 0.342 ts8 2.782 1.268 -1.030 0.377 nre1 2.595 1.367 -0.957 0.535 nre2 2.695 1.188 -0.643 0.608 nre3 2.759 1.356 -1.147 0.313 nre4 2.868 1.316 -1.072 0.510 nre5 2.782 1.348 -1.083 0.415 nre6 2.832 1.219 -0.908 0.417 nre7 2.841 1.299 -1.013 0.311 nre8 2.700 1.269 -0.971 0.405 pec1 2.682 1.246 -0.963 0.268 pec2 2.864 1.07 -0.830 0.409 pec3 2.727 1.246 -0.897 0.303 pec4 2.914 1.231 -0.980 0.269 pec5 2.736 1.237 -0.951 0.281 tbi1 2.7 1.269 -0.971 0.405 tbi2 2.673 1.199 -0.787 0.399 tbi3 2.427 1.221 -0.647 0.615 tbi4 2.709 1.139 -0.703 0.46 tbi5 2.732 1.403 -1.302 0.169 table 3 given below reports results of respondent’s profile including their gender, marital status, age, academic qualifications, and their native provinces. majority of the respondents (69%) were male while the remaining 31% were female belonged to all provinces including gb. the highest number of participants were from panjab (41%). journal of applied economics and business studies, volume.4, issue 4 (2020) 55-74 https://doi.org/10.34260/jaebs.443 63 these participants were all educated and some of them were employed (50%) while 30% of the respondents were self-employed and 17% of the participants were students from different universities of the country. table 3: respondent’s profile demographic characteristics frequency percent (%) gender male 152 69.09 female 68 30.91 marital status married 122 55.45 single 98 44.54 divorced 00 0.00 age under 29 years 98 44.54 30 39 years 67 30.45 40 49 years 38 17.27 above 49 years 17 7.72 qualification illiterate 00 0.00 primary 10 4.54 secondary 22 10.00 higher secondary 53 24.09 bachelors 57 25.90 masters and above 78 35.45 occupation employed 111 50.45 unemployed 04 1.81 self employed 66 30.00 students 39 17.72 province punjab 92 41.81 sindh 22 10.00 kpk 57 25.90 baluchistan 04 1.81 ajk 12 5.45 gb 33 15.00 others 00 0.00 n = 220 4.1 evaluation of measurement model the assessment of the measurement model is carried out using the criterion suggested by hair et al. (2017, 2013). table 4 given below shows item loadings, cronbach’s alpha, composite reliability and average variance extracted. the estimated values of items loadings, cronbach’s alpha and composite reliability exceeded the recommended value 0.7 (hair et al., 2013 latan & noonan, 2017). similarly, calculated value of average variance extracted also exceeded its recommended value 0.5 (hair et al., 2013; wong, 2013). table 4. validity and reliability of constructs (outer loadings, cronbach's alpha, average variance extracted and composite reliability) faqeer muhammad, kifayat ullah & rehmat karim 64 constructs items loadings cronbach's alpha ave cr nre1 0.789 0.931 0.674 0.943 nre2 0.806 nre3 0.861 nre4 0.791 nre5 0.853 nre6 0.848 nre7 0.837 nre8 0.779 pec1 0.905 0.932 0.785 0.948 pec2 0.889 pec3 0.912 pec4 0.865 pec5 0.859 tbi1 0.852 0.852 0.629 0.894 tbi2 0.822 tbi3 0.772 tbi4 0.798 tbi5 0.715 ts1 0.851 0.943 0.716 0.953 ts2 0.84 ts3 0.867 ts4 0.868 ts5 0.853 ts6 0.854 ts7 0.832 ts8 0.803 figure 2. pls-sem algorithm (measurement model) journal of applied economics and business studies, volume.4, issue 4 (2020) 55-74 https://doi.org/10.34260/jaebs.443 65 criterions like forner-lacker criterion and heterotrait-monotrait are commonly used to test discriminant validity of the measurement model (franke & sarstedt, 2019; hair et al., 2017). fornell-larcker criterion ensures discriminant validity through the examination of square root of average variance extracted for every latent variable included in the model. this criterion suggested that square root of average variance extracted for every latent variable must exceed than its correlation with other latent variable (hair et al., 2013). results reported in table 5 confirmed fornell-larcker criterion of discriminant validity. table 5: fornell-larker criterion constructs nre pec tbi ts nre 0.821 pec 0.788 0.886 tbi 0.785 0.758 0.793 ts 0.692 0.701 0.698 0.846 note: “values on the diagonal (bolded) are square root of the ave while the offdiagonals are correlations” henseler, ringle, and sarstedt, 2015 suggests that fornell and larcker method is not a reliable technique to validate the discriminant validity. therefore, henseler et al. (2015) proposed a reliable method i.e. heterotrait-monotrait ratio (htmt) method. the results reported in the table shows that the values of the heterotrait-monotrait ratio (htmt) method are less than 0.90 which is threshold value (henseler et al., 2015). table 6. heterotrait-monotrait (htmt) constructs nre pec tbi ts nre pec 0.844 tbi 0.861 0.848 ts 0.734 0.745 0.775 4.2 evaluation of structural model table 7: collinearity assessment (inner vif values) constructs nre pec tbi ts nre 2.934 2.639 pec 3.011 2.639 tbi ts 2.188 the results in table 7 indicate the absence of collinearity among independent variables. it explains overall variations in the dependent variables/endogenous constructs due to any change in independent variables/ exogenous constructs in the model. the value of r2 is considered substantial at 0.75, moderate at 0.50, and weak at faqeer muhammad, kifayat ullah & rehmat karim 66 0.26, respectively (hair et al., 2017; henseler et al., 2015). the endogenous latent constructs i.e. tourist’s behavioral intensions and tourist’s satisfaction have r2 values 0.771 and 0. 543 respectively. it means that 77 percent variation in the model’s dependent construct tourist’s behavioral intension (tbi) is caused by the model’s independent constructs i.e. natural resources & environment (nre), political & economic conditions (pec) and the mediator tourists satisfaction (ts). similarly, two independent variables (nre & pec) caused 54 percent variation in the dependent variable tourist satisfaction (ts). for the present study, we found significant values of r2. table 8: results coefficient of determination (r2) constructs r square r square adjusted tbi 0.774 0.771 ts 0.543 0.539 4.3 mediation test according to hair et al., (2013) three basic condition are required to meet for variable to accts a mediator. first, a significant relationship must be identified between dependent and independent variables without including the mediator. it implies that direct path relationship from nre and pec to pbi must be significant without including ts. second, the indirect relationship between dependent, mediator and independent variable (after inclusion of mediator variable in the model) must be significant. it means that indirect relationship between nre-ts-tbi and pec-ts-tbi must be significant. significance level of relationships among different variables in a model is tested through the calculation of beta (β) and associated (t) and (p) values. results reported in table 9 revealed that our mediation model fulfilled first two crucial conditions for mediation analysis. the direct effect from nre to tbi without including ts is highly significant (β = 0.867, t = 50.051 and p = 0.000). the indirect effect from nre-tbi through ts (after including mediator ts) is also significant at 5% (β = 0.055, t = 2.077 and p = 0.038). in the similar lines, direct effect from pec to pbi (β = 0.757, t = 27.978 and p = 0.000) and statically significant at 1 %. the indirect effect from pec -tbi through ts is also statically significant at 5 % (β = 0.062, t = 2.271 and p = 0.024). table 9: mediation test results constructs beta (β) t-value p-value nre tbi 0.867 50.051 0.000 pec tbi 0.757 27.978 0.000 nre ts tbi 0.055 2.077 0.038 pec ts tbi 0.062 2.271 0.024 notes: *(p < 0.01); **(p < 0.05) journal of applied economics and business studies, volume.4, issue 4 (2020) 55-74 https://doi.org/10.34260/jaebs.443 67 4.4 structural estimates (hypotheses testing) mostly beta (β) value is used to test the significance of hypotheses in a model. the value of beta (β) in a model shows variation in the dependent variable to a unit change in independent variable. to test whether beta (β) value is significant or not we use ttest and p-value. table 10 given below shows analysis of the structural model i.e. proposed hypotheses and their decisions along with beta (β), (t) and (p) values (also see figure 3). table 10. hypotheses testing hypotheses beta (β) t-value p-value decision h1: nre tbi 0.652 10.156 0.000 supported h2: nre ts tbi 0.055 2.077 0.038 supported h3: pec tbi 0.14 2.078 0.038 supported h4: pec ts tbi 0.062 2.271 0.024 supported notes: *(p < 0.01); **(p < 0.05) figure 3: bootstrapping results faqeer muhammad, kifayat ullah & rehmat karim 68 the results reported in table 10 revealed that natural resources & environment have a significant positive impact on tourists’ behavioral intension (h1) and was significant statically at 1% level of significance (β = 0.652, t = 10.156 and p = 0.000). thus, the results supported hypothesis 1. similarly, the results also supported hypothesis 3, which stated that political and economic conditions (pec) have a significant positive effect on tourists’ behavioral intensions ((β = 0.14, t = 2.078 and p = 0.038). this relationship was also significant statically at 5% level of significance. thus, based on study results we conclude that the condition of natural resources & environment (nre) and stable political & economic conditions are the key factors, which further attracts tourists and future growth of tourism industry in the region. 4.5 mediation analysis the smart pls-sem results also endorsed our hypothesis (h2) and (h4). h2 stated that tourist’s satisfaction (ts) would mediate the positive relationship between natural resources & environment (nre) and tourists’ behavioral intensions (tbi). the results of the indirect effect from nre-ts-tbi (β = 0.055, t = 2.077 and p = 0.038) in table 10 supported this hypothesis. to test the strength of mediation the researchers have uses variance accounted for (vaf) method. the vaf value is 0.224 (0.055/0.243=0.224) which indicates that partial mediation because vaf value lies between 0.20 and 0.80 shows partial mediation (hair et al., 2013). similarly, (h4) described that tourists’ satisfaction (ts) will mediate the positive relationship between political & economic conditions (pec) and tourists’ behavioral intensions (tbi). the results presented in table 11 also supported this hypothesis i.e. the indirect effect from pec-ts-tbi (β = 0.062, t = 2.271 and p = 0.024). the strength of mediation in this case was (0.062/0.201=0.308). therefore, we conclude that tourist’s satisfaction (ts) partially mediate the positive relationships between natural resources and environment (nre), political & economic conditions (pec) and tourist’s behavioral intensions (tbi). the results reported in table 11 has shown that the relationship between natural resources and environment (nre) and tourist’s behavioral intensions (tbi) is strong (0.64) while all other variables included in the model have moderate relationships among each other. according to the (cohen, 2013) the size effect (f2) value at 0.35, 0.15 and 0.02 shows strong, moderate and weak effect respectively. table 11: effect size (f2) constructs tbi ts nre 0.64 0.112 pec 0.030 0.141 ts 0.045 journal of applied economics and business studies, volume.4, issue 4 (2020) 55-74 https://doi.org/10.34260/jaebs.443 69 quality of path model’s endogenous latent constructs is assessed by predictive relevance (q2). predictive relevance is estimated via blindfolding technique (chin, peterson, & brown, 2008; tenenhaus et al., 2005). if the calculated value of predictive relevance exceeds zero, then predictive relevance of a model is guaranteed (aman et al., 2019). the value of predictive relevance less than zero is an indication of lacking model’s predictive relevance (ali et al., 2016). it is evident from q2 value reported in table 12 (also see fig. 4) that our endogenous variables have acceptable values of predictive relevance. table 12. results of blindfolding constructs sso sse q² (=1-sse/sso) tbi 1100 602.597 0.452 ts 1760 1125.87 0.36 figure 4. pls-sem blindfolding (construct cross-validated redundancy) 4.6 model fit standardized root mean square residual (srmr) measures goodness of fit of a projected and estimated model (brown, 2006). an srmr value less than or equal to 0.08 indicates that the designed model is good fit and acceptable (aman et al., 2019). faqeer muhammad, kifayat ullah & rehmat karim 70 the results reported in table 13 showed an srmr value 0.061, which is less than the threshold value 0.08. thus based on srmr value, we conclude that our designed model is well fitted. table 13: model fit (standardized root mean square residual) test saturated model estimated model srmr 0.061 0.061 5. discussion tourism development in a country positively contributes to human functioning (beings and doings) which ultimately lead to improved quality of life in a society through the provision of better education, higher standards of health and nutrition, more equality of opportunities, higher incomes and less poverty, improved infrastructure facilities and greater individual freedom etc. tourists satisfaction can influence the perception level of tourist regarding tourism development (aman et al. 2019). we developed three hypotheses to test our “model”. the first hypothesis (h1) stated that natural resources and environment (nre) have a significant positive effect on tourists behavioral intentions (tbi). the study results supported this hypothesis (β = 0.652, t = 10.156, p = 0.000). study results also endorsed hypothesis 2 (h2) which postulated that tourists satisfaction (ts) will mediate the positive relationship between natural resources and environment (nre) and tourists’ behavioral intentions (tbi) (β = 0.055, t = 2.077, p = 0.038). similarly study results also recognized, hypotheses 3 (h3) which stated that politico-economic conditions (pec) have a significant positive effect on tourists behavioral intentions (tbi) (β = 0.14, t = 2.078, p = 0.038). hypothesis 4 which postulates that tourists satisfaction (ts) will mediate the positive relationship between politico-economic conditions (pec) and tourists’ behavioral intentions (tbi). the results of the study also endorsed this hypotheses i.e. (β = 0.62, t = 2.271, p = 0.0024). from the study results we established that higher level of tourist satisfaction in hunza valley favored tourism development and expansion, therefore, tourists’ satisfaction is an important indicator of sustainable tourism development in the study area. 6. conclusions this study explored the key factors, which influence tourists’ satisfaction, which ultimately has effect on tourists’ behavioral intention or “word to mouth” in hunza a famous tourist destination for foreign and domestic tourists. furthermore, the current research also investigated the mediating effect of the tourism satisfaction in the study area. recently, the initiatives of china pakistan economic corridor and the journal of applied economics and business studies, volume.4, issue 4 (2020) 55-74 https://doi.org/10.34260/jaebs.443 71 improvement in law and order situation caused attracting influx of domestic tourists to the region. the large number of tourists’ inflow to the region has on one hand created many business and employment opportunities while on the other hand, the negative externalities include unplanned development, congestion and conflicts with residents etc. the outcomes of the study have shown that the significant elements of the tourists’ satisfaction (ts) and tourist behavioural intentions (tbi) are natural resources environment (nre), political and economic conditions (pec). in addition, tourists’ satisfaction is mediating the relationship between nre and tbi. similarly, tourists’ satisfaction is mediating the relationship between pec and tbi. therefore, the findings suggest that all stakeholders’ i.e. local people, hotel owners and government should take care of the natural environment. likewise, government should improve safety and security of the tourist. lastly, maintaining and regulating the prices of commodities and room rents of hotel are also important to control prices in peak season. references ali, f., kim, w. g., & ryu, k. (2016). the effect of physical environment on passenger delight and satisfaction: moderating effect of national identity. tourism management, 57, 213-224. alim, m, a., ray, r., & hossain, m, e. (2016). visitors’ perception towards tour destinations: a study on padma garden. journal of business studies, 9. aman, j., abbas, j., mahmood, s., nurunnabi, m., & bano, s. (2019). the influence of islamic religiosity on the perceived socio-cultural impact of sustainable tourism development in pakistan: a structural equation modeling approach. sustainability, 11(11), 3039. ariya, g., wishitemi, b., & sitati, n. (2017). tourism destination attractiveness as perceived by tourists visiting lake nakuru national park, kenya. international journal of research in tourism and hospitality (ijrth). 3(4), 1-13. doi: 10.20431/2455-0043.0304001. assaf, a. g. and josiassen, a. (2012). identifying and ranking the determinants of tourism performance: a global investigation. journal of travel research. 51(4) 388–399. doi: 10.1177/0047287511426337. baker, d, a., & crompton, j, l. (2000). quality, satisfaction and behaviour intention. annals of tourism research, 27 (3), 785-804. pii: s0160-7383(99)00108-5. banki, m, b., ismail, h, n., dalil, m., & kawy, a. (2014). moderating role of affective destination image on the relationship between tourists satisfaction faqeer muhammad, kifayat ullah & rehmat karim 72 and behavioural intention: evidence from obudu mountain resort. journal of environment and earth science, 4(4). beerli, a., & martın, j, d. (2004a). factors influencing the destination image. annals of tourism research, 31 (3), 657–681. doi:10.1016/j.annals.2004.01.010. beerli, a., & martin, j.d. (2004b). tourists’ characteristics and the perceived image of tourist destinations: a quantitative analysis—a case study of lanzarote, spain. tourism management, 25,623–636. doi:10.1016/j.tourman.2003.06.004. bertan, s., & altintaş, v. (2013). visitors perception of a tourism destination. the case of a pamukkale. tourismos: an international multidisciplinary journal of tourism, 8, (1), 115-132. brown, t.a. confirmatory factor analysis for applied research; guilford publications: new york, ny, usa, 2006. canny, i, u. (2013). an empirical investigation of service quality, tourist satisfaction and future behavioral intentions among domestic local tourist at borobudur temple. international journal of trade, economics and finance, 4 (2). chin, w. w., peterson, r. a., & brown, p. s. (2008). structural equation modelling in marketing: some practical reminders. journal of marketing theory and practice, 16(4), 287e298. cohen, j. statistical power analysis for the behavioral sciences; taylor & francis: abingdon, uk, 2013. dragicevic, v., stankov, u., & аrsenovic, d. (2011). the attractiveness of vojvodina province as a tourist destination on the basis of domestic tourist’s perception, geographica timisiensis, 20(1), 59-67. d’silva, b., & d’silva, s. (2008). perception of tourists towards india as a preferred tourist destination: an empirical research. conference on tourism in india – challenges ahead, iimk. fornell, c., & larcker, d. f. (1981). evaluating structural equation models with unobservable variables and measurement error. journal of marketing research, 18, 39-50. franke, g.r., sarstedt, m., 2019. heuristics versus statistics in discriminant validity testing: a comparison of four procedures. internet res. in press. gnanapala, w, k, a. (2015) .tourists perception and satisfaction: implications for destination management. american journal of marketing research, 1 (1), 7-19. gallup pakistan (2019). “cultural heritage and museum visits in pakistan”. journal of applied economics and business studies, volume.4, issue 4 (2020) 55-74 https://doi.org/10.34260/jaebs.443 73 haenlein, m. & kaplan, a. m. (2004). a beginner’s guide to partial least squares analysis, understanding statistics, 3(4), 283–297. hair, j. f., hult, g. t. m., ringle, c., & sarstedt, m. (2013). a primer on partial least squares structural equation modelling (pls-sem). los angeles: sage publications. hair, j.f., risher, j.j., sarstedt, m., ringle, c.m. ,2019. when to use and how to report the results of pls-sem. eur. bus. rev. 31, 2–24. hair, j.f.; sarstedt, m.; ringle, c.m.; gudergan, s.p. (2017). advanced issues in partial least squares structural equation modeling. thousand oaks, ca, usa: sage publications. henseler, j., ringle, c. m., & sarstedt, m. (2015). a new criterion for assessing discriminant validity in variance-based structural equation modeling. journal of the academy of marketing science, 43, 115-135. lai, i, k, w., hitchcock, m., lu, d., & liu, y. (2018).the influence of word of mouth on tourism destination choice: tourist–resident relationship and safety perception among mainland chinese tourists visiting macau. sustainability, 10, 2114. doi: 10.3390/su10072114. latan, h., noonan, r. (eds.), 2017. partial least squares structural equation model ing: basic concepts, methodological issues and applications. springer, heidel berg. li, l, zhang, j., nian, s., & zhang, h. (2017). tourists’ perceptions of crowding, attractiveness, and satisfaction: a second-order structural model, asia pacific journal of tourism research, 22:12, 1250-1260, doi: 10.1080/10941665.2017.1391305. mohamad, m., ali, a, m., & ghani, n, i, a. (2011). a structural model of destination image, tourists satisfaction and destination loyalty. international journal of business and management studies, 3 (2). neuts, b., & nijkamp, p. (2012). tourist crowding perception and acceptability in cities: an applied modelling study on bruges. annals of tourism research, vol. 39, no. 4, pp. 2133–2153, 2012. doi.org/10.1016/j.annals.2012.07.016. petter, s., straub, d., and rai, a. (2007). specifying formative constructs in information systems research, mis quarterly, 31 (4), 623-656. philemon, j, r, m. (2015). assessment of tourist’s perception and satisfaction of the tanzanian destinations. european scientific journal, 11 (13). faqeer muhammad, kifayat ullah & rehmat karim 74 rahman, m.s. (2012). exploring the tourist’s perception: the case of bangladesh. tourismos: an international multidisciplinary journal of tourism, 7 (1), 81-98. tapachai, n., & waryszak, r. (2000). an examination of the role of beneficial image in tourist destination selection. journal of travel research, 39 (37). doi: 10.1177/004728750003900105. tenenhaus, m.; vinzi, v.e.; chatelin, y.m.; lauro, c. pls path modeling. comput. stat. data anal. 2005, 48, 159–205. tukamushaba, e, k., xiao, h., & ladkin, a. (2016). the effect of tourists’ perceptions of a tourism product on memorable travel experience: implications for destination branding. ejthr, 7(1):2-12. doi 10.1515/ejthr-2016-0001. wong, j, y., & yeh, c. (2009). tourist hesitation in destination decision-making. annals of tourism research, 36 (1), 6–23. doi: 10.1016/j.annals.2008.09.005. journal of applied economics and business studies, volume. 6, issue 3 (2022) 1-22 https://doi.org/10.34260/jaebs.631 1 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x creating entrepreneurs via social media platforms and employment generation in nigeria egunjobi, t. adenike1*1, 1 economics department, university of lagos, akoka lagos state, nigeria abstract this paper explores the potential need and benefits inherent in the use of social media as an instrument for creating entrepreneurs and generating employment amongst youths in nigeria given the high youth unemployment rate. 92 youths who are social media users selected through survey responded to structured questionnaire via google forms. the data was analysed using statistical descriptive tools and chi square test. findings revealed that in nigeria, social media can promote on-line trading and create a source of livelihood for youths though, the enormous potentials of doing business through social media platforms locally and internationally are fairly explored. also, trading via social media significantly impact employment creation and such trading does not significantly increase profits and reduce cost of doing business. the study recommended a reliable, affordable telecommunication network and combined efforts of government and the private sector to facilitate online trading and training in ict entrepreneurship skills. keywords entrepreneurship, online training, social media, youth unemployment jel classification j24, o15, o31, e24 1. introduction advancement in information, communication and technology (ict) is rapidly transforming the international business setting from the traditional business practices to better, faster, far reaching and most effective means of doing business. this has led to the enormous opportunities for employment creation thereby reducing unemployment among the youths 1* aegunjobi@unilag.edu.ng mailto:aegunjobi@unilag.edu.ng egunjobi, t. adenike 2 who are the greatest users of icts. such opportunities include the use of mobile technologies, internet facilities, computing and social media platforms (oestreicher-singer and zdalmanson, 2013). social media are interactive web based technologies that fosters establishment and sharing of information, concepts and thoughts through virtual networks and communities (kietzmann and kristopher, 2011). users can create and share content electronically, quickly and easily which makes it very suitable for business purposes. according to global social media statistics (2022) the number of people who make use of social media by july 2022, is 4.7 billion, which buttresses the amazing potentials for business that can be harnessed through social media platforms especially in present times when covid 19 pandemic restrictions has limited and constrained the traditional means of doing business. statistics showed that by january 2020, total world population stood at 7.75 billion, out of which 5.19 billion are smart mobile users, 4.54 billion are internet users while 3.80 million are active social media users. obviously, billions of people globally make use of social media apps because of its wider reach, accessibility and prompt response, though its use as a business tool and employment creation remains not substantially explored (global social media statistics, 2021). in 2010, the world recorded a 12.6% youth unemployment rate, this is almost three times the 4.8% adult unemployment rate (undesa, 2011). youths that are under the age of 34 forms more than 60% of the population of nigeria and half of this youthful population are unemployed (nbs, 2021). with this high level of youth unemployment, the number of youths involved in e-commerce activities should undoubtedly be higher than witnessed. in nigeria, yearly, about 300,000 youths graduate from tertiary institutions, the problem is escalated considering the population growth rate, reduced private investment and reduced quality of education and skill acquisition (imoisi et al, 2017). also, the global financial crisis and the consequences of covid 19 pandemic on the labour market and the economy had aggravated the problem. these advance the need for a new avenue for employment creation. thus, a great army of unemployed youth evolves with some participating in various acts of crime and deviant behaviour (egunjobi, 2021). amidst all the personal, social, economic, political and security challenges posed by youth unemployment, most research works had been dedicated towards entrepreneurship training and skill development programmes as possible ways of reducing youth unemployment (egunjobi, 2021; okafor, 2019; adelekan et al., 2018; ekong and ekong, 2016 and aribaba et al., 2013) there exists a dearth of knowledge on the enormous potentials that can be journal of applied economics and business studies, volume. 6, issue 3 (2022) 1-22 https://doi.org/10.34260/jaebs.631 3 harnessed in solving the unemployment problem via social media platforms. this study thereby fills this gap as the pilot empirical study in nigeria by providing a means of curbing youth unemployment that has not been really explored in nigeria which is creating employment and doing business through the social media platforms which has enormous local and international benefits for the youths and the nation. this arises out of the high rise in unemployment especially amongst the youths, the business opportunities yet untapped in the growing social media market as evident in the small percentage of those involved in on line trading compared to the size of the population and the need to redirect the enormous time spent by the youths on various social media platforms for business purposes, training and sustainable development. thus, the questions asked in this study are, what is the effect of trading via social media platforms on employment creation amongst youths in nigeria? is trading via social media platforms profitable and reduce cost of doing business? and what is the influence of online training programmes on employment creation? this paper test the hypotheses of a significant relationship between entrepreneurship trading via social media platforms and employment creation and that trading via social media platforms significantly increase profit and reduce cost of doing business using descriptive tools and chi squares analysis. the above is the introductory part of this paper, followed by review of literature, then theoretical framework, methodology, analysis and discussion of result and then, summary, conclusion and recommendations. 2. literature review 2.1 review of concepts entrepreneurship entrepreneurship is a dynamic progression of creation, vision and change. the person in charge of arranging and running an enterprise is called an entrepreneur (akanwa and agu, 2005). entrepreneurship requires the generation and implementation of new ideas and solutions and the commonest type of entrepreneurship occurs when a person starts a business (adelekan and dansu, 2016). hence, it can be concluded that an entrepreneur is that person who takes advantage of business prospects despite limited resources available in order to make profit. he could be employed or self-employed. unemployment a situation where a person is without a job is called unemployment. as a result, the unemployed are those actively searching for jobs and unable to secure one over a specified period of time (olubukola, 2013). according to the national bureau of statistics (2009), “the egunjobi, t. adenike 4 unemployed are the individuals with no work, but are looking for work”. globally, the rate of unemployment seems to be rising, even in developed nations. statistics in nigeria showed that 65.5% of the youthful population are unemployed, this represents more than half of the youths in nigeria. also, between 2014 and 2020, youth unemployment averaged 25.87% recording the highest percentage of 53.40% and the lowest percentage of 11.7% in the fourth quarter of 2020 and 2014 respectively (nbs, 2020). social media social media is a mixture of two words; social, which means exchange of information and media which is the medium of communication. it involves series of websites and applications that enables human beings to distribute content like links, pictures, posts, texts, etc. quickly, on time and efficiently (roopak, 2021). social media is an internet based technology that is interactive and user generated content which accelerates the exchange of thoughts and material through virtual systems (dollarhide, 2021) which can be used either by big or small firms as a means of creating and improving business activities (rosenbaum and cronin, 2013). the medium of connectivity is referred to as social media platforms. online training programmes online education is described as when teachers and learners are separated from each other while learning is taking place, this is distinct from self-study and face to face instruction. it uses computer to exchange and transfer information amongst business operators thus facilitating business transactions (keegan, 1988). online learning according to benson (2002) is an improved type of distance learning which produces better assess to educational opportunities and has greater flexibility and connectivity to promote wide-ranging interface (hiltz and turoff, 2005 and ally, 2004). social media, employment opportunities and national development the earliest forms of social media were newspapers, magazines, television, radios etc. with limited forms of interaction and feedback, it then progressed to a means whereby family and friends could interact without local or international boundaries and was later adapted for business purposes. in business, it is used to market goods, promote brands, connect to customers and foster new businesses. build customer confidence as they can easily interact with the sellers, helps business to respond rapidly to positive and negative feedback (fadi and racha, 2012) social media with ict is not only a source of career guidance but also potential for new employment and prospects. also, its contribution to economic growth is high and has assisted in reducing unemployment, creating better employment conditions and guaranteeing job journal of applied economics and business studies, volume. 6, issue 3 (2022) 1-22 https://doi.org/10.34260/jaebs.631 5 security (undesa, 2011; fadi and racha, 2012). again, usage of social media platforms and technology promotes reduction in cost of doing business and educational investment and reforms though favorable business environment is required (world economic forum, 2011and ilo, 2011) the crucial support given by icts to the promotion, progression and expansion of ebusiness through social media platforms cannot be quantified. social media is used as promotional tools, for marketing, public relations, advertising, and foundation for techstart up in the work place and support the learning of entrepreneurial skills in the classrooms (fadi and racha, 2012). governments also encourage entrepreneurship via social media because it helps to foster development of the economy (nelson et al., 2014) such as cultural, human, social, political and economic development (oso , 2011) and is a market place of ideas and creativity which promotes business and development of the economy (danbatta, 2017). however, government must be ready to provide the requisite infrastructure, training and skills needed to promote such investments (unctad, 2012). other benefits as advanced by unctad, (2012) are; business costs are reduced, communication is improved and higher degree of transparency is promoted. thus, to better tap these benefits the private and public sectors must work together to establish an ict driven investment and business environment. essentially, social media recorded over 90% growth from 2000 to 2008 and this upward trend had continued in nigeria to 2016. again, social media fact sheet (2021) revealed that users of social media are youths of which 90% are between the ages of 18 and 29, are better educated and earned over $75,000 per year. as ayo et al., (2011) discovered that though many of the websites in nigeria received patronage however, most customers use the business websites to access information as concerns comparison of prices and brands but hardly make purchases online. 2.2 empirical literature as observed, the causes and consequences of unemployment in nigeria are detrimental to individual and national development. the paper by kayode et al., (2016) identified bad governance, deterioration of industries, wide scale corruption and unfavourable business environment as the factors liable for high unemployment in nigeria and the resultant implications include youth restlessness, poverty and high societal evils. the paper recommended urgent intervention in power, industry, and agricultural sectors of the economy. in addition, adebayo, (1999) recognised poverty and food insecurity and this resorts to various unlawful behaviors which distorts the nation’s economic progress because it lowers standard of living, breeds anxiety about the future, threat to peace and stability and encourages egunjobi, t. adenike 6 brain drain (egunjobi, 2014). egunjobi (2021) study on nigeria described unemployment as the root cause of insecurity and poverty using granger causality test, in the same vein, tourtou (2018) discovered that the root cause of poverty and deviant behaviors in nigeria is unemployment, especially amongst the youths. the paper advised that rather than spending huge money on military artillery and personnel, unemployment should be seen as a collective problem and tackled with emphasis on the role of the mass media, provision of industrially friendly environment and restructuring of the educational system. evidences showed that a major way of reducing unemployment is through entrepreneurship education and training of youths physically and/or online. according to anyadike et al., (2012) the rising unemployment situation dwindles the potentials of the country and the way out is that government should make skills development for entrepreneurial training paramount, this can be achieved by funding and restructuring the educational system to promote employment and encourage practical/ online training and skill development (egunjobi, 2014). especially because of the mismatch between “gown and work” in nigeria, thus the need for training programmes to facilitate employment and development (olubukola, 2013) the study by ekong and ekong, (2016) in akwa ibom state on the national directorate of employment (nde) made use of descriptive survey and found that, there is a direct link between acquisition of skills by nde trainees and unemployment reduction. similar studies by charney and liecap, (2000) and luthje and frank, (2002) also discovered that a positive relationship existed between education and training and creation of businesses. in addition, khursheed (2012); aribaba, et al, (2013) and aribaba, et al., (2011) noted that entrepreneurship training programmes had direct impact on capital formation and employment using empirical data, but okafor (2019) study which assessed the effectiveness of government in reducing unemployment using the educational institutions of learning revealed that government has fallen short of expectation thus increasing running costs and adversely affecting training. since, entrepreneurship brings about self-sufficiency and economic prosperity for individuals and the economy at large. government should pursue it as an employment strategy because it does not only create jobs but also makes more people self-employed (idemobi, 2015) social media platforms, apart from creating and providing jobs have enormous advantages derivable. the enormous advantages of social media for business includes; it permits focus journal of applied economics and business studies, volume. 6, issue 3 (2022) 1-22 https://doi.org/10.34260/jaebs.631 7 on targeted marketing and outreach at minimal cost, promotes social connections, increases social networking and collaborations, creates new apparatus for businesses and facilitates quick distribution (economist, 2011 and needleman, 2010). however, reduced cost of marketing is the greatest advantage of social media (gupta 2021). in addition, communication is improved, business costs are reduced and higher degree of transparency is achieved (unctad, 2012) assessing the benefits of the use of social media through interview, wantchami (2020) revealed that marketing via social media had increased sales, generated awareness and improved customer relationships with benefits outweighing cost incurred. however, a limitation is the lower level of trust exhibited by consumers. empirically, leung et al., 2015; idota et al., 2017 and aladwani, 2015 alluded the use of social media as having a positive and significant effect on business sales. in the same vein, kumar et al., (2013) indicated that social media can generate greater revenue via strong customer relationships. jagongo and kinyua (2013) showed that when social media is properly harnessed it can promote brand and increase sales and profit., though economic conditions in nigeria, seems to make this difficult to achieve. however, the benefits of social media is not only local but international as asserted by bouranta et al., (2019) while highlighting key deliverables which are promotion of business and customer communication that help businesses to attain clients all over the world. although, most businesses especially in nigeria has not fully harnessed this international option. buttressing social media as a source of employment is being grossly under explored. urbiati et al., (2020) employing data gathered from youths on the effective use of social media for creating employment in italy, discovered that though the unemployed seemed to use social media basically for news and entertainment purposes, the employed use it majorly for social connectivity and not for business. the study by okande, (2015) concluded that the effect of social media in employment creation has been exaggerated, because kenyan youths largely use the social media for interaction socially and has not fully explored it as a means to participate in employment creation initiatives. the study however stressed that social media can potentially and significantly affect employment creation amongst youths. as opined by kane (2000) more economies are concerned about encouraging entrepreneurship via social media as avenue for job creation because during recession and instability, entrepreneurs /small businesses continue to create more jobs while reputable companies retrench employees. in a similar vein, uzochukwu et al., (2015) noted that egunjobi, t. adenike 8 entrepreneurship is the solution to most societal problems. thus, given the scarcity of literature on the use of social media as a means of creating entrepreneurs and because of the abundant unexplored opportunities in the use of the social media as employment generation and creation instrument in nigeria since it is still a growing market and yet to be fully tapped. this research thus intends to fill this gap by assessing the effectiveness of social media in employment creation and generation in nigeria. 3. theoretical framework and methodology 3.1 theoretical framework the entrepreneurial effect professor schumpeter has contributed in no small measure to entrepreneurship study. he is referred to as the first researcher to postulate about entrepreneurship, he contended that the invention and technological change of a nation come from the entrepreneurs, or wild spirits which is referred to as "entrepreneur-spirit". he affirmed that "... the doing of new things or the doing of things that have already been done in a different way stemmed directly from the efforts of entrepreneurs”. schumpeter argued against popular opinion and asserted that the drivers of innovation and economy progress are businesses which not only invest in research and development of new products and services but are also able to sell them at cheaper prices, thereby increasing standard of living. the procedure through which entrepreneurship undertakings decreases unemployment problem in a country is termed “schumpeter effect” (pull effect). research showed that unemployment is indirectly related to new firm startups, that is, employment is stimulated when new businesses are established and unemployment becomes substantively reduced (garofoli, 1994; andretsch and fritsch, 1994). oladele et al. (2011) noted that countries with lower degree of entrepreneurial activities, having low propensity to set up enterprises, recorded high rate of unemployment. the reason for this development is because of the deficiency in manpower capital and entrepreneurial skills required to effective mange such businesses/jobs and consequently they remain unemployed. thus to promote industrial growth and reduce unemployment, a certain level of manpower prowess is required (andretsch and fritsch, 1994). 3.2 data the survey research design was employed in this study because it enables the researcher to select respondents relevant to the research and provides effective data summary and journal of applied economics and business studies, volume. 6, issue 3 (2022) 1-22 https://doi.org/10.34260/jaebs.631 9 reporting with little observer subjectivity (cohen et al., 2007). google forms were sent out randomly to 115 online users adjudged to be youths via their online profile with an introductory letter specifying that only youths below 35 years, involved in online trading and who possess basic ict skills in nigeria should participate. the sample size consisted of 92 respondents (80% response rate) across the nation who met these criteria and correctly filled the questionnaire sent out via email and whatsapp. the research instrument used was the structured questionnaire by converse and presser (1986), modified by the researcher and was chosen because it requires low cognitive knowledge and provides a more responsive and wider coverage allowing same information to be gathered from the respondents (turner, 2010). it had two sections. the first focused on the demographic characteristics of the respondents and the second section focused on issues concerning trading using social media platforms, online training and employment creation in nigeria. the respondents provided answers based on their own opinions. the questions asked in the second section were measured using a 5 point likert scale. statistical methods used for analyzing data collected was descriptive statistics (frequencies and percentages) and inferential statistics (chi square test) was used in drawing inferences as well as conducting test of the hypotheses. 4. data analysis and discussion 4.1 demographic characteristics of the respondents table 1: demographic characteristics of the respondents variables frequency (f) percentage (%) gender male 40 43.5 female 52 56.5 total 92 100.0 age bracket below 29yrs 88 95.7 30 39yrs 4 4.3 40 49yrs 50yrs and above total 92 100.0 status self employed 28 30.4 joint owner employer 64 69.6 total 92 100.0 monthly profit / net income range n20,000 – n30,000 32 34.8 n30,000 – n40,000 4 4.4 n40,000 – n50,000 12 13 above n50,000 44 47.8 egunjobi, t. adenike 10 total 92 100.0 nature of business services 62 67.4 food and beverages 10 11 clothing and jewelry 12 13 electronics 2 2.1 others 6 6.5 total 92 100.0 source: author’s computation table 1 above depicted that 43.5% of the respondents analysed in the study are male and 56.5 are female. thus, the study captured more number of females than males. on the issue of age, 95.7% of the respondents are below age 29 while 4.3% of the respondents are between 30 and 39. this conforms to the requirement of the study, since it centers on youths. also, 30.4% of the respondents are selfemployed while a greater percentage, 69.6% are employees in entrepreneurship internet related business. on the profit / net monthly income earned 34.8% earn incomes between n20,000 and n30,000, 4.4% earn incomes between n30,000 – n40,000, 13% earn between n40,000 – n50,000 and 47.8% which are in the majority earn n50,000 and above. this depicts that with the right tools online trading is lucrative. on the nature of business, majority of the respondents (61.8% ) are in the service industry, followed by those in clothing and jewelry (13%) then, food and beverages (11%), others (6.5%) and electronics (2.1%). in this section the descriptive statistics of the variables of interest to the study are analysed using simple percentage table. the code used is strongly agree (sa), agree (a), neutral (n), disagree (d) and strongly disagree (d). table 2: the effect of social media on youth employment creation variables likert scale frequency (f) percentage (%) the spread of social media among the youths has largely promoted online trading as well as created more employment opportunities for the youth sa 62 67.4 a 30 32.6 n d sd total 92 100.0 the introduction of google adsense clicks through which youth are able to monetize their contents via adverts on youtube, instagram and blogs has helped create sources of livelihood for the youth sa 44 47.8 a 44 47.8 n 4 4.40 d sd journal of applied economics and business studies, volume. 6, issue 3 (2022) 1-22 https://doi.org/10.34260/jaebs.631 11 total 92 100.0 trading via social media platforms has not only created entrepreneurs but has also made the youths employers of labour sa 54 58.7 a 28 30.4 n 4 4.4 d 6 6.5 sd total 92 100.0 social media platforms has helped to boost the establishment of retail outlets requiring little capital and investment among the youths. sa 58 63 a 28 30.4 n 4 4.4 d 2 2.2 sd total 92 100.0 the establishment of customer services by organizations to enhance interaction between consumers and sellers through social media especially on twitter, facebook, whatsapp and other social media platforms has generated more employment opportunities for the youth. sa 46 50 a 42 45.7 n 4 4.3 d sd total 92 100.0 the adoption of social media platform in work places has created employment opportunities for youths with ict prowess sa 34 37 a 54 58.6 n 2 2.2 d 2 2.2 sd total 92 100.0 online trading has brought the opportunity to further penetrate local and international markets, thereby creating more employment. sa 52 56.6 a 38 41.3 n d 2 2.1 sd total 92 100.0 source: author’s computation as concerns the question, if the spread of social media among the youths has largely promoted online trading and created more employment opportunities for the youth, majority of the respondents (67.4 %) strongly agreed while (32.6%) respondents agreed. this asserts the role of social media as an employment generation tool (idemobi, 2005). the introduction of google adsense clicks through which youth are able to monetize their contents via adverts on youtube, instagram and blogs has helped created sources of livelihood for the youth, 47.8% of the respondents strongly agreed and also agreed. concerning, if trading via social media platforms has not only created entrepreneurs but has also made the youths employers of labour, more than half of the respondents (58.7%) strongly agreed. finding out if social media platforms have helped to boost the establishment of retail outlets, a larger percentage egunjobi, t. adenike 12 of the respondents (63%) strongly agreed. the establishment of customer services by organizations to enhance interaction between consumers and sellers through social media platforms has generated more employment opportunities for the youth, half of the respondents (50%) strongly agreed while (45.7%) agreed to this assertion. asking if the adoption of social media platform in work places has created employment opportunities for youths with ict prowess, majority of the respondents, 58.7% agreed. online trading has brought the opportunity to further penetrate local and international markets, (56.5%) of the respondents strongly agreed and (41.3%) agreed, this is collaborated by bouranta, (2019). thus, it can be concurred that trading using social media platforms has not only created job opportunities, but it has made youths not only employees especially in customer care organizations which uses ict but made them employers of labour as well. it has become a source of livelihood for youths and created opportunities to penetrate both local and international markets though this option has not been fully explored. table 3: profitability and cost of doing business via social media platforms variables likert scale frequency (f) percentage (%) online trading has brought the opportunity to further penetrate the markets, greater visibility, increased customer base and thereby creating more employment and% generate greater profits sa 50 54.3 a 34 37 n 6 6.6 d sd 2 2.1 total 92 100.0 businesses are becoming more innovative, convenient and flexible to operate with electronic products and services both for the buyers and sellers. sa 52 56.5 a 32 34.8 n 6 6.6 d sd 2 2.1 total 92 100.0 online trading when compared to non online trading reduces overhead costs like renting of shop, office space, cleaning services, buying of office equipment and furniture, waste of time, etc sa 50 54.3 a 26 28.3 n 14 15.2 d sd 2 2.2 total 92 100.0 online trading gives immediate feedback from customers which helps in building and retaining customers leading to more profit sa 42 45.7 a 34 37 n 12 13 d 2 2.1 journal of applied economics and business studies, volume. 6, issue 3 (2022) 1-22 https://doi.org/10.34260/jaebs.631 13 sd 2 2.1 total 92 100.0 many of these businesses are informal and thus reduces cost of doing business like payment of multiple taxes, stress of incorporation, hassle of government inspection of business and bureaucracy, difficulty in accessing credit facilities etc sa 44 47.8 a 34 37 n 8 8.7 d 4 4.4 sd 2 2.1 total 92 100.0 online trading eliminates the problem of physically attending to customers which limits the scope of business, thus reduces the cost of marketing and ensures more profit sa 44 47.8 a 28 30.4 n 14 15.2 d 2 2.2 sd 4 4.4 total 92 100.0 source: author’s computation responding to the question, whether, online trading has brought the opportunity to further penetrate the markets, greater profits and so on, 54.3% respondents strongly agreed. are businesses becoming more innovative, convenient and flexible to operate using electronic products and services, 56.5% strongly agreed while 34.8% agreed to this assertion? online trading when compared to non online trading reduces overhead costs, 54.3% strongly agreed. 45.7%, 47.8% and 47.8% respondents all strongly agreed that online trading gives immediate feedback from customers, reduces cost and ensures more profit respectively. thus, though many online businesses may be informal and small scale, benefits derivable include reduction in the overhead costs and cost of doing business, greater clientage, visibility and outreach to customers locally and internationally, generates employment and greater profits as ascertained by needleman, (2020); aladwani, (2015) and idota et al., (2016) table 4: online training programmes and employment creation variables likert scale frequency (f) percentage (%) e-learning/training affords youths the opportunity of becoming selfemployed and starting their own businesses sa 52 56.6 a 32 34.8 n 4 4.3 d sd 4 4.3 total 92 100.0 online training programmes equip youths with technical skills and business acumen required to manage business sa 40 43.5 a 30 32.6 n 14 15.2 d 4 4.3 egunjobi, t. adenike 14 sd 4 4.3 total 92 100.0 e-learning offers a more efficient, accessible, convenient relaxed and flexible way of training. sa 21 45.7 a 16 34.8 n 8 17.4 d sd 1 2.1 total 92 100.0 online training programmes have improved the technical skills and knowledge of the youths in ict proficiency making them more employable. sa 21 45.7 a 22 47.8 n 2 4.3 d sd 1 2.1 total 92 100.0 online line training has enhanced the entrepreneurship ability of the youths beyond that which has been provided by the secondary and tertiary educational institutions sa 24 52.2 a 16 34.8 n 3 6.5 d 2 4.3 sd 1 2.1 total 92 100.0 author’s compilation regarding the enquiry if e-learning/training affords youths the opportunity of becoming selfemployed, more than half of the respondents (56.34%) strongly agreed as attested to by idemobi, (2015). also, (43.5%) strongly agreed and (32.6%) agreed that online training programmes equipped youths with technical skills and business acumen. again, (45.7%) strongly agreed and (34.8%) agreed that e-learning offers a more efficient, accessible, convenient and flexible way of training. this is in line with the work of ally, (2004). online training programmes have improved the technical skills and knowledge of the youths in ict proficiency making them more employable, (45.7%) strongly agreed and (47.8%) agreed to this assertion. on, if online line training has enhanced the entrepreneurship ability of the youths beyond that which had been provided by secondary and tertiary educational institutions, (52.2%) strongly agreed and (34.8%) agreed. as concerns inquiry on the three most preferred social media platform for online trading, they are, in order of preference; instagram, whatsapp and facebook. it is thus affirmed that online training programmes, especially in ict proficiency are required and necessary, so youths can acquire skills needed to start and manage business and/ or make them more employable. also, in order to turn their hobbies into profitable ventures. journal of applied economics and business studies, volume. 6, issue 3 (2022) 1-22 https://doi.org/10.34260/jaebs.631 15 chi square results: chi square was employed to investigate the hypotheses. the chi square was chosen because of its validity as the instrument to evaluate test of independence, ease of computation and the robustness of analysing results with respect to distribution of data. hypothesis i: there is no significant relationship between entrepreneurship trading via social media platforms and employment creation. the variables tested are trading via social media as the independent variable and employment creation as the dependent variable as specified in table 2. table 5: chi square result on hypothesis 1 observed frequency expected frequency error chi-sqr stat. 50.0 18.4 31.6 121.69 37.7 18.4 19.3 2.6 18.4 -15.8 1.7 18.4 -16.7 0.0 18.4 -18.4 92.0 92.0 2 cal x = 121.69, 2 tab x = n-1, α 2 tab x = 92-1, 0.05, which is 91= 114.27 the result of the chi-square test presented in table 5, depicted that the value of chi-square calculated is greater than the value of the chi-square critical value at 5% significance level. hence, the null hypothesis is rejected. this means that entrepreneurship trading via social media platforms has a significant impact on employment creation in nigeria. hypothesis ii: trading via social media platforms (online trading) does not significantly increase profit and reduce cost of doing business. the variables tested are trading via social media as the independent variable and profitability and cost of doing business as dependent variable as specified in table 3 table 6: chi square result on hypothesis 2 observed frequency expected frequency error chi-sqr stat. 47.6 18.4 29.2 88.92 30.8 18.4 12.4 10.4 18.4 -8.0 0.8 18.4 -17.6 2.4 18.4 -16.0 92.0 92.0 2 cal x = 88.92, 2 tab x = n-1, α 2 tab x = 92-1, 00.5, which is 91= 114.27 from the result of the chi-square test presented in table 6. the result of the chi-square test egunjobi, t. adenike 16 depicted that the value of chi-square calculated is lesser than the value of the chi-square critical value at 5% significance level. hence, the fail to reject the null hypothesis is upheld. thus, it is asserted that trading via social media platforms does not significantly increase profit and reduce cost of doing business in nigeria. this is contrary to some research findings (idota et al., 2017; jagongo and kinyua, (2013) and kumai et al., (2013) which discovered the existence of a significant relationship. this could be as a result of the high cost of electricity and data in nigeria. summary, conclusion and recommendations summary the study inspected trading entrepreneurs via social media platforms impact on employment creation in nigeria using a statistically descriptive survey with the aid of the questionnaire and chi square tests. the study revealed that many of the respondents are females, youths below age 29, employed in internet related businesses and earning n50,000 and above as profit / net monthly income. it also discovered that social media platforms and ict training has created many job opportunities for the youths as entrepreneurs either as employees or employers of labour (selfemployment inclusive). trading via social media platforms is not only a lucrative business but it’s also very flexible, innovative, provides greater clientage, reduces stress, overhead costs and cost of doing business, enhances visibility of the business and extends reach to customers locally and internationally. the three foremost social media platforms used for trading are instagram, whatsapp and face book. finally, the study established that there is a significant impact of entrepreneurship trading via social media platforms on employment creation in nigeria, though social media platforms does not significantly increase profit and reduce cost of doing business in nigeria. conclusion given the rising rate of unemployment in nigeria and the aggregate time spent by youths on social media platforms coupled with the fact that the greatest choice of business investment is to invest in your passion/ hobby, there is great potentials and profit to be attained if many youths can turn the time and money used on social media platforms from tool of recreation, waste of time and human resources, perpetuating crime and gambling to a means of creating journal of applied economics and business studies, volume. 6, issue 3 (2022) 1-22 https://doi.org/10.34260/jaebs.631 17 employment and wealth especially in nigeria where this kind of business opportunities are yet to be fully exploited. in short, trading via social media platforms is an opportunity for youths to turn their hobby into employment and wealth creating profitable ventures ultimately reducing unemployment and enhancing economic development. recommendations the study has discovered that online trading is profitable, in order to further harness these potentials there needs to be good telecommunication network and better power supply in the country and if possible at the minimum price. this is also necessary to enable youths assess the various elearning training platforms. government must educate and provide alternatives required to divert youths from making use of social media platforms for illegal and fraudulent practices to legal business. such education can be achieved by organizing free seminars and workshops for youths, depicting the enormous benefits of legal business online transactions. such alternatives can be through rendering financial and material assistance. furthermore, government need to provide more educational opportunities more than been presently offered in terms of entrepreneurship skills, training and ict proficiency in our institutions of learning so as to harness the enormous benefits of online trading in providing employment. this presupposes that much more is required from government in terms of training in ict and development of entrepreneurship skills and this should be the collective efforts of government, private sector, voluntary and international organizations so as to get more youths involved in business. majority of the online businesses are in the service sector, there is the need for government to create means of encouraging nigerian produced and manufactured tangible goods online, this if developed will not only generate more employment but will open up our market potentials and the much needed foreign exchange as online business makes international trading easy, flexible and convenient. egunjobi, t. adenike 18 references adebayo, a. (1990) ‘youth unemployment and national directorate of employment. selfemployment programmes’, nigerian journal of economics and social studies, 41(1) adelekan s., majekodunmi, s and a. omotayo (2018) ‘impact of entrepreneurship development programme on economic development of lagos state’, iosr journal of business and management, 20(9) ver. 111, pp. 74-81. www.iosrjounals.org adelekan, s. and o. dansu (2016) ‘impact of entrepreneurship ventures on development of small and medium enterprises in nigeria: case study of three senatorial district in osun state’, nigeria. international journal of economics, commerce and management, 6(9), pp. 423441 akanwa, b. and c. agu (2005) entrepreneurship: theory and practice in nigeria. first edition, resource development centre, owerri. aladwani, a. (2015). facilitators, ‘characteristics, and impacts of twitter use: theoretical analysis and empirical illustration’, international journal of information management, 35(1), pp. 15-25. andretsch, d. and m. fritsch (1994) ‘on the measurement of entry rates’, empirical 21, pp. 105-113 aribaba f. (2013) ‘the examination of the effect of participation on entrepreneurial development programmes on enterprise performance: a case of small businesses in nigeria’, mediterranean journal of social science, 4(1), pp. 237242. aribaba f., ashaolu, t. and o. olaopa (2011) ‘an evaluation of the impact of technological innovative entrepreneurial development programmes on the performance of small scale businesses in nigeria’, global journal of business management and accounting, 1(1), pp. 001009 ayo, c., adewoye, j. and a. oni (2011) ‘business-to-consumer e-commerce in nigeria: benson, a. (2002) ‘using online learning to meet workforce demand: a case study of stakeholder influence’, available at www.researchgate.net/publication/234660106 bouranta, n., tsampra, m and sklavos, m (2019) ‘digital practices of greek small entrepreneurs: social media and self employment’, 10 pages available at www.academia.edu/33357972 cesaroni, f.and d. consoli (2015) are small businesses really able to take advantage journal of applied economics and business studies, volume. 6, issue 3 (2022) 1-22 https://doi.org/10.34260/jaebs.631 19 of social media? the electronic journal of knowledge management, 13(4), pp. 257-268. charmey, a. and libelap, d. (2000) ‘impact of entrepreneurship education: an evaluation of berger entrepreneurship programme at the university of avizona, 1985 – 1999. the kauffman centre for entrepreneurial leadership, kanbas city. cohen, l., manion, l. and k. morrison (2007) research methods in education. routledge, london. converse, j and s. presser (1986). survey questions: handcrafting the standardized questionnaire. sage publisher, london. danbatta, u. (2017) role of social media in economic, political and social development of nigeria. first annual lecture of ckn news in lagos on june 13, 2017. dollarhide, m. (2021) ‘social media’, available at www.investopedia.com economist (2011) ‘too much buzz social media provides huge opportunities, but will bring huge problems’, the economist. egunjobi, t adenike (2014) ‘poverty and unemployment paradox in nigeria’, iosr journal of humanities and social science (iosr-jhss), 19(5), ver. iv, pp. 106116. www.iosrjournals.org egunjobi, t adenike (2021) ‘poverty, unemployment and insecurity challenges in nigeria’, tanzania economic review, 11(1), pp. 115-136. ekong, u. and c. ekong (2016) ‘skills acquisition and unemployment reduction in nigeria: a study of national directorate of employment (nde) in akwa ibom state’, international journal of economics and management sciences, 5(4), pp. 210. fadi, s and m. racha (2012) social media, employment and entrepreneurship; new frontiers for the economic empowerment of arab youth. united arab emirate: dubai school of government’s governance and innovation programme. garofoli, g. (1994) ‘new firm formation and regional development: the case study of italy’, regional studies, 28, pp. 381-393 global media statistics (2022) ‘global media statistics research summary’, available at www.smartinsights.com gupta, r. (2021) hureka technologies inc. available at www.hurekatek.com. idemobi, e. (2015) reducing unemployment rate in nigeria: the role of entrepreneurship. 1st faculty of management sciences public lecture. igbariam egunjobi, t. adenike 20 campus: chukwuemeka odumegwu ojukwu univeristy. idota, h., bunno, t and m. tsuji (2017). ‘the effectiveness of social media use in japanese firms’. the review of social networks strategies, 11(1). imoisi, a., amba, e. and i. okon (2017) ‘unemployment rate and economic growth in nigeria: an empirical analysis, 1980–2016’, international journal of development and sustainability, 6(7), pp. 369-384. international labour organisation, (2011) statistical updates on arab states and territories and north african countries. international labor organization. jagongo, a. and c. kinyua (2013) ‘the social media and entrepreneurship growth: a new business communication paradigm among smes in nairobi’, international journal of humanities and social science, 3, (10), pp. 213-227. kane, t. (2010) the importance of startups in job creation and job destruction. kauffman foundation research series: firm formation and economic growth. ewing marion kauffman foundation. kayode a., arome, s and s, anyio (2016) ‘rising unemployment in nigeria: the social – economic and political implications’, global business and economic research journal 3(1), pp. 68 – 89 khursheed, a. (2012) ‘cost benefit analysis of entrepreneurship development programs in j & k state’, annual summit on business and entrepreneurial studies (asbes 2011) proceeding 043 – 051 kietzmann, j. and h. kristopher (2011) ‘social media? get serious! understanding the functional building blocks of social media’, business horizons, 54(3), pp. 241 251. kumar, v., bhaskaran, v and r. mirchandani (2013) ‘creating a measurable social media marketing strategy: increasing the value and roi of intangibles and tangibles for hokey pokey’, marketing science, 32(2), pp. 194-212. langcuster, j. (2010). ‘social media changing extension work’. southeast farm press, 37(22), pp. 1723. leung, x., bai, b and k. stahura 2015. ‘the marketing effectiveness of social media in the hotel industry’. journal of hospitality & tourism research, 39 (2), pp. 147-169. luthje, c. and n. fran (2002) ‘fostering entrepreneurship through university education and training. lessons from massachusetts institute of technology’, proceeding of journal of applied economics and business studies, volume. 6, issue 3 (2022) 1-22 https://doi.org/10.34260/jaebs.631 21 the european academy of management, stockholm, sweden, 9 – 11 may. national bureau of statistics, (2009) and (2021). social statistics in nigeria. abuja: the nbs publication. needleman, s. (2010) entrepreneurs question value of social media. wall street journal. nelson, o., kehinde, o., oyero, o., williams, o., and o. soola (2014) ‘effective use of information sources for breast cancer care: interplay of mass media and interpersonal channels’, journal of african research in business and technology, 3 (1), pp. 1-12 oestreicher-singer, g. and l. zalmanson (2013) ‘content or community? a digital business strategy for content providers in the social age’, management information systems quarterly, 37(2), pp. 591-616. okafor, c. (2019) ‘reducing unemployment rate in nigeria through entrepreneurship development: a study of selected small businesses in anambra state’, european journal of research and reflection in management sciences, 7(1), pp. 21-34. okande, j. (2015) ‘the influence of media in creating awareness of job uon repository’, available at http:// repository.uonbi.ac.ke olubukola, s. (2013) ‘unemployment and security challenges in nigeria’, international journal of humanities and social science, 3(7), pp. 60-69. oso, l. (2011) mass media and society in nigeria. 2nd edition, african books collective, oxford: uk. prospects and challenges’, african journal of business management. 5(13) rosenbaum, h. and b. cronin (2013) ‘digital entrepreneurship: doing business on the information superhighway’. international journal of information management, 13(6), pp. 461463. social media fact sheet (2021) social media. a publication of pew research center. available at www.pewresearch.org turner, d. (2010). ‘qualitative interview design: a practical guide for novice investigators’. the qualitative report, 15(3), pp. 754 – 760. unctad, (2012) information economy report 2012. in report no unctad/ier/2012. united nations conference on trade and development undesa (2011) united nations world youth report. united nations. urbinati, a., kalimeri, k., bonanaomi a., rosina, a and c. cattuto (2020) unemployment egunjobi, t. adenike 22 through the lens of social media. available at https://www.researchgate.net/publication/344603068_unemployment_through_th e_lens_of_social_media uzochukwu, o., orogbu, o and m. oyigbo (2015) ‘entrepreneurship development and job creation in selected local government areas in enugu state, nigeria’, international journal of managerial studies and research, 3(7), pp. 41-53 wantchami, n., ngange, k., and b. efosi (2020). ‘nexus of social media marketing and small scale businesses’ performance in buea, cameroon’. advances in social sciences research journal, 7(2), pp. 214-223. world economic forum (wef) (2011) economic growth and job creation in the arab world. geneva: world economic forum journal of applied economics and business studies, volume. 4, issue 1 (2020) 29-52 https://doi.org/10.34260/jaebs.412 29 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x examining the relationship between banking competition and solvency, liquidity and credit risks in pakistan sheikh muhammad umer farooq1 & muhammad zubair mumtaz2,* 1 school of social sciences & humanities, national university of sciences & technology, islamabad 2 associate professor, school of social sciences & humanities, national university of sciences & technology, islamabad abstract this study empirically examines the relationship between the banking competition and the risks faced by the financial sector (i.e. solvency, liquidity, and credit risks) considering 31 banks for the period 2001 to 2018. banks are further sub-divided into three categories i.e. state-owned banks, foreign banks, and private/commercial banks. the results reveal that pakistan’s banking industry is relatively elastic and an increase in competition is directly associated with solvency risk, liquidity risk and credit risk of financial institutions and these findings corroborate the competition fragility theory. besides, state-owned banks have a lesser probability to cope with solvency risk, however, foreign banks appear to face the least liquidity risk whereas private banks appear to face the least credit risk among the entire cluster. keywords banking competition, market power, solvency risk, liquidity risk, credit risk, pakistan. jel classification g21, g28, k22, l11 1. introduction the financial sector acts as a backbone of the economy that controls the money supply and plays a pivotal role in the financial development and economic growth of an economy. banks usually take funds from surplus monetary assets and lends as advances. industrial, agricultural and economic development would not be possible without an efficient banking system (babar, 2011). banking ordinance (1962) of pakistan defines banking as “accepting for lending or investment of deposits of money from the public, repayable on demand or otherwise and withdrawal by cheque, draft, order or otherwise”. in this competitive world, the banking industry faces various risks either external or internal and it becomes difficult for the banks to compete within and outside the system (beck et al., 2013). external risks arise due to difficulty in making enough returns on investments/equity, regulatory pressures issued from the central authorities, and failure in delivering the services that customers demanded like the use of advanced technologies. however, the increasing popularity of financial technology (fintech) companies who __________________ * corresponding author: zubair@s3h.nust.edu.pk sheikh muhammad umer farooq & muhammad zubair mumtaz 30 provide financial services to customers like the banking system but are not banks further poses a threat to the banking industry. internal risks come in the form of competition within the banking system because they have to compete with each other to earn higher profits (thorsten et al., 2012). to increase profit, banks usually take some risk-taking initiatives which could have a severe impact within the banking industry and in the overall economy (zuzana et al., 2013). to address this issue, several questions arise like how does the banking competition affect the risks faced by the banks? (jimenez et al., 2008). how the policy-makers aim at an increase in competition among banks; affecting their risk-taking initiatives at an individual level and at system; as a whole? riaz (2009) argue the relationship between competition and stability in the context of market concentration, zuzana et al. (2015) analyze in the context of their impact on lending channels, and schaeck et al. (2006) examine that up to what level of competition is good for the economy under given circumstances. to protect this global issue, an international agreement was signed in december 1987 between central banking authorities (i.e. basal accord). this agreement contains two important clauses – a minimum risk asset ratio requirement and a minimum leverage requirement. the minimum risk asset ratio requirement suggests a set of procedures for deciding how much capital a bank should have to support its risk-weighted assets, offbalance sheet assets, straight lending, and bank’s domestic and offshores assets and operations. a minimum leverage ratio requirement is based on the bank’s capital that cannot be less than a certain proportion of total liabilities. to address the aforesaid ratios, capital adequacy sub-divided into credit risk, operational risk, market risk, liquidity risk, and legal risk. the rules and regulations to determine these risks have been set out under the basal accord. thus, central banks through their policies enforce the bis implementations and ensure that the commercial banks abide by those set of procedures1. among these risks, solvency, liquidity, and credit risks are of prime concern for the banking industry and such risks are inter-related. therefore, this study aims to determine the competitiveness of the banking industry in pakistan by analyzing the relationship between competition and the risks faced by the financial institutions using the sample of 31 banks in pakistan during the period lasting from 2001 to 2018. 2. a review of the earlier studies 2.1 banking competition the relationship between banking competition and financial sector stability has been widely argued in theoretical and empirical literature with contradictory views that are, competition fragility theory and competition stability theory (berger et al., 2006). 1 solvency: models, assessment and regulation (2005) by stanstrom (first edition) and a comparative assessment of basel ii/iii and solvency ii (2012) by gatzert & wesker. journal of applied economics and business studies, volume. 4, issue 1 (2020) 29-52 https://doi.org/10.34260/jaebs.412 31 competition fragility theory is based on the charter value hypothesis (keeley, 1990) which debated that an increase in banking competition leads to a decrease in lending rates and an increase in deposits rates that cause the bank’s profit margin to decrease which eventually decreases bank’s franchise value. this enforces that banks to take risk-taking initiatives, causing instability of the individual institution and the economy as a whole. on the other hand, the competition stability approach (klaus et al., 2006) predicts that an increase in banking competition results in a decrease in interest rates which implies better credit options for borrowers. this provides the ease to repay their debt obligation which reduces the risk of the overall loan portfolio and thereby improving the financial sector stability of the individual institution and the economy as a whole (elmas, 2010). ijaz et al., (2020) studied 38 european countries for 2001 to 2017 and concluded that banking competition in terms of boone indicator supports economic growth and increases overall financial stability. juan et al., (2020) explored the baltic countries over the period 2001 to 2014 and found inverse u-shaped relationship between competition in terms of lerner index and financial stability in terms of z-score. riaz (2009) measures the association between banking competition and financial sector stability of the banking industry in pakistan using a non-structural approach and reported that the banking sector of pakistan as a whole is consistent with a monopolistic competitive market structure. in a recent study, tahir et al. (2016) employed panzer and rose methodology and found that the banking industry of pakistan is operating under the characteristics of monopolistic competition. also, he confirmed that massive deregulations and financial liberalization contributed positively to improving the competition condition of pakistan’s banking sector. subsequently, mehmood et al. (2017) analyzes the banking industry using boone indicators and found that the banking industry of pakistan is more competitive. however, inefficient banks are impairing their market shares to efficient banks. in another study, faisal (2016) use panzer and rose methodology and suggested that liberalization and deregulation of pakistan’s banking industry is successful in providing a competitive environment. 2.2 solvency risk almarzoqi et al. (2015) report an inverse relationship between profitability and competition. this suggests that an increase in market competition leads to a decrease in bank lending rates, which eventually affects the profitability of banks. this also implies that banks having large market power are more profitable. if profitability is considered only as a key determinant of competition, then competition may hurt profitability and so may reduce solvency. allan et al. (2011) decompose the relationship between competition and capital stock into two parts: (a) micro-prudential structure envisages a minimum capital requirement for unexpected losses, thus, treating the capital stock as exogenous. in this case, sheikh muhammad umer farooq & muhammad zubair mumtaz 32 market competition does not affect the capital stock of a given bank. whereas, (b) some banks’ despite minimum capital requirement hold more than the regulatory requirements. considering this argument, it is conferred that price competition may affect the capital stock of a bank (increase in market competition reduces the bank profitability and retained earnings which may induce credit institutions to set high targets for capital ratios and to ensure accessibility of capital in bank’s distress. thus, if the bank is raising capital in the short run, an increase in competition may imply more capital stock. in this case, competition has two counteracting effects on the bank’s solvency; reduction in profit margins and an increase in the bank’s capital. hence, the overall result depends on which factor dominates the other, that is, if a fall in bank’s profit is small than the rise in bank’s capital, the overall impact of increasing competition is positive and vice versa. selvi et al. (2009) argues that the debate of a positive or negative relationship depends upon the construction of variables chosen for competition and solvency. each variable is equally important but has its pros and cons. kashif et al. (2016) examined that increase in loan growth along with competition was the result of insolvency for pakistani banks due to weak prudential regulation among competitors, asymmetric information of the borrowers, and underestimation of the risk of lending during credit booms. 2.3 liquidity risk liquidity risk refers that a bank may not fulfill its short-term debt obligations, either due to the non-accumulation of enough funds or securities or investments that cannot be converted into cash to obtain good market price. marcus et al. (2008) divide liquidity risk into the funding liquidity and market liquidity and established a link with margin requirement. if margin requirements are higher, funding liquidity will also be higher than market liquidity and vice versa. dian & moshe (2014) analyzed a bank's decision to borrow in interbank markets or sell assets and reported that banks prefer to borrow rather than selling assets because of lower chances of adverse selection on the overall portfolio than on some of its assets. on the market liquidity front, it is emphasized that higher liquid assets have low returns and illiquid assets have a higher return. as banks with low market power, which is high competition, may invest in highly liquid assets to increase the availability of funds in terms of their short-term borrowings and to forgo return benefits that investing in illiquid assets and vice versa. banks diversified their portfolio by investing in both local and foreign securities to get the advantage from the market (lapteacru, 2014). ahsen (2012) examined the market liquidity risk of pakistani banks during the period between 2004 and 2009 and found that liquidity risk significantly affects banking profitability whereas liquidity gap and npls are exacerbating liquidity problems. farhan & amir (2013) analyzed the liquidity position and identified that the liquidity of pakistani banks is positively associated with the policy rate and negatively related to inflation. journal of applied economics and business studies, volume. 4, issue 1 (2020) 29-52 https://doi.org/10.34260/jaebs.412 33 2.4 credit risk beck et al. (2006) argued that surge in market competition causes a reduction in lending rates which distress the credit risk of the overall loan portfolio in two different perspectives: (a) decrease in lending rates enables the borrowers to fulfill their obligation which reduces the probability of default and increase the profitability thereby reducing the risk of the overall loan portfolio, and (b) decrease in lending rates reduces profitability from the provision of credit. this reduces franchise value and induces managers to take risk-taking initiatives by allowing credit to riskier borrowers, thus increases the average credit risk of the loan portfolio. emiliano et al. (2000) suggested that credit risk management and measurement is one of the areas where the difference between theory and practice is wider. according to emiliano, banks have diversified loan portfolios comprising different sectors of the economy, and fluctuations in one sector may affect the further availability of loans of a particular industry of the other sector. cheron (2013) reported that interest rate, net income, tax, inflation, and discount rate increase the npl ratio in pakistani banks. usman (2014) studied the relationship between loan growth and risk-taking behavior during the expansionary periods and identified that loan growth significantly affects npls and decrease the solvency of banks with a time lag of many years. besides, he reported that weak prudential regulation, the asymmetric information of the borrowers, and underestimation of lending risks in boom phases leads to higher npls. kashif et al. (2016) argued that an increase in loan growth correspondingly increases non-performing loans and increasing solvency risks for pakistani banks due to their weak prudential regulation. 3. banking competition in pakistan prior studies indicate that the pakistan banking industry operates under a less competitive environment (riaz, 2009; ayesha, 2012). since the last decade, there had been a decrease in corporate tax from 58% to 35% which enabled the banks with more market share to generate more revenue and this helped them to expand their scope of business and stability (bilal et al., 2011). the banking sector is one of the most important sectors of the economy not only in terms of employment opportunity but also in terms of its operations. since the 1980s, an increase in banking competition and an increase in nonperforming loans induces risks in the banking industry in pakistan. in pakistan, there is no significant research carried out by using the lerner index to assess the competition and relationship between competition and financial sector stability. furthermore, there is no significant research that decomposed the pakistani banking industry into state-owned banks, private banks, and foreign banks for research analysis. some scholars like alam et al. (2011) compared the financial performance of public and private banks. babar et al (2016) used the camel approach to evaluate the performance and stability of commercial banks in pakistan. usman (2014) made a comparison of islamic and non-islamic banks in terms of stability. riaz (2009) uses the hershman herfindahl index (hhi) to determine the stability of pakistan’s banking industry. sheikh muhammad umer farooq & muhammad zubair mumtaz 34 4. data and methodology 4.1 data we take the entire baking sector as a population, which includes state-owned, private, and foreign banks. to maintain symmetry and harmony among banking operations and comparability, islamic banks, agricultural banks, micro-finance banks, and other lending institutions are not considered in this study. the list of banks along with their date of establishment in terms of business operation is presented in appendix i. to circumvent the problem of heterogeneity and autocorrelation, panel fgls technique is employed to estimate our findings. this study employs 31 banks operating in pakistan and the data is gathered from their financial statements during the period from 2001 to 2018. banks are sub-divided into three categories which include 11 commercial banks, 5 state owned-banks, and 15 foreign banks. the era of the early 1970s comprises only five big banks because small banks were amalgamated into them due to the nationalization policy of that time. hereinafter, 1990 was a period of policy shift where banks were privatized and foreign banks were allowed to be operated in pakistan. later 2000 was a period with technical innovations and implementation of international accords, therefore, such period was a dream period to study the impact of competition on stability and the way they interact with each other along with the sector-wise in-depth analysis. 4.2 empirical models to examine the relationship of banking competition/market power on the financial sector stability at the institution-level, model (1) is used to evaluate the financial stability of a bank concerning various sources of risk (the solvency risk, liquidity risk, and credit risk of the assets portfolio). 𝑅𝑖𝑠𝑘𝑖𝑡 = 𝛼𝑖 + 𝛽1𝐶𝑜𝑚𝑝𝑒𝑡𝑖𝑡𝑖𝑜𝑛𝑖𝑡−1 + 𝛽2𝐵𝑎𝑛𝑘𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡−1 + 𝜀𝑖𝑡 … … … . (1) where i represents the bank and t indicates the time in the year. we segregate the risks of assets portfolio into solvency, liquidity, and credibility. the main explanatory variables comprised: price competition at the bank-level (the lerner index), and banking risks. the control variables include (a) loan to asset ratio (net loans/total assets), (b) government bonds ratio (government bonds/total assets), (c) asset growth (percentage change in assets), (d) non-performing loan ratio (non-performing loans/total advances), (e) non-interest income ratio (non-interest income/ total revenues. the independent variables are taken as one period lag because changes in the current year do not affect stability in the same period but appear after a lag (almarzoqi, et al., 2015). we include a dummy variable in equation (1) to examine the behavior of each sector on different risks. model (2) can be expressed as: 𝑅𝑖𝑠𝑘𝑖𝑡 = 𝛼𝑖 + 𝛽1𝐶𝑜𝑚𝑝𝑒𝑡𝑖𝑡𝑖𝑜𝑛𝑖𝑡−1 + 𝛽2𝐵𝑎𝑛𝑘𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡−1 + 𝛽3𝐷𝑢𝑚𝑚𝑦𝑖 + 𝜀𝑖𝑡 … . (2) equation (2) is estimated thrice separately by assigning 1 for state-owned dummy variable banks and 0 otherwise; 1 for foreign banks dummy variable and 0 otherwise and 1 for private banks and 0 otherwise. journal of applied economics and business studies, volume. 4, issue 1 (2020) 29-52 https://doi.org/10.34260/jaebs.412 35 𝑅𝑖𝑠𝑘𝑖𝑡 = 𝛼𝑖 + 𝛽1𝐶𝑜𝑚𝑝𝑒𝑡𝑖𝑡𝑖𝑜𝑛𝑖𝑡−1 + 𝛽2𝐵𝑎𝑛𝑘𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡−1 + 𝛽3𝐶𝑜𝑚𝑝𝑒𝑡𝑖𝑡𝑖𝑜𝑛𝑖𝑡−1 ∗ 𝐷𝑢𝑚𝑚𝑦𝑖 + 𝜀𝑖𝑡 ………………………(3) to extend our analysis, we add an interaction term by multiplying banking competition and each sector to determine how this variable influences the bank risk. in model (4), we include both the dummy variable and interacting dummy variable were incorporated to determine the combined impact of market power/competition on various risks. 𝑅𝑖𝑠𝑘𝑖𝑡 = 𝛼𝑖 + 𝛽1𝐶𝑜𝑚𝑝𝑒𝑡𝑖𝑡𝑖𝑜𝑛𝑖𝑡−1 + 𝛽2𝐵𝑎𝑛𝑘𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡−1 + 𝛽3𝐶𝑜𝑚𝑝𝑒𝑡𝑖𝑡𝑖𝑜𝑛𝑖𝑡−1 ∗ 𝐷𝑢𝑚𝑚𝑦𝑖 + 𝛽4𝐷𝑢𝑚𝑚𝑦𝑖 + 𝜀𝑖𝑡 ………………………………..(4) 4.3 variables earlier studies identified two main variables namely; banking competition and financial stability. in this study, banking competition is measured through the lerner index, and financial stability is estimated in the shape of solvency (z-score), liquidity (liquidity ratio), and credibility (non-performing loan ratio). a brief description of these variables is explained in the ensuing paragraphs. (a) bank competition: in the empirical literature, the banking competition is measured through the structural and non-structural approaches. structural measures are based on the competition that can be inferred from indirect proxies like market structure or market share whereas non-structural measures are measured directly. structural measures include hirschman herfindahl index (hhi) and cr5 (structure-conduct-performance paradigm) whereas non-structural measures comprise lerner index, panzar, and ross h-statistics and boone indicator. all these measures are based on the analysis of the effective behavior of institutions on the market (zuzana, 2013). the level of banking competition is analyzed using various measures, for instance, price competition, market contestability, and market concentration. lerner index is used as a proxy for bank competition and its value ranges between zero and one. the zero value specifies perfect competition and one shows pure monopoly. the lerner index is calculated as: 𝐿𝑒𝑎𝑟𝑛𝑒𝑟 𝐼𝑛𝑑𝑒𝑥𝑖𝑡 = 𝑃𝑖𝑡+ 𝑀𝐶𝑖𝑡 𝑃𝑖𝑡 ……………………. (5) whereas 𝑃𝑖𝑡 is the price of bank i at t time, 𝑀𝐶𝑖𝑡 is the marginal cost of bank i at time t. the price of banking products on a single output is determined as a proxy of the ratio of total revenues to total earning assets. total revenues comprise interest and non-interest income, and equity-accounted profit/loss operating income. 𝑃𝑟𝑖𝑐𝑒𝑖𝑡 = 𝑇𝑅𝑖𝑡 𝑇𝐸𝐴𝑖𝑡 … … … … … … … … … (6) sheikh muhammad umer farooq & muhammad zubair mumtaz 36 where 𝑇𝑅𝑖𝑡 is the total revenue of bank i at time t and 𝑇𝐸𝐴𝑖𝑡 is the total earning asset of bank i at time t. in this study, the total cost function is calculated by using a translog functional form (khalil, 2005, hector et al. 2007): 𝑙𝑛𝑡𝑐𝑖𝑡 = 𝛼0 + 𝛼1𝑙𝑛𝑞𝑖𝑡 + 𝛼2(𝑙𝑛𝑞𝑖𝑡 ) 2 + ∑ 𝛼3 3 𝑗=1 𝑙𝑛𝐼𝑗𝑖𝑡 . 𝑙𝑛𝑞𝑖𝑡 + ∑ ∑ 𝛼4 3 𝑘=1 3 𝑗=1 𝑙𝑛𝐼𝑗𝑖𝑡 . 𝑙𝑛𝐼𝑘𝑖𝑡 + ∑ 𝛼5 3 𝑗=1 𝑙𝑛𝐼𝑗𝑖𝑡 + 𝜀𝑖𝑡 ………………………………(7) where lntcit is the log of the total cost of bank i at time t, lnqit is the output (total assets) of bank i at time t and lnijit is the input price of bank i at time t whereas the value of j ranges from 1 to 3. j=1 refers input labor; j=2 for input physical capital and j=3 for input borrowed funds and k is the input prices of other banks. the price of labor is the unit price of labor which is calculated as the proportion of personnel expenses to the number of personnel in a bank for a given time. the price of physical capital is calculated as the ratio of other noninterest expenses to fixed assets. finally, the price of funds is the unit price of funds, and it is calculated as the ratio of interest expenses to the total deposit: (a) input labor = personal expenses/no. of persons, (b) input physical capital = non-interest expense/fixed assets, and (c) input funds = interest expense/total deposits (elmas, 2011). the marginal cost is derived from the total cost function by driving equation (7), w.r.t. lnqit as 𝑀𝐶𝑖𝑡 = 𝑑𝑙𝑛𝑡𝑐𝑖𝑡 𝑑𝑙𝑛𝑞𝑖𝑡 = 𝑡𝑐𝑖𝑡 𝑞𝑖𝑡 ∗ (𝛼1 + 2𝛼2𝑙𝑛𝑞𝑖𝑡 + ∑ 𝑙𝑛𝐼𝑘𝑖𝑡 3 𝑖=1 ) ………….(8) (b) solvency risk the solvency of a bank is generally measured by three parameters: profitability of the bank; availability of adequate capital stock; and economic conditions. the first two indicators are computed by z-score because it is calculated as the sum of equity-asset ratio (covering the first factor) and the return on assets (covering the second factor) divided by the standard deviation of return on assets (covering the change in the second factor). the third factor is the capital adequacy ratio (car) which is measured by the weighted average combination of on-balance sheet items, off-balance sheet items, market risk, credit risk, and operational risk. in this research, solvency risk is proxied by the z-score which is an indirect measure of the bank from insolvency. 𝑍 − 𝑠𝑐𝑜𝑟𝑒𝑖𝑡 = 𝐸 𝐴𝑖𝑡⁄ + 𝑅𝑂𝐴𝑖𝑡 𝑆. 𝐷. (𝑅𝑂𝐴𝑖𝑡 ) … … … … … … (9) where higher the value of z-score refers to better solvency of a bank. e/ait is the equity to asset ratio for bank i in year t, roait is the return on asset for bank i in year t, and s.d. (roait) is the standard deviation of return on asset. z-score shows an inverse measure of solvency risk which illustrates that higher value shows the lower solvency risk and vice versa. journal of applied economics and business studies, volume. 4, issue 1 (2020) 29-52 https://doi.org/10.34260/jaebs.412 37 (c) liquidity risk there are two broad measures of measuring liquidity risk (i.e. liquidity gap and liquidity risk elasticity). the liquidity gap means the net liquid assets of a firm that represent the excess value of the firm's liquid assets over its volatile liabilities (almarzoqi et al., 2015). liquidity risk elasticity is the change of net of assets over-funded liabilities that occur when the liquidity premium on the bank's marginal funding cost rises by a small amount. liquidity ratio is the ratio between liquid assets and short-term borrowing which illustrates the size of the liquid assets buffer of a bank; at its disposal and liquidity index which is calculated by discounting the present values of liquid assets (diana & moshe, 2014). 𝐿𝑅𝑖𝑡 = 𝐿𝐴𝑖𝑡 𝑆𝑇𝐵𝑖𝑡 … … … … … … . . (10) where 𝐿𝑅𝑖𝑡 is the liquidity risk of bank i for year t, 𝐿𝐴𝑖𝑡 is the liquid assets for bank i for the year t and 𝑆𝑇𝐵𝑖𝑡 is the short term borrowing of bank i for the year t. liquidity ratio is an inverse measure for liquidity risk, that is, higher the liquidity ratio, lower will be the liquidity risk. (d) credit risk for measuring credit risk, allen et al. (2011) classified four broad categories of measuring the credit risk i.e. external ratings approach, financial analysis models, the merton / kmv structural model, and the transition-based models of credit metrics and credit portfolio view and provide relevant pros and cons of each method. edward et al. (1998) measures the credit risk by establishing a new model known as mortality risk framework to measure the risk and returns on loans and analyze the risk-return structures of portfolios of credit-risk exposed debt instruments. the credit risk represents the loans created by the bank of the securities held on the balance sheet. a good measure for the asset portfolio risk is the non-performing loans (npls) ratio. 𝐶𝑅𝑖𝑡 = 𝑁𝑃𝐿𝑖𝑡 𝐺𝐿𝑖𝑡⁄ … … … . (11) where crit is the credit risk of bank i for year t, nplit is the non-performing loan of bank i for year t and glit is the gross loan of bank i for year t. npl is defined as a loan that is in default or close to being in default. credit ratio shows a direct link to credit risk i.e. higher the credit ratio higher will be the credit risk. 5. empirical results 5.1 descriptive statistics table 1 shows the descriptive statistics of the variable used in this study. lerner index (l) is used as a proxy for banking competition/market power had a minimum value of 0.468 (lerner index is calculated in short-run and the negative value of lerner index could be possible in short-run, (sofronis et al., 2013) and a maximum value of 0.581 with mean 0.12 and standard deviation 0.30. this shows that the banking industry in pakistan is sheikh muhammad umer farooq & muhammad zubair mumtaz 38 operated in a competitive environment. z-score is used as an inverse proxy for solvency measures had a minimum value of 0.2%, the maximum value of 71.3% with mean 30%, and standard deviation 24% showing overall better solvency position. liquidity ratio (lr) has a minimum value of 0.8%, maximum value 223% with mean 107%, and standard deviation 0.9 which predicts that assets of the banking industry are highly liquid. non-performing loan ratio (npl) has a minimum value of 3%, maximum value 30% with mean 9.9%, and standard deviation 0.08 indicating that the non-performing loans are less likely to occur in the banking industry of pakistan. loan to assets ratio (lar) has a minimum value of 15.3%, maximum value 65.9% with mean 42.3%, and standard deviation 0.43. non-interest income ratio (niir) has a minimum value of 7%, maximum value of 57.7% with mean 24% and standard deviation of 0.137. asset growth (ag) has a minimum value of -13.3%, a maximum value of 66.4% with an average growth of 17.7% and a standard deviation of 0.19. return on average equity (roae) has a minimum value of -37.7%, maximum value 98.4% with mean 24%, and standard deviation 0.298. government bonds ratio (gbr) has a minimum value of 0.3%, a maximum value of 88.3% with a mean 27% and a standard deviation of 0.146. the graphical representation of the variables for the sample period can be shown in appendix ii. for understanding, the graphs are further split into the overall national level, state-owned, private, and foreign banks. before regression, some further tests like cross-sectional dependence, panel unit root, serial correlation, and heteroskedasticity tests are conducted. the results of these tests shown in appendix iii, iv, v, and vi respectively. the results indicate that there is no unit root and cross-section dependency however, there is a problem of serial correlation and heteroskedasticity that could generate inconsistent coefficients with usual panel regression. therefore, the models are regressed using the fgls technique that corrects the heteroskedasticity and serial correlation of observations within a cross-section. for a large sample, fgls is preferred over gls under the presence of heteroskedasticity or serial correlation and all the properties of fgls are common as compared to gls (mantobaye, 2017). table 1: descriptive statistics variable mean median max. min. std. dev. obs. leaner index 0.120 0.159 0.581 -0.468 0.307 454 z-score 0.305 0.213 0.713 0.002 0.240 454 liquidity ratio 1.070 0.653 2.230 0.008 0.921 454 npl ratio 0.099 0.085 0.327 0.003 0.086 454 loan to asset ratio 0.423 0.437 0.659 0.153 0.128 454 non-interest income ratio 0.240 0.210 0.577 0.071 0.137 454 asset growth 0.177 0.146 0.664 -0.133 0.194 454 return on average equity 0.240 0.239 0.984 -0.377 0.298 454 govt. bond ratio 0.270 0.255 0.883 0.003 0.146 434 journal of applied economics and business studies, volume. 4, issue 1 (2020) 29-52 https://doi.org/10.34260/jaebs.412 39 this table presents the descriptive statistics of 31 pakistani banks during the period from 2001 to 2018. 5.2 effect of solvency risks on banking competition table 2 exhibits the empirical estimation between solvency risk and banking competition. model (1) shows that the lerner index increases the market power (decrease in market competition), increases the z-score (i.e. improves the solvency of banks). this finding is in line with the competition fragility theory which envisages that banks with more market power are less likely to take risks; increases profit margins and therefore increase stability (almarzoqi et al., 2015). the negative association of loans to assets ratio indicates that an increase in advances decreases stability through an increase in risk-taking initiatives of the banks. banks with more shares of government bonds among their assets show higher solvency, however, this study negates on the plea that despite government bonds are highquality assets but yield low return as compared to advances lent, so it is a rationale for a stable bank to go for riskier assets and get more return. thus, stable banks hold less share of government bonds in their asset’s portfolio. assets growth show expansion in banks’ balance sheet. in this context, banks disproportionately increase their assets (extend new credits to more risky borrowers that may increase solvency risk of banks showing a negative relationship with the z-score). npl ratio is directly associated with solvency risk (franklin et al., 2011), i.e., higher the npl ratio; higher will be the solvency risk (lower will be the z-score). non-interest income ratio (niir) comprises banks’ fees, services, etc., representing the diversified sources of banks’ income. whenever banks rely more on their non-interest income, it increases the borrowing cost of loans and unusual burdens on the customers who then search for the alternatives and withdrawing their investments thereby decreasing the z-score. model (2), (5), and (8) are estimated by incorporating the interacting dummy variable for state-owned, foreign, and private banks respectively. model (2) suggests that the effect of market power on state-owned banks is 16.5 times more than that of foreign and private banks. models (5) and (8) show that market power for foreign banks is 1.70 times less than state-owned and private banks and private banks have the market power of 5.3 times less than state-owned and foreign banks respectively. more market power for state-owned banks indicates more faith from the general public. furthermore, the government supervised the banks for public welfare, not for its profit motive. however, a decrease in market power for foreign and private banks mainly comprises due to lack of confidence by the general public and lack of branch operations especially by foreign banks who are confined to just big cities. sheikh muhammad umer farooq & muhammad zubair mumtaz 40 model (3), (6), and (9) are further classified as state-owned, foreign, and private banks to determine the impact of each sectors’ competition; on banks’ stability; by introducing dummy variables. model (3) indicates the coefficient for state-owned dummy variable to be 8.89 which means that for a given level of market power, the average solvency for stateowned banks will be higher by 8.89 times then foreign and private banks. similarly, models (6) and (9) show that for a given level of market power, the average solvency of foreign banks is decreased by 2.49 times and for private banks, it is decreased by 3.06 times respectively. state-owned banks have support from the government and enjoy better solvency irrespective of the market power. the decrease in solvency for private and foreign banks is due to strict rules and regulations from the central bank. models (4), (7), and (10) include the results of both dummy and interacting term which is multiplied by the dummy for state-owned, foreign, and private banks with learner index. model (4) predicts that on average state-owned banks have 6.29 times higher solvency by keeping overall market power constant and their market power is 14.14 times more than foreign and private banks. likewise, model (7) suggests that on average foreign banks have 1.68 times less solvency by keeping overall market power constant and their market power is 2.62 times less than state-owned and private banks. the results of model (10) present that on average commercial banks have 2.67 times less solvency by keeping overall market power constant and their market power is decreased by 3.5 times relative to state-owned and foreign banks. 5.3 effect of liquidity risks on banking competition table (3) demonstrates the relationship between liquidity risks and banking competition. in model (1), the positive coefficient of the lerner index indicates that an increase in market power (decrease in market competition) increases the liquidity ratio. this illustrates the existence of lower liquidity risk which is consistent with the competition fragility theory. this is because the banks with more market power have passive management style, that is, their liquidity buffer is a function of their cash flows and profits whereas banks with less market power have active management style i.e. their liquid buffer is a function of cash flows only (tend to adjust their liquid buffer according to the cash flow). the rationale for active management is that whenever there is a dire need for liquid assets, instead of access to the interbank market for obtaining loans and paying interest, they opt to use their cash flows to build their buffer of liquid assets (almarzoqi et al., 2015). banks’ liquidity position mainly depends both on the quality and quantity of lending activities. the negative relationship of lar shows that an increase in banks’ lending activities decreases the liquidity ratio and it is obvious that banks have either two options i.e. invest in illiquid assets (yields more return) and invest in liquid assets (yields low return). banks opt to choose a combination of these two but normally made more investment in illiquid assets to journal of applied economics and business studies, volume. 4, issue 1 (2020) 29-52 https://doi.org/10.34260/jaebs.412 41 generate more revenue thus decreases the liquidity ratio. the coefficient of niir is negative with liquidity ratio because income from non-interest sources may be important for liquidity but are rigid and less volatile. therefore, banks who rely more on non-interest income or have high non-interest earning may need to hold less liquidity. npl is directly associated with liquidity risk, that is, higher the npl ratio; higher will be the risk and vice versa, therefore, the negative sign indicates that whenever the npl ratio increases; banks tend to convert their liquid assets into ready cash and thereby decreases their liquidity ratio. government bond ratio is negatively associated with liquidity ratio because although the government bonds appear to be stable but yield less return therefore banks continue to invest in riskier bonds to get more return. assets growth show expansion in banks’ balance sheet because the increase in assets is positively related to an increase in liquidity ratio. model 2, 5, and 8 are estimated by incorporating the interacting dummy variable for state-owned, foreign, and private banks respectively. model (2) shows that effect of market power for state-owned banks is 0.15 times lower than that of foreign and private banks whereas market power for foreign banks is 0.15 times higher than state-owned and private banks and private banks have the market power of 0.13 times lower than state-owned and foreign banks (models 5 and 8). model (3) indicates the coefficient for state-owned dummy variable to be 0.4 which means that for a given level of market power, the average liquidity for state-owned banks will be higher by 0.4 times than foreign and private banks. for a given level of market power, the average liquidity of foreign banks is 0.23 times less than state-owned and private banks and private banks have 0.02 times more liquidity than stateowned and foreign banks (models 6 and 9). state-owned banks have more investment in government securities i.e. t-bills and pibs and are thus more liquid. low liquidity for foreign banks is mainly due to an increase in advances to increase their market power rather than investing in liquid assets. liquidity for private banks is less the state-owned banks because they adjust their liquidity buffer according to the given situation. model 4 shows that on average state-owned banks have 0.46 times higher liquidity by keeping overall market power constant and their market power is 0.46 times less than foreign and private banks. in model 7, on average foreign banks have 0.26 times less liquid by keeping overall market power constant and their market power is 0.31 times higher than state-owned and private banks. however, model 10 reflects that on average commercial banks have 0.05 times more liquidity by keeping overall market power constant and their market power is less than 0.21 times from state-owned and foreign banks. sheikh muhammad umer farooq & muhammad zubair mumtaz 42 5.4 effect of credit risk on banking competition table (4) displays the relationship between credit risk and banking competition we report an inverse relationship between the lerner index and market power (model i). we argue this evidence that banks with more market power are risk-averse and due to their already increased profit margin, they are reluctant to invest in risky assets. prior studies argue that an increase in banking competition decreases lending rate which influences the dominancy of either way (a decrease in lending rates makes it easy for the borrowers to pay back their debt hence, npl ratio decreases or decrease in lending rates decreases profit margins for the banks and to increase their profit margin, banks lent more advances even to riskier customers which could increase npl ratio, (almarzoqi, 2015). in model 2, the effect of market power for state-owned banks is 0.016 times more than that of foreign and private banks while the market power for foreign banks is 0.001 times more than state-owned (model 5) and private banks and private banks have the market power of 0.008 times less than state-owned and foreign banks (model 8). model (3) elaborates the coefficient for state-owned dummy variable to be 0.037 which means that for a given level of market power, the average npl ratio for state-owned banks will be higher by 0.03 times then foreign and private banks. model 4 shows that on average state-owned banks have 0.04 times more npl ratio by keeping overall market power constant and their market power is 0.029 times less than foreign and private banks. model 7 suggests that on average foreign banks have 0.018 times less npl ratio by keeping overall market power constant and their market power is 0.019 times more than state-owned and private banks. model 10 indicates that on average private/commercial banks have 0.012 times less npl ratio by keeping overall market power constant and their market power is increased by 0.006 times than state-owned and foreign banks. journal of applied economics and business studies, volume. 4, issue 1 (2020) 29-52 https://doi.org/10.34260/jaebs.412 43 table 2 effect of solvency risk on banking competition variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) lerner(-1) 4.759*** 2.033 4.068*** 1.945*** 5.492*** 4.804*** 6.199*** 6.709*** 5.609*** 6.793*** (2.9163) (1.0956) (2.4683) (1.0404) (2.3074) (2.8890) (2.517) (3.333) (3.352) (3.358) loan to asset ratio(-1) -29.833*** -29.277*** -27.878*** -28.057*** -30.007*** -29.913*** -29.887*** -29.648*** -26.921*** -27.183*** (-7.165) (-7.046) (-6.702) (-6.718) (-7.188) (-7.137) (-7.061) (-7.132) (-6.247) (-6.327) government bonds ratio(-1) 0.536 0.408 0.018 -0.044 0.305 -0.141 -0.318 0.446 1.739 1.246 (0.188) (0.142) (0.006) (-0.015) (0.105) (-0.048) (-0.105) (0.153) (0.579) (0.408) asset growth(-1) -2.573** -2.411* -2.260 -2.057 -2.526** -2.054 -1.772 -2.707** -2.077** -2.213 (-1.827) (-1.652) (-1.578) (-1.389) (-1.776) (-1.460) (-1.230) (-1.910) (-1.457) (-1.551) non-performing loan ratio(1) -11.163** -11.826*** -13.050*** -11.878*** -11.445*** -10.993** -10.466** -11.473*** -10.123** -10.554*** (-1.965) (-1.967) (-2.284) (-1.968) (-1.968) (-1.908) (-1.730) (-2.015) (-1.708) (-1.784) non-interest income ratio(1) -2.789 -2.305 -2.317 -2.773 -2.765 -3.421 -3.852 -2.589 -4.141 -3.797 (-0.960) (-0.758) (-0.799) (-0.909) (-0.946) (-1.169) (-1.281) (-0.882) (-1.384) (-1.269) lerner(-1)*state owned bank 16.507*** 14.142*** (3.515) (2.896) state owned banks 8.893*** 6.294*** (3.205) (2.184) lerner(-1)*foreign banks -1.705 -2.625 (-0.521) (-0.783) foreign banks -2.493 -1.683 (-1.174) (-0.799) lerner(-1)*private banks -5.300 -3.504 (-1.517) (-0.991) private banks -3.069 -2.673 (-1.412) (-1.217) constant 42.147*** 41.805*** 40.182*** 40.464*** 42.252*** 43.582*** 43.216*** 42.214*** 42.078*** 42.202*** (15.913) (15.587) (14.850) (14.655) (15.745) (14.423) (14.015) (15.858) (15.483) (15.568) t-values in parenthesis ***p<0.01, **p<0.05, *p<0.1 sheikh muhammad umer farooq & muhammad zubair mumtaz 44 table-3 effect of liquidity risk on banking competition variable (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) lerner(-1) 0.347*** 0.369*** 0.281*** 0.341*** 0.267*** 0.270*** 0.093*** 0.405*** 0.342*** 0.417*** (4.733) (4.790) (3.926) (4.472) (2.854) (2.567) (0.984) (4.629) (4.679) (4.775) loan to asset ratio(-1) -1.64*** -1.68*** -1.58*** -1.71*** -1.64*** -1.77*** -1.77*** -1.64*** -1.67*** -1.67*** (-8.727) (-8.901) (-8.567) (-9.342) (-8.684) (-9.630) (-9.569) (-8.714) (-8.780) (-8.733) govt bond ratio(-1) -0.76*** -0.74*** -0.70*** -0.61*** -0.76*** -0.60*** -0.57*** -0.75*** -0.72*** -0.71*** (-5.610) (-5.370) (-5.227) (-4.493) (-5.574) (-4.426) (-4.191) (-5.438) (-5.272) (-5.185) asset growth(-1) 1.04*** 1.04*** 0.918*** 0.865*** 1.03*** 0.89*** 0.89*** 1.02*** 1.02*** 1.025*** (3.621) (3.597) (3.222) (3.068) (3.584) (3.176) (3.181) (3.552) (3.582) (3.585) non-performing loan ratio(-1) -0.014 -0.018 0.050 0.069 -0.010 -0.024 -0.007 -0.029 -0.037 -0.044 (-0.172) (-0.213) (0.610) (0.847) (-0.122) (-0.288) (-0.079) (-0.348) (-0.442) (-0.522) non interest income ratio (-1) -0.073 -0.063 -0.070 -0.047 -0.085 -0.078 -0.102** -0.085 -0.072 -0.090 (-1.064) (-0.915) (-1.005) (-0.688) (-1.216) (-1.120) (-1.569) (-1.211) (-1.040) (-1.284) lerner(-1)*state owned banks -0.152 -0.59*** (-0.937) (-3.666) state owned banks 0.405*** 0.461*** (3.719) (4.167) lerner(-1)*foreign banks 0.158 0.319*** (1.315) (2.658) foreign banks -0.23*** -0.26*** (-2.786) (-3.158) lerner(-1)*private banks -0.138 -0.212*** (-1.092) (-1.652) private banks 0.023 0.056*** (0.264) (0.629) constant 1.769*** 1.778*** 1.659*** 1.686*** 1.773*** 1.927*** 1.942*** 1.778*** 1.775*** 1.769*** 16.297 16.309 14.903 15.300 16.311 17.036 17.017 16.369 (16.292) (16.240) t-values in parenthesis ***p<0.01, **p<0.05, *p<0. journal of applied economics and business studies, volume. 4, issue 1 (2020) 29-52 https://doi.org/10.34260/jaebs.412 45 table-4 effect of credit risk on banking competition variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) lerner(-1) -0.054*** -0.055*** -0.056*** -0.053*** -0.055*** -0.058*** -0.069*** -0.050*** -0.051*** -0.053*** (-7.975) (-7.217) (-8.075) (-7.141) (-5.688) (-8.368) (-6.871) (-6.059) (-7.418) (-6.430) loan to assets ratio(-1) -0.028** -0.026 -0.017 -0.018 -0.028* -0.036** -0.036** -0.027 -0.020 -0.020 (-1.705) (-1.529) (-1.052) (-1.119) (-1.676) (-2.150) (-2.142) (-1.608) (-1.183) (-1.201) government bonds ratio(-1) -0.022** -0.023** -0.016 -0.014 -0.022* -0.016 -0.014 -0.023** -0.025*** -0.026*** (-1.803) (-1.831) (-1.292) (-1.133) (-1.749) (-1.278) (-1.054) (-1.829) (-2.023) (-2.076) asset growth(-1) -0.041*** -0.041*** -0.042*** -0.042*** -0.041*** -0.043*** -0.045*** -0.042*** -0.040*** -0.040*** (-6.812) (-6.745) (-6.787) (-6.908) (-6.711) (-7.009) (-7.195) (-6.926) (-6.654) (-6.669) lerner(-1)*state owned banks 0.016 -0.029*** (0.851) (0.019) state owned banks 0.037*** 0.042 (2.880) (0.013) lerner(-1)*foreign banks 0.001 0.019 (0.076) (1.376) foreign banks -0.013 -0.018** (-1.328) (-1.708) lerner(-1)*private banks -0.008 0.006 (-0.562) (0.412) private banks -0.010 -0.012 (-0.993) (-1.073) constant 0.109*** 0.107*** 0.102*** 0.102*** 0.109*** 0.122*** 0.126*** 0.108*** 0.107*** 0.107*** (10.922) (10.564) (9.969) (10.052) (10.857) (10.125) (10.233) (10.809) (10.731) (10.789) t-values in parenthesis ***p<0.01, **p<0.05, *p<0.1 journal of applied economics and business studies, volume. 4, issue 1 (2020) 29-52 https://doi.org/10.34260/jaebs.412 46 6. conclusions pakistan’s banking industry is diverse and rich with extreme policy measures. it had tasted the fruits of extreme nationalization policy on one hand and open-ended privatization at the other. before 2000, the number of private/foreign banks was too low and their operations were also very limited but after 2000, there was a period when the banking industry of pakistan had started and incorporated technological innovations and international accords. therefore, despite the overall look, a sector-wise detailed study is necessary to investigate the performance of key financial variables along with the way they interact with each other. the banking industry shows a relatively high level of competition and a low level of market concentration. the overall stability of the banking industry is quite satisfactory especially during the global financial crisis. when we compare pakistan’s banking system (pbs) with global banking, pbs shows a high level of solvency, a satisfactory buffer of liquid assets concerning deposits and short-term borrowings, and relatively low nonperforming loans. this study employs four models to determine the relationship between market power and stability. the first model explains that how the market power affects stability using three different types of risks i.e. solvency, liquidity, and credit risk after accounting the onbalance sheet variables – which explains the positive relationship between market power and stability. second, third and fourth models are segregated into state-owned, private, and foreign banks by introducing separate dummy variables, separate interacting dummy variables, and both to determine the impact of each sector on stability. state-owned banks appear to be more solvent because of their high z-score whereas the foreign banks have the least solvency. foreign banks appear to be more liquid while state-owned banks have the least due to their low liquidity ratio and private banks have more credibility whereas stateowned banks are the least creditworthy due to their high npls. banks’ primary function is to convert liquid deposit (liabilities) to illiquid assets such as loans, which makes them inherently vulnerable to liquidity risk. management of such risks is to preserve their ability to fulfill their role. while some outflows are known with certainty, however, others depend on the external events which become the sources of liquidity risk (alam, 2012). liquidity risks could be reduced if a bank incorporates liquidity costs in their internal pricing and in case of a new product approval process, which alienates the risk-taking incentives with the liquidity risk exposures. a bank should actively monitor and control liquidity risk exposures within and across legal entities, business lines, and currencies. funding strategies should be established that provide effective diversification of resources and tenor of funding. according to the sbp, the npl ratio is 8.6% in december 2019 which seems much higher. npl ratio could not be reduced by merely opening the credit control bureaus. such bureaus or institutions should perform their due operations and allow commercials banks to journal of applied economics and business studies, volume. 4, issue 1 (2020) 29-52 https://doi.org/10.34260/jaebs.412 47 increase credit creation under guidance and supervision. no sector-wise study has conducted before for the pakistani banking industry; however, this study retrains to certain limitation i.e. microfinance, islamic, leasing, and other lending institutions were not incorporated in this research to maintain symmetry among the banks’ under-sample study. for future research, the effects of other lending institutions can be accounted for. furthermore, the lerner index was used as a proxy for market power/competition whereas several other variables may be used to determine the impact of competition on banking stability. sheikh muhammad umer farooq & muhammad zubair mumtaz 48 references almarzoqi, r., naceur, s. b., & scopelliti, a. d. (2015). how does bank competition affect solvency, liquidity, and credit risk? evidence from the mena countries. imf working paper wp/15/210. alam, h. m., raza, a., & akram, m. (2011). a financial performance comparison of public vs private banks: the case of commercial banking sector of pakistan. international journal of business and social science 2, 56-64. ayesha a., & nwazish m. (2012). size, diversification, and risk: preliminary evidence from commercial banks in pakistan. pakistan journal of commerce & social sciences 6(2), 282296. babar, h. z, & zeb, g. (2011). does “camels” system provide similar ratings as pacra system in assessing the performance of banks in pakistan? master thesis, umea school of business, sweden. beck, t., a. demirguc-kunt & r. levine (2006). bank concentration, competition, and crises: first results. journal of banking and finance. 30 (5), 1581-1603. beck t., de jonghe o. & schepens g. (2013). bank competition and stability: cross-country heterogeneity. journal of financial intermediation 22, 218–244. berger, a.n., l. klapper & r. turk-ariss (2009). bank competition and financial stability. journal of financial services 35, 99–118. bilal, z., asif, n., imran k. & rizwan a. (2011). the rapid growth of banking sector in pakistan and its impacts on revenue generation. syndicate group no. 2, faculty of the dot academy, lahore. cheron (2013). impact of non-performing loans on pakistan banks’ uk essays 2013. diana, b. & moshe k. (2014). liquidity risk in banking: is there herding? european banking center discussion paper no. 2012-024. d. e. allen, r. j. powel, (2011). credit risk measurement methodologies, http://mssanz.org.au/modsim2011 edward, i. altman, & anthony s. (1998). credit risk measurement: developments over the last 20 years. journal of banking & finance 21, 1721-1742. elmas, y. (2010). competition-fragility or competition-stability? evidence from the turkish banking system. working paper from cifrem seminar presentation, turkey. elmas, y. (2011). market power and riskiness in the turkish banking system. working paper from cifrem seminar presentation. emiliano, l., & valter l. (2000). credit risk measurement for a banks’ loan portfolio. sas community, seugi-2000. faisal, m. m., olvar b. & isma k. (2016). measuring the degree of competition in pakistan’s banking industry: an empirical analysis. applied economics 48(53), 5138-5151. franklin, a., elena c., & robert m. (2011). credit market competition and capital regulation. divisions of research & statistics and monetary affairs, federal reserve board, washington, d.c. hector, s. b. & lorenzo e. & bernal v. (2007). translog cost function for the mexican manufacturing industry. mexican working paper 07-08. journal of applied economics and business studies, volume. 4, issue 1 (2020) 29-52 https://doi.org/10.34260/jaebs.412 49 ijaz, s., hassan, a., tarazi, a., & fraz, a. (2020). linking bank competition, financial stability, and economic growth. journal of business economics and management, 21(1), 200-221. jimenez, g, j. a. lopez & j. saurina (2008). how does competition impact bank risk-taking? working paper, the federal reserve bank of san francisco. juan, c. c., lucotte, y. & reigl, n. (2020) banking sector concentration, competition, and financial stability: the case of the baltic countries, post-communist economies, 32:2, 215-249 khalil, m. a. (2005). a cross-section estimation of translog production function; jordanian manufacturing industry. meea and luc 7, 1-14. klaus s., martin s. & simone w. (2006). are more competitive banking sector more stable?” imf working paper wp/06/143. lapteacru, i. (2014). do more competitive banks have less market power? the evidence from central and eastern europe. journal of international money and finance 46(c), 41-60. mantobaye, m., william, s. r, and w. robert r. (2017), ‘which panel data estimator should i use? a corrigendum and extension’, discussion paper no. 2017-58 markus k brunnermeier and laise hege pederson (2008), ‘market liquidity and funding liquidity”, rfs advance kashif, m., faizan, s. i., & khurram i. (2016). loan growth and bank solvency: evidence from the pakistani banking sector. financial innovation 2:22, 1-13. farhan, m., amir, r. (2013). achieving the optimum level of liquidity is crucial for every bank. the romanian economic journal, xvi no. 48. tahir, m., sadaqat, m. a. s., & asim, m. a. (2016). assessing nature of competition in banking sector of pakistan. the journal of finance and data science 2(4), 244-253. riaz r. (2009). concentration and competition in banking sector of pakistan: empirical evidence. sbp working paper series no. 28. schaeck, k., m. čihák & s. wolfe (2006). are more competitive banking systems more stable? imf working paper, wp/06/143. selvi, ek k. (2009). bank competition and banking system stability: evidence from turkey. department of economics, metu econ 2009. sofronis c., menthos d., & sotirois k. (2013). a new dataset on bank competition in national banking market. university of cyprus, working paper no. 08-2013. zuzana f., laura s. & laurent w. (2013). does bank competition influence the lending channel in the euro area?” bofit discussion papers no. 17. zuzana f., anastasiya s., & laurent w. (2015). does bank competition reduce cost of credit? crosscountry evidence from europe. journal of banking and finance 83, 104-120. https://ideas.repec.org/a/eee/jimfin/v46y2014icp41-60.html https://ideas.repec.org/a/eee/jimfin/v46y2014icp41-60.html https://ideas.repec.org/s/eee/jimfin.html sheikh muhammad umer farooq & muhammad zubair mumtaz 50 appendix i: list of banks state owned banks 1) bank of punjab september 19, 1994 2) bank of khyber 1994 3) first women bank november 21, 1989 4) national bank of pakistan 1961 5) sindh bank october 29, 2010 foreign banks 1) atlas bank limited july 17, 2003 2) the bank of tokyo-mitsubishi ufj limited 1996 3) barclays bank plc july 23, 2008 4) citi bank n.a. april 1992 5) habib metropolitan bank limited august 3, 1992 6) royal bank of scotland september 30, 1991 7) bank alfalah limited june 21, 1992 8) samba bank limited february 12, 1980 9) deutsche bank ag october 7, 2011 10) faysal bank limited october 3, 1994 11) hsbc bank middle east limited 2008 12) hsbc bank oman november 01, 2008 13) nib bank limited october 2, 2003 14) silkbank limited april 4, 1994 15) standard chartered bank limited july 19, 2006 domestic (private) banks 1) allied bank limited august 14, 1947 2) askari bank limited october 09, 1991 3) abn amro bank n.v. september 30, 1991 4) bank al habib limited october 15, 1991 5) habib bank limited august 14, 1947 6) mcb bank limited july 9, 1947 7) js bank limited july 13, 1987 8) mybank limited october 7, 1991 9) soneri bank limited september 28, 1991 10) summit bank limited december 9, 2005 11) united bank limited march 15, 1959 source: state bank of pakistan journal of applied economics and business studies, volume. 4, issue 1 (2020) 29-52 https://doi.org/10.34260/jaebs.412 51 appendix ii: graphical representation of the variables 0.00 1.00 lerner index national 0.00 1.00 z-score national 0.00 1.00 2.00 liquidity ratio national 0.00 1.00 npl national state owned -1.000 0.000 1.000 ag national state owned 0.000 1.000 gbr national state owned -1.00 0.00 1.00 roae national state owned 0.00 1.00 lar national state owned sheikh muhammad umer farooq & muhammad zubair mumtaz 52 appendix iii: cross-section dependence test methods solvency model liquidity model credit model breusch pagam lm 467.45 393.05 422.91 pesaran scaled lm 0.08 02.351 -1.379 pesaran cd 3.678 2.939* 1.269 p-values ***p<0.01, **p<0.05, *p<0.1 appendix iv: unit root test variable levin, lin & chu im, pesaran and shin adf test pp test leaner index -3.833*** (0.000) -1.624* (0.052) 75.67** (0.027) 82.80*** (0.007) z-score -4.188*** (0.000) 0.843 (0.207) 45.76* (0.079) 77.45** (0.034) liquidity ratio -2.244** (0.012) 0.091 (0.536) 61.98 (0.119) 75.82** (0.017) npl ratio -10.50*** (0.000) -2.212** (0.016) 78.151** (0.011) 82.62*** (0.004) loan to asset ratio -29.91*** (0.000) -0.014*** (0.000) 84.86*** (0.004) 96.71*** (0.000) non-interest income ratio -5.28*** (0.000) -2.44*** (0.007) 81.89*** (0.008) 98.84*** (0.000) asset growth (0.000) (0.007) (0.008) (0.000) return on average equity -2.23*** (0.009) -2.71*** (0.003) 81.42*** (0.000) 119.5*** (0.000) govt. bond ratio -5.9*** (0.000) -0.935* (0.082) 52.26** (0.038) 45.27** (0.066) p-values are in parenthesis. ***p<0.01, **p<0.05, *p<0.1 appendix v: serial correlation test tests solvency model liquidity model credit model f-stats f-stats f-stats breusch godfrey lm test 77.29*** 41.32*** 101.61*** p-values ***p<0.01, **p<0.05, *p<0.1 appendix vi: heteroskedasticity tests p-values ***p<0.01, **p<0.05, *p<0.1 tests summary solvency model liquidity model credit model f-stats f-stats f-stats breusch pagan godfrey test 1.478 0.557 4.94*** harvey test 4.34*** 0.751 3.88*** glejser test 3.07*** 0.434 6.191*** arch test 63.5*** 9.82*** 58.67*** white test 1.93* 5.65*** 2.65*** journal of applied economics and business studies, volume. 4, issue 1 (2020) 53-76 https://doi.org/10.34260/jaebs.413 53 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online): 2663-693x service quality dimensions and customer satisfaction in online shopping: a customer’s perspective amna khan1, syed sohaib zubair2*, sobia khurram2 & mukaram ali khan2 1 institute of quality & technology management, university of the punjab, lahore, pakistan 2 institute of administrative sciences, university of the punjab, lahore, pakistan abstract with the expansion of e-commerce, e-service quality is becoming increasingly important. the study aims to investigate the relationship between e-service quality dimensions and customer satisfaction in b2c online shopping in pakistan emphasizing the customer’s perspective. a total of five dimensions of e-service quality have been tested with endogenous variable i.e. customer satisfaction. using a quantitative research strategy, data from 225 participants was collected using online questionnaires where 88% of the respondents already had the experience of online shopping that made a valid sample of 199. smart pls was used to conduct the data analysis. following reliability and validity analysis, out of all five dimensions, reliability showed the strongest path coefficient of 0.323. results also show that all the five hypotheses were supported except for responsiveness. as far as implications of the study are concerned, online shopping stores should devote valuable resource to enhance the significant e-service quality attributes that are emphasized by this study. moreover, e-services are becoming increasingly important in b2c e-commerce for improving customer relations and increasing sales and developing countries are of no exception to it. lastly, as the study is conducted in pakistani context, generalizability to other countries particularly outside south asia may be limited due to difference in customers’ shopping behaviours and attitudes. keywords e-service quality, ecommerce, customer satisfaction, b2c, cash on delivery jel classification l0, l20, l81, m10, m15 1. introduction e-commerce is the use of internet in conducting business transactions. it is widely spread in e-commerce models and applications that started a new movement called the dynamic e-business, which is the emerging technique of simplifying business transactions over the internet (chen, chen, & shao, 2003; sharma & lijuan, 2014). today, internet and web-based technologies have gained remarkable importance in business and organizations, especially those organizations that deal primarily in online business have increased their * corresponding author: sohaib.iqtm@pu.edu.pk amna khan, syed sohaib zubair, sobia khurram & mukaram ali khan 54 market share rapidly. the transition in marketplace is also encouraging conventional companies used to operate in old fashioned ways for carrying out their operations online. consequently, internet serves as a crucial medium for buying and selling products by using online sites (amin, 2016; kamhawi & gunasekaran, 2009; khan & faisal, 2015; parthasarathy, 2012). the digital dawn has led the e-service quality to become an important element for companies to retain old customers and to engage new ones, thus making it their utmost business strategy (khan, 2013; liu & yang, 2012). organizations not only apprehend the significance and effectiveness of online sites and their services for business ventures but also face difficulties in comprehending the perception of quality from customer’s point of view (awan, khan, & zhang, 2012; hirmukhe, 2013). lack of adequate service quality in e-business and services shows failures and demerits which are supported by wide evidences, also research of service delivery by means of online sites authenticates how service quality contributes to perception of customer towards overall service quality (chen & chen, 2014; lennon & harris, 2002). according to cronin and taylor (1992), service quality is a theoretical construct which is hard to quantify. the first model of service quality developed by parasurman et al. (1998) known as servqual has been extensively tested as a device for calculating customer perception towards service quality. the model of serqual has five different dimensions; “responsiveness, tangibles, assurance, reliability and empathy”. model of servqual has been widely used since past decades in measuring the e-service quality (devaraj, fan, & kohli, 2002). it has full potential for measuring service quality in physical marketplace but in e-commerce setting there is a need for changing original scale items, only rewording them is not the solution. therefore, researchers need to pay more attention in this area of research (jeong, oh, & gregoire, 2003). the rationale for this study emanates from the researcher’s urge to unearth customer perspective for e-service quality in e-shopping and to emphasize the factors that have significant role in attaining customer satisfaction (cs) in e-shopping. this will help in better adoption of business-to-customer (b2c) transactions in pakistan and help online store managers to grasp potential customers by improving the services they desire and keeping quality as their utmost priority. the e-retail business in pakistan is still very low, being less than 1% of cumulative retail sale (business recorder, 2018). the significance of service quality and the challenges faced by the e-services makes it necessary to gather insights about what factors customers utilize in the evaluation of online service quality. this study provides the theoretical and academic evidence regarding what attributes significantly influence the perceived customer service quality. therefore, this research is imperative in developing a journal of applied economics and business studies, volume. 4, issue 1 (2020) 53-76 https://doi.org/10.34260/jaebs.413 55 smooth transaction between service provider and customer which not only benefits the online service providers but also potentially benefits the society at large. although there have been various studies conducted on customer satisfaction and eservice quality, however, this study specifically tests each dimension of e-service quality in connection with cs unlike others including khan et al. (2019). even in the south asian context, broadly e-sq has been studied as a single latent construct, however, as mentioned earlier, this study informs the literature by testing each dimension of e-sq against customer satisfaction. moreover, unlike earlier studies in local context, this study also highlights various factors that impede the use of online shopping platforms. the online shopping or ecommerce industry is a rapidly growing industry, therefore attention should be given to issues related to this industry from the customers’ perspective. although several projects are conducted around e-service quality, but it seldom focuses on online shopping. therefore, it is not yet evident from the literature whether or not service quality of online shopping is adapting to the challenges it is facing in the course of emerging b2c online shopping concept, consequently it is necessary to scrutinize factors contributing to the evaluation process of service quality of online shopping by consumers and it is particularly a curious topic for academics and practitioners. as far as theoretical underpinnings or contribution is concerned, several theories are linked with the idea of e-service quality and customer satisfaction from varied angles, however, utility theory here explains the “prioritization” of attributes of customer satisfaction and service quality that helps in strengthening the overall serqual idea (li et al., 2002), precisely the theory aids in prioritizing those attributes of servqual that maximize expected utility of customers. this study also attempts to explore the most important dimension or attribute of service quality that aids in improving customer satisfaction. precisely, prime objective of the study is to analyze the dimensions of e-service quality in online shopping and to measure their effects on customer satisfaction. along with this, an overall snapshot of customer profile of b2c online shopping sites with reference to various categorical factors such as age, gender, education, employment status and income level and the common problems faced by customers in online shopping are also being explored. 2. literature review e-service quality can be defined as the “customer’s assessment of the virtual marketplace by evaluating the quality of e-service”. since past decades, service quality measures have been implemented in virtual market places to evaluate the effectiveness of websites (kuo, 2003) and to measure customer’s satisfaction with these e-commerce channels and websites and also to regulate the factors necessary for the success of online amna khan, syed sohaib zubair, sobia khurram & mukaram ali khan 56 sites (devaraj et al., 2002). a complex phenomenon made up of intricate processes is known as online shopping, it can be segregated into various sub-processes: browsing, online transaction, navigation or customer interactions. sub process will not likely be assessed by customers in much depth during their first encounter, rather services will be perceived as an overall shopping experience of the online platform (van riel et al., 2001). besides, online customers prefer high standards of service, therefore e-service quality is the only source which can be targeted towards the customer to build their perception regarding realization of the benefits the internet has to offer via online shopping sites (al-debei et al, 2015; yang, 2001). firstly, sq has been conceptualized by parasuraman as the gap between expectations of consumers & evaluations of service experienced by them. for measuring service-quality, a multi-item scale model has been developed and known as servqual (parasuraman et al., 1988). servqual consists of five dimensions namely “reliability, responsiveness tangibility, assurance and empathy”. this model has been extensively used in measuring service quality of information technology system (dyke et al., 1999; pitt et al., 1997). keeping in view previous studies conducted for conventional in-store shopping, two prominent scholars parasuraman and zeithaml conducted research on online service quality which resulted in development of new measuring scales for e-services known as “e-sq scale”, defined as “the extent to which a particular website facilitates efficient & effective business or purchase”. in the beginning, e-sq scale consisted of eleven dimensions according to the study by zeithamal et al. (2000). later, further studies decreased this number to only seven dimensions (parasuraman et al., 2005). however, modified dimensions of e-sq has been adapted for this study from lee and lin (2005) and parasuraman et al. (2005). first dimension is the website service which is related to the fundamental hub of information that helps customers to search for the product/service they are interested in and make purchase decisions (jeon & jeong, 2017). the design, aesthetics and characteristic of an online page have an influence on customer’s purchase (shergill & chen, 2005). literature review regarding website service quality indicates that several dimensions have been proposed depending upon website properties. virtual scale for website service quality has been put forth by kuo (2003), where he suggested that webpage design, online quality, information safety, customer service management, and content can be used for the assessment of quality of a website (kuo, 2003). web portals perceived quality information has been measured by yang et al. (2005) using accessibility, interaction, adequacy of information, usefulness of content and usability. next, reliability being the second dimension of servqual has been defined as “the ability to perform the service commitments consistently and accurately”. it has two facets: one includes the functions of journal of applied economics and business studies, volume. 4, issue 1 (2020) 53-76 https://doi.org/10.34260/jaebs.413 57 websites such as mode of payment and search engines. other facet includes the product information, financial information and service commitment (if it is reliable as promised by the website) ( yang et al., 2003; parasuraman et al., 2005). the reliability consists of performing the promised services in the correct way at the very first time, and employees or organizations having sincere customer interest always provide the right services at the first instance for solving customers’ problems. therefore, as per parasuraman et al. (1991), consumers not only evaluate dependability and accuracy of the services delivered but also how these services are being delivered. the third aspect is responsiveness, which is the desire of service providers to assist the customers and increase their service level swiftly. this dimension includes different items as follows: providing timely service, give right information to the customer and on time, willingness of staff in helping customers, providing prompt service to the customers (alqeed, 2013; huang et al., 2018). responsiveness measures firms willingness and ability in providing prompt services to the consumers during time of queries or problems (zeithaml et al., 2002). fourth dimension is trust which is defined as “the willingness to be vulnerable to something in spite of having risks and uncertainties of the situation” (rousseau et al., 1998, p.395). hence in e-commerce setting e-retailers helps the consumers to overcome the perceptions of risks in online interactions. mcknight et al. (2002) established that trust is vital for e-commerce, and following this argument building of trust is even more critical and important in e-commerce. yet it has been observed that trust and the antecedents of trust are indefinable and cannot be comprehended effortlessly. lastly, privacy is defined as “the degree to which the site is safe and protects customer information” (khan et al., 2018, p.150). it is the dimension of security that originates from dimension of assurance of servqual, and has following items: trustworthy employees, providing ease to the customers during transactions, and support to the employees from company (zavareh et al., 2012). moreover, customer satisfaction has been conceptualized as “the evolution of an emotion” (hunt, 1977). it is suggested that customer satisfaction is a degree of possession of a service, leading to a customer believing that he/she has gained something good and evokes positive feeling. another notable aspect is the conceptualization of customer satisfaction into two broad ways: one is when an emotional response is triggered because of specific performance attribute of a particular service experience, it is known as the transaction specific customer satisfaction. on contrary, when satisfaction depends on various factors that occur again and again during transactions, it conceptualizes as overall customer satisfaction or cumulative outcome (shankar et al., 2003). hence, in an eenvironment, when a customer makes a purchase for the very first time from a new service provider, then the “website satisfaction” will be transaction specific. alternatively, when a amna khan, syed sohaib zubair, sobia khurram & mukaram ali khan 58 customer purchases a product repeatedly from same online retailer then it will be known as cumulative outcome (chang, 2005). it has been established that service quality and customer satisfaction are different constructs with high disparity whereas satisfaction and service quality are the constructs with high similarity and correlation (richard, 1997). according to putrianti and semuel (2018), among all predictors including e-sq, trust and perceived value, e-sq was found to be the most influential predictor towards escalation of buying intention of customer. similarly, rico et al. (2019) found positive association between service quality and customer satisfaction. in a recent study conducted by khan et al. (2019), researchers developed a model using covariance based sem and concluded a significant positive relationship between e-sq and e-cs. ali (2016) reports a positive relationship between customer satisfaction and website service. similar finding was observed in online tourism as well (pereira et al., 2017). according to jauhari et al. (2019), it is good to have better website design to enhance the customer satisfaction. as far as reliability is concerned, a study was conducted in kigali branches with 498 respondents using cross sectional research design. a positive relationship between customer satisfaction and reliability was reported (felix, 2017). similar findings were found in malaysian banks (kashif et al., 2015). according to al-dweeri et al. (2018), reliability was found to be the positive predictor contributing towards customer satisfaction. responsiveness was found to be positively related with customer satisfaction as well (agnihotri, 2016; felix, 2017). trust seems to be a significant dimension of service quality as well in terms of escalating customer satisfaction (dimyati, 2018). similarly, putrianti and semuel (2018) found a positive relationship between e-sq and escalation of buying intention of customer. rico et al. (2019) studied similar relationships using sem and found positive relationship between trust and customer satisfaction. lastly, privacy also has a positive relation with customer satisfaction in internet banking (ling et al., 2016). moreover, a similar study was conducted on jordanian students and found positive relationship between customer satisfaction and privacy (al-dweeri et al., 2018). e-services are very significant in b2c e-commerce for the sake of managing customer relations and boosting sales. in today’s electronic world the customer and the online shopping provider do not meet face to face, and customers are more sensitive to increasing products and various options available to them. similarly it is easy for them to switch from one website to another with a single click of mouse (singh, 2002). a study by nazir et al. (2012) was conducted to find out the factors affecting consumers in online shopping. trust was found to be the most important factor as customer can hesitate in shopping because of insecure transaction and payment system. another research conducted by bashir (2013) focused on the consumer behaviors in online shopping and was found that price, convenience and time saving are the factors that influence online shopping, and their service journal of applied economics and business studies, volume. 4, issue 1 (2020) 53-76 https://doi.org/10.34260/jaebs.413 59 quality should be improved. a study by zafar et al.(2011) conducted study in e-banking and findings suggested that e-servqual dimensions and customer satisfaction had a significant relation. moreover, web service quality should be improved to gain high levels of customer satisfaction. it is not clear from the literature whether service quality of online shopping is adopting the challenges it faced during b2c online shopping, consequently it is necessary to scrutinize factors contributing to the evaluation process of service quality of online shopping by consumers and it is particularly a curious topic for academics and practitioners. thus, servqual model should be reformulated in context of online shopping characteristics. this study adapts e-service quality dimensions developed by lee and lin (2005) and parasuraman et al. (2005). the conceptual framework of the study is developed on the basis of the relationships between the exogenous variables (e-sq dimensions) and endogenous variable (customer satisfaction) as shown in figure 1. figure 1: conceptual framework on the basis of the literature reviewed and framework developed, following hypotheses are formulated; h1: privacy in b2c shopping has a positive impact on cs. h2: reliability in b2c shopping has a positive impact on cs h3: responsiveness in b2c shopping has a positive impact on cs. h4: trust in b2c shopping has a positive impact on cs. h5: website design in b2c shopping has a positive impact on cs. website design responsiveness trust privacy reliability customer satisfaction amna khan, syed sohaib zubair, sobia khurram & mukaram ali khan 60 3. research methodology a quantitative strategy is followed here using a descriptive research approach. this approach is designed to determine and identify the participants accurately. moreover, in this study, survey design is used to collect primary data from a sample of large population. the survey was conducted in pakistan and filled by online shopping consumers. the data was collected using google forms platform. the link to the designed google survey was shared on various promotional pages of various social media sites. for that matter, convenience sampling strategy was used because the population frame is not available. population frame could have been available if a case analysis of one website was conducted, but due to the already critical issue of data privacy, no organization was willing to share their customer profiles/data. this study consisted of a sample of 225 respondents. this is the total number of subjects who participated in the research by filling the online questionnaire. however, 26 individuals had never conducted an online transaction, therefore, the data of only 199 respondents was used for conducting hypotheses testing. as mentioned above, the sampling strategy used in this research is homogenous convenience sampling, which is the most common type of non-probability sampling, and offers clearer generalizability as compared to conventional convenience sampling. it can be defined as the availability of a sampling to a researcher by means of its accessibility (bryman, 2006). the questionnaire used for the study has three sections. section one was designed to find out the demographic and socio-economic factors of the customers. section two was designed to collect information about shoppers and non-shoppers, their preferences, frequency and hurdles they faced during online shopping. section three was based on modified servqual model and was developed to collect data about e-service quality dimensions, and was to be filled in a manner of five-point likert scale, selecting between strongly disagree to strongly agree. the definitions and source of measures adapted for each construct are mentioned in table 1. three items each were used to measure website design (e.g, “the online store is visually appealing”), reliability (e.g, “transactions with the online store are error-free”), responsiveness (e.g, “i believe the online store is always willing to help customers”) and trust (e.g, “i trust the online stores”), whereas, four items each were used to measure privacy (e.g, “it protects information about my web-shopping behavior”) and endogenous variable i.e. customer satisfaction (e.g, “overall, i am very satisfied with my shopping”). journal of applied economics and business studies, volume. 4, issue 1 (2020) 53-76 https://doi.org/10.34260/jaebs.413 61 table 1: constructs definition construct definition references website design it is the degree of user friendliness of an online store as perceived by customer. (lee and lin, 2005) reliability it is the degree of accurate and persistent services provided by an online store as perceived by customer. (kim and lin, 2002) responsiveness it is the degree of willingness and helpfulness of services provided by an online store as perceived by customer. (lee and lin, 2005) trust it is the degree of trust and overcoming the perception of uncertainties of an online store as perceived by customer. (kimery and mccord, 2002) privacy it is the degree of protection and safety of an online store as perceived by customer. (tsai and yeh, 2010) customer satisfaction it is the degree of satisfaction of an online store as perceived by customer. (wen et al., 2014) to explore the objectives of the study and for testing hypotheses, smart pls was employed on the collected data due to its non-normal distribution. sample characteristics were analyzed by drawing frequency tables. 4. results and discussion the data analysis begins with the brief demographic profile of the respondents. the study revealed that both male and female customers are inclined towards online shopping with almost equal frequencies, more than half of the respondents are between the ages of 18-25 years. eighty three percent of the respondents have acquired bachelors or ms/mphil degrees indicating that majority of the users are highly educated and are aware of the benefits and drawbacks of online shopping. employment status indicated that almost half of the respondents are employed, having monthly income between the bracket of pkr 20,000-40,000. amna khan, syed sohaib zubair, sobia khurram & mukaram ali khan 62 table 2: descriptive statistics of the demographics variable frequency percentage gender female 115 51.1 male 110 48.9 total 225 100.0 age (years) 18 to 25 117 52.0 26 to 35 79 35.1 36 to 45 15 6.7 above 45 6 2.7 under 18 8 3.6 total 225 100.0 education level bachelors 95 42.2 intermediate 18 8.0 ms/m.phil. 92 40.9 other 15 6.7 phd 5 2.2 total 225 100.0 employment status employed 99 44.0 homemaker 13 5.8 others: housewife 1 .4 retired 2 .9 self-employed 18 8.0 student 67 29.8 unemployed 25 11.1 total 225 100.0 monthly income (pkr) 21,000 – 40,000 51 22.7 41,000 – 60,000 42 18.7 61,000 – 80,000 17 7.6 81,000 – 100,000 17 7.6 above 100,000 22 9.8 below 20,000 76 33.8 total 225 100.0 in evaluating customer’s profile, it is necessary to carry out the frequency analysis. factors indicating customer’s perceptions led to the conclusion that 88% of the respondents have experienced online shopping and majority of them prefer online shopping because it saves time. most of the customers are already using the medium of online shopping for over period of 12 months. cash on delivery was their most preferred mode of payment. journal of applied economics and business studies, volume. 4, issue 1 (2020) 53-76 https://doi.org/10.34260/jaebs.413 63 table 3: descriptive statistics of the factors in online shopping variable frequency percent online shopping yes 199 88.4 no 26 11.6 total 225 100.0 shopping time never less than 3 months 26 63 11.56 28.0 3–6 months 32 14.2 6–12 months 35 15.6 more than 12 months 69 30.6 total 225 100.0 shopping frequency never less than 3 times 26 106 11.56 47.1 3–6 times 58 25.8 more than 6 times 35 15.6 total 225 100.0 preferred payment mode never debit card 26 24 11.56 10.7 credit card 19 8.4 online banking e-payments 9 4.0 cash on delivery 145 64.4 other total 2 225 0.9 100.0 shopping preference saves time 105 46.8 convenient 68 30.2 low price/discounts 30 13.3 product variety 15 6.6 other 7 3.1 total 225 100.0 furthermore, to highlight factors that discourage and impede a consumer’s decision to shop online, it was deemed necessary to add an open-ended question to elicit participants’ thoughts about the factors they considered as impediments to their online shopping. the most pressing factor was found to be lack of product quality reported by 37% (n=139) of the respondents, indicating that product quality was not the same as it appeared online. another important factor to hindrance was the return and repair of the faulty products having lengthy procedures, which overall makes online shopping an unpleasant experience as compared to conventional shopping for 20% (n=74) respondents. other factors with more or less equal response rate focused on additional charges, lack of security and warranty 12% (n=46), 11% (n=43) and 10% (n=36) respectively. it showed an almost equal importance in the eyes of customers. delivery and logistics have a response rate of 8% (n=29) which is again noteworthy. some respondents did not participate much in adding other factors for hindrance yielding a response rate of 2% (n=6) as shown in figure 2. amna khan, syed sohaib zubair, sobia khurram & mukaram ali khan 64 figure 1: factors hindering online shopping next, inferential hypotheses testing starts with the measurement model in order to calculate the reliability and validity. however, this study found out that all the loadings were above threshold value of 0.70 which meant that all the items were retained in the study as per the following figure 3. figure 3: measurement model 37% 8% 10% 20% 12% 11% 2% factors hinders shopping product quality delivery and logistics warranty return or repair policies additional charges lack of security other journal of applied economics and business studies, volume. 4, issue 1 (2020) 53-76 https://doi.org/10.34260/jaebs.413 65 construct validity is calculated by using smart pls. in case of construct validity, it is mandatory to have factor loadings of more than 0.60 (chin, 1998) and in this study, all the factor loadings were way beyond the threshold i.e. more than 0.80. as it is clear from table 4, factor loadings for customer satisfaction are above 0.90 except one item that is cs-4 which is above 0.80. in case of privacy, all factor loadings are above 0.80. similar pattern has been observed for reliability, responsiveness, trust and website design with factor loadings between 0.80 and 0.90. cross loadings are shown in table 4. table 4: cross loadings cs pri rel res tru wd cs_1 0.900 0.510 0.603 0.438 0.500 0.572 cs_2 0.933 0.436 0.579 0.431 0.521 0.527 cs_3 0.917 0.401 0.582 0.465 0.542 0.426 cs_4 0.858 0.357 0.536 0.462 0.508 0.447 pri-1 0.442 0.849 0.414 0.239 0.473 0.299 pri-2 0.277 0.815 0.249 0.104 0.245 0.185 pri-3 0.416 0.863 0.297 0.185 0.354 0.288 pri-4 0.417 0.822 0.337 0.236 0.342 0.244 rel_1 0.545 0.306 0.868 0.416 0.439 0.399 rel_2 0.557 0.298 0.849 0.462 0.481 0.467 rel_3 0.544 0.418 0.862 0.387 0.418 0.372 res_3 0.424 0.258 0.374 0.829 0.452 0.462 res_1 0.421 0.220 0.522 0.868 0.452 0.510 res_2 0.444 0.144 0.383 0.899 0.491 0.548 tru_1 0.435 0.365 0.448 0.431 0.878 0.259 tru_2 0.472 0.397 0.462 0.468 0.876 0.361 tru_3 0.574 0.376 0.450 0.499 0.867 0.347 wd_2 0.501 0.337 0.372 0.436 0.346 0.828 wd_1 0.389 0.253 0.407 0.545 0.325 0.816 wd_3 0.440 0.158 0.402 0.466 0.238 0.801 second condition related to construct validity is composited reliability value, whose threshold value should be more than 0.70. third condition related to construct validity is average variance extracted (ave) whose threshold value is 0.50, and any value beyond 0.50 is considered to be good for construct validity (hair et al., 2006). following table 5 shows the values of composite reliability (between 0.856 and 0,946) and ave (between 0.664 and 0.814). all of the values are beyond threshold values. amna khan, syed sohaib zubair, sobia khurram & mukaram ali khan 66 table 5: reliability and validity of constructs latent variables cronbach’s alpha composite reliability average variance extracted cs 0.924 0.946 0.814 pri 0.859 0.904 0.701 rel 0.824 0.895 0.739 res 0.832 0.900 0.750 tru 0.847 0.906 0.763 wd 0.749 0.856 0.664 composite reliability and ave are also presented through graphs in the following figures 4 and 5. composite reliability shows inter consistency among items of the scale and it can be seen in figure 4 that all the values are beyond the threshold value of 0.70. figure 4: composite reliability ave is used for construct validity and assumes that if the ave values are higher than the threshold value of 0.70, then it is considered to have existence of construct validity as shown in figure 5. journal of applied economics and business studies, volume. 4, issue 1 (2020) 53-76 https://doi.org/10.34260/jaebs.413 67 figure 5: average variance extracted for divergent validity, an old yet mostly used principle of “fornell-larcker” is used and it shows that “all of the construct items are distinct” (hair et al., 2013). as per this principle, √ave should be greater than the correlations. table 6 shows that discriminant validity is present in this study as √ave ranges from 0.815 to 0.902 whereas correlation’s maximum value is 0.638 which shows that participants were able to discriminate among latent constructs. table 6: discriminant validity cs pri rel res tru wd cs 0.902* pri 0.475 0.837* rel 0.638 0.396 0.860* res 0.497 0.238 0.491 0.866* tru 0.573 0.435 0.519 0.538 0.874* wd 0.549 0.311 0.481 0.586 0.373 0.815* *√ave there is another technique used to test the discriminant validity i.e. heterotrait monotrait ratio (htmt). as per this technique, values of all the constructs should not exceed 0.90 and if this is the case then one can say that discriminant validity is present in the data. according to table 7, it can be seen that all in-between construct values are less than the threshold value i.e. 0.90 with highest htmt value of 0.748. amna khan, syed sohaib zubair, sobia khurram & mukaram ali khan 68 table 7: heterotrait monotrait ratio cs pri rel res tru wd cs pri 0.516 rel 0.731 0.460 res 0.568 0.271 0.594 tru 0.639 0.493 0.621 0.635 wd 0.650 0.372 0.613 0.748 0.461 moreover, structural model is developed that depicts path coefficients, r-square and effect size. “path coefficient” is used to determine the significance of each hypothesis. it can be seen in table 7 that path coefficients of pri→cs, rel→cs, res→cs, tru→cs, wd→cs are 0.168, 0.323, 0.040, 0.223, 0.235. all paths are having significant values i.e. p-value < 0.05 except res→cs whose p-value is 0.710 which shows that all the hypotheses are accepted except h3. positive coefficient (0.168) of pri→cs shows that privacy helps customers in increasing satisfaction while engaging in b2c online shopping. similarly, positive coefficient (0.323) for rel→cs implies that customers give relatively high priority to platforms which have consistent and error free service. referring to the path coefficient (0.223) of tru→cs, it can be observed that trust plays an important role in increasing confidence of the customers. among accepted hypotheses, path coefficient (0.235) of wd→cs implies that easy user interface and appealing outlook of the website also positively affects customer satisfaction of online customers. on the other hand, path coefficient (0.040) of res→cs was found to be insignificant, i.e. responsiveness does not contribute towards escalating customer satisfaction. furthermore, r2 value is tested using “pls-algorithm”. its value should be more than 0.10 in order to accept “model’s predictive relevance” (falk & miller, 1992). in this case, r2 value is quite significant as it is more than the threshold value of 0.10 i.e. 0.563 or 56.3%. moreover, as per table 8, “variance inflation factor” (vif) is accessed for the paths pri→cs, rel→cs, res→cs, tru→cs, wd→cs and is concluded that all the vif values are within the threshold value of 5 (ringle et al., 2015), which indicates that collinearity does not exist among independent variables. journal of applied economics and business studies, volume. 4, issue 1 (2020) 53-76 https://doi.org/10.34260/jaebs.413 69 figure 6: structural model literature highlights that “effect sizes of 0.02, 0.15, and 0.35” are taken as “small, medium and large effects” (cohen, 1988). f2 values are used in order to find the effect sizes. the rule of thumb in this case is that if f2 value is less than 0.10 then it is considered as ‘small’ effect and if the f2 values are between 0.10 and 0.35 then it is taken as ‘medium’ effect. however, if f2 is found to be more than 0.35 then effect is considered as ‘large’ (hair et al., 2013). in this case, the effect size of pri→cs, tru→cs and wd→cs is 0.048, 0.065 and 0.075 with significant p-values of less than 0.05 showing small size effects. however, the effect size of rel→cs is 0.153 which is taken as the medium level effect. lastly, there is also insignificant effect of res→cs found with f2 of 0.002. lastly, the fifth technique for model fit namely q² is used to predict the “predictive relevance of path model”. fornell & cha (1993) is of the view that threshold value for q2 should be zero and if value of q2 is positive it shows the predictive relevance of the model however if the value is negative; it shows that predictive relevance of model is lacking. table 8 shows the value of q2 of 0.418 which is quite significant to predict model’s relevance. amna khan, syed sohaib zubair, sobia khurram & mukaram ali khan 70 table 8: structural model analysis *p<0.05 5. conclusions, recommendations and implications the objective of this study is to achieve a better comprehend e-service quality dimensions affecting customer satisfaction in online shopping context. the findings of smart pls suggest that all e-service quality dimensions positively effect on customer satisfaction except for responsiveness. moreover, out of all five dimensions, rel→cs showed the strongest path coefficient of 0.323. wd→cs showed the second strong and significant path coefficient of 0.235. moreover, tru→cs path coefficient was 0.223 and last significant path coefficient was 0.168 for pri→cs. the acceptance of hypotheses are supported through literature as well (ali (2016; dimyati, 2018; felix, 2017; kashif et al., 2015; ling et al., 2016; pereira et al., 2017). moreover, it was also observed that res→cs path coefficient was insignificant. hence, it is concluded that all the hypotheses are supported except h3. this finding is very critical in the present study as people in pakistan may tolerate or accept any responsiveness related issues in b2c online shopping. similar pattern was observed in a study where responsiveness was a significant factor in banking industry of pakistan in terms of service quality, but its significance level was lower among all dimensions of e-service quality (ali & raza, 2015; raza et al., 2015). this finding needs further exploration and in-depth analysis. lastly, cash on delivery is a very crucial component in pakistan as identified in this study, as many as 64.4% of respondents preferred cash on delivery option as their preferred mode of payment. this finding is not something out of blue, rather many developing countries focus on cash on delivery due to the transaction security and privacy issues (rahman et al., 2018). data also revealed that online shopping is also done for time saving as 46.8% of the respondents do online shopping merely for saving time. this finding can also be validated with the findings where researchers believe that online shopping can save a lot of time as compared to conventional shopping (shanthi & desti, 2015). thirty seven percent of the respondents were of the view that product quality is the issue in online shopping as when product is delivered it is not the same as displayed on website. this hypothesis paths path coef. pvalue results vif f2 r2 q2 h1 pri→cs 0.168 0.047* supported 1.33 0.054* 0.56 0.42 h2 rel→cs 0.323 0.002* supported 1.67 0.153* h3 res→cs 0.040 0.710 rejected 1.93 0.002* h4 tru→cs 0.223 0.026* supported 1.76 0.065* h5 wd→cs 0.235 0.043* supported 1.68 0.075* journal of applied economics and business studies, volume. 4, issue 1 (2020) 53-76 https://doi.org/10.34260/jaebs.413 71 finding can also be linked with the perceived risks in purchasing online, where product quality is considered as one of the significant risks while buying online (pappas, 2016). the swift development of e-commerce has led to the realization that key determinant of success is the delivery of high-quality services to the customers. if the management of online businesses wish to maintain e-service quality at high levels then they must pay attention to dimensions of e-service quality as highlighted in this study. results highlight that among esq dimensions, reliability has the highest influence on customer satisfaction; therefore, online shopping stores should regard it as their utmost priority to enhance reliability by developing stringent marketing strategy focusing on time management, customer services and transactions based on authenticity and accuracy every time a customer uses the platform. this study recommends that managers should also emphasize on other dimensions of eservice quality because of their positive influence on customer satisfaction. trust is seen to be the second largest contributors towards e-sq which managers should use as an opportunity to capitalize in their e-businesses, and this may be done by providing timely customer services, easy return and refund systems. website design should also be the focus of managers and these designs should be customer-friendly with ease of use ensured to increase customer satisfaction. satisfaction through website design can be achieved by keeping it simple yet comprehensive, also the website should always be updated, and it should enable customers to navigate through various pages easily. similarly, privacy and security of personal data has been one of the key concerns for customers in pakistan due to increasingly reported fraudulent cases, hence, managers should focus on privacy dimension as well in order to capture the customer satisfaction, and this also reflects in why cash on delivery (cod) is the preferred mode for payment of majority customers in pakistan. due to the growing nature of this industry, the e-commerce entrepreneurs and state regulatory bodies should sit together and develop mechanisms to overcome issues associated with data privacy and online transactions. lastly, online shopping stores should devote valuable resources to enhance the significant e-service quality attributes that are emphasized by this research. it will be helpful in attracting new customers and retaining the existing customers. 6. future research directions and limitations the study can be replicated further by using various research methodologies. interviews and focus groups sessions can be conducted to have an in-depth evaluation of relationship between customer satisfaction and service quality. future studies can undergo comparisons with other countries and can be reproduced in different cultures by using the same measuring instrument; as this will provide a much needed comparison of b2c online shoppers across cultures. the research can be enriched by the expansion of the current model. further depth can be added in future studies by focusing on differences in other markets and geographic amna khan, syed sohaib zubair, sobia khurram & mukaram ali khan 72 regions. future studies can discover how the quality dimensions (other than e-service quality) suggested by other researchers affect the customer’s perspective in online shopping. the current study is conducted from the perspective of a customer, one may explore and evaluate e-service quality in b2b dealings as well. although the study has provided sufficient information and relevant insights for the customer’s online shopping adoption in pakistan, but there are a few limitations which are associated with the study. the first and the foremost limitation is the sample employed, as the study is conducted by taking convenience sample through online medium in pakistani cities only, which may not be generalized as representative of online shopper’s population. the current research is generic by nature which limits its scope to be applicable to all segments of industry. moreover, the relationship between exogenous and endogenous variables can be further tested by taking various hurdles identified earlier as control variables. finally, since the study is conducted in pakistan, generalizability to other countries, outside south asia may be limited due to major difference in customers’ shopping behavior. references agnihotri, r., dingus, r., hu, m. y., and krush, m. t. (2016). ‘social media: influencing customer satisfaction in b2b sales’, industrial marketing management, 53, pp.172-180. al-debei, m. m., akroush, m. n., and ashouri, m. i. (2015). ‘consumer attitudes towards online shopping: the effects of trust, perceived benefits, and perceived web quality’, internet research, 25(5), pp.707-733. al-dweeri, r. m., ruiz moreno, a., montes, f. j. l., obeidat, z. m., and al-dwairi, k. m. (2019). ‘the effect of e-service quality on jordanian student’s e-loyalty: an empirical study in online retailing’, industrial management & data systems, 119(4), pp.902-923. ali, f. (2016). ‘hotel website quality, perceived flow, customer satisfaction and purchase intention’, journal of hospitality and tourism technology, 7(2), pp.213-228. ali, m., and raza, s. a. (2017). ‘service quality perception and customer satisfaction in islamic banks of pakistan: the modified servqual model’, total quality management & business excellence, 28(5-6), pp.559-577. alqeed, m. a. (2013). ‘service quality relationship with customer satisfaction and business profitability (a case study of royal jordanian)’, european journal of business and management, 5(1), pp.210-216. amin, m. (2016). ‘internet banking service quality and its implication on e-customer satisfaction and e-customer loyalty’, international journal of bank marketing, 34(3), pp.280-306. awan, m. a., khan, h. u., and zhang, w. (2012). ‘a comparative study on online service quality perception of two major regional economies’, international journal of eeducation, e-business, e-management and e-learning, 2(6), pp.529-533. journal of applied economics and business studies, volume. 4, issue 1 (2020) 53-76 https://doi.org/10.34260/jaebs.413 73 ayyash, m. m., ahmad, k., and singh, d. (2012). ‘a questionnaire approach for user trust adoption in palestinian e-government initiative’, american journal of applied sciences, 9(1), pp.1202-1208. bashir, a. (2013). consumer behavior towards online shopping of electronics in pakistan (master). seinäjoki university of applied sciences. bryman, a. (2006). ‘integrating quantitative and qualitative research: how is it done?’, qualitative research, 6(1), pp.97–113. business recorder. (2019). e-commerce in fy18 | business recorder. [online] available at: https://www.brecorder.com/2018/11/28/455272/e-commerce-in-fy18/ [accessed 22 sep. 2019]. chang, h. h. (2005). ‘the influence of continuous improvement and performance factors in total quality organization’. total quality management & business excellence, 16(3), pp. 413–437. chen, f.-y., and chen, s.-h. (2014). ‘application of importance and satisfaction indicators for service quality improvement of customer satisfaction’, international journal of services technology and management, 20(1–3), pp.108–122. chen, m., chen, a. n., & shao, b. b. (2003). ‘the implications and impacts of web services to electronic commerce research and practices’, j. electron. commerce res., 4(4), pp.128–139. chin, w. w. (1998). ‘issues and opinions on structural equation modelling’, mis quarterly, 22(1), pp.7–16. cohen, j. (1988). statistical power analysis for the behavioral sciences mahwah. lawrence erlbaum, nj. cronin jr, j. j., and taylor, s. a. (1992). ‘measuring service quality: a reexamination and extension’, the journal of marketing, 56(3), pp.55–68. devaraj, s., fan, m., and kohli, r. (2002). ‘antecedents of b2c channel satisfaction and preference: validating e-commerce metrics’, information systems research, 13(3), pp.316–333. dimyati, m. (2018). customer satisfaction and switching cost toward trust in the brand and customer retention of simpati in jember. dyke, t. p., prybutok, v. r., and kappelman, l. a. (1999). ‘cautions on the use of the servqual measure to assess the quality of information systems services’, decision sciences, 30(3), pp.877–891. falk, r. f., and miller, n. b. (1992). a primer for soft modeling. university of akron press. felix, r. (2017). ‘service quality and customer satisfaction in selected banks in rwanda’, journal of business & financial affairs, 6(1), pp.246-256. fornell, c. and cha, j. (1993). partial least squares (pls) (unpublished working paper). ann arbor. university of michigan business school, mi. hair, j. f., hult, g.t, ringle, c. m., and sarstedt, m. (2013). a primer on partial least squares structural equation modeling (pls-sem). thousand. sage, oaks. hair, j., black, w., babin, b., anderson, r., and tatham, r. (2006). multivariate data analysis upper saddle. pearson prentice hall, river, nj. amna khan, syed sohaib zubair, sobia khurram & mukaram ali khan 74 hirmukhe, j. (2013). ‘measuring the customer’s perception of service quality using servqual in public services’, international journal of services technology and management, 19(4–6), pp.267–277. huang, z., luo, y., and wang, d. (2018). ‘online customer service quality of online shopping: evidence from dangdang. com’, cluster computing, pp.1–9. hunt, h. k. (1977). conceptualization and measurement of consumer satisfaction and dissatisfaction. marketing science institute. jauhari, m. t., kusumawati, a., and nuralam, i. p. (2019). ‘the impact of website quality on consumer satisfaction and purchase intention (study case of e-commerce lazada indonesia in malang city)’, jurnal administrasi bisnis, 67(1), pp.54-61. jeon, m. m., and jeong, m. (2017). ‘customers’ perceived website service quality and its effects on e-loyalty’, international journal of contemporary hospitality management, 29(1), pp.438-457. jeong, m., oh, h., and gregoire, m. (2003). ‘conceptualizing web site quality and its consequences in the lodging industry’, international journal of hospitality management, 22(2), pp.161–175. kamhawi, e. m., and gunasekaran, a. (2009). ‘erp systems implementation success factors: is and non-is managers’ perceptions’, international journal of business information systems, 4(6), pp.688–704. kashif, m., wan shukran, s. s., rehman, m. a., and sarifuddin, s. (2015). ‘customer satisfaction and loyalty in malaysian islamic banks: a pakserv investigation’, international journal of bank marketing, 33(1), pp.23-40. khan, h. u., awan, m. a., and ho, h. c. (2018). ‘how do chinese and saudi customers perceive online service quality? a comparative study’, the journal of business inquiry, 13(2), pp.146–161. khan, h., and faisal, m. n. (2015). ‘a grey-based approach for erp vendor selection in small and medium enterprises in qatar’, international journal of business information systems, 19(4), pp.465–487. khan, h. u. (2013). ‘use of e-learning tools to solve group work problems in higher education: a case study of gulf country’, advances in computer science: an international journal, 2(3), pp.90–96. khan, m., zubair, s. and malik, m. (2019). an assessment of e-service quality, esatisfaction and e-loyalty. south asian journal of business studies, 8(3), 283-302. kim, j., and lee, j. (2002). ‘critical design factors for successful e-commerce systems’, behaviour & information technology, 21(3), pp.185–199. kimery, k. m., and mccord, m. (2002). ‘third-party assurances: mapping the road to trust in eretailing’, journal of information technology theory and application, 4(2), pp.63-82. kuo, y.f. (2003). ‘a study on service quality of virtual community websites’, total quality management & business excellence, 14(4), pp.461–473. lee, g., and lin, h. (2005). ‘customer perceptions of e-service quality in online shopping’, international journal of retail & distribution management, 33(2), pp.161–176. lennon, r., and harris, j. (2002). ‘customer service on the web: a cross-industry investigation’, journal of targeting, measurement and analysis for marketing, 10(4), pp.325–338. journal of applied economics and business studies, volume. 4, issue 1 (2020) 53-76 https://doi.org/10.34260/jaebs.413 75 li, y.n., tan, k.c. and xie, m. (2003), "managing service quality: applying utility theory in the prioritization of service attributes", international journal of quality & reliability management, 20 (4), pp. 417-435. ling, g. m., fern, y. s., boon, l. k., and huat, t. s. (2016). ‘understanding customer satisfaction of internet banking: a case study in malacca’, procedia economics and finance, 37, pp.80-85. liu, x., and yang, j. (2012). ‘social buying metanetwork modelling and analysis’, international journal of services technology and management, 18(1–2), pp.46–60. mcknight, d. h., choudhury, v., and kacmar, c. (2002). ‘the impact of initial consumer trust on intentions to transact with a web site: a trust building model’, the journal of strategic information systems, 11(3–4), pp.297–323. nazir, s., tayyab, a., sajid, a., ur rashid, h., and javed, i. (2012). ‘how online shopping is affecting consumers buying behavior in pakistan?’, international journal of computer science issues (ijcsi), 9(3), pp.486-495. oliver richard, l. (1997). satisfaction: a behavioral perspective on the consumer. new york ˈ ny: irwin-mcgraw-hill. pappas, n. (2016). ‘marketing strategies, perceived risks, and consumer trust in online buying behaviour’, journal of retailing and consumer services, 29, pp.92-103. parasuraman, a., zeithaml, v. a., and malhotra, a. (2005). ‘es-qual: a multiple-item scale for assessing electronic service quality’, journal of service research, 7(3), pp.213–233. parasuraman, arun, berry, l. l., and zeithaml, v. a. (1991). ‘understanding customer expectations of service’, sloan management review, 32(3), pp.39–48. parasuraman, a., zeithaml, v. a., and berry, l. l. (1988). ‘servqual: a multiple-item scale for measuring consumer perc’, journal of retailing, 64(1), pp.12-40. parthasarathy, s. (2012). ‘research directions for enterprise resource planning (erp) projects’, international journal of business information systems, 9(2), pp.202–221. pereira, h. g., de fátima salgueiro, m., and rita, p. (2017). ‘online determinants of ecustomer satisfaction: application to website purchases in tourism’, service business, 11(2), pp.375-403. pitt, l. f., watson, r. t., and kavan, c. b. (1997). ‘measuring information systems service quality: concerns for a complete canvas’, mis quarterly, pp.209–221. putrianti, a. s., and semuel, h. (2018). ‘the analysis of e-service quality, customer trust, perceived value, and behavioral intention on online transportation in surabaya’, petra international journal of business studies, 1(1), pp.1-10. rahman, m. a., islam, m. a., esha, b. h., sultana, n., and chakravorty, s. (2018). ‘consumer buying behavior towards online shopping: an empirical study on dhaka city, bangladesh’, cogent business & management, 5(1), 1514940. raza, s. a., jawaid, s. t., and hassan, a. (2015). ‘internet banking and customer satisfaction in pakistan’, qualitative research in financial markets, 7(1), pp.24-36. rico, m. t., wahyoedi, s., and purnama, e. d. (2019). the effects of trust, service quality and perceived value on satisfaction and their impact on loyalty. in proceedings of the 7th international conference on entrepreneurship and business management (icebm untar), pp.325-330 amna khan, syed sohaib zubair, sobia khurram & mukaram ali khan 76 ringle, c.m., wende, s. and becker, j.m. (2015). smartpls. smartpls gmbh, boenningstedt. rousseau, d. m., sitkin, s. b., burt, r. s., and camerer, c. (1998). ‘not so different after all: a cross-discipline view of trust’, academy of management review, 23(3), pp.393–404. shankar, v., smith, a. k., and rangaswamy, a. (2003). ‘customer satisfaction and loyalty in online and offline environments’, international journal of research in marketing, 20(2), pp.153–175. shanthi, r., and desti, k. (2015). ‘consumers' perception on online shopping’, journal of marketing and consumer research, 13, pp.14-21. sharma, g., and lijuan, w. (2014). ‘ethical perspectives on e-commerce: an empirical investigation’, internet research, 24(4), pp.414–435. shergill, g. s., and chen, z. (2005). ‘web-based shopping: consumers’ attitudes towards online shopping in new zealand’, journal of electronic commerce research, 6(2), pp.79-94. singh, m. (2002). ‘e‐services and their role in b2c e‐commerce’, managing service quality: an international journal, 12(6), pp.434–446. tsai, y. c., & yeh, j. c. (2010). perceived risk of information security and privacy in online shopping: a study of environmentally sustainable products. african journal of business management, 4(18), pp. 4057-4066. van riel, a. c., liljander, v., and jurriens, p. (2001). ‘exploring consumer evaluations of e-services: a portal site’, international journal of service industry management, 12(4), pp.359–377. wen, c., prybutok, v. r., blankson, c., & fang, j. (2014). the role of e-quality within the consumer decision making process. international journal of operations & production management, 34(12), pp. 1506-1536. yang, z. (2001). „consumer perceptions of service quality in internet-based electronic commerce‟ proceedings of the 30th emac conference, 8-11 may. bergen. yang, zhilin, cai, s., zhou, z., and zhou, n. (2005). ‘development and validation of an instrument to measure user perceived service quality of information presenting web portals’, information & management, 42(4), pp.575–589. yang, zhilin, peterson, r. t., and cai, s. (2003). ‘services quality dimensions of internet retailing: an exploratory analysis’, journal of services marketing, 17(7), pp.685–700. zafar, m., zaheer, a., and rehman, k. (2011). ‘impact of online service quality on customer satisfaction in banking sector of pakistan’, african journal of business management, 5(30), pp.11786–11793. zavareh, f. b., ariff, m. s. m., jusoh, a., zakuan, n., bahari, a. z., and ashourian, m. (2012). ‘e-service quality dimensions and their effects on e-customer satisfaction in internet banking services’, procedia-social and behavioral sciences, 40, pp.441–445. zeithaml, v. a., parasuraman, a., and malhotra, a. (2000). conceptual framework for understanding e-service quality: implications for future research and managerial practice. zeithaml, v. a., parasuraman, a., and malhotra, a. (2002). an empirical examination of the service quality-value-loyalty chain in an electronic channel. university of north caroline, chapel hill, nc, working paper. journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 53 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x cost of nutritious diet for children in pakistan and effects of imminent afghan refugees on existing consumption pattern rahema obaid1, tehseen ahmed qureshi2, stephen davies3 & abdul wajid rana4 1 rahema obaid is a research analyst at international food policy research institute, pakistan 2 tehseen ahmed qureshi is a research analyst at international food policy research institute, pakistan 3 stephen davies is a senior research fellow international food policy research institute, pakistan 4 abdul wajid rana is the program leader of pakistan agriculture capacity enhancement project (pace) in international food policy research institute (ifpri), islamabad, pakistan. abstract the fall of kabul after the us withdrawal from afghanistan may have serious repercussions for pakistan if the influx of refugees accelerates, as additional demand for food items and resulting inflationary pressures can jeopardize the diets of children. to examine this issue, we estimate the gap between actual and desired per capita expenditures on a least cost nutritious diet consumed by pakistani children aged 3-10 years. the study finds the gap across income groups and regions for two provinces of pakistan i.e., kp and balochistan. we find that under-consumption is highest in children of balochistan urban and kp urban, as households in both locations spend only 21 percent of the recommended dietary expenditures. we estimate the decrease in dietary expenditures coming from an increase in food prices after the refugees’ influx. we found the dietary gap is high if the influx of refugees exceeds one million by the first quarter of 2022. however, the dietary gap does not increase much if the number of refugee arrival remains under 700,000, albeit the large nutrition gap does not improve. keywords nutrition, children’s diet, refugees, food prices jel classification i1, i3, q1 1. introduction access to a healthy diet remains a major global concern, even in the 21st century. the rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 54 nexus between nutrition and food security remains largely dependent upon diet quality and food consumption, but a healthy diet remains unaffordable for many1. the state of food and nutrition security report (2020) defines three types of diets: an “energy sufficient diet,” which provides adequate calories for the energy needed for work each day. using mainly the basic starchy staple (e.g., maize, wheat or rice only)2 . a “nutrient adequate diet” not only provides adequate calories (per the energy sufficient diet above), but also relevant nutrient intake values of 23 macroand micronutrients to prevent deficiencies and avoid toxicity. a “healthy diet” provides adequate calories and nutrients (per the energy sufficient and nutrient adequate diets above), but also includes a more diverse intake of foods from different food groups to help prevent malnutrition in all its forms, including diet-related non-communicable diseases. the nutritious diet discussed in this paper falls under the “healthy diet” category because it is based on country-specific food-based dietary guidelines (fbdgs). the food and nutrition security report (2020) argues that a ‘healthy diet’ costs about 60 percent more than a ‘nutrient adequate diet’.3 however, a healthy diet costs on average 500 percent more than an ‘energy sufficient diet’. it is estimated that three billion people worldwide cannot afford the healthy diet, including 57 percent of the population in south asia and africa4. the less-than-optimal consumption of nutritious and healthy foods results in greater stunting and wasting in children. an estimated 149 million children under the age of five suffer from stunting around the world, while 45.4 million children under five years old suffer from wasting5. therefore, availability and affordability of a healthy diet for children is significant in determining their future productivity and living standards. this paper estimates the gap between the actual and recommended nutritious diets for children between the ages of 3-10 at the subnational level and across different income groups for pakistan. it is a stark fact that 68 percent of the pakistani population cannot afford a 1 dizon, f., wang, z., & mulmi, p. (2021). the cost of a nutritious diet in bangladesh, bhutan, india, and nepal. world bank policy research working paper 9578, washington, dc: world bank 2 fao, ifad, unicef, wfp and who. (2020). the state of food security and nutrition in the world 2020. transforming food systems for affordable healthy diets. rome, fao. https://doi.org/10.4060/ca9692en 3 ibid. 4 ibid. 5 fao, ifad, unicef, wfp and who. 2021. the state of food security and nutrition in the world 2021. transforming food systems for food security, improved nutrition and affordable healthy diets for all. rome, fao.https://doi.org/10.4060/cb4474en journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 55 ‘nutritionally adequate diet’ and ‘healthy diet,’ whereas about five percent of the population still cannot afford an ‘energy sufficient’ diet6. the target of this paper is pakistani children as they are likely to be more affected by the impact of positive and negative shocks to consumption, such as an influx of refugees. pakistan has remained a top host for displaced afghans since the 1980s when, with the advent of afghan-soviet war in 1979, over 3 million afghan refugees crossed to pakistan. after the russian invasion of afghanistan in 1979, afghan refugees were initially held in camps in the border areas of kp and balochistan. however, later refugees began to settle in urban areas and started small-scale businesses or took jobs in the informal sector. today, about 1.6 million afghan refugees are dispersed across pakistan, and they have acquired legal refugee status, so no major refugee camps are left in pakistan7. thus, they are integrated with the pakistani population and share families and businesses—and thus face the same prices for goods and services as the local population does. the taliban takeover of afghanistan on august 15, 2021 may now trigger a new wave of out-migration to different countries including pakistan. it is estimated that around 700,000 afghans might migrate to pakistan in the coming months if the internal security situation in afghanistan deteriorates. the hosting of ‘externally displaced afghans (eda)’ for three years may cost around usd 2.2 billion8. however, while the above direct cost may be correct, the indirect costs could be much higher through inflationary pressures on food products due to increased demand by refugees. furthermore, effects on population and consequent demands for education and health, employment, property, as well as negative impacts on forests, other environmental and land degradation may arise with the higher demand for the country’s limited resources. this paper will restrict its analysis to the expected rise in food inflation due to the arrival of these migrants, which can result in even lower than desired expenditure on a nutritious diet by the households—thereby compromising the future of affected children. the economic cost of stunting and wasting is estimated to be seven percent of annual per capita income (galasso and wagstaff, 2018). as a large part of this cost is related to deficiencies in diets, this paper estimates the possible linkage between prices of a 6 global alliance for improved nutrition. (n.d.) pakistan. retrieved from https://www.gainhealth.org/impact/countries/pakistan 7 united nations high commissioner for refugees. (2016). after decades in pakistan, more afghan refugees set to return. retrieved fromhttps://www.unhcr.org/news/latest/%202016/6/576bd0a84/decades-pakistan-afghan-refugees-setreturn.htm 8 rana, s. (2021, july 18). hosting 700k afghans will cost $2.2 b for 3 years. the express tribune. retrieved from https://tribune.com.pk/story/2311234/hosting-700k-afghans-will-cost-22-b-for-3-years https://www.gainhealth.org/impact/countries/pakistan https://www.unhcr.org/news/latest/%202016/6/576bd0a84/decades-pakistan-afghan-refugees-set-return.htm https://www.unhcr.org/news/latest/%202016/6/576bd0a84/decades-pakistan-afghan-refugees-set-return.htm rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 56 nutritious diet and actual versus needed per capita expenditures on nutritious diets in a business-as-usual case and then with the arrival of refugees. 1.1 purpose of the study the main purpose of this study is to estimate the gap between food intake based on required food-based dietary guidelines (fbdgs) and actual consumption. we study the gap for pakistan for two of its provinces – khyber pakhtunkhwa (kp) and balochistan – and within these, for all income quintiles. the focus of the study is children between 310 years old. the study is novel because a) it measures the gap for children who are just old enough to require a varied diet rather than adolescents; b) it also estimates the gap at the sub-national level across income quintiles; and c) it is based on the most recent available household survey. the household survey does not explicitly report consumption according to age groups. this is one of the limitations that is addressed in detail under the section 3.1. the major objective of this study is to understand the impact of food prices on the potential to consume a nutritious diet for all five income quintiles. further, we introduce shocks to compute the impact of a possible refugee influx from afghanistan in the provinces of balochistan and kp. a negative shock would be an inflow of refugees leading to higher demand for food—causing an upward spiral in food prices. on the flip side, a positive but less likely shock in the near term would be the repatriation of existing refugees back to afghanistan should the security and political situation stabilize, thereby generating economic opportunities and reduced food demand—causing a downward shock in food prices. as majority of refugees are expected to settle in kp and balochistan, only these two provinces are the focus of this paper. fbdgs encompass nutritious food items with the objective of promoting healthy diets for children, adolescents, and adults. this study can be helpful for policymaking as it estimates the diversion from recommended intake and computes the effect of a future refugee influx on the diet of the local population. the results of the study can support strategies to eliminate the existing gap in expenditure on nutritious diets using various policy interventions. furthermore, the research is also supportive of policy directions on refugee intake as it affects food security due to higher demand and increased food prices. 2. literature review the literature on the cost of a nutritious diet is still at a nascent stage but is not entirely new. there is an extensive use of linear programming (lp) models to estimate the cost of a diet that provides the most nutrition at least cost, which goes as far back as journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 57 1945 when it was proposed by george stigler (stigler, 1945). it has been further presented as a better measure of poverty compared to the conventional world bank poverty line (allen, 2017). moreover, lp analyses have been used to investigate the effect of food inflation on a least-cost diet in various countries alone, as well as relative to actual expenditures (o'brien-place and tomek, 1983; håkansson, 2015; omiat and shively, 2017; jensen and miller, 2010; maillot et al., 2017). in a similar vein, comparisons between countries have also been studied (chastre et al., 2007). the emerging literature often makes use of linear programming software to calculate the least-cost diet. the “cost of the diet” (or cod) created by save the children computes the least costly dietary combination of locally available food meeting average energy requirements and recommended ingestion of micronutrients, protein, and fat for one or more persons (deptford et al., 2017). the software also can model the impact of food and nutrition-based interventions. another similar linear programming software is optifood, which goes one step ahead in that its computation of nutritious diet is directed at a given population segment (food and nutrition technical assistance, n.d.). according to dizon et al. (2019), one obvious drawback of these approaches is the need for specific data that may not be easily available. the calculation of the cost of the recommended diet (cord) on the other hand is straightforward and can be computed based on available food price data. moreover, while cod estimates are based on securing important minerals and vitamins, cord incorporates food-based dietary guidelines (fbdgs) that aim to strengthen general health rather than meeting selected nutrients. in this regard, masters et al. (2018) developed a least-cost dietary diversity index based on the food groupings used in the minimum dietary diversity for women (mdd-w) indicator9. mulik and haynes-maslow (2017) utilize the us department of agriculture’s (usda’s) myplate dietary standards to estimate the cost of diet for various population groups in the us. recent work by dizon et al. (2019) uses fbdgs to calculate the cord in four south asian countries, where cord is matched against actual expenditures using a simple methodology rather than more complex linear programming. extending their analysis, this paper focuses on pakistan and the implications on the estimated cord in the light of the potential refugee influx following the takeover of afghanistan by the taliban on the diet of 3-10 years old 9 masters, w., bai, y., herforth, a., sarpong, d. b., mishili, f., kinabo, j., & coates, j. c. (2018). measuring the affordability of nutritious diets in africa: price indexes for diet diversity and the cost of nutrient adequacy. american journal of agricultural economics, 100(5), 1285-1301 rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 58 as of 2018, nearly 37 percent of the population in pakistan was food insecure (government of pakistan and unicef, 2018). the rate of stunting and wasting of children under the age of five years stood at 40.2 percent and 17.7 percent respectively. also, 28.9 percent of children under five are underweight and 9.5 percent are overweight. pakistan ranks 88th out of 107 countries in the 2020 global hunger index and its score of 24.6 is categorized as “serious”10, with the score just short of being in the “alarming” category (global hunger index, 2020). being a net food importer with a population of 221 million to feed, the influx of refugees would put a further strain on food security in pakistan (br research, 2021). the soviet invasion of afghanistan in december 1979 spurred the movement of afghans to other countries as refugees (lapping et al., 2002). since then, there has been a constant flow of refugees to pakistan due to the persistent political instability in afghanistan, with the cumulative number of afghan migrants to pakistan reaching nearly 3.3 million in 1989 (grare and maley, 2011). this number corresponded to more than 3 percent of pakistan’s population at that point, and, during 1980-2002, the afghan asylum seekers in pakistan formed the largest refugee group in the world. pakistan currently hosts 1.6 million registered afghan refugees of which 58 percent are located in kp, while balochistan holds 23 percent– the two provinces where they have traditionally been concentrated (unhcr, 2021). as such, this paper looks at the impact of the inflow of afghan refugees on the cord in these provinces. as mentioned before, the influx of refugees following the taliban takeover of afghanistan is likely to further squeeze food availability and accessibility for the local population, driving up the cord. this could operate through the channel where increased demand drives up local food prices (baloch et al., 2018; chambers, 1986).there exists some evidence on the impact of refugees’ influx on food inflation. akgündüz et al. (2015) examine the effect of syrian refugees in turkey following the syrian crisis in 2013. the study estimates that an incursion of syrian refugees equal to one percent of the turkish population would result in a food price hike of 2.2 percent. enghoff et al. (2010) also reported a decline in food prices with added refugee camps in kenya. these anomalous results however were driven by the re-selling of the world food program (wfp) food rations; registration of some locals as refugees to claim 10 countries with a score lesser or equal than 9.9 are categorized as having “low” level of hunger whereas countries with a score of greater than or equal to 50 are categorized to have “extremely alarming” hunger level. journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 59 rations; and smuggling low-cost food items from somalia which translated into usd 10 million per annum savings for kenyan consumers. in summary, refugees’ influx in various countries is found to be associated with changes in food prices. these effects depend upon the similarity between diets of locals and refugees, integration of markets, whether the refugees source their food from markets or direct food transfers and the extent to which aid is available. empirically, the upward pressure on prices is due to the increased demand for food, mainly the staple items consumed equally by both the local population and refugees. a possible future exodus of afghans to pakistan without significant food aid may result in a similar price hike for food items in the markets of kp and balochistan (where afghan refugees mainly settled after the soviet invasion). this is more likely to be true for items such as wheat and dairy products as these are staples for both afghans and pakistanis residing in kp and balochistan 2.1 present refugee situation in pakistan following recent political developments in afghanistan, there has been concern for a possible refugee influx, but which has so far remained under control because of fencing of the pakistan-afghanistan border and effective border management arrangements in pakistan, which implemented the government’s strategy that keeps refugees in border camps (amparado et al., 2021). the government’s stance on the afghan situation is categorical: pakistan is not in a state to welcome further refugees (furqan, 2021). nevertheless, as the inexperienced taliban government runs into economic difficulty, the exodus of refugees could accelerate. according to unhcr, 500,000 more afghan refugees were likely to flee to bordering countries by the end of 2021 (dawn, 2021). while the movement of afghan refugees in pakistan seems controlled at present, it may only be a matter of time before an additional population has to be hosted. 3. methodology this study applies a two-step methodology. the first part of the paper develops estimates of the gap between actual and required expenditures, where the latter is the cost of a nutritious diet (cord). these gaps are calculated for different household quintiles in kp and balochistan. the second approach makes use of price and income elasticities drawn from recent work using the quadratic almost ideal demand system (quaids). details are given below. rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 60 3.1 calculating the gap between actual and required household expenditures for a nutritious diet first, we calculate the minimum cost of meeting food-based dietary guidelines (fbdgs) using the household integrated economic survey (hies) 2018-19 for pakistan, from which we developed a basket of items that reflect the fbdgs published by the pakistan ministry of planning, development & reforms and the food and agriculture organization (2018). to estimate the prices of food items, two approaches are commonly used. first, the monthly cpi index can be used to extract prices at the city or national level, while a second method estimates prices using a ratio of per capita monthly expenditure and quantities consumed of each food item found in hies. we used the second method as it is largely followed in the literature of cord and incorporates actual market prices paid by different consumers (dizon et al., 2019). a further advantage of using hies data is that it reports consumption and expenditure across all income quintiles instead of using aggregated data. deaton (1988) states that an advantage of using survey data for prices is that it incorporates household variation influenced by the demographic characteristics such as age, male/female ratio in a household, income quintiles, seasonal effects, quality of products, and others. this study classified the nutrition basket into the following groups: a) milk and milk products, b) cereals, c) vegetables, d) fruits and e) meat and pulses. these account for between 40 and 50 percent of household expenditures in the different provincial and income groups, but they represent those commodities that most likely to face greater competition and increased prices with an afghan influx. each food group initially contains all commodities specified by the fbdgs except those which are not reported in the hies. then, we calculate the least cost of a healthy and nutritious diet for children based on the fbdg for children aged 3-10 years, using the lowest cost item in the group. the purpose of selecting age group 3-10 is that it is the most important age group for nourishment that determines the stunting and wasting of children in the future. although the first 3 years of a child life are also important for nutrition, however, their diet plan does not include any other item other than milk. in this way, the most important group is 3-10 years that also has to consume market items in their daily intake and are thus affected by the prices. journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 61 we implicitly assume that households pick the least expensive item within each food group and so pursue a type of cost minimization. the food items selected for each group are those for which we were able to compute prices based on data availability in hies. for instance, bananas are the lowest cost item in our fruit basket—however, it is possible that another fruit is cheaper, but we could not use it, as the hies does not report per capita consumption and expenditure for all commodities. the final commodities, with the lowest costs per serving, selected to represent each group are a) fresh milk, b) wheat, c) onions, d) bananas and e) chicken. thus, a major difference between our research and prior work is that we do not take an average cost but rather estimate the diet based on the lowest cost items (mulik and haynes-maslow, 2017). dizon et al (2019) used the average of two lowest cost items in each category for households at the provincial level, and then estimated a median value across households to calculate cord for pakistan as a whole. they chose more than one low-cost item as fbdg recommends diversity within food groups. instead, we use the lowest-cost option and estimate the provincial gaps separately rather than using a median value of provinces to arrive at a national estimate. we have only taken one least-cost commodity as our focus is the affordability of the nutritious diet. further, to the best of our knowledge, this is the only study in pakistan that estimates the least-cost at the quintile level using a diet appropriate for 3to 10-yearold children. the next step is to estimate the difference between actual per capita expenditure on the food items found in the nutritious diet and the cord, in other words ‘the gap’. this gap is compared at the sub-national level, specifically for rural and urban kp and balochistan compared to urban punjab. urban punjab is the most prosperous region of the country and can be taken as a baseline.11 3.2 simulations of the nutrition effects from changing prices with selected price elasticities computed by haider and zaidi (2017), who used the quadratic almost ideal demand system (quaids), we estimate the change in quantity demanded when consumers face changing prices. the estimates contain both own and cross-price elasticities of various food products at the provincial level for pakistan, including those that are in the fbdgs. the basic quaids model provides estimates of a system of demand equations, where the dependent variables are quantities expressed as budget shares. these equations determine the effects of a full set of prices for the 11 as per the punjab growth strategy 2023 report, the per capita income of punjab in 2017-18 was two percent higher than the national average, i.e., usd 1673 rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 62 food items within the system, along with total expenditures, on the allocation of budget shares to those items. a set of standard theoretical restrictions of homogeneity, adding up, and symmetry are imposed on the estimation. we use these calculated unconditional elasticities, along with varying levels of price changes of commodities affected by an influx of afghan refugees to find the effects on nutrition in 3to 10-year-old pakistani children. the process can be seen in the following equation, which shows how we translate price shocks into changes in quantities consumed and thus to the impact on nutrition. the effect on a quantity consumed of food item i, is expressed as 𝜕𝑄𝑖 = 𝜀𝑖𝑖 𝜕𝑃𝑖 + ∑ 𝜀𝑖𝑗𝑗≠𝑖 𝜕𝑃𝑗 + 𝜀𝑖𝑦 𝜕𝑌ℎ ……(i) eq (i) refers to the quaids model used to estimate the elasticities of food items. where ∂qi is the percentage change in quantity consumed for the ith commodity, εii is the own price elasticity and ∂pi is the percentage change assumed for the ith price. additionally, the qi is affected by changes in prices of the other commodities in the system via cross price elasticities (εij) and changes in the other prices (pj). the last term in the equation describes effects of changes in income of the relevant households, translated through the expenditure elasticity (εiy). most of our analysis focuses on the price effects but we make occasional reference to possible income changes from a greater influx of afghan refugees, and so we include this term in the above equation. there are several important assumptions in this approach. the elasticities used are for the overall household and there may be differences in responses for 3to 10-yearolds. there may be either less sensitivity to prices for younger children, as mothers worry about their health, or they could be a “residual claimant” on food, with more variable consumption and perhaps more sensitivity to prices. also, the elasticities are the same for both baluchistan and kp. finally, the changes in prices are assumed to be real changes and do not include the normal inflation rate in the country. it is also pertinent to note that influx of refugees is expected to be greater in balochistan compared to kp. this is because the pak-afghan border in the kp province is now almost completely fenced and any illegal arrival is not possible,12 whereas the pak-afghan border fence in balochistan is still not completely finished and about 12 yousaf, m. (2021, august 04). pakistan army completes 90% of fence along afghan border. ap news. retrieved from https://apnews.com/article/pakistan-bd8165697772792b69d65c8509633cd9 https://apnews.com/article/pakistan-bd8165697772792b69d65c8509633cd9 journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 63 20,000 afghans are entering pakistan from the chaman border daily.13 currently, government of pakistan is only allowing afghans who are coming for medical treatment or who possess a valid refugee identity—still this number is three times the daily average of pre-taliban takeover era.14 the predicted number of total refugees’ influx in pakistan is based on government estimates, whereas the number of refugees expected to settle in kp is based on the current distribution of afghan refugees across provinces (see annex a). 3.3 assessing outcomes for 3–10-year-old children there are two explicit limitations in this study. first, the hies survey only reports per capita expenditure on food items by household, but they not disaggregated into gender and age groups. the actual expenditure on children aged 3-10 years old in the study is thus likely to be overstated and the gap vis-à-vis cord understated because children are likely to be consuming less than adults on average. table 1: differences in dietary guidelines for children and adults as per fbdgs items number of servings per day for 3-10 years old number of servings per day for 19-60 years old milk and dairy 2-3 2-3 cereal grains & grains products 3-5 4-5 vegetables 1-2 2-3 fruits 1-2 2-3 meat, pulses& eggs 3-4 2-3 source: based dietary guidelines (2019), planning commission of pakistan in kp and balochistan, approximately 25 and 28 percent of the population is between the ages of 3-10 years respectively. out of this, 58 percent of the children in balochistan and 52 percent in kp are male. it is also important to mention that the dietary guidelines of children aged 3-10 are not much different than that of adolescents and adults (table 1). if we compare the minimum dietary requirements of adults and children as per pakistan’s fbdgs, the daily servings of milk and dairy and cereals remains the same. however, there is increase of one serving per day in fruits and vegetables for adults along with a decrease of one serving per day in meat and pulses. this shows that dietary requirements of adults and children in a typical pakistani 13 baloch, s.m. & ellis-petersen, h. (2021, august 27). ‘unprecedented’ numbers crossing from afghanistan to pakistan. the guardian. retrieved from https://www.theguardian.com/world/2021/aug/27/unprecedented-numbers-crossing-afghanistanpakistan 14 ibid https://www.theguardian.com/world/2021/aug/27/unprecedented-numbers-crossing-afghanistan-pakistan https://www.theguardian.com/world/2021/aug/27/unprecedented-numbers-crossing-afghanistan-pakistan rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 64 household are quite the same. hence, the assumption that per capita expenditure on each commodity would approximately be same on children as well as on adults may be quite true. the estimation also does not consider gender dynamics. for instance, argaw (2020) found a high gender inequality in dietary diversity in ethiopia, which in turn depends upon where the food is sourced from – the market or home produced.15 the gender inequality was found to be higher in case of market sourced food. we assume that the per capita expenditure on food items reported by household is same across all age groups and gender, including 3 to 10-year-olds. however, as proportion of boys is relatively higher in balochistan, some gender inequality might exist in the household diet. second, we do not know what the magnitude of the increased population on individual food items could be. at the moment, we assume that there would be a 1:2 impact of refugees’ influx on food prices, based on akgündüz et al. (2015) estimates for turkey. thus, a 10 percent increase in population would result in food prices increasing by 20 percent. these effects will also not be uniform but will vary according to the extent of market integration with the wider economy. thus, the prices of wheat, which tend to be nationally tied together, will increase less than for milk, which may be more likely to see locally separated markets. 4. results the study results are presented in three subsections, with the first providing a context on current expenditures for important food items. secondly, comparisons between the cord and current expenditures are presented, with the gap between those two measures as the main investigation. next, we present simulations of price shocks on nutritional outcomes using price elasticities from the quaids model. lastly, we discuss the market integration and price variability across provinces. 4.1 current expenditure proportions, prices, and per capita income differentials figure 1 presents the daily expenditures made on relevant food groups in the lowest and highest quintiles in each region. we see that overall, wheat and milk account for most expenditures among quintile 1 households, usually accounting for 33 to 37 percent of the total food expenditures. vegetables, meat, and fruits together are no more than 15 the same may hold true for pakistan which ranks 153 out 156 in gender parity index published by the world economic forum (2021). journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 65 five percent of the total expenditure for these households. balochistan quintile 1 households consume less milk proportionally and more chicken than do the other poor households included in the figure. while in quintile 5, families also spent considerable amount on meat consumption (chicken in this case). the comparisons between quintile 1 and 5 in these regions provide a perspective on how increased family income affects the consumption of cord food products. first, wheat on average accounts for an expenditure of pkr 9.03 in quintile 1 (about 16 percent of the total food expenditure in that quintile), and pkr 10.18 for households in quintile 5 but that is 9.9 percent of their total food expenditures. similarly, we can see that when moving from quintile 1 to quintile 5, there is a clear diminishing proportion of wheat with a rising one for milk—implying better dietary diversity as income increases. this analysis shows that the households in kp and balochistan are spending much of their food expenditure on wheat and milk and less on fruits, vegetables, and meat. this is not surprising given that wheat meets more than fifty percent of the caloric requirement for the poor in pakistan16. this is also related to the food culture and preferences in these provinces, where meat along with bread (chapatis) are preferred meals. also, in terms of daily serving requirements, wheat is the cheapest amongst all lowest cost items included in cord to meet caloric requirements—thus the price effect cannot be ignored. as seen in figure 1, households in quintile 1 are deriving much of their caloric requirement from wheat. as per pakistan’s fbdgs, children aged 3-10 years old are required to consume 1365 calories per day. this can be called a mid-point as 3 years old require 1000 calories per day while 10 years should consume 1800 calories a day. thus, fbdgs prepared the cord based on mid-point calorie requirement. out of the 1350 required calories, wheat has a share of 400 calories. further, the average cost per calorie of wheat for quintile 1 households is pkr 0.017 and 0.018 for quintile 5 households, cheapest across all food groups, while milk is the most expensive with pkr 0.12 and pkr 0.14 cost per calorie for quintile 1 and 5 respectively. this is followed by chicken at 0.060 and 0.062 for quintiles 1 and 5. thus, cost can be a major reason for the poorest households having a large intake of wheat in their daily diet. however, households in quintile 5 have a more balanced diet, with higher proportions of milk and 16 malik, s. j., nazli, h. l., & whitney, e. (2014, december). food consumption patterns and implications for poverty reduction in pakistan. 30th agm & conference of pakistan society of development economists. paper submitted to the 30th agm & conference of pakistan society of development economists, islamabad, pakistan. rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 66 chicken—yet much lower than the recommended intake and are still higher than fbdg requirements for wheat. for the ease of understanding, we have not included households from quintile 2 to 4 in this graph as there is a linear increase with every quintile. figure 1: cord vs actual daily expenditure on least cost items source: authors’ own estimations however, we do dig deeper into them in the later analysis. the importance of cereals also holds true the diet of afghan children. more than 70 percent of afghan children under the age of 5 were reported to consume cereals (72.9 percent) and fats (70.6 percent) as per the 2013 national nutrition survey of afghanistan (government of afghanistan, 2013). dairy and sugar also feature prominently in their diet, consumed by 67.1 and 61.7 percent of children under the age of 5 respectively. the proportion of children who consume tubers, fruits, and pulses and nuts was estimated to be 48.9, 41.5, and 27.3 percent respectively. around 30 percent children under the age of 5 consumed meat, fish, and eggs. this shows that similar to pakistan, afghan children diet also comprises mainly of cereals and milk, however, we are unable to find any data on the per capita consumption of these items by afghan children. the diet of afghans generally is comprised of halal meat, bread, rice, dairy produce, eggs, onions, peas, beans, tomatoes, dried fruit, and nuts (khakpor et al., 2019). as per the afghanistan living conditions survey 2016-17, poor afghans very rarely consume meat, eggs, dairy and dairy products, so their diet consists mainly of cereals, oil and sugar. rich afghans consume meat, eggs, and pulses two days per week, and dairy 0 10 20 30 40 50 60 70 80 90 100 e x p e n d it u r e i n p k r milk wheat onion banana chicken journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 67 products for about 5 days a week, but they eat vegetables only 3 times a week. moreover, according to dizon et al. (2019), about 35 percent of afghan households’ total food expenditure is for wheat, 23 percent is protein and 11.7 percent is for dairy and oils. although it varies by income quintile, this is similar to the kp and balochistan diet in terms of preferences, where wheat takes about 17 percent and dairy has a 15 percent share in the total food expenditures on average across both provinces. 4.2 disaggregated cord estimates for balochistan and kp the results of the estimated cord and the actual daily per capita expenditure on the nutritious diet food groups are shown in table 2. we see that, overall, the cord increases with income quintile, which shows that prices vary by households belonging to the different income quintiles, as the associated quantities are fixed for all households. this could be because low-income households generally purchase groceries from kiryana stores (small grocery kiosks) or street vendors, whereas richer households tend to purchase from supermarkets. thus, the cost of the cord differs across income groups and is captured in our estimation. in the presence of such variation, the use of survey data is a better measure than cpi. it might be expected that the overall proportions spent on key food items in the cord would differ significantly across rural and urban areas, but figure 3 suggests they do not. part of the reason is that the per capita incomes in the different regions of a province do not differ much in most quintiles. as per the hies 2018-19, the following monthly per capita incomes of all quintiles in urban and rural regions show no major differences in levels between rural and urban households in quintiles 1 to 4 (table 2). however, a significant jump occurs in quintile 5, where average per capita income in urban households exceeded that in rural ones by nearly one quarter, which probably shows up in the much larger proportion of expenditures on readymade food in urban areas. however, this is partly because these foods are less available in rural areas. table 2: quintile-wise per capita monthly incomes urban rural quintile 1 2,865 2,838 quintile 2 3,903 4,167 quintile 3 5,051 5,106 quintile 4 6,717 6,904 quintile 5 15,537 12,478 source: hies 2018-19 the cord as percentage of monthly income is inversely proportional to the quintile, as expected. as seen in figure 2, the cord for quintile 1 is highest for children aged rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 68 3-10 years in the balochistan urban region, where on average, a household would have to spend 107 percent of per capita monthly income to provide the cord food items for their child. the cord for quintile 5 is found to be highest in kpk rural, i.e., at pkr 105.6 per capita per day. this translates into only 25 percent of total per capita monthly income for that quintile (see figure 3). in other words, an individual falling in quintile 1 from balochistan and kp (both urban and rural) will have to spend around 100 percent of their monthly income only to cover the expense of a recommended diet for a child alone (figure 2). almost all regions have a similar situation when moving from quintile 1 to quintile 5—the proportion of income required for maintaining cord decreases with the increase in income. these declines are largest as households move from quintile 1 to 2, or from quintile 4 to 5. so, households in quintile 1 are in significantly worse shape than quintile 2. punjab urban households in quintile 5 spend 19 percent less of their per capita income than those in quintile 4, an increment that is similar in urban areas of balochistan and kp. figure 2: cord as a percentage of per capita income source: authors’ own estimations these results can also be interpreted as measuring the state of competition in various provinces, and in urban versus rural areas. the cord is lower for punjab urban for all quintiles, implying lower food prices there —likely because that province is the country’s supply center for many staple foods, as well as due to higher connectivity between farms and urban areas it is logical that in balochistan, with many rural areas being very remote, prices are higher, while in punjab, extensive farm to market linkages 0% 20% 40% 60% 80% 100% 120% punjab urban kpk urban balochistan urban kpk rural balochistan rural quintile 1 quintile 2 quintile 3 quintile 4 quintile 5 journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 69 have developed and there are likely competitive food markets throughout the province. similarly, prices are higher in rural areas in balochistan and kp as food commodities from punjab and sindh are transported to those areas, incurring added costs that are reflected in the market prices. for instance, poultry farms are generally located across urban peripheries and thus, rural areas face higher poultry prices due to transportation costs. similarly, households in quintile 5 face higher prices because the quality of their consumed products might be superior. although, it cannot be validated from the given data, it is likely that individuals with higher incomes purchase better quality products, which in turn increases the cost of their recommended diet. given the low purchasing power and higher prices, quintile 1 households in balochistan rural are only spending 21 percent of the expenditure needed to reach a nutritious diet. if they consume the recommended diet, they will have very little left to cover other expenses such as the child’s education, health, clothing, and others. further, we calculate the actual expenditure as a percentage of cord across regions, where differences among groups is driven by price and/or income differentials. as seen in figure 3, the expenditure as a percentage of cord is highest in quintile 5, punjab urban, at 48 percent. however, the expenditure is still 52 percent lower than the required amount needed to reach the cord (figure 3). similarly, we found that underconsumption is highest in balochistan urban and kp urban, i.e., both spend only 17 and 18 percent respectively of the recommended cord, equivalent to a 30-percentage point gap between these regions and the baseline (punjab urban quintile 5). this shows that balochistan and kp are lagging significantly behind the baseline—even as baseline is not an ideal scenario. moreover, quintile 1 households in balochistan rural and kp rural are spending 19 of the required expenditure. the slightly better consumption in rural areas could be attributed to home-grown staples, and hence lesser vulnerability to market price. however, even quintile 5 households in kpk and balochistan are spending between 30 to 37 percent of the recommended diet—much lower than required for healthy growth of children. these findings show that there is little divergence between regions in each quintile, and a dire situation exists for all in quintile 1 households. figure 3: actual expenditure per capita as percentage of cord rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 70 source: authors’ own estimations the findings suggest that households belonging to quintiles 1-3 are all in poor shape, as their expenditure per capita on children aged 3 to 10 years accounts for between 21 and 36 percent of cord respectively, regardless of geography. in quintiles 4 and 5, however, region appears affect the divergence in nutritious consumption. for instance, in quintile 4, both balochistan and kp lag punjab urban, which is even more obvious for quintile 5. the high disparity between the expenditure of households in quintile 5 relative to other quintiles in punjab urban could be a reflection of wealth concentration in lahore. overall, none of the households are spending the desired amount on the nutritious diet recommended by the fbdgs, and thus most children between the ages of 3-10 years are consuming less than a nutritious diet, likely leading to development challenges. because of the extremely high proportion of per capita income covered in the cord, behavioral decisions that lower consumption for 3–10year-olds will likely have even higher proportional impacts on such children with rising prices due to the influx of refugees expenditures on specific commodities: figures 4 and 5 show the expenditures on specific food groups as a percentage of the recommended diet for children aged 310 years. the figures show obvious heterogeneity across food groups, reflecting consumption preferences and costs per calorie for the pakistani population (children aged 3-10 years) as a whole, as well as across regions and income groups. first, we analyze the expenditure on cereals, which has the highest share in consumption 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% punjab urban balochistan rural kpk rural balochistan urban kpk urban quintile 1 quintile 2 quintile 3 quintile 4 quintile 5 journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 71 expenditure for quintile 1 across regions in our data (figure 4). (only milk has a greater expenditure proportion for quintile 5 households. see figure 5). wheat, which is the national staple, was found to be the lowest cost item among cereals and hence features prominently in the consumption pattern described below. the highest consumption of cereals in quintile 1 is in kp rural, i.e., at about 140 percent of the recommended daily intake, followed by kp urban and balochistan rural, each consuming 135 percent of the recommended cereal intake. punjab urban is spending relatively lesser on cereals (wheat) as a percentage of the required diet, but still consuming three percent less than the required diet. this shows that there is overconsumption of cereals by almost 30 to 40 percent in the poorest households. cereals are the only category in the nutrition basket in which balochistan and kp are consuming at or over the required intake. if households in kp and balochistan follow the pattern set by punjab urban ones, they will reduce their expenditures on cereals by forty percent or more as income grows to punjab levels figure 4: expenditure on cereals as a percentage of the required diet source: authors’ own estimations however, moving back to cord, the lower consumption of cereals by children aged 3-10 years in quintile 5 households, relative to quintile 1, can be explained by significant substitution towards meat (i.e., chicken) or milk and dairy (see figure 5). for instance, chicken consumption of punjab in quintile 1 is 05 percent of the required expenditure, compared with 20 percent in quintile 5. the corresponding figures for milk consumption are 20 percent for quintile 1, compared to 60 percent for quintile 5 in punjab. this pattern also is evident in the earlier figure 1, where the proportions of expenditures on wheat and milk vary clearly across the two locations. 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% 120.0% 140.0% 160.0% 180.0% punjab urban kpk rural kpk urban bal rural bal urban q1 q2 q3 q4 q5 rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 72 overall, the gradual increase in the consumption of these food groups can be seen across regions as we move from quintile 1 to quintile 5, with the highest jump from quintile 4 to quintile 5, particularly in the case of meat and fruits. all of the graphs show positive relationships between dietary diversity and income across all regions. journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 73 figure 5: expenditure as a percentage of required dietother than cereals 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% punjab urban kpk rural kpk urban bal rural bal urban meat & pulses q1 q2 q3 q4 q5 0.0% 20.0% 40.0% 60.0% 80.0% punjab urban kpk rural kpk urban bal rural bal urban milk and dairy q1 q2 q3 q4 q5 0.0% 10.0% 20.0% 30.0% 40.0% punjab urban kpk rural kpk urban bal rural bal urban vegetables q1 q2 q3 q4 q5 0.0% 10.0% 20.0% 30.0% punjab urban kpk rural kpk urban bal rural bal urban fruits q1 q2 q3 q4 q5 rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 74 one common pattern across regions in quintile 1 is that no region spends more than 21 percent of the amounts needed for a nutritious diet in any food group. this highlights the overconsumption of cereals and relatively significant underconsumption of other important food groups – which can be explained by their low-income and dependence on cereals for calorie adequacy. the analysis shows great variation in consumption among food groups – for instance vegetables (i.e., for which onions had the least cost) are consumed the most across regions on average, whereas fruits (i.e., bananas) are consumed the least. moreover, within regions, punjab urban leads in the consumption of milk and dairy, and fruits, while balochistan urban is consuming more vegetables, and meat and pulses than punjab and kp. kp lags punjab urban across all food groups except meat and pulses, with the deficiency highest in milk and dairy, i.e., by about nine percentage points. the highest dispersion among regions across food groups is found in milk and dairy, where on average other regions are consuming 30 percentage points less than punjab urban. 4.3 scenario analysis: post-refugees influx in this section, we provide estimates for three different scenarios of afghan refugee influx into pakistan. the scenario analysis is based on the refugees’ influx anticipated by the government, with a fifty percent deviation. the government of pakistan is anticipating 700,000 new afghan refugees by the first quarter of 2022.17 nevertheless, given border controls and lack of a clear policy at present regarding accepting further refugees, the actual number may be less than the anticipated number. on the other hand, however, the difficult political and economic situation in afghanistan in the future may result in lax policy and greater humanitarian effort, leading to a higher than anticipated influx of refugees. thus, we have also estimated a conservative (350,000) and liberal (1,050,000) figure of refugees’ influx (scenarios 2 and 3). table 2 breaks down dietary expenditure estimates for kp based on region and income groups for all three scenarios. the table shows estimates for pre and post refugees’ influx, where the post-influx results are the expenditures on a nutritious diet by the local population following an increase in food prices from increased demand. 17 rana, s. (2021, july 18). hosting 700k afghans will cost $2.2 b for 3 years. the express tribune. retrieved from https://tribune.com.pk/story/2311234/hosting-700k-afghans-will-cost-22-b-for-3-years journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 75 the results show the law of demand at work, in that an increase in general price level results in a decrease in quantity demanded, and hence expenditure across food items. as expected, a higher number of refugees creates a lower expenditure on a nutritious diet. in scenario 1, we introduced a shock of 700,000 refugees which results in a 2.07 percent increase in prices across food items (see annex a). the corresponding percentage increase in prices for scenario 2 and 3 is 1.04 and 3.11 percent respectively. as shown in table 3, we found that the impact of the arrival of refugees on dietary expenditure of local children aged 3-10 increases with the income of the household. however, the gap increase remains under one percentage point – which translates to just pkr 9 per capita per month on average – under scenario 1. in all scenarios, a common finding is that rural households face a higher impact of increased food prices and the biggest increase in gap is seen in q5 rural households i.e., 3.28 percent. however, shortages are likely to be perceived as more pressing among the lower quintiles, leading to higher more deficiencies in the 3–10-year-old’ diets. overall, if the number of refugees remains under 700,000, the adverse impact on existing dietary expenditure, even for children aged 3-10, would not be high. however, if the number of refugees exceeds one million, then the adverse impact on children’s nutritious diet could be relatively higher. as a fiscal response to the increase in food prices, government can opt for various interventions. these could include providing subsidies on staple items, decreasing income or indirect taxes, and/or increasing purchasing power through transfer payments. pakistan has already run a successful ehsaas cash disbursement program during the first wave of covid-19 in the second quarter of 2020. however, depending upon the own and cross price elasticities and income elasticities, each intervention for a particular commodity will give different results. table 3: impact of refugee influx on existing dietary expenditures of children aged 3-10 years in kp scenarios income group change in existing dietary gap rural (pp) change in existing dietary gap urban (pp) scenario 1700,000 refugees q1 0.74 0.70 q2 0.96 0.83 q3 1.17 0.99 q4 1.47 1.15 q5 1.99 1.6 scenario 2350,000 refugees q1 0.30 0.29 q2 0.39 0.34 rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 76 q3 0.48 0.41 q4 0.61 0.47 q5 0.77 0.66 scenario 3 1,050,000 refugees q1 1.30 1.22 q2 -1.68 1.46 q3 2.05 1.74 q4 2.59 2.03 q5 3.28 2.81 source: authors’ own estimations note: gap refers to the difference between cord and actual expenditure by the households and is estimated in terms of change in percentage points (pp) with the help of available price and income elasticities, our simulations found that a subsidy of more than seven percent in wheat could in fact increase the gap in dietary expenditures.18 this could be explained by the fact that the existing consumption of wheat is already one and a half times higher than the recommended diet across all income groups and regions. therefore, lowering the price of wheat can be counterproductive and could eventually increase the gap. on the flip side, providing a similar subsidy to other commodities may result in a decrease in the expenditure gap, something that is desirable. for example, we found that the highest decrease in the gap would be when a subsidy is provided to fresh milk. a seven percent subsidy could nearly eliminate the increase in gap caused by the increased food prices post-influx. lastly, the price changes that are developed for our analyses are generally similar across all commodities, as we use estimates reflecting overall price effects from rising immigration. however, it could be that if markets are not well integrated, prices could vary more than the average in our simulations for selected commodities. to assess this possibility, in figure 7 we show that prices of chicken and milk vary considerably across households in the different regions, both for the quintile 1 and 5 households. in contrast, markets for wheat, onion, and banana, are relatively well integrated and the prices only vary by small amounts regardless of where households are located. this implies that the shocks and the difference in consumption expenditures across regions are not mainly due to prices but rather to income and preferences. even in the case of chicken, where price variation is high, the quantities consumed are so low with respect to the recommended diet that these differences are not likely due to price differences. the 18 the seven percent subsidy number is derived from the government of pakistan’s plan to provide a subsidy equal to pkr 2.5 per kg, or about seven percent of the market price of wheat per kg. see: rana, s. (2021, september 14). pm okays lifting of 32% flour subsidy. the express tribune. retrieved from https://tribune.com.pk/story/2320136/pm-okays-lifting-of-32-flour-subsidy journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 77 main commodity where local markets might have distinctly different price behavior is milk where the percentage variation around the households means in hies is high. figure 7: price differentials across commodities for quintile 1 and 5 households source: authors’ own estimations 5. conclusion and policy implications this paper highlighted the existing malnutrition of children between the ages of 310 years at the subnational level for pakistan. we estimated the least cost for a recommended nutritious diet using the hies 2018-19 data. the results suggest a significantly larger gap in the actual expenditures on a nutritious diet in the provinces of kp and balochistan when compared to the relatively prosperous region of urban punjab. we also found that the expenditures are significantly higher in cereals but drastically lower in fruits and vegetables. overall, the results suggest that cord is very high relative to the per capita income of quintile 1; however, even children in households from higher quintiles are consuming on average under 40 percent of the recommended diet in the provinces of kp and balochistan. moreover, the percentage impact of refugee influx on dietary expenditures of children in kp is expected to be higher for quintile 5 households in the rural regions while almost insignificant for households in lower quintiles in both rural and urban areas. however, as indicated earlier, if children 3-10 years of age have less priority in household food allocation, they are likely to be more at risk in the lower quintiles, where the challenges of meeting a nutritious diet are even higher. in expectation of the arrival of afghan refugees in pakistan, wfp has prepared a six-month contingency plan to provide support to the refugees and their host 0 5 10 15 20 25 30 milk (ltr) wheat (kg) onion (kg) banana (num) chicken (kg) s ta n d a rd d e vi a ti o n rural q1 urban q1 rural q5 urban q5 rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 78 population19. to finance the plan, the wfp has sought funding of usd 23.2 million, but only usd 2.1 million had been secured as of september 2021. however, even if wheat is donated to all of the refugees, other dietary products such as meat, milk, fruits and vegetables are to be purchased by the market at the market prices. this would put pressure on the prices due to increased demand even in the presence of international food transfers. in terms of policy implications, the first and foremost policy should be to improve the existing dietary diversity of children, especially for the marginalized households falling in quintile 1. there is a significant overconsumption of wheat by more than 1.5 times, resulting in significant underconsumption of other necessary items such as milk, fruits, and vegetables. there could be three possible policy interventions to balance out this asymmetry in consumption: (a) cash transfers to quintile 1 households; (b) providing direct food transfers with the help of international donors; or (c) subsidizing dietary items for all households. if (a) is adopted, the income elasticity of various food items for kp indicates that households are likely to spend more on poultry, beef, eggs, and milk than on wheat. for example, given the existing income elasticities of kp rural, a 10 percent increase in income may result in 10 percent increase in poultry, 9.3 percent in milk and 14 percent in fruits compared with just 3.6 percent in wheat. the same is true for kp urban households. this implies that any further increase in income through cash transfers, such as the ehsaas program, may lead towards balancing of the diet. refugees are not eligible for the ehsaas program but receive aid from international donors from time to time. in case of (b), there exists evidence of similar interventions in the past. the local population has historically benefitted from food related aid, such as direct food transfers, cash transfers to purchase food, and through food-for-work programs from international donors. for instance, following drought conditions in the last few years, the wfp made a provision of 4,200 metric tons of food and cash transfers worth usd 88,000 to nearly 315,000 people in pakistan in august 2019 alone, which went to vulnerable groups in sindh and balochistan and supported internally displaced people in kp.20 19 wfp. (2021). wfp pakistan country brief. retrieved from https://reliefweb.int/sites/reliefweb.int/files/resources/wfp-0000132944.pdf 20 usaid. (2019). pakistan – complex emergency and drought. retrieved from https://reliefweb.int/sites/reliefweb.int/files/resources/wfp-0000132944.pdf journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 79 additionally, almost 401,000 vulnerable individuals impacted by the harsh winter in azad jammu and kashmir (ajk) and balochistan were provided direct food transfers by wfp from february to june 2020.21 more recently, food transfers were provided to 6,870 households including flood affectees in the lower kohistan district of kp and to the poor families suffering from malnourishment in upper dir and khyber districts.22 on the other hand, an across-the-board commodity-specific subsidy will benefit both poor and rich households as it is difficult to exclude the rich from availing food subsidies until and unless the government initiates a food stamp program—something that would require a greater administrative cost. governments in pakistan have been providing a wheat subsidy to rural and urban households for many years and even in 2021, the government provided pkr 14.3 per kg subsidy on wheat to urban households.23 as wheat remains one of the most highly regulated crops in pakistan, administering the subsidy on it is relatively easy. however, any further provision of a subsidy on wheat may lead to a more imbalanced diet, whereas, a same subsidy on milk and poultry could partially offset the existing imbalance in dietary intake of children. 21 usaid. (2020). food assistance factsheet: pakistan. retrieved from https://www.usaid.gov/sites/default/files/documents/1866/ffp_pakistan_fact_sheet.pdf 22 wfp. (2021). wfp pakistan country brief. retrieved from https://reliefweb.int/sites/reliefweb.int/files/resources/wfp-0000132944.pdf 23 rana, s. (2021, september 14). pm okays lifting of 32% flour subsidy. the express tribune. retrieved from https://tribune.com.pk/story/2320136/pm-okays-lifting-of-32-flour-subsidy rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 80 references akgündüz, y. e., van den berg, m., & hassink, w. (2015). the impact of refugee crises on host labor markets: the case of the syrian refugee crisis in turkey (no. 8841). iza discussion papers alix-garcia, j., & saah, d. (2010). the effect of refugee inflows on host communities: evidence from tanzania. the world bank economic review, 24(1), 148-170 allen, r. c. (2017). absolute poverty: when necessity displaces desire. american economic review, 107(12), 3690-3721 amparado d., dempster, h., & laghari, i. k. (2021, august 25). with us withdrawal, rights of afghan refugees in pakistan hang in the balance. retrieved from https://www.cgdev.org/blog/us-withdrawal-rights-afghan-refugees-pakistanhang-balance argaw, t. l., phimister, e., & roberts, d. (2021). from farm to kitchen: how gender affects production diversity and the dietary intake of farm households in ethiopia. journal of agricultural economics, 72(1), 268-292. baloch, a., shah, s. z., noor, z. m., & lacheheb, m. (2017). the economic effect of refugee crises on neighbouring host countries: empirical evidence from pakistan. international migration, 55(6), 90-106 br research. (2021, june 26). net food importer – since 1990. business recorder. retrieved from https://www.brecorder.com/news/40102732 chambers, r. (1986). hidden losers? the impact of rural refugees and refugee programs on poorer hosts. international migration review, 20(2), 245-263 chastre, c., duffield, a., kindness, h., lejeune, s., & taylor, a. (2007). the minimum cost of a healthy diet: findings from piloting a new methodology in four study locations. london: save the children uk. dawn. (2021, august 28). un warns of up to 500,000 more afghan refugees by year-end. retrieved from https://www.dawn.com/news/1643034/un-warns-of-up-to500000-more-afghan-refugees-by-year-end deaton, a. (1988). quality, quantity, and spatial variation of price. american economic review, 78, 418-430 journal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 81 deptford, a., allieri, t., childs, r., damu, c., ferguson, e., hilton, j., parham, p., perry, a., rees, a., seddon, j., & hall, a. (2017). cost of the diet: a method and software to calculate the lowest cost of meeting recommended intakes of energy and nutrients from local foods, bmc nutrition, 3 dizon, f., wang, z., & mulmi, p. (2021). the cost of a nutritious diet in bangladesh, bhutan, india, and nepal. world bank policy research working paper 9578, washington, dc: world bank dizon, f., herforth, a., & wang, z. (2019). the cost of a nutritious diet in afghanistan, bangladesh, pakistan, and sri lanka, global food security, 21, 38-51 ecker, o., & comstock, a.r. (2021). dataset of income and price elasticities of food demand: estimates from complete demand systems for selected african and asian countries. washington, dc: international food policy research institute(ifpri). https://ebrary.ifpri.org/utils/getfile/collection/p15738coll2/id/134675/filename/134884.p df enghoff, m., hansen, b., umar, a., gildestad, b., owen, m., & obara, a. (2010). in search of protection and livelihoods: socio-economic and environmental impacts of dadaab refugee camps on host communities. nairobi: royal danish embassy haider, a. zaida, m. (2017). food consumption patterns and nutrition disparity in pakistan, mpra working paper no. 83522 food and nutrition technical assistance. (n.d.). optifood. retrieved from https://www.fantaproject.org/tools/optifood furqan, a. (2021, july 13). pakistan refuses to host additional afghan refugees. retrieved from https://www.voanews.com/a/south-central-asia_pakistan-refuseshost-additional-afghan-refugees/6208191.html galasso, e and wagstaff. a. (2018). the aggregate income losses from childhood stunting and the returns to a nutrition intervention aimed at reducing stunting. world bank policy research working paper 8536, washington, dc: world bank global hunger index. (2020). pakistan. retrieved from https://www.globalhungerindex.org/pakistan.html government of pakistan, & unicef. (2018). national nutrition survey 2018: key findings report. ministry of national health services pakistan grare, f., & maley, w. (2011). the afghan refugees in pakistan. j middle east institute. refugees cooperation rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 82 håkansson, a. (2015). has it become increasingly expensive to follow a nutritious diet? insights from a new price index for nutritious diets in sweden 1980–2012. food & nutrition research, 59(1), 26932 jensen, r. t., & miller, n. h. (2010). a revealed preference approach to measuring hunger and undernutrition (no. w16555). national bureau of economic research khakpor, m., iqbal, r., hussain, n.g., engler-stringer, r., koc, m., garcea, j., farag, m., henry, c., & vatanparast, h. (2019). facilitators and barriers toward food security of afghan refugees residing in karachi, pakistan. ecology of food and nutrition, 58 (4) lapping, k., schroeder, d., marsh, d., albalak, r., & jabarkhil, m. z. (2002). comparison of a positive deviant inquiry with a case-control study to identify factors associated with nutritional status among afghan refugee children in pakistan. food and nutrition bulletin, 23(4_suppl_1), 26-33 maillot, m., vieux, f., delaere, f., lluch, a., & darmon, n. (2017). dietary changes needed to reach nutritional adequacy without increasing diet cost according to income: an analysis among french adults. plos one, 12(3), e0174679 masters, w., bai, y., herforth, a., sarpong, d. b., mishili, f., kinabo, j., & coates, j. c. (2018). measuring the affordability of nutritious diets in africa: price indexes for diet diversity and the cost of nutrient adequacy. american journal of agricultural economics, 100(5), 1285-1301 mulik, k., & haynes-maslow, l. (2017). the affordability of myplate: an analysis of snap benefits and the actual cost of eating according to the dietary guidelines. journal of nutrition education and behavior, 49(8), 623-631 o'brien-place, p. m., & tomek, w. g. (1983). inflation in food prices as measured by least-cost diets. american journal of agricultural economics, 65(4), 781-784 omiat, g., & shively, g. (2017). charting the cost of nutritionally-adequate diets in uganda, 2000-2011. african journal of food, agriculture, nutrition and development, 17(1), 11571-11591 poi, b. p. (2012). easy demand-system estimation with quaids stata journal, 12(3):433–446(14). retrieved from http://www.statajournal of applied economics and business studies, volume. 6, issue 1 (2022) 53-84 https://doi.org/10.34260/jaebs.613 83 journal.com/article.html?article=st0268. stigler, g. j. (1945). the cost of subsistence. journal of farm economics, 27(2), 303-314 unhcr. (2021, august 31). pakistan (islamic republic of). retrieved from https://data2.unhcr.org/en/country/pak#category-4-258975 rahema obaid, tehseen ahmed qureshi, stephen davies & abdul wajid rana 84 annex a: estimation of refugees’ scenario scenario 1 total expected refugee influx 700,000 bal kp refugees 159,600 406,700 existing population 14,083,862 39,272,000 population after refugees 14,243,462 39,678,700 % increase in population 1.13% 1.04% impact on inflation 2.3% 2.07% scenario 2 total expected refugee influx 350,000 bal kp refugees 79,800 203,350 existing population 14,083,862 39,272,000 population after refugees 14,163,662 39,475,350 % increase in population 0.57% 0.52% impact on inflation 1.1% 1.04% scenario 3 total expected refugee influx 1,050,000 bal kp refugees 239,400 610,050 existing population 14,083,862 39,272,000 population after refugees 14,323,262 39,882,050 % increase in population 1.70% 1.55% impact on inflation 3.4% 3.11% journal of applied economics and business studies, volume.4, issue 3 (2020) 137-158 https://doi.org/10.34260/jaebs.436 137 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x does economic integration affect the co-movement between financial markets? evidence from pakistani equity market with its global partners ahmad fraz*, arshad hassan & sumayya chughtai 1 assistant professor, pakistan institute of development economics islamabad 2 professor, capital university of science and technology islamabad 3 assistant professor, international islamic university, islamabad abstract the study investigates the impact of bilateral trade, economic fundamentals and financial crisis on the equity market integration (emi) of pakistan’s equity market with its major global trading partners (china, india, usa and uk) for the period 1998 to 2016. the findings of the study indicate that bilateral trade and economic conditions have a significant impact on emi, the export dependence of two economies may increase the emi and import dependence reduces the emi of two economies. moreover, inflation differential and volatility in the bilateral exchange rate have a negative impact on emi. it implies that inflation rates in pakistan’s equity market are higher as compare to other markets and volatility in bilateral exchange rate may reduce trade flows and its tendency to follow other market (bracker, docking , & koch, 1999). furthermore, the financial crisis in an economy may reduce the emi with its trading partners and emi between different markets is affected by their bilateral economic fundamentals. the results imply that financial integration between different markets is affected by their bilateral economic fundamentals. the study has strong implications for international investors who need to assess risks and benefits associated with international portfolio diversification. keywords bilateral trade; bilateral economic conditions, equity market integration; developed markets, regional markets jel classification f36, g11, g15 1. introduction * aahmadfraz@gmail.com mailto:aahmadfraz@gmail.com ahmad fraz, arshad hassan & sumayya chughtai 138 the integration between the capital markets is an important domain in financial economics that covers various facets of the interrelationship across equity markets. the current study is an attempt to explore the bilateral macroeconomic factors that explains the co-movement between different equity markets. in last couple of decades, the financial and economic linkages among various economies have increased and resulted in the synchronization among equity markets. numerous studies have explored emi between different equity markets in a linear setting by using cointegration analysis and var framework. most of the work in this domain has been done in the usa (johnson & soenen, 2002; tai, 2007; dimitriou & simos , 2014) but in recent past the emerging markets have also grabbed the attention of the researchers due to their growing importance (wang & moore, 2008; simpson, 2008; karagoz & ergun, 2010; peir´o, 2016; paramati et al., 2018). although, recent work does not focus on determining the economic factors behind co-movement of equity markets and a few studies have been conducted to explore different macroeconomic forces that impact emi of developed and emerging markets i.e. robert and soenen (2003) explore the impact of trading relations on co-movement between south american markets and the usa. existing studies attribute this growing emi to the rise in financial integration between world economies but appear to ignore the presence of real economic integration. financial integration covers markets within the same region or the existence of common attributes of the markets (norén, 2011; bouët et al., 2012; peir´o, 2016; thomas et al., 2017). on the other hand, economic integration specifically deals with bilateral economic ties and cross-border investment. literature also cites some other reasons behind economic integration such as providing convenience to market participants, relaxing rules, technological advances, and removals of over regulations. these dynamics of the market contribute to increase emi between economies, which, in turn, give rise to synchronization among the equity markets. büttner and hayo (2011) study the role of exchange rate risk in weakening the integration among equity markets in the european union region. the degree of emi and economic integration between two countries is generally revealed through the degree of equity market synchronization (joyo & lefen, 2019). this particular study is an attempt to contribute to the second strand of literature by investigating the economic factors that influence emi. the study focuses on pakistan that is an emerging market and has been identified as one of the best performing markets during the current decade. the integration of the pakistani market with the us and the uk is interesting from the historical trade and finance perspective. the us and the uk journal of applied economics and business studies, volume.4, issue 3 (2020) 137-158 https://doi.org/10.34260/jaebs.436 139 have economic and social ties with pakistan in areas ranging from education to energy, trade, and investment. the us is pakistan’s largest export destination country followed by the uk. during the last decade, it has been observed that economic power is shifting towards china and india. it is noteworthy that both countries share the border with pakistan. china has a long history of economic relations with pakistan and has contributed significantly in infrastructure development. statistics highlight some interesting facts regarding economic bonding with china. imports from china have touched to the level of 28% of total cumulative imports of pakistan. the role of china has further increased after china pakistan economic corridor (cpec hereafter) initiative, as more than $50 billion are allocated for energy and infrastructure related projects. these long-term projects are indicators of potential economic integration between the two countries. india is becoming one of the major emerging market of the world. it is the largest economy of south asia and a member of the south asian free trade area (safta). however, it has a long history of political issues with pakistan that dampens the opportunities for economic cooperation. despite these adverse relationships with india, pakistan’s export to india has significantly increased from 2003 to $403 million in 2013 and then decreases to $312 million in 2015. on the flip side, imports have also increased from 2003 to 2013 from $184 million to $2.18 billion and reported at $1.96 billion in 2015. it is further added that informal trade between these counties is $ 4.71 billion. these global economic trend calls into question whether these markets are integrated with pakistan or not. keeping in view the above scenario, it is imperative to unfold the causes and drivers of the integration of pakistan’s equity market with two global and two regional giants by using bilateral trade relations and economic fundamentals, such as inflation differentials, real interest rate differential, ∆ bilateral exchange rate, variation in the exchange rate of these countries and gdp growth rate differential. the research on determinants of integration among the equity market has significant implications for the investors who want to seek portfolio diversification and for the policymakers and regulators who are keen to know about factors that may affect the national equity market and financial stability of the country (chen, 2018). it is hypothesized that exports dependence has positive impact on emi whereas a negative impact of imports dependence is expected. the macroeconomic variables including inflation, real interest rate differential, and ∆ bilateral exchange rate, annual growth rate differential are expected to have negative impact on emi. the financial ahmad fraz, arshad hassan & sumayya chughtai 140 crisis is expected to decrease the integration between the markets and during the time of volatility, it may have a positive and negative impact on emi. the rest of the study is organized as under: section ii explains the literature on the integration of equity markets and its determinants, section iii covers the data description and methodology employed for analyzing the data, section iv reports results of data analysis and section v concludes the study and offers remarks for future research implications. 2. review of literature economic integration has long been debated over the past couple of decades in the context of trade creation and trade diversion. integration among equity markets provides diversification opportunities for the investor to mitigate unsystematic risk. as the portfolio managers use global diversification strategy as a risk-hedging tool, therefore, international portfolio managers need to understand the behavior of the market such as co-movement, sensitivity, and exogenous shocks. economic theory suggests that there is an expectation of co-movement between different pairs of financial time-series in the long-run. the relationship between these series may not hold in the short run, but this deviation does not hold due to the market forces, government regulations and investor’s preferences (ghosh, saidi, & johnson, 1999; dhanaraj, gopalaswamy, & babu, 2017; ge, wu, zhang, & zou, 2019). numerous studies have documented the dynamic linkages between different equity markets and volatility transmission mechanisms based on such interdependence (hasan, saleem, & abdullah, 2008; abbas, khan, & shah, 2013; fraz & hasan, 2016; paramati et al., 2018; kim, choi, & kim, 2019). roll (1992) investigates the disparate behavior of the different equity market indices and report that some market indices are highly diversified as compared to the others. bodurtha, chinhyung and senbet (1989) suggest that some macroeconomic factors are persistent and are the prime source of information, which leads co-variation among the returns of different financial time-series. campbell and hamao (1992) argue that degree of economic integration keeps on changing between any two pairs of countries and it is expected that the emi between two markets may also change. bracker, docking and koch (1999) have reported that emi between two markets may depend upon certain macroeconomic factors such as bilateral trade, bilateral exchange rate, bilateral gdp, etc. the characteristics of these macroeconomic variables may increase/decrease the degree of economic integration between these countries. vithessonthi and kumarasinghe (2016) argue that financial liberalization between two journal of applied economics and business studies, volume.4, issue 3 (2020) 137-158 https://doi.org/10.34260/jaebs.436 141 countries provide an opportunity for local firms to borrow from the other countries and investors may invest in portfolios of the equities or import to the competing firms in another country. the study suggests that export-based economies can improve their bilateral export with other countries and their financial integration has a significant and positive relationship with international trade. johnson and soenen (2002) have investigated whether the financial markets of american region are integrated with the us equity market. the study suggests that bilateral imports, bilateral exchange rate volatility are the important factors behind emi. yu, fung and tam (2010) explore the emi of ten asian markets by using high frequency indicators. the findings of the study document emi between emerging and developed markets is not the same, due to the different attributes like institutional settings, economic and political factors across asia. the study further implies that no such definition of financial integration is available, which is unanimously accepted in the literature. different studies have used correlations and global betas as a proxy of the financial integration between the markets (morana & beltratti, 2008; wälti, 2011; bekaert, erb, & harvey, 1997; vo, 2018) and r2 obtained from the asset-pricing model (schotman & zalewska, 2006; pukthuanthong & roll, 2009; vithessonthi & kumarasinghe, 2016). pukthuanthong and roll (2009) have suggested that the correlation between two equity markets is a weak indicator of financial integration. even though the index of each of country has different responsiveness towards these factors, it may not reveal the same perfect correlation. further states that the beta may vary due to leverage, level of production, industrial structure and representation of equity market. emi has been measured by using different measure, a lot of studies measures emi by using asset pricing model’s r2as a proxy of equity price synchronicity (morck, yeung, & yu, 2000; dasgupta, gan, & gao, 2010; vithessonthi & kumarasinghe, 2016; farooq, ahmed, & bouaddi , 2018). vithessonthi and kumarasinghe (2016) examine the effect of financial development and economic integration (by using international trade integration as a proxy) on emi and employ the same r2 approach based on a multifactor model of pukthuanthong and roll (2009) by using the sample of fifteen developed and developing countries of asia. the findings suggest that financial development has a significant and positive relationship with emi, whereas economic integration is not associated with bilateral emi. during the last two decades’ globalization, cross border flows and economic interdependency reshaped the dynamics of integration among equity markets. pakistani ahmad fraz, arshad hassan & sumayya chughtai 142 market as observed significant changes in socio-political and economic front during this period. therefore, it is need of the time to revisit the intensity of the integration along with other factors influencing the phenomena. 3. research methodology this section discusses the framework for data analysis from data collection to sample framework and employed methodology for the analysis. 3.1 data collection and sample description the study employs monthly closing prices of equity market indices of pakistan i.e., pakistan stock exchange (psx), china i.e., sse composite index (sse), india i.e., bombay stock exchange (bse), us i.e., standard and poor’s 500 (s&p), uk i.e., financial times stock exchange (ftse) 100 index and msci global index. the data for this study ranging from june 1998 to june 2016 to test the emi between pakistan’s equity market index along with msci global index with sample countries equity markets. however, data from 1998 to 2001 is consumed for the calculation of emi. the study takes into the account emerging regional markets of pakistan, india and china and developed markets of the u.s. and the u.k. as sample countries. prior to the partition, both india and pakistan as sub-continent have remained colonies of the great britten. even, after independence, both countries are linked with the u.k. in terms of trade. however, with the passage, of time the u.s. dominates asian markets as a global leader due to the liberalization of the capital markets. it is worthwhile to shed light on the increasing influence of china in the recent past due to its trade magnitude and announcements of different mega infrastructure project within the region. due to the above-mentioned reasons, the sample for these countries has been selected to test the phenomenon that how the world developed markets and regional emerging markets have integrated with pakistan’s market. to examine the inter-linkages between psx with each of the countries monthly index return is calculated as: monthly index return = ln ( 𝑃𝑡 𝑃𝑡−1 ) (1) where, ln is natural logarithm considers continuous compounding assumption, pt is the closing index price at the end of the month t and pt-1 is the closing index price at the end of the montht-1. the study uses two sets of economic determinants. first, potential economic determinants of international emi are based on bilateral trade including exports and journal of applied economics and business studies, volume.4, issue 3 (2020) 137-158 https://doi.org/10.34260/jaebs.436 143 imports contribution of the countries. second, set of the variables is used to estimate the economic conditions inflation, real interest rate, gdp and exchange rate. the data of macroeconomic variables is obtained from the comtrade, wdi and ifs database for the period of 2001 to 2016. 3.1.1 variables description 3.1.1.1 dependent variable multifactor model using r2, the proxy of sps is used as emi and it is measured by using the capital asset pricing model. to estimate, emi with other trading partners the degree of association is calculated by using the return of stock index of trading partner and msci global index with the return of pakistan’s stock index (an & zhang, 2013; vithessonthi & kumarasinghe, 2016; fraz & hassan, 2017; farooq, ahmed, & bouaddi, 2018).the study uses monthly market returns of pakistan’s equity market and msci global index with the returns of trading partners’ equity market for36 months to estimate r2 for each year from 1998 to 2016. the following model is used to estimate r2. rjt = α0 + β itr it + β wtrwt + et (2) rj,t is the return of stock index of country j at time t, ri,t is the return of stock index of country i at time t, rw,t is return of msci global index w at time t and ej,t is the error term. for bilateral financial integration of country j with each of the trading partner country,i is measured by using r2 measure proposed by vithessonthi and kumarasinghe (2016)and the log transformation of this r2is used in the second step to address the problem of boundedness of r2. synchjit = ln ( r2𝑗𝑖𝑡 1−r2𝑗𝑖𝑡 ) (3) r2 in equation 3 is the r2 estimated from the equation 2. this log transformation of the ratio of explained portion (of returns using market bilateral country return and world index return) versus unexplained variance component (of returns using market bilateral country return and world index return) of the returns. 3.1.1.1 independent variable as discussed earlier, the degree of emi between two countries is dependent upon the nature of bilateral trade and different economic conditions. in the second step, the degree of association between emi and economic fundamentals has been examined by using potential economic factors and their economic market conditions. ahmad fraz, arshad hassan & sumayya chughtai 144 a) measures of bilateral trade the potential economic determinants of emi are exij/exi, exji/exj, imij/imiand imji/imj. the first set of two proxies for bilateral trade is based on exports, exij is the dependence of country i on country j as a potential export market. if the exports of country i have a higher portion of its total exports to country j (exij/exi), in this case, the sensitivity of equity market of the country i should be higher to the movements in the equity market of the country j (bracker, docking , & koch, 1999; johnson & soenen, 2002).it will lead to the prospects of higher or continuous export demand in the future and may have a positive relationship with the emi. similarly, the exports of country j to country i and a positive coefficient is predicted for exji/exj. the second set of two proxies for bilateral trade is based on imports, imij is the dependence of country i on country j for the imports. if the imports of country i have a greater portion of its total imports from country j (imij/imi), then the equity market of the country i should be more concerned and it is expected that due to lower cost, consumers and producer of country i continue to import more goods from country j (bracker, docking , & koch, 1999). according to the future economic perspective of country j, the more demand of products by country i generates a signal for country j about the capacity of country i’s buyers to carry on purchasing the import goods of country j. the larger dependency of country i imports from the country j lead to the emi of country i and j. if the economy of the country i is going well and consumers of the country i obtain more goods from its trading partners. in that case, the economy and equity market of county j will perform well and the positive coefficient relationship between imij/imi and emi is expected (johnson & soenen, 2002). at the same time, with the growing opportunities for exporting companies, it will be a great concern for country i that their exporting firms to compete with the exporting firms of country j in the international market. therefore, the decrease in the imij/imi will generate a signal for better future predictions for the exporting firms in country i, and country i’s equity market may react adversely against the decrease in imij/imi (bracker, docking , & koch, 1999). hence, the competing ability of country i's exporting firms with country j’s firms might be inversely related due to the import dependence with country j (johnson & soenen, 2002). in the context of relative import dependence, both positive and negative coefficient is expected for imij/imi and imji/imj with emi. b) measures of economic conditions that affect bilateral trade relationship journal of applied economics and business studies, volume.4, issue 3 (2020) 137-158 https://doi.org/10.34260/jaebs.436 145 the second set of potential economic determinants are the measures of economic conditions that affect the bilateral trade, these include inflation differential, real interest rate differential, ∆ bilateral exchange rate, volatility in the bilateral exchange rate and the difference in gdp growth rate. the first three variables i.e., inflation differential, real interest rate differential and ∆ bilateral exchange rate are used to encompass the deviation from purchasing power parity (ppp) and interest rate parity (irp). under the assumption of ppp, the relative changes between the inflation of two trading partners can be balanced by the changes in exchange rate. the inflation differential and ∆ bilateral exchange rate between county i and country j are collectively representing the terms of trade, which shows the relative competitiveness of two economies. therefore, these two variables inflation differential and percentage ∆ bilateral exchange rate between county i and j characterize the possible deviation from ppp. likewise, the two variables, real interest rate differential and ∆ bilateral exchange rate collectively represent the potential deviation from irp. these deviations regarding ppp and irp may have an effect on trade and capital flows of two countries and have an impact on the well beings of the firms in both countries (bodurtha, cho, & senbet , 1989; bracker, docking , & koch, 1999). the changes in the inflation rate differential, real interest rate differential and the bilateral exchange rates affect the trade flows and create a divergence in the co-movement between equity markets of country i and country j. therefore, the negative relationship between inflation differential, real interest rate differential, ∆ bilateral exchange rate and emi is expected. the volatility in the bilateral exchange rate is the fourth economic condition variable that influences the bilateral trade relationships. it measures the uncertainty in economic costs that are imposed on individuals and businesses in county i and country j. this uncertainty may reduce the economic and emi between the two countries. büttner and hayo (2011) suggest that if two countries are in the same region then the bilateral exchange rate risk may have a negative impact on emi. the study reports bilateral exchange rate volatility results in a decline in the emi of regional markets and asymmetric effect of bilateral exchange rate volatility with other equity markets (büttner & hayo, 2011). in this current study, two regional markets are taken from asia and two developed markets international markets are taken for the analysis. therefore, it is expected that there may be a positive or negative coefficient for the bilateral exchange rate. the difference in gdp growth rate of the two countries is intended to measure the growth ahmad fraz, arshad hassan & sumayya chughtai 146 dynamics, it is expected that the relative gdp growth rate of one country might influence the economy of the other country. if the growth rate in the economy is due to the economic integration of the other country. on the flipside, faster-growing economies are more independent than large economies and may have less co-movement with other economies. therefore, the negative relationship between annual growth rate gdp differential and emi is expected. during the last half of the previous decade, the pakistani equity market has remained exposed to crashes and it has unprecedently reported three major downward movements during 2005, 2008 and 2010. during the mentioned period, a dramatic decline in the market is observed across a substantial cross-section of the equity market. these crashes are generally driven by fear and greed of the market participants. the pakistani equity market has downed sharply in 2005 with the news of inquiry against insider trading in various stocks. in 2008, the pakistani equity market has faced another worse crash when the panic sale of stocks hit the market. in order to control the collapse of the market, kse management has taken the decision to freeze the market. this decision has created a panic and caused a steep fall in the market which resulted in a depreciation of 68% in the market index. the ‘floor’ remained in place for 108 days. the great crash of 2008 swept billions of rupees from the market. a similar episode is repeated in 2010 when average volumes decreased to 162 million shares in the pakistani equity market. chen, chen and lee (2014) examine the emi among frontier and leading world equity markets before and after the global financial crisis. the study further investigates the impact of macroeconomic variables on emi. the results indicate that the global financial crisis impacts the relationship between these markets and changes the macroeconomic determinantsof the emi. frijns, tourani-rad and indriawan (2012) exmaine the emi of 19 emerging equity markets including south asia, east asia, latin america, central eupore and eastern europe in the presence of political crises around the world. the results suggest that during the period of 1991 to 2006 the crises reduce, the integration among these emerging markets. the study of wang, yang and bessler (2003) concludes that the degree of emi of african markets with the us market decreases after the financial crisis. these market specific incidents affect the dynamics of integration across equity markets. the effect of the financial crisis may not be uniform. its influence may differ from market to market, where as the integation attribution show decreasing trends after the crisis (wu, journal of applied economics and business studies, volume.4, issue 3 (2020) 137-158 https://doi.org/10.34260/jaebs.436 147 2020). this evidence shows that the crisis weakens the integration of markets, therefore, it is expected that emi may decrease during the crisis. 3.2 statistical modeling the study has used panel data analysis using pooled ols. the study employs twostep procedures to explore financial and economic integration. in first step degree of association between two markets has been examined by using r2 procedure as reported above in equation 3. this reflects the co-movement between pakistan’s equity market with the trading partners’ markets. in the second step, pooled ols is employed to test the impact of economic fundamentals on emi. 3.2.1 econometric model for emi and economic integration the following equations are used to estimate the impact of bilateral trade and economic conditions on emi: synchij,t = β0 + β1 ( exij exi ) ij,t + β2 ( exji exj ) ij,t + β3 ( imij imi ) ij,t + β4 ( imji imj ) ij,t + β5(infi − infj )ij,t + β6 (riri − rirj )ij,t + β7(∆xr)ij,t + β8σ(xr)ij,t + β9(gdpgri − gdpgrj)ij,t + 𝐶𝑟𝑠𝑡 + δij,t (4) four different measures are used to determine bilateral trade, the first measure is based on exports i.e., exij/exi at time t, which is exports from country i to country j divided by total exports of country i at time t. the second measure of bilateral trade is also based on exports and is measured as exji/exj at time t, which are exports from country j to country i divided by total exports of country j at time t. the third measure is based on imports and it is measured as imij/imi at time t, which is imports of country i from country j divided by total imports of country i at time t. the fourth measure of bilateral trade is also based on imports and it is measured as imji/imj at time t, which are imports of country j from country i divided by total imports of country j at time t. the data of bilateral trade flow variables i.e. imports and exports are taken from the comtrade database. the second set of potential economic determinants of emi is the economic conditions that influence the bilateral trade relationships. this study employs five variables to estimate the economic conditions of the bilateral trade. the first is variable is the inflation differential (infiinfj)t, which is the inflation differential between country i and county j at time t. the second variable of economic conditions is the real interest rate differential (riririrj)t, which is the real interest rate differential between country i and county j at time t. the third variable of economic conditions is the difference in gdp growth rate (gdpgrigdpgrj)t, which is an annual growth rate differential between country i and county j at time t. the fourth variable of economic conditions is ahmad fraz, arshad hassan & sumayya chughtai 148 the average ∆ bilateral exchange rate between country i and county j at time t using daily bilateral exchange rate data. the fifth variable of economic conditions is volatility in the bilateral exchange rate between country i and county j at time t using the standard deviation of daily bilateral exchange rate data. the last crst is the dummy used for the global financial crisis of 2005, 2008 and 2010. 4. results and discussion this section reports the results of emi and macroeconomic variables for the period 2001 to 2016. to check the behavior of the data, the statistical behavior of the data is given in table 1. it reports first, second, third, fourth moment of variables, which are mean, median, standard deviation, skewness and kurtosis respectively along with the minimum and maximum values of all variables. the mean and median explain the average and central value of the data, while the standard deviation shows the dispersion. the maximum and minimum values report the range of the data and location of the data is reported through skewness and kurtosis. table 1: descriptive statistics of all variables for the period 2001-2016 mean standar d error median standard deviation kurtosis skewness min max r2 0.101 0.009 0.089 0.073 0.211 0.845 0.001 0.278 ln(r2/1r2) -2.601 0.158 -2.326 1.266 4.032 -1.718 -7.348 -0.956 exij/exi 0.003 0.000 0.002 0.002 0.532 1.148 0.001 0.009 exji/exj 0.084 0.009 0.059 0.075 0.206 1.118 0.004 0.273 imij/imi 0.001 0.000 0.002 0.001 -1.147 -0.072 0.001 0.002 imji/imj 0.061 0.007 0.040 0.056 5.176 2.147 0.012 0.291 infiinfj -0.049 0.006 -0.047 0.045 0.383 -0.822 -0.167 0.032 riririrj 0.010 0.005 0.008 0.039 1.113 0.918 -0.051 0.127 ∆xr 0.000 0.000 0.000 0.000 0.100 0.249 0.000 0.001 σ(xr) 0.004 0.000 0.003 0.002 1.963 1.182 0.001 0.011 gdpgrigdpgrj 0.008 0.005 0.009 0.039 -0.478 0.084 -0.078 0.086 note: * descriptive statistics is calculated for each variable from 2001 through 2016 ** r2 is bilateral r2 of multifactor model using county i to country j along with msci world market index, ln(r2/1 r2) is log transformation of bilateral r2,exij/exi is exports from country i to country j divided total exports of country i, exji/exjis exports from country j to country i divided by total exports of country j, imij/imi is imports of country i from country j divided by total imports of country i, imji/imjis imports of country j from country i divided by total imports of country j, infiinfj is inflation differential between country i and county j,riririrj is real interest rate differential between country i and county j, ∆xr is average percentage change in bilateral exchange rate between country i and county j, σ(xr)is volatility in bilateral exchange rate between country i and county j and gdpgrigdpgrj is the difference in gdp growth rate between country i and county j. table 1 reports the descriptive statistics of all the variables. the average r2for emi of pakistan with trading partners’ equity market is 0.101, which explains a 10.1% variation in overall markets over the sample period. the highest r2 for any market is 0.278 and the lowest r2 is 0.001 with a standard deviation of 0.073. the synchronization of international markets with the pakistani market is low as the value of r2on average journal of applied economics and business studies, volume.4, issue 3 (2020) 137-158 https://doi.org/10.34260/jaebs.436 149 is 10% whereas it touches a maximum level of 27%. it means that pakistani market may have idiosyncratic factors that influence its alignment with global markets. the countryspecific and firm-specific variables may contribute to the synchronization of different markets (campbell, lettau, malkiel, & xu, 2001). table 1 also shows that the export ratio of country i to country j divided by total exports of country i (exij/exi) is 0.003 and the highest ratio is0.009 and the lowest ratio is 0.001 with a standard deviation of 0.002. the average ratio of exports from country j to country i divided by total exports of country j (exji/exj) is 0.084, which is higher from the average (exij/exi). that export ratio shows that the average exports from pakistan to other sample countries are the higher and have highest ratio with any country in any year is 0.273 and the minimum ratio is .004 with a higher standard deviation of 0.075 as compared to the ratio (exij/exi). the results of the imports ratio show that the average import ratio of country i from country j divided total imports of country i (imij/imi)is 0.001 and the highest ratio is 0.002 and the lowest ratio is 0.001 with a standard deviation of 0.001. the average ratio of imports from country j to country i divided by total imports of country j (imji/imj) is 0.091, which is higher from average (imij/imi). that import ratio shows that the average imports from the sample countries are higher and have the highest ratio with any country in any year is 0.291 and the minimum ratio is .021 with a higher standard deviation of 0.056 as compared to the ratio (imji/imj). the descriptive statistics for economic conditions are also reported in the above table. the inflation rate differential has an average value of -0.049, which shows that the average inflation in pakistan as compared to the sample countries is higher in the sample period. whereas, the real interest rate differential has a positive mean value of 0.010, which shows that the average real interest rate in pakistan for the sample period is lower as compared to sample countries. therefore, it is concluded that in recent past higher nominal interest rates in pakistan are not due to the higher real interest rate but due to the high inflation rate. the average percentage of bilateral exchange ratio is .01%, with the highest value of .001 and has standard deviation of .0002. the descriptive of volatility in bilateral exchange rate shows an average value of .004 with a highest value of .011 and the lowest value of .001 along with a standard deviation of .002. the positive value of the growth rate in gdp differential shows that pakistan has a higher average growth rate in gdp during the sample period. whereas, the highest and lowest value of the bilateral growth rate in gdp is observed 0.086 and -0.078, which shows in some years the gdp growth rate of sample countries is higher ahmad fraz, arshad hassan & sumayya chughtai 150 as compared to pakistan. table 2 reports the bilateral r2 of pakistan and selected countries. table 2: r2 between pakistan and other markets 2001-2016 2001 2002 2003 2004 2005 2006 2007 2008 china→ pk 0.003 0.036 0.09 0.015 0.052 0.122 0.186 0.107 india→pk 0.076 0.143 0.099 0.108 0.045 0.14 0.186 0.175 uk→pk 0.028 0.076 0.061 0.044 0.041 0.164 0.278 0.13 us→pk 0.088 0.04 0.078 0.055 0.037 0.141 0.206 0.069 2009 2010 2011 2012 2013 2014 2015 2016 china→ pk 0.016 0.023 0.046 0.271 0.111 0.073 0.001 0.099 india→pk 0.007 0.024 0.073 0.271 0.125 0.085 0.001 0.111 uk→pk 0.018 0.029 0.064 0.269 0.174 0.177 0.077 0.11 us→pk 0.114 0.145 0.134 0.27 0.182 0.107 0.023 0.109 table 2 shows the value of r2 for the period of 2001 to 2016 for sample countries. the bilateral r2 is calculated each year and is used as a proxy of emi. the results reveal that the bilateral emi of the pakistani equity market and the uk market is higher and the highest value of r20.278 is observed in 2007. whereas, the lowest value of r2 0.001 is observed with china and india in 2015. historically, the emi with pakistan and trading partners is lower and all markets report the lowest value of r2 in 2005 due to the hit of the first financial crisis in the pakistani equity market. it is the period of three major crises in the pakistani equity market and is evident of downward movements during 2005, 2008 and 2010 as shown in fig. 1. after the first crisis, the revival of emi is started and the second crisis of 2008 followed by the 2010 crisis results a decrease in the emi. figure 1 shows these variations of the degree of emi from 2001 to 2016. fig 1: r2 between pakistan and other markets 2001-2016 the graphical representation is showing a decrease that after 2012 the emi of the pakistani market with other global partners. however, it is worth mentioning that after 2015 the integration with all trading partners is increasing. it can also observe from the -0.1000 0.0000 0.1000 0.2000 0.3000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 china→ pk india→pk uk→pk us→pk journal of applied economics and business studies, volume.4, issue 3 (2020) 137-158 https://doi.org/10.34260/jaebs.436 151 graph below that the degree of emi of pakistani equity market with its trading partners varies across countries over time. table 3: impact of economic fundamentals on emi from 2001 to 2016 variable s β s.e tstat β s.e tstat β s.e t-stat model 1 model 2 model 3 c -1.43 0.83 1.73 -1.68 0.81 2.08 -2.08 0.54 -3.88 exij/exi 126.92 119.9 9 1.06 206.10* 121.6 0 1.69 209.51* * 107.3 3 1.95 exji/exj -0.89 3.37 0.26 -0.14 3.28 0.04 imij/imi -236.22 402.7 0 0.59 -245.68 389.4 6 0.63 imji/imj -7.82** 3.34 2.34 -9.63*** 3.33 2.89 -9.52*** 3.04 -3.13 infiinfj -7.28* 4.28 1.70 14.20** * 5.22 2.72 14.33** * 4.86 -2.95 riririrj -5.40 5.82 0.93 -7.43 5.70 1.30 -6.44 5.12 -1.26 ∆xr 310.34 877.0 5 0.35 177.48 850.3 5 0.21 σ(xr) 250.52* ** 92.77 2.70 265.41* ** 89.97 2.95 267.80* ** 77.67 -3.45 gdpgr i gdpgr j -10.45 6.50 1.61 -9.07 6.32 1.43 -6.71 4.90 -1.37 crisis -1.05** 0.48 2.18 -1.07** 0.47 -2.28 adj. r2 0.12 0.18 0.21 f-stat 1.95 2.35 13.46 fsig 0.06 0.02 0.00 note:(1) *** significance at 0.01 level ** significance at 0.05 level and * significance at 0.10 level. (2) the table represents three models; model 1 is based on the bilateral trade and economic fundamentals. model 2 represents the impact of bilateral trade and economic fundamentals on emi along with dummy of 2005, 2008 and 2010 crisis observed in pakistan. model 3 is the final model represents the impact of bilateral trade and economic fundamentals on emi along with the dummy of 2005, 2008 and 2010 crisis observed in pakistan by excluding the variables burden on the model 2. the table 3 reports the results of pooled regression of model 1, model 2 and model 3 for the entire sample period. model 1 explains 12% variation in emi due to macroeconomic fundamentals and goodness of the fit statistics shows the attributes of bilateral trade and economic conditions significantly explains the variations in emi.as earlier discussed that different types of the crisis which are associated to any market may influence differently to the markets. the pakistani equity market has exposed to three crashes of 2005, 2008 and 2010. in the second model dummy of crisis for 2005, 2008 and 2010 has been introduced and it is observed that r2 increases from 12% to 18%. the coefficient of crisis in model 2 reports a significant value of -1.05, which indicates the crisis weakens the emi of pakistan with global equity markets. finally, the general to specific approach is used to extract a parsimonious model, therefore, the variables that are a burden on the model excluded. the parsimonious model is proposed as model 3 that indicates 21% variation in emi due to macroeconomic fundamentals. the explanatory power of model 3 is higher than model ahmad fraz, arshad hassan & sumayya chughtai 152 1 and model 2. exports from country i (trading partner)to country j (pakistan)as a percentage of total exports of trading partner (exij/exi), imports of country j from country i as a percentage of total imports of country j (imji/imj), inflation rate differential, volatility in bilateral exchange rate and crisis in pakistan show significant impact on emi. all the variables have the expected signs of coefficients. the bilateral trade proxy based on exports (exij/exi) has a significant and positive impact on emi. it indicates the dependence of trading partners’ on the pakistani market as a potential export market. if the portion of trading partners total exports to pakistan increases the sensitivity of equity market co-movement should increase, it may lead to the increasing export demand in future (bracker, docking , & koch, 1999).this relationship indicates that the dependence of the export of two countries can increase the emi. the bilateral trade proxy based on imports (imji/imj) has a significant and negative impact on emi. it shows that an increase in the portion of pakistan’s total imports with its trading partners will reduce emi. bracker, docking and koch (1999) argue that growing exporting opportunities for all the firms globally the firms of both country i, and country compete with each other in the international market as well as in-home countries. in such case these firms may be inversely related and have low emi due to import dependence, which means import dependence of two countries can decrease the emi. the results of model 3 report a significant and negative impact of inflation differential on emi. the inflation differential between the two countries characterizes the possible deviation from ppp which affects the trade flows between two countries and represents the competitiveness of two economies (bodurtha, cho, & senbet , 1989).the relative changes in the inflation difference affect the term of trade and result in the divergence between the co-movement of two economies. the results suggest that an increase in the inflation differential reduces the emi. the volatility in bilateral exchange has a significant and negative impact on emi. the volatility in the bilateral exchange rate captures the uncertainty in economic costs imposed on individuals and businesses (büttner & hayo, 2011). the results indicate that the increase in the uncertainty in economic cost will reduce the emi between two markets. the dummy of the financial market crisis of 2005, 2008 and 2010 in the pakistani equity market has reported a significant and negative impact on emi. it indicates that the market-specific incidents affect the dynamics of integration across equity markets (frijns, tourani-rad, & indriawan, 2012; chen, chen, & lee, 2014; mobarek, journal of applied economics and business studies, volume.4, issue 3 (2020) 137-158 https://doi.org/10.34260/jaebs.436 153 muradoglu, mollah, & hou, 2016).the negative sign of the financial crisis shows that the crisis weakens the integration of markets. 5. conclusion the study examines the impact of economic variables on emi of pakistani equity market with two developed market (the u.s. and the u.k.) and two regional emerging markets (india and china) for the period of 1998 to 2016. the study employs two-step procedures, in the first step the co-movement of equity markets is estimated by using the multifactor model, and in the second step the impact of economic variables on emi is investigated. the high co-movement between two markets is considered a proxy of emi (vithessonthi & kumarasinghe, 2016). the results of co-movement indicate that bilateral emi of the pakistani equity market and the uk market is higher as compared to the other countries and the highest value of r2 0.278 is observed in 2007.in the second step, the study examines the impact of economic fundamentals on emi. the results of this study show that bilateral trade and economic conditions have a significant impact on emi. the findings of the study suggest that the export dependence of two economies may increase the emi, whereas the import dependence may reduce the equity market dependence of two economies. two variables of economic conditions inflation differential and volatility in the bilateral exchange rate have a negative impact on emi. the higher share of import of a country along with high inflation differential and volatility in the bilateral exchange rate have a negative impact on equity market comovement (bracker, docking , & koch, 1999). further findings suggest the financial crisis in an economy may reduce the emi with its trading partners. the study implies that the co-movement of two equity markets statistically depends upon bilateral trade, economic conditions and tend to change during the crisis period. references abbas, q., khan, s., & shah, s. a. (2013). volatility transmission in regional asian stock markets. emerging markets review, 16, 66-77. doi:https://doi.org/10.1016/j.ememar.2013.04.004 an, h., & zhang, t. (2013). stock price synchronicity, crash risk, and institutional investors. journal of corporate finance, 21, 1-15. doi:https://doi.org/10.1016/j.jcorpfin.2013.01.001 beck, k. g., finnerty, j. e., & tucker, a. l. (1992). the intraday interdependence structure between u.s. and japanese equity markets. journal of financial research, 15, 27-37. doi:https://doi.org/10.1111/j.1475-6803.1992.tb00784.x ahmad fraz, arshad hassan & sumayya chughtai 154 bekaert, g., erb, c. b., & harvey, c. r. (1997). what matters for emerging equity market investments. emerging markets quarterly, 2, 17-46. retrieved from https://faculty.fuqua.duke.edu/~charvey/research/published_papers/p45_what _matters_for.pdf bodurtha, j. j., cho, d. c., & senbet , l. w. (1989). economic forces and the stock market: an international perspective. the global finance journal, 1, 21-46. retrieved from https://pdfs.semanticscholar.org/cc8e/8cdf1e78f24e0c6fe3a3bfea218fa485c551 .pdf bouët, a., berisha-krasniqia, v., estradesa, c., & labord, d. (2012). trade and investment in latin america and asia: perspectives from further integration. journal of policy modeling, 34, 193-210. doi:https://doi.org/10.1016/j.jpolmod.2011.09.004 bracker, k., docking , d. s., & koch, p. d. (1999). economic determinants of evolution in international stock market integration. journal of empirical finance, 6, 1-27. doi:https://doi.org/10.1016/s0927-5398(98)00007-3 büttner, d., & hayo, b. (2011). determinants of european stock market integration. economic systems, 35, 574–585. doi:https://doi.org/10.1016/j.ecosys.2010.10.004 campbell, j. y., & hamao, y. (1992). predictable stock returns in the united states and japan: a study of long-tern capital market integration. journal of finance, 47, 43-69. doi:http://onlinelibrary.wiley.com/doi/10.1111/j.15406261.1992.tb03978.x/full campbell, j. y., lettau, m., malkiel, b. g., & xu, y. (2001). have individual stocks become more volatile? an empirical exploration of idiosyncratic risk. the journal of finance, 56, 1-43. doi:http://onlinelibrary.wiley.com/doi/10.1111/j.15406261.1992.tb03978.x/full chen, m.-p., chen, p.-f., & lee, c.-c. (2014). frontier stock market integration and the global financial crisis. the north american journal of economics and finance, 29, 84-103. doi:https://doi.org/10.1016/j.najef.2014.05.004 chen, p. (2018). understanding international stock market comovements: a comparison of developed and emerging markets. international review of journal of applied economics and business studies, volume.4, issue 3 (2020) 137-158 https://doi.org/10.34260/jaebs.436 155 economics & finance, 56, chen, p. (2018). understanding international stock market comovements: a comparison of developed and emerging markets. international review of economics & finance, 56, 451-464. doi:https://doi.org/10.1016/j.iref.2017.12.004 dasgupta, s., gan, j., & gao, n. (2010). transparency, price informativeness, and stock return synchronicity: theory and evidence. journal of financial and quantitative analysis, 45, 1189–1220. doi:https://doi.org/10.1017/s0022109010000505 dhanaraj, s., gopalaswamy, a. k., & babu, m. s. (2017). trade, financial flows and stock market interdependence: evidence from asian markets. empirical economics review, 7, 15-34. dimitriou, d. i., & simos , t. m. (2014). contagion effects on stock and fx markets: a dcc analysis among usa and emu. studies in economics and finance, 31(3), 246-254. doi:doi: https://doi.org/10.1108/sef-07-2012-0075 farooq, o., ahmed, n., & bouaddi , m. (2018). stock price synchronicity and its effect on stock market volatility: evidence from the mena region. american journal of finance and accounting, 5(3), 276-292. doi:https://doi.org/10.1504/ajfa.2018.093042 fraz, a., & hasan, a. (2016). dragon’s entry in south asia and its impact on financial market. the pakistan development review(special issue), 445-454. fraz, a., & hassan, a. (2017). stock price synchronicity and information environment. business & economic review, 9(4), 213-232. doi:doi: dx.doi.org/10.22547/ber/9.4.10 frijns, b., tourani-rad, a., & indriawan, i. (2012). political crises and the stock market integration of emerging markets. journal of banking & finance, 36, 644-653. doi:https://doi.org/10.1016/j.jbankfin.2011.05.007 ge, y., wu, j., zhang, j., & zou, l. (2019). the role of chinese economic variables on australian and new zealand equity returns. accounting & finance, 59(5), 30533103. doi:https://doi.org/10.1111/acfi.12573 ghosh, a., saidi, r., & johnson, k. h. (1999). who moves the asia-pacific stock markets-us or japan? empirical evidence based on the theory of cointegration. the financial review, 34, 159-170. doi:https://doi.org/10.1111/j.1540-6288.1999.tb00450.x ahmad fraz, arshad hassan & sumayya chughtai 156 hasan, a., saleem, h., & abdullah, m. s. (2008). long-run relationships between an emerging equity market and equity markets of the developed world an empirical analysis of karachi stock exchange. international research journal of finance and economics, 52-62. retrieved from https://www.researchgate.net/publication/239849976 johnson, r., & soenen, l. (2002). asian economic integration and stock market comovement. journal of financial research, 25, 141-157. doi:https://doi.org/10.1111/1475-6803.00009 joyo , a. s., & lefen, l. (2019). stock market integration of pakistan with its trading partners: a multivariate dcc-garch model approac. sustainability, 11(2), 303. doi:doi:10.3390/su11020303 karagoz, k., & ergun, s. (2010). stock market integration among balkan countries. mibes transactions, 4, 49-59. retrieved from http://mtol.teithessaly.gr/vol4_2010/karagoz-ergun.pdf kim, s.-w., choi, m. j., & kim, y.-m. (2019). does intra-regional trade matter in regional stock markets?: new evidence from asia-pacific region. asian economic journal, 33(3), 253-280. doi:https://doi.org/10.1111/asej.12186 kim, y. h. (2005). the optimal path of regional economic integration between asymmetric countries in the north east asia. journal of policy modeling, 27, 673-687. doi:https://doi.org/10.1016/j.jpolmod.2005.05.004 mobarek, a., muradoglu, g., mollah, s., & hou, a. j. (2016). determinants of time varying co-movements among international stock markets during crisis and noncrisis periods. journal of financial stability, 24, 1-11. doi:https://doi.org/10.1016/j.jfs.2016.03.003 morana , c., & beltratti, a. (2008). comovements in international stock markets. journal of international financial markets, institutions and money, 18, 31-45. doi:https://doi.org/10.1016/j.intfin.2006.05.001 morck, r., yeung, b., & yu, w. (2000). the information content of stock markets: why do emerging markets have synchronous stock price movements? journal of financial economics, 58, 215-260. doi:https://doi.org/10.1016/s0304405x(00)00071-4 journal of applied economics and business studies, volume.4, issue 3 (2020) 137-158 https://doi.org/10.34260/jaebs.436 157 norén, r. (2011). towards a more integrated, symmetric and viable emu. journal of policy modeling, 33, 821-830. doi:https://doi.org/10.1016/j.jpolmod.2011.08.003 paramati, s. r., zakari, a., jalle, m., kale, s., & begari, p. (2018). the dynamic impact of bilateral trade linkages on stock market correlations of australia and china. applied economics letters, 25(3), 141-145. doi:https://doi.org/10.1080/13504851.2017.1305067 peir´o, a. (2016). stock prices and macroeconomic factors: some european evidence. international review of economics and finance, 41, 287-294. doi:https://doi.org/10.1016/j.iref.2015.08.004 pukthuanthong, k., & roll, r. (2009). global market integration: an alternative measure and its application. journal of financial economics, 94, 214-232. doi:https://doi.org/10.1016/j.jfineco.2008.12.004 robert, j., & soenen, l. (2003). economic integration and stock market comovement in the americas. journal of multinational financial management, 13, 85-100. doi:https://doi.org/10.1016/s1042-444x(02)00035-x roll, r. (1992). industrial structure and the comparative behavior of international stock market indices. the journal of finance, 47, 3-43. doi:https://doi.org/10.1111/j.1540-6261.1992.tb03977.x schotman , p. c., & zalewska, a. (2006). non-synchronous trading and testing for market integration in central european emerging markets. journal of empirical finance, 13, 462–494. doi:https://doi.org/10.1016/j.jempfin.2006.04.002 simpson, j. l. (2008). financial integration in the gcc stock markets: evidence from the early 2000s development phase. journal of economic cooperation, 1-28. doi:https://dx.doi.org/10.2139/ssrn.994706 tai, c. s. (2007). market integration and contagion: evidence from asian emerging stock and foreign exchange markets. emerging markets review, 8, 264-283. doi:https://doi.org/10.1016/j.ememar.2006.09.011 thomas, n. m., kashiramka, s., & yadav, s. s. (2017). dynamic linkages among developed, emerging and frontier capital markets of asia-pacific region. journal of advances in management research, 14(3), 332-351. doi:doi: https://doi.org/10.1108/jamr-10-2016-0088 ahmad fraz, arshad hassan & sumayya chughtai 158 vithessonthi, c., & kumarasinghe, s. (2016). financial development, international trade integration, and stock market integration: evidence from asia. journal of multinational financial management, 35, 79-92. doi:https://doi.org/10.1016/j.mulfin.2016.03.001 vo, x. v. (2018). trade integration and international financial integration: evidence from asia. the singapore economic review, 1-12. doi:https://doi.org/10.1142/s0217590818500248 wälti, s. (2011). stock market synchronization and monetary integration. journal of international money and finance, 30, 96–110. doi:https://doi.org/10.1016/j.jimonfin.2010.07.004 wang, p., & moore, t. (2008). stock market integration for the transition economies: time‐varying conditional correlation approach. the manchester school, 76, 116-133. doi:https://doi.org/10.1111/j.1467-9957.2008.01083.x wang, z., yang, j., & bessler, d. (2003). financial crisis and african stock market integration. applied economics letters, 10, 527-533. doi:https://doi.org/10.1080/1350485032000100198 wu, f. (2020). stock market integration in east and southeast asia: the role of global factors. international review of financial analysis, 67, 1-11. doi:https://doi.org/10.1016/j.irfa.2019.101416 yu, i. w., fung, k. p., & tam, c. s. (2010). assessing financial market integration in asia – equity markets. journal of banking & finance, 34, 2874–2885. doi:https://doi.org/10.1016/j.jbankfin.2010.02.010 journal of applied economics and business studies, volume. 6, issue 1 (2022) 1-24 https://doi.org/10.34260/jaebs.611 1 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x comparative analysis of private equity investments in europe and the asia-pacific region amin ullah khan1 muhammad zubair khan2, zafir ullah khan3 1 institute of management science, university of science & technology bannu, khyber pakhtunkhwa, pakistan 2 institute of management sciences, university of science & technology bannu, khyber pakhtunkhwa, pakistan 3 department of economics, university of science & technology bannu, khyber pakhtunkhwa, pakistan abstract the article presents a comparative analysis of private equity (pe) investments in 13 asian-pacific and 19 european economies over the period of twenty-nine years from 1990 to 2017 using the fixed effects estimation technique. results show that ict, investment profile, human capital, and market capitalization are the strong positive determinants of pe in all the samples. for the rest of the variables, there are significant differences between the two regions. population growth and real interest rate exert significant influence in the overall samples as well as in the asia-pacific region while real exchange rate, unemployment, and tax burden play significant role in the europe. the article contributes to the pe literature by filling the regional gap and offering new insights into the previously contested results. keywords private equity, market capitalization, human capital, ict jel classification b26, b29 1. introduction previous studies have investigated determinants of private equity (pe) and venture capital (vc) using different aspects namely stages of investment (khan, khan and hameed, 2020; khan et al., 2021), vc availability (li and zahra, 2012), investment size (ning, wang and yu, 2015), performance (hege, palomino and schwienbacher, 2011), networks (guler and guillén, 2010; milosevic, 2018) governance (hearn, randøy and oxelheim, 2016) fundraising (gompers and lerner, 1998; oberli, 2014) and exit options (cumming and fleming, 2002; cumming and macintosh, 2003; cumming, fleming and schwienbacher, 2006). other facets examined include decisions to cross border (hain et al., 2016; devigne et al., 2018; ragosa and warren, 2019), contractual choices and syndications (lerner, 1994; da rin et al., 2006; tykvová, 2018) structures and deal processes (karsai, wright and filatotchev, 1997; karsai et al., 1998; smolarski amin ullah khan, muhammad zubair khan, & zafir ullah khan 2 et al., 2005; ahlstrom, bruton and yeh, 2007; klonowski, 2007) and post-investment processes (wright et al., 1999; da rin, hege, et al., 2006; cumming, schmidt and walz, 2008). the studies have found that institutional (cumming and knill, 2012; groh and wallmeroth, 2016), macroeconomic (baygan, and freudenberg, 2000; romain and pottelsberghe, 2004), technological environment (schertler, 2003a; da rin et al., 2006b; schertler, 2007) entrepreneurial environment (gilson and black, 2000; jeng and wells, 2000; cherif and gazdar, 2011; bozkaya and kerr, 2014), political factors (bonini and alkan, 2012) and oil prices and media coverage (cumming et al., 2010) exert impact over pe/vc. these studies have paid much attention to the us and european pe/vc industries. in a sample of 18 studies, eight articles have incorporated some asia-pacific economies (japan and australia) in their sample while there is no study focusing exclusively on asia-pacific or comparing asia with other continents. in their analysis of 314 vc/pe research articles since 2011, 52% of all the articles still rely on us data (tykvova, 2017). this is, despite the fact, that asian pe market has grown tremendously after the 2000 high-tech market crash. for example, asia-pacific region experienced pe investment twice than its european counterpart in 2015. growth pattern of asia-pacific pe is not much different than its european counterpart in the overall period under consideration. both markets have been affected by the financial crises. however, asia-pacific market was less vulnerable to these crises and has surpassed europe in terms of total pe investments in consequence of the 2008 financial crisis. asia-pacific region has also maintained higher pe investments to gdp ratio compared to europe in the last two decades. cumming and zhang (2016) point out that vc determinants literature still give conflicting results. among several reasons, one could be that previous studies lack comparative analyses to examine if determinants of pe differ as stage of counties’ development, context and geography varies. comparative analysis will help reduce the conflicting results in the literature and assist policies actors to consider the country or region-specific effects while framing policies to promote pe markets. moreover, most of the previous studies have used data before dot-com era while much pe/vc investments in regions outside us and western europe flourished after the bubble was burst in 2000. moreover, previous studies have used relatively short time periods mostly involving older periods. fresh research with long-term data incorporating more countries particularly that have flourished during this period will bring more insight to the literature. we need more cross-country comparisons and inter-disciplinary research on venture capital due to the existing home biases and discipline specific biases where finance and management disciplines are not open to each other when it comes to publishing or citing each other’s papers (cumming and vismara, 2017). the current article examines the influences on pe investment in 32 economies spanning 1990-2017 using fixed effects. it presents a comparative analysis of pe investments in asia-pacific and europe. before proceeding to next section, it is important to clarify the concept of pe and vc. vc is journal of applied economics and business studies, volume. 6, issue 1 (2022) 1-24 https://doi.org/10.34260/jaebs.611 3 the mix of early and later stage privately held equity investments whereas pe consists of vc and other privately held equity investments including expansion, replacement, growth, and buyout capital. as vc is a component of pe, previous literature on vc cannot be ignored while discussing pe. the current article examines pe in its analysis and cites both vc and pe literature. this is a standard practice to do so (kumar and orleck, 2002). 2. literature review previous research shows that institutional factors, entrepreneurial environment, general macroeconomic factors, technological opportunity, and political factors play role in vc/pe development. 2.1 entrepreneurial environment an important element of an entrepreneurial environment is the presence of a healthy stock market (black and gilson, 1999; jeng and wells, 2000). liquid stock markets provide an exit route from vc investment through an ipo (schröder, 2009). another factor that affects the entrepreneurial environment is the taxation regime with some showing that taxation has a negative effect (cherif and gazdar, 2011) while others demonstrate that it has no or little importance (romain and van pottelsberghe, 2004). research shows that taxation is more a powerful determinant in the european countries compared to the us and asia reflect a regional disparity. labor market also influence a pe market. studies show that labor market rigidities exert a negative effect on vc/pe market (jeng and wells, 2000; cherif and gazdar, 2011; bonini and alkan, 2012) yet some show a positive impact of labor market rigidities (schertler, 2003a; bozkaya and kerr, 2014).the level of entrepreneurship represents the favorability of the entrepreneurial environment. research shows that total entrepreneurial activity (romain and pottelsberghe, 2004; bonini and alkan, 2012), high-tech investments and a reduction in barriers to entrepreneurship (da rin et al., 2006b), and investment opportunities play an important role in the development of venture capital (avnimelech et al., 2004). moreover, a strong entrepreneurial culture positively influences the effectiveness of the r&d stock of knowledge for vc (romain and pottelsberghe, 2004). according to baygan and freudenberg (2000) framework of conditions to spur research and innovation, incentives in exchange of risk, entrepreneurial activities, investment opportunities, and expertise are the keys to vc development across oecd countries. groh and von liechtenstein (2009) show that ‘entrepreneurial opportunities’ are less likely to attract foreign vcs in the cee than in eu-15. 2.2 general economy business cycles such as the dot com bubble and 2008 financial crisis affect risk preferences and investment strategies in the vc markets (ning, wang and yu, 2015). however, the effect of gdp on pe/vc is mixed. for example, cherif and gazdar (2011) demonstrate that gdp growth has a positive effect on vc/pe but others show that gdp is not an important factor (jeng and wells, 2000; kumar and orleck, 2002). previous studies also demonstrate that interest rate negatively affects vc investment (bonini and alkan, 2012). but füss and schweizer (2008) find amin ullah khan, muhammad zubair khan, & zafir ullah khan 4 positive relationship between long-term interest rates and vc investment while a negative relationship between short-term interest rates and vc investment. if cost of capital increases and bank loans get expensive, demand for pe should increase due to substitute effect. on the other hand, low cost of capital lead to economic growth and economic growth lead to more pe investments. however, more economic growth can cause inflation and in response central banks usually enhance interest rates. finally exchange rate might affect the pe/vc investments particularly cross-border investments as foreign investors may meet losses while converting the local currencies into us dollars after exit from an investment. in their study of brazilian vc/pe industry, minardi et al. (2017) finds that currency exchange rates does not have a significant influence on pe returns in long term. since total pe investments also include cross border investments and international investors calculate their investments and returns in us dollars, although, they convert us dollars into local currencies while investing in portfolio companies and back into us dollars to receive proceeds upon exits (minardi et al., 2017). however, it is still not clear how much is the effect of the exchange rate over years on country’s pe investment in a cross-country analysis. we expect negative effect of exchange rate. 2.3 technological opportunity romain and van pottelsberghe (2004) find that the impact of gdp growth rate and r&d stock of knowledge on vc intensity is subject to a country’s labour market rigidities. the lower the labour market rigidities, the higher the positive impact of the gdp growth rate and r&d stock of knowledge on the vc intensity and vice versa. however, the effect decreases with an increase in labour market rigidities. similarly, the impact of r&d capital stock on vc intensity is positively influenced by a strong entrepreneurial culture and intense technological opportunities. schertler (2007), using a panel dataset of 15 european countries, spanning 1991-2001, finds a strong connection between various measures of total knowledge capital (sum of government and business-financed knowledge capital) and vc investments. the number of patents, number of r&d researchers, and the gross domestic expenditures on r&d scaled by gdp (gerd), have a significant and positive impact on the number of early-stage and earlyand expansion-stage vc investments. however, the impact of total knowledge capital in terms of volume is not greatly significant and government-financed knowledge capital has weak explanatory power for the size of vc investments, probably due to commercial applications (keuschnigg and nielsen, 2003) or due to its long-term focus which is translated into commercial applications in more than two years. schertler (2007), however, finds that business-financed gerd does not affect volume of investments, which contradicts the results of romain and van pottelsberghe (2004), probably due to differences in trends in their data. schertler (2003) also finds that human capital endowment plays a positive role in vc development. however, according to da rin, nicodano, et al. (2006) r&d has no effect on early stage and high-tech vc investment. they argued that policies intended to increase expected return in investment in new ventures attract vc toward high-tech and earlyjournal of applied economics and business studies, volume. 6, issue 1 (2022) 1-24 https://doi.org/10.34260/jaebs.611 5 stage investments. 2.4 political and factors vc research suggests that politically stable environments grow more vc. bonini and alkan (2006) find that investment profile, socioeconomic conditions and corruption affect vc. eleswarapu and venkataraman (2006) show that the stability of political systems explains international differences in transaction costs of equity and hence the liquidity of stock markets, expected return on equity, and ultimately vc differs across countries. institutions are one of the powerful determinants of vc/pe. for instance, us and european investors display different investment patters (bertoni and colombo, 2015). groh and von liechtenstein (2009) demonstrate that vc/pe investors are attracted by strong protection of investors’ rights and efficient corporate governance mechanisms. institutional proxies having effect on vc/pe include freedom from corruption (cherif and gazdar, 2011); merges & acquisitions investment volume, the shareholder suits index, and the legal rights index (groh and wallmeroth, 2016) and temperate bankruptcy laws for entrepreneurs (armour and cumming, 2006). cumming, henriques and sadorsky (2016), on the other hand, show that cleantech vc activity is negatively influenced by the rule of law and government effectiveness. legal institutions also influence international allocations in pe/vc. for example, ragosa and warren (2019) exhibit that regulatory support measures and feed-in tariffs, coupled with political stability, are strong drivers of cross-border investment in renewable energy in developing countries. moreover, strong legal frameworks of host economies attract foreign pe investors as such environments protect property rights and reduce agency risks (guler and guillén, 2010). groh and von liechtenstein (2009) also show that the protection of property rights is the key consideration while allocating private equity globally. 2.5 population growth population growth has several adverse effects which include environmental degradation, resource depletion, human health, human behaviors, and quality of life (ehrilich and holdren, 1971). the downside of population growth is that households’ spending on dependent children reduces the ability of the families to save money that ultimately reduces savings. for example, households able to control birth rate in bangladesh villages experienced higher income, more assets and better health compared to their counterparts in control villages (joshi and schultz, 2007). it reduces the pace of development and reduces the ability of families to escape poverty (klasen and lawson, 2007). the allocation of resources at the macro level by governments from population growth perspective has also implications on pe. economies that have achieved birth control have achieved high and sustained economic growth and reduced poverty as governments’ per capita allocation to health and education improve (sinding, 2009). others suggest that low birth rate in high income countries could be problematic while high birth rate in low income economies can slow down their development (peterson, 2017). in this context, it is expected that environments where birth control has been achieved are likely to attract vc/pe investments. amin ullah khan, muhammad zubair khan, & zafir ullah khan 6 2.6 ict internet, computing and telecommunication (ict) has emerged as one of the dominant forces shaping the landscape of business around the globe (khan, khan and hameed, 2020; khan et al., 2021). ict has resulted in improved financial participation, financial inclusion, financial access, and the overall financial development (pradhan, arvin and norman, 2015; asongu and moulin, 2016; asongu, s. a., & acha-anyi, 2017; gabor, d., & brooks, 2017; pradhan et al., 2017, 2018; lechman and marszk, 2019). considering financial systems as information systems, ict has changed the information asymmetry in vc/pe markets that has bearing on transaction and agency costs (ocampo, 2018). vc/pe is an information-dependent industry where investors are largely dependent upon market information for deal selection, deal origination, and postinvestment monitoring (carey, prowse, rea & udell, 1993) that the extent that investors have to rely on informal informants (fiet, 1995). in this context, ict is not only a tool making processes more efficient but also a means to creating an entrepreneurial environment by creating enormous amount of business opportunities and thus a positive effect of ict is expected. 3. data and methodology the article takes data from different sources. data for dependent variable i.e., private equity investment in asian-pacific countries has been taken from asian venture capital journal and pe investment in european countries has been taken from eurostat and evca yearbooks. in european data, pe investment data from 1990 to 2015 was taken from eurostat while 2016 to 2017 data was taken from evca yearbooks. ict, patents, ict exports, market capitalization, gdp growth, unemployment, real interest rates, real exchange rates, tertiary education and investment profile have been used as independent variables and data of these variables are taken world bank databank. data on investment profile has been taken from the international country risk guide (icrg) whereas tax burden data has been taken from freedom of the world, heritage foundation (see appendix 1). the article uses standard hausman test to decide which method is more appropriate (greene, 2008). the test rejects the null hypothesis that the errors are not correlated with the regressors, hence, fixed effects (fe) estimation technique was chosen as the estimation technique. we also found heteroskedasticity in the data, but we run a robust option in stata to obtain heteroskedasticity-robust standard errors. the dependent variable was also checked for autocorrelation using lagrange-multiplier. to cope with the autocorrelation, the lag of dependent variable was included in the models that removed autocorrelation. journal of applied economics and business studies, volume. 6, issue 1 (2022) 1-24 https://doi.org/10.34260/jaebs.611 7 table 1: summary statistics variable observations mean standard deviation minimum maximum private equity % gdp 805 0.30 0.50 0.00 6.34 ict 857 39.52 33.58 0.00 97.30 market capitalization 724 90.38 142.26 0.86 1273.39 exchange rate 771 99.29 14.89 42.10 165.88 gdp growth 796 4.00 2.74 0.01 25.12 unemployment 680 6.19 3.80 0.49 26.09 patents 869 8149.40 17161.48 17.00 135885.00 population growth 780 0.89 0.71 0.00 5.32 tax burden 729 63.39 15.50 29.80 93.80 investment profile 864 8.97 2.39 3.00 12.00 real interest rate 469 4.71 2.79 0.03 15.61 human capital (tertiary) 730 48.19 23.34 1.53 121.86 r&d expenditure 562 1.65 0.94 0.05 4.29 source: author own calculation 3.1 fixed effects estimation to examine the impact of variables related to entrepreneurial environment, general economy, technological opportunity and political and institutional factors on private equity (pe) investment, the article uses the following regression model: , 1 , 1 2 , 3 , 4 , , , i t i t i t i t i t i t i t privateequity privateequity internetuse marketcap exchangerate     − = + + + + + where it i it   = + i  represents the individual country specific effects while it  is independently and identically random variable term. this allows the regressors to be correlated with the time-invariant component of the error captured by i  but uncorrelated with the idiosyncratic error captured by it  (cameron and trivedi, 2009). subscript i= 1,2,3,……… signify country while t=1,2,3…….. represent period. all the variables have been expressed in log form. , 1i t privateequity − is the initial (previous year) volume of pe investments across countries. the expected sign of initial pe investment is positive as already invested pe investment encourages further investment in that country. ,i t internetuse measures the use of internet over the years by sampled countries. amin ullah khan, muhammad zubair khan, & zafir ullah khan 8 table 2: pair wise correlation of the variables analyzed variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (1) private equity %gdp 1.000 (2) ict 0.534 1.000 (3) market capitalization 0.396 0.258 1.000 (4) reer 0.024 0.157 0.276 1.000 (5) gdp growth 0.015 -0.208 0.083 -0.008 1.000 (6) unemployment 0.044 0.144 -0.259 -0.115 -0.280 1.000 (7) patents 0.163 -0.107 0.260 0.120 0.199 -0.410 1.000 (8) population growth 0.068 -0.080 0.215 0.006 0.276 -0.314 0.166 1.000 (9) tax burden 0.023 -0.007 0.212 0.106 0.279 -0.349 0.308 0.362 1.000 (10) investment profile 0.488 0.707 0.254 0.191 -0.143 0.032 -0.055 -0.008 -0.101 1.000 (11) real interest rate -0.228 -0.344 -0.160 0.065 0.105 0.198 -0.003 0.106 0.114 -0.287 1.000 (12) human capital 0.311 0.707 0.125 0.359 -0.359 0.346 -0.235 -0.336 -0.260 0.548 -0.227 1.000 (13) r&d expenditure 0.247 0.502 0.188 0.214 -0.270 0.229 0.019 -0.248 -0.420 0.357 -0.231 0.539 1.000 journal of applied economics and business studies, volume. 6, issue 1 (2022) 1-24 https://doi.org/10.34260/jaebs.611 9 this has expected positive sign as internet represents the technological environment and at the same time facilitates entrepreneurial and financial environment denotes market capitalization in different countries over different periods. it is the market value of shares outstanding for listed companies at the end of respective years. market capitalization is expected to have positive impact as developed stocks facilities exit from vc (black and gilson, 1998; jeng and wells, 2002) denotes real effective exchange rate (reer) and have been part of the baseline regression and represents all the other independent variables to be put one by one separately. they include gdp growth, unemployment, patents (resident), tax burden, population growth, investment profile, real interest rate, human capital tertiary and r&d expenditure. the institutional variables will not be examined in this research because they do not frequently vary across time to be captured by the fixed effects models. 3.2 summary statistics summary statistics of pe investment, macroeconomic variables and innovation indicators variables are presented in table 1. total maximum observations are 864. the number of observations varies from one variable to another because of missing observations. the mean values of pe as percentage of gdp is 0.2965 with a minimum value of zero and maximum value 0f 6.33. the mean value of pe shows that on the average, the share of pe in gdp is 0.30%. variation in pe data is the smallest among all the variables as is clear from the table value of standard deviation of pe as percentage of gdp which is 0.503. the average number of patent applications by resident is 8149.397 which minimum value 17 and maximum value is 135885. patent variable shows the largest variation which is clear from the standard deviation value 17161.74. the correlation matrix in table 2 shows that ict, market capitalization and investment profile exhibit high correlation with pe investment. gdp growth, tax burden and exchange rate display least correlation. table 2 also shows that pe and real interest rate are negatively correlated; however, the correlation coefficient is smaller in magnitude. among all the variables, the coefficient of interest rate and pe capital shows strong negative correlation. however, higher correlation does not guarantee dependence of a variable on another variable. 4. empirical results and discussion results of overall sample have been presented in table 3, europe in table 4 and asia-pacific in table 5. lagged pe investment show strong positive significant effect on pe in all the speciation used in the fixed effects estimation. this confirms the notion that previous year investment facilities more investment. a population ecology approach suggests that every new entry to the market enhances legitimacy, expertise, and networking opportunities in the market, which further eases new founding (manigart, 1994). in the overall sample, ict, investment profile, university education, r&d expenditure, and market capitalization have strong positive impact on pe investment whereas population growth has strong negative effect. moreover, real interest rates and real exchange rates are also significant determinants displaying negative sign as expected. rest of the variables such as gdp growth, patents, and tax burden have insignificant amin ullah khan, muhammad zubair khan, & zafir ullah khan 10 impact on pe while unemployment is unexpectedly positive but insignificant. region-wise analysis shows that european pe investment is strongly sensitive to ict, real exchange rate, and r&d expenditure and tax burden and moderately responsive to market capitalization, investment profile, and unemployment and university education. tax burden, real exchange rate and unemployment have significant negative effect on pe investment while ict, market capitalization, investment profile, r&d expenditure and university education have significant positive impact on pe. population growth, real interest rate, gdp growth and patents play insignificant role while the negative sign displayed by patent is unexpected. in asia-pacific region, investment profile is the strongest determinant of pe investment. moreover, ict, and university education have significant positive influence on pe while population growth and real interest rate have significant negative bearing on pe investment. rest of the factors is insignificant with unemployment and tax burden displaying unexpected positive signs. it could be inferred that despite the unemployment on the rise and increasing taxes in those economies, business still attract vc investments in asian context. ict is the strongest determinant of pe investment over time in all the samples, although it is more pronounced in europe than in the asia-pacific region. pe investment represents more the demand side potential compared to fundraising that represents the efficiency of an economy to provide supply of pe, thus it can be argued that internet has created ample entrepreneurial opportunities in these economies. however, more investment could also be due to more local fundraising because major proportion of such investments is raised locally in most of the economies leaving possibility that internet has also positively facilitated the fundraising processes and that is why its scale which triggers more investment. strong impact of internet also shows that it might have facilitated pe investment processes particularly information for deal origination, business evaluation, monitoring and exits. volatility of real exchange rate is strong negative determinant of pe investment in europe compared to asia-pacific. one could expect that high reer would result in high pe investment because cross border pe investments are made in terms of us dollars which are converted to destination currencies and any loss of value of destination currency over the years would mean loss of value in terms of us dollars when the proceeds are converted back from the undervalued local currency. however, minardi et al. (2017) show that in the long-term exchange risk does not affect international investors decisions to investors’ returns in brazil because good cycles compensate bad cycles. moreover, domestic pe investments, which are a major part, are not affected by currency exchange directly as they affect cross border investors. an increase in reer means that exports become costlier, and imports get cheaper or in other words the loss of a country’s trade competitiveness in comparison to its major trading partners . for example, us exports of automobiles, aluminum, iron and steel, wood and pulp, and paper plunge when dollar gets strength (thorbecke, 2018). thus, countries losing trade efficiency would certainly lose business opportunities as well and leave narrow space for pe/vc actors. increase in exchange rate reflects a country’s lack of trade competitiveness and therefore entrepreneurial opportunities, deal journal of applied economics and business studies, volume. 6, issue 1 (2022) 1-24 https://doi.org/10.34260/jaebs.611 11 flow and demand side of pe/vc. rodrik (2008) argues that currency undervaluation stimulates economic growth. the question why it is more pronounced in europe than in asia-pacific region might be due to the flexible and sustained interventions by asian governments that have effectively achieved growth through stable and competitive real exchange rate policies (guzman, ocampo and stiglitz, 2018). amin ullah khan, muhammad zubair khan, & zafir ullah khan 12 table 3: fixed effects estimation results overall: dependent variable is pe investment (1) (2) (3) (4) (5) (6) (7) (8) (9) pe investmentt-1 0.521*** 0.468*** 0.497*** 0.514*** 0.388*** 0.573*** 0.338*** 0.458*** 0.369*** ict 0.200*** 0.259*** 0.215*** 0.204*** 0.323*** ---- market cap. 0.261** 0.280*** 0.266*** 0.260*** 0.338*** 0.393*** 0.374** 0.328*** 0.375*** reer -0.741* -0.896* -1.068** -0.722 -0.964* -0.973* -1.630** -1.874** -0.617 gdp growth 0.060 -------- unemployment -0.044 ------- patents --0.100 ------ pop. growth ----0.166*** ---- tax burden -----0.396 ---- invest. profile -----0.746*** --0.813*** real interest rate -------0.267** -- human capital ------1.172*** 1.173*** - rd expenditure --------0.705*** constant 0.782 1.096 1.474 0.647 2.454 0.402 0.642 1.602 -1.933 observations 516 437 568 498 514 557 184 447 400 r-squared 0.556 0.533 0.583 0.613 0.479 0.572 0.462 .592 0.393 no. of countries 28 27 28 26 28 27 16 27 27 *** p<0.01, ** p<0.05, * p<0.10 journal of applied economics and business studies, volume. 6, issue 1 (2022) 1-24 https://doi.org/10.34260/jaebs.611 13 table 4: fixed effects estimations results europe: dependent variable is pe investment (1) (2) (3) (4) (5) (6) (7) (8) (9) pe investmentt-1 0.572*** 0.222* 0.558*** 0.386*** 0.629*** 0.618*** 0.239 0.550*** 0.358*** ict 0.147*** 0.196*** 0.133*** 0.279*** 0.114*** market capitalization 0.205 0.295*** 0.248** 0.245** 0.267* 0.365*** 0.276 0.288** 0.229** reer -1.544*** -2.428*** -1.778*** -2.143*** -1.324*** -1.552*** -2.733 2.131*** -2.051*** gdp growth 0.049 -------- unemployment --0.359** ------- patents ---0.007 ------ tax burden ----0.608*** ----- population growth -----0.095 ---- investment profile -----0.462** --0.849*** real interest rate -------0.165 -- human capital ------1.018* 0.908*** - rd expenditure --------0.687** constant 4.932** 8.513*** 5.925** 9.096*** 3.869* 3.905** 6.294 4.081* 5.019* observations 346 258 383 331 322 389 96 348 289 r-squared 0.697 0.595 0.707 0.573 0.773 0.688 .361 .681 0.462 no. of countries 19 18 19 19 17 19 8 19 19 *** p<0.01, ** p<0.05, * p<0.10 amin ullah khan, muhammad zubair khan, & zafir ullah khan 14 table 5: fixed effects estimation results asia-pacific: dependent variable is pe investment (1) (2) (3) (4) (5) (6) (7) (8) (9) pe investmentt-1 0.389*** 0.370*** 0.384*** 0.298*** 0.385*** 0.407*** 0.328** 0.347** 0.310** ict 0.480*** 0.638*** 0.502** 0.376* 0.484** market capitalization 0.407* 0.365** 0.262 0.446* 0.240* 0.452* 0.433* 0.444** 0.752* reer -0.242 -0.288 -0.729 -0.555 -0.327 -0.082 -1.326 -1.438 0.802 gdp growth 0.064 -------- unemployment -0.725 ------- patents --0.226 ------ tax burden ---4.842 ----- population growth -----0.338** ---- investment profile -----3.246*** --1.951 real interest rate -------0.286** -- human capital ------1.463** 1.655** - rd expenditure --------0.707 constant -3.297 -4.465 -2.556 -22.583 -2.092 -9.889** -2.015 -2.474 -12.717 observations 170 179 185 183 176 168 88 99 111 r-squared 0.488 0.580 0.533 0.479 0.551 0.520 .506 .478 0.393 no. of countries 9 9 9 9 9 8 8 8 8 *** p<0.01, ** p<0.05, * p<0.10 journal of applied economics and business studies, volume. 6, issue 1 (2022) 1-24 https://doi.org/10.34260/jaebs.611 15 table 6: lag effects of gdp growth and patents: dependent variable is pe investment (1) (2) (4) (5) pe investmentt-1 0.521*** 0.537*** 0.497*** 0.493*** ict 0.200*** 0.214*** 0.215*** 0.222*** market capitalization 0.261** 0.205** 0.266*** 0.262*** reer -0.741* -1.219** -1.068** -1.056** gdp growth 0.060 --- lagged gdp growth -0.059* -- patents --0.100 - lagged patents ---0.123* constant 0.782 3.185 1.474 1.227 observations 516 518 568 567 r-squared 0.556 0.606 0.583 0.584 number of countries 28 28 28 28 *** p<0.01, ** p<0.05, * p<0.10 amin ullah khan, muhammad zubair khan, & zafir ullah khan 16 market capitalization is strong in overall sample than the regional samples and european investors are more sensitive to stock market signals compared to their asia-pacific counterparts in within regressions. in terms of political risk factors, asia-pacific pe market is more sensitive to investment profile compared to its european counterpart in within regressions although both are significant. the tax results are interesting in the sense that tax burden does not refrain pe investors in asia-pacific from investment quite opposite to their european counterparts who are strongly sensitive to taxation. this is also consistent with the previous research. studies showing that taxation is not significant determinant have taken countries such as us, australia, and japan along with european counties (jeng and wells, 2000; romain and van pottelsberghe, 2004) while studies showing that taxation is significant have taken only european countries in their sample (cherif and gazdar, 2011, schroder, 2009; groh and von liechtenstein, 2009). the results of unemployment confirm the previous studies that it has significant negative effect on pe/vc investment in europe unemployment (cherif and gazdar, 2011). we could not find studies on asia-pacific region studying unemployment as independent variable. there are certain features that cause a rise in unemployment which are unemployment benefits, high unionisation, high payroll taxes and poor educational standards and usually these features are found in the european economies (nickell, 1997). thus, the significant negative impact in european context is comprehendable and the results support the overall sense that labour market have negative effect on vc investment. showing positive sign, gdp growth does not exert significant influence on the pe in all the three samples in within regressions. however, when we took the lag gdp growth, it displays significance (see table 6). similarly, lag of patent was taken and again it got significance. reason for this is that both gdp and patents should have influence on the pe investments in the following year. although, projected gpd growth is reported during the year from time to time, the annual gdp growth is determined and published after the completion of financial year. if gdp growth has any signaling value, then the previous year growth rate should be meaningful compared to current year because investors would not be knowing the current year growth rate as such. this is different from other variables such as taxation, interest rates or exchange rate which have impact after they are in effect. similarly, patents should have slow effect whether we talk about signaling influence or actual influence. strong positive effect of r&d expenditure on vc investment in the european countries was expected. results of asia-pacific countries are surprising though. r&d expenditure is not only insignificant, but it also reduces the effect of other variables in the specification i.e., investment journal of applied economics and business studies, volume. 6, issue 1 (2022) 1-24 https://doi.org/10.34260/jaebs.611 17 profile, market capitalization, and real exchange rate. the previous research on effect of r&d on vc investment has been conducted in europe (schertler, 2007; romain and van pottelsberghe, 2004; to da rin et al., 2006b), and hardly any research on asian context has taken this variable. the fact that r&d expenditure take time to fuel entrepreneurial activities, we take log1 and log 2 to see its impact on pe. the results confirmed that it is still insignificant. population growth exhibit interesting results. population growth significantly and negatively influences asia-pacific pe but not european pe. overall, it has negative impact on venture capital and pe investments. we can conclude that international risk investors choose destinations with increasing prospects for prosperity (which is linked with low birth rates). high domestic pe investment in low birth rate countries reflect their rising household savings at micro level and high human capital investment for sustained economic growth and reduction in poverty at macro level (sinding, 2009). asian economies display strong negative impact of interest rate on pe investment compared to european countries during the period under consideration. overall, the results support the findings of bonini and alkan (2012) that interest rates play negative role opposite to romain and pottelsberghe (2004) and füss and schweizer (2008) who found positive relationship between interest rates and venture capital investment. it can be inferred that as interest rates increase, economic growth weakens and fundraising reduces, both affecting pe investment (bonini and alkan, 2012). however, romain and pottelsberghe (2004) found that cost of capital has more impact on demand side than on fundraising. the asian markets might be more responsive to interest rates because of fluctuation. as figure 3 shows that interest rates fluctuate more in asian context and the fluctuation goes against the fluctuation in pe investment in figure 2 or in other words interest rates push pe investments in its opposite direction. overall, the trend analysis also show that interest rates have gradually decreased while pe investments have increased both in europe and asia attesting a negative relationship in long term. finally, the university enrollment demonstrating strong positive sign in all samples point to the importance of science and technological base or human capital that fuels an entrepreneurial economy. 5. conclusion the study has analyzed the impact of different factors influencing pe investment using data of 32 european and asia-pacific countries over the period 1990-2017. results show that ict, stock market capitalization, investment profile and human capital are strongly linked to pe investments in the overall sample as well as in europe and asia-pacific. in addition, population growth and real interest rate have significant negative impacts while r&d expenditure has strong positive impact on pe investment in the overall sample. region-wise, pe investment significantly and positively respond to ict and market capitalization in europe and to investment profile in the asia-pacific region. real exchange rate, unemployment and tax burden exert significant negative impact on european pe whereas the population growth and real interest rates have significant negative impact in asia-pacific context. amin ullah khan, muhammad zubair khan, & zafir ullah khan 18 the article claims important contributions to the research on private equity. ict, population growth, and real exchange rate have been identified new determinates in the empirical investigation which also have theoretical relevance. moreover, the regional comparisons reveal important differences that helps interpret the conflicted results in previous research. the article also fills the regional gap in literature by focusing on asia-pacific region. in future, more comparative studies are required that will remove the conflicting accounts in pe/vc research that will result into broader consensus and convergence. in future more research with latest data will be helpful because much of the developing pe has gained momentum after the 2000 dot com bubble and is still growing and spreading to other developing economies. references ahlstrom, d., bruton, g. d. and yeh, k. s. (2007) ‘venture capital in china: past, present, and future’, asia pacific journal of management, 24(3), pp. 247–268. doi: 10.1007/s10490006-9032-1. asongu, s. a., & acha-anyi, p. n. (2017) ‘ict, conflicts in financial intermediation and financial access: evidence of synergy and threshold effects’, netnomics, 18(2–3), pp. 131–168. asongu, s. a. and moulin, b. (2016) ‘the role of ict in reducing information asymmetry for financial access’, research in international business and finance. elsevier ltd, 38, pp. 202–213. doi: 10.1016/j.ribaf.2016.04.011. baygan, g. and freudenberg, m. (2000) the internationalisation of venture capital activity in oecd countries: implications for measurement and policy. oecd science, technology and industry working papers, doc(2000)7. doi: 10.1787/084236411045. bertoni, f. and colombo, m. g. (2015) ‘the patterns of venture capital investment in europe’, small business economics. springer us, 45(3), pp. 543–560. doi: 10.1007/s11187-0159662-0. black, b. s. and gilson, r. j. (1999) ‘does venture capital require an active stock market: banks versus stock markets’, journal of applied corporate finance. elsevier, 11(4), pp. 36–48. bonini, s. and alkan, s. (2012) ‘the political and legal determinants of venture capital investments around the world’, small business economics. springer science and business media llc, 39(4), pp. 997–1016. doi: 10.1007/s11187-011-9323-x. bozkaya, a. and kerr, w. r. (2014) ‘labor regulations and european venture capital’. blackwell publishing inc., 23(4), pp. 776–810. doi: 10.1111/jems.12077. cameron, a. c. and trivedi, p. k. (2009) microeconometrics using stata. texas: a stata press publication statacorp lp. carey, m. et al. (1993) ‘the economics of private placements: a new look’, financial markets, institutions and instruments. cambridge ma: blackwell, 2(3), pp. 1–66. cherif, m. and gazdar, k. (2011) ‘what drives venture capital investments in europe? new results from a panel data analysis’, journal of applied business and economics, 12(3), pp. 122–139. cumming, d. et al. (2010) ‘legal protection , corruption and private equity returns in asia’, journal of business ethics. springer, 95(2), pp. 173–193. doi: 10.1007/sl0551-011-08536. journal of applied economics and business studies, volume. 6, issue 1 (2022) 1-24 https://doi.org/10.34260/jaebs.611 19 cumming, d. and fleming, g. (2002) a law and finance analysis of venture capital exits in emerging markets. the australian national university working paper series in finance 02-03. cumming, d., fleming, g. and schwienbacher, a. (2006) ‘legality and venture capital exits’, journal of corporate finance. elsevier, 12(2), pp. 214–245. doi: 10.1016/j.jcorpfin.2004.12.004. cumming, d., henriques, i. and sadorsky, p. (2016) ‘cleantech’ venture capital around the world’, international review of financial analysis. elsevier b.v. doi: 10.1016/j.irfa.2016.01.015. cumming, d. j. and macintosh, j. g. (2003) ‘a cross-country comparison of full and partial venture capital exits’, journal of banking and finance. elsevier, 27(3), pp. 511–548. doi: 10.1016/s0378-4266(02)00389-8. cumming, d. j. and vismara, s. (2017) ‘de-segmenting research in entrepreneurial finance’, venture capital. routledge, 1066(january), pp. 1–11. doi: 10.1080/13691066.2016.1225910. cumming, d. and knill, a. (2012) ‘disclosure, venture capital and entrepreneurial spawning’, journal of international business studies. springer, 43(6), pp. 563–590. doi: 10.1057/jibs.2012.9. cumming, d., schmidt, d. and walz, u. (2008) ‘legality and venture capital governance around the world’, journal of business venturing. elsevier, 25(1), pp. 54–72. doi: 10.1016/j.jbusvent.2008.07.001. cumming, d. and zhang, y. (2016) ‘alternative investments in emerging markets: a review and new trends’, emerging markets review. elsevier b.v., 29, pp. 1–23. doi: 10.1016/j.ememar.2016.08.022. devigne, d. et al. (2018) ‘venture capital internationalization: synthesis and future research directions’, journal of economic surveys, 32(5), pp. 1414–1445. doi: 10.1111/joes.12276. ehrilich, p. r. and holdren, j. p. (1971) ‘impact of population growth on jstor’, science, 171(3977), pp. 1212–1217. available at: https://www.jstor.org/stable/1731166?seq=3#metadata_info_tab_contents. eleswarapu, v. r. and venkataraman, k. (2006) ‘the impact of legal and political institutions on equity trading costs : a cross-country analysis’, the review of financial studies, 19(3), pp. 1081–1111. doi: 10.1093/rfs/hhj026. fiet, j. o. (1995) ‘reliance upon informants in the venture capital industry’, journal of business venturing. elsevier, 10(3), pp. 195–223. doi: 10.1016/0883-9026(94)00039-w. gabor, d., & brooks, s. (2017) ‘the digital revolution in financial inclusion: international development in the fintech era’, new political economy, 22(4), pp. 423–436. gilson, r. j. and black, b. s. (2000) ‘does venture capital require an active stock market?’, ssrn, 48(166). doi: 10.2139/ssrn.219174. gompers, p. and lerner, j. (1998) what drives venture capital fundraising? brooking institutions press, brookings working paper on economic activity, microeconomics. greene, w. h. (2008) no title, econometric analysis. 6th ed.,. upper saddle river, n.j.: prentice hall. groh, a. p. and von liechtenstein, h. (2009) ‘how attractive is central eastern europe for risk capital investors?’, journal of international money and finance. elsevier ltd, 28(4), pp. 625–647. doi: 10.1016/j.jimonfin.2009.01.006. groh, a. p. and wallmeroth, j. (2016) ‘determinants of venture capital investments in emerging amin ullah khan, muhammad zubair khan, & zafir ullah khan 20 markets’, emerging markets review. elsevier b.v., 29, pp. 104–132. doi: 10.1016/j.ememar.2016.08.020. guler, i. and guillén, m. f. (2010) ‘home country networks and foreign expansion: evidence from the venture capital industry’, academy of management journal. academy of management briarcliff manor, ny , 53(2), pp. 390–410. doi: 10.5465/amj.2010.49389027. guzman, m., ocampo, j. a. and stiglitz, j. e. (2018) ‘real exchange rate policies for economic development’, world development. the authors, 110, pp. 51–62. doi: 10.1016/j.worlddev.2018.05.017. hain, d. et al. (2016) ‘determinants of cross-border venture capital investments in emerging and developed economies : the effects of relational and institutional trust’, journal of business ethics. springer netherlands, 138(4), pp. 743–764. doi: 10.1007/s10551-0152772-4. hearn, b., randøy, t. and oxelheim, l. (2016) ‘the institutional determinants of private equity involvement in business groups — the case of africa’, journal of world business. elsevier inc. doi: 10.1016/j.jwb.2016.02.002. hege, u., palomino, f. and schwienbacher, a. (2011) ‘venture capital performance: the disparity between europe and the united states’, ssrn electronic journal. doi: 10.2139/ssrn.482322. jeng, l. a. and wells, p. c. (2000) ‘the determinants of venture capital funding: evidence across countries’, journal of corporate finance. elsevier, 6(3), pp. 241–289. joshi, s. and schultz, t. p. (2007) ‘family planning as an investment in development: evaluation of a program’s consequences in matlab, bangladesh’. karsai, j. et al. (1998) ‘screening and valuing venture capital investments: evidence from hungary, poland and slovakia’, entrepreneurship and regional development, 10(3), pp. 203–224. doi: 10.1080/08985629800000012. karsai, j., wright, m. and filatotchev, i. (1997) ‘venture capital in transition economies: the case of hungary’, entrepreneurship theory and practice. sage publications, 21(4), pp. 93–110. doi: 10.1177/104225879702100407. khan, m. z. et al. (2021) ‘on the upside or flipside: where is venture capital positioned in the era of disruptions?’, technology in society, xxx(xxx), p. xxx. khan, m. z., ferrier, w. and khan, z. u. (2020) ‘institutions, financial development and private equity fundraising’, submitted manuscript. khan, z. m., khan, z. u. and hameed, a. (2020) ‘institutions, digitization, innovation and venture capital: evidence from europe and the asia-pacific’, journal of applied economics and business studies. awkum, 2(2), p. 41. doi: 10.34260/jaebs.423. klasen, s. and lawson, d. (2007) the impact of population growth on economic growth and poverty reduction in uganda. klonowski, d. (2007) ‘the venture capital investment process in emerging markets: evidence from central and eastern europe’, international journal of emerging markets, 2(4), pp. 361– 382. doi: 10.1108/17468800710824518. kumar, v. and orleck, s. (2002) ‘why does private equity vary across countries and time ?’, working paper, pp. 1–38. lechman, e. and marszk, a. (2019) ict-driven economic and financial development : analyses of oecd countries. academic press, an imprint of elsevier. journal of applied economics and business studies, volume. 6, issue 1 (2022) 1-24 https://doi.org/10.34260/jaebs.611 21 lerner, j. (1994) ‘the syndication of venture capital investments’, financial management, 23(3), p. 16. doi: 10.2307/3665618. li, y. and zahra, s. a. (2012) ‘formal institutions, culture, and venture capital activity: a crosscountry analysis’, journal of business venturing. elsevier b.v., 27(1), pp. 95–111. doi: 10.1016/j.jbusvent.2010.06.003. manigart, s. (1994) ‘the founding rate of venture capital firms in three european countries (1970– 1990)’, journal of business venturing. elsevier, 9(6), pp. 525–541. doi: 10.1016/08839026(94)90020-5. milosevic, m. (2018) ‘skills or networks? success and fundraising determinants in a low performing venture capital market’, research policy. elsevier, 47(1), pp. 49–60. doi: 10.1016/j.respol.2017.09.009. minardi, a. m. a. f. et al. (2017) ‘currency impact in brazilian pe/vc deals’ performance’, journal of private equity, 20(3), pp. 45–51. doi: 10.3905/jpe.2017.20.3.045. nickell, s. (1997) ‘unemployment and labor market rigidities : europe versus north america’, 11(3), pp. 55–74. ning, y., wang, w. and yu, b. (2015) ‘the driving forces of venture capital investments’, small business economics. springer international publishing, 44, pp. 315–344. doi: 10.1007/s11187-014-9591-3. oberli, a. (2014) ‘private equity in emerging markets: drivers in asia compared with developed countries’, the journal of private equity. institutional investor journals group, 17(3), pp. 45–61. doi: 10.3905/jpe.2014.17.3.045. ocampo, j. a. (2018) ‘international asymmetries and the design of the international financial system 1’, in critical issues in international financial reform. routledge, pp. 45–74. doi: 10.4324/9781351323765-2. peterson, e. w. f. (2017) ‘the role of population in economic growth’, sage open, 7(4), p. 215824401773609. doi: 10.1177/2158244017736094. pradhan, r. p. et al. (2017) ‘ict-finance-growth nexus: empirical evidence from the next-11 countries’, cuadernos de economia. elsevier españa, s.l.u, 40(113), pp. 115–134. doi: 10.1016/j.cesjef.2016.02.003. pradhan, r. p. et al. (2018) ‘mobile telephony, economic growth, financial development, foreign direct investment, and imports of ict goods: the case of the g-20 countries’, economia e politica industriale. springer international publishing, 45(2), pp. 279–310. doi: 10.1007/s40812-017-0084-7. pradhan, r. p., arvin, m. b. and norman, n. r. (2015) ‘the dynamics of information and communications technologies infrastructure, economic growth, and financial development: evidence from asian countries’, technology in society. elsevier ltd, 42, pp. 135–149. doi: 10.1016/j.techsoc.2015.04.002. r., f. and denis, s. (2008) ‘dynamic interactions between venture capital returns and the macroeconomy : theoretical and empirical evidence from the united states’. working paper, european business school (ebs) and whu., (june). ragosa, g. and warren, p. (2019) ‘unpacking the determinants of cross-border private investment in renewable energy in developing countries’, journal of cleaner production. elsevier ltd, 235, pp. 854–865. doi: 10.1016/j.jclepro.2019.06.166. da rin, m., nicodano, g., et al. (2006) ‘public policy and the creation of active venture capital markets’, journal of public economics. north-holland, 90(8–9), pp. 1699–1723. doi: amin ullah khan, muhammad zubair khan, & zafir ullah khan 22 10.1016/j.jpubeco.2005.09.013. da rin, m., hege, u., et al. (2006) ‘the law and finance of venture capital financing in europe: findings from the ricafe research project’, european business organization law review, 7(2), pp. 525–547. doi: 10.1017/s1566752906005258. rodrik, d. (2008) ‘the real exchange rate and economic growth’, brookings papers on economic activity, (2), pp. 365–412. doi: 10.7312/columbia/9780231175081.003.0012. romain, a. and pottelsberghe, b. van (2004) the determinants of venture capital: additional evidence, discussion paper series 1: studies of the economic research centre. 19/2004. deutsche bundesbank. schertler, a. (2003) ‘a comparative overview of venture capital in europe and the united states’, international journal of entrepreneurship education, 1(4), pp. 539–584. schertler, a (2003) ‘driving forces of venture capital investments in europe: a dynamic panel data analysis’, working paper, (03). available at: http://www.econstor.eu/handle/10419/17695. schertler, a (2007) ‘knowledge capital and venture capital investments: new evidence from european panel data’, german economic review. john wiley & sons, ltd (10.1111), 8(1), pp. 64–88. doi: 10.1111/j.1468-0475.2007.00134.x. schertler, andrea (2007) ‘knowledge capital and venture capital investments: new evidence from european panel data’, german economic review. john wiley & sons, ltd (10.1111), 8(1), pp. 64–88. doi: 10.1111/j.1468-0475.2007.00134.x. schröder, c. (2009) financial system and innovationsdeterminants of early stage venture capital in europe. schumpeter discussion papers, sdp 2009-04 issn 1867-5352. sinding, s. w. (2009) ‘population, poverty and economic development’, philosophical transactions of the royal society b: biological sciences. royal society, 364(1532), pp. 3023–3030. doi: 10.1098/rstb.2009.0145. smolarski, j. et al. (2005) ‘risk management in indian venture capital and private equity firms: a comparative study’, thunderbird international business review, pp. 469–488. doi: 10.1002/tie.20063. thorbecke, w. (2018) ‘the exposure of u.s. manufacturing industries to exchange rates’, international review of economics and finance. elsevier ltd, 58(may), pp. 538–549. doi: 10.1016/j.iref.2018.06.002. tykvova, t. (2017) ‘venture capital and private equity financing : an overview of recent literature and an agenda for future research’, journal of business economics, volume 88, pp. 325– 362. doi: 10.1007/s11573-017-0874-4. tykvová, t. (2018) ‘legal framework quality and success of ( different types of ) venture capital investments’, journal of banking and finance. elsevier, 87, pp. 333–350. doi: 10.1016/j.jbankfin.2017.10.015. wright, m. et al. (1999) ‘transition and active investors: venture capital in hungary, poland and slovakia’, post-communist economies, 11(1), pp. 26–46. doi: 10.1080/14631379996039. journal of applied economics and business studies, volume. 6, issue 1 (2022) 1-24 https://doi.org/10.34260/jaebs.611 23 appendix 1: variable descriptions and data sources variables description source private equity investment private equity investment as percentage of gdp. it includes all stages of venture capital, expansion, growth, and buyout investments. avcj, eurostat, evca yearbooks 19902017 ict ict represents individuals who have used the internet (from any location) in the last 3 months. the internet can be used via a computer, mobile phone, personal digital assistant, games machine, digital tv etc. world development indicators, world bank databank 19902017 market capitalization market capitalization (also known as market value) is the share price times the number of shares outstanding (including their several classes) for listed domestic companies. investment funds, unit trusts, and companies whose only business goal is to hold shares of other listed companies are excluded. data are end of year values. world development indicators, world bank databank 19902017 real effective exchange rate (reer) real effective exchange rate is the nominal effective exchange rate (a measure of the value of a currency against a weighted average of several foreign currencies) divided by a price deflator or index of costs. world development indicators, world bank databank 19902017 gdp growth annual percentage growth rate of gdp. aggregates are based on constant 2010 u.s. dollars. world development indicators, world bank databank 19902017 unemployment (national) unemployment refers to the share of the labor force that is without work but available for and seeking employment. definitions of labor force and unemployment differ by country. world development indicators, world bank databank 19902017 patent (residents) patent applications are worldwide patent applications filed through the patent cooperation treaty procedure or with a national patent office for exclusive rights for an invention--a product or process that provides a new way of doing something or offers a new technical solution to a problem. a patent provides protection for the invention to the owner of the patent for a limited period, generally 20 years. world development indicators, world bank databank 19902017 population growth annual population growth rate for year t is the exponential rate of growth of midyear population from year t-1 to t, expressed as a percentage. population is based on the de facto definition of population, which counts all residents regardless of legal status or citizenship. world development indicators, world bank databank 19902017 tax burden tax burden is a composite measure that reflects marginal tax rates on both personal and corporate freedom of the world (heritage 19952017 amin ullah khan, muhammad zubair khan, & zafir ullah khan 24 income and the overall level of taxation (including direct and indirect taxes imposed by all levels of government) as a percentage of gross domestic product (gdp). foundation) real interest rate it is the lending interest rate adjusted for inflation as measured by the gdp deflator. the terms and conditions attached to lending rates differ by country, however, limiting their comparability. world development indicators, world bank databank 19902017 human capital gross enrollment ratio is the ratio of total enrollment, regardless of age, to the population of the age group that officially corresponds to the level of education shown. tertiary education, whether to an advanced research qualification, normally requires, as a minimum condition of admission, the successful completion of education at the secondary level. world development indicators, world bank databank 19902017 r&d expenditures (% of gdp) gloss domestic expenditures on research and development (r&d), expressed as a percent of gdp. they include both capital and current expenditures in the four main sectors: business enterprise, government, human capital and private non-profit. r&d covers basic research, applied research, and experimental development. world development indicators, world bank databank 19962016 investment profile this is an assessment of factors affecting the risk to investment that are not covered by other political, economic and financial risk components. the risk rating assigned is the sum of three subcomponents, each with a maximum score of four points and a minimum score of 0 points. a score of 4 points equates to very low risk and a score of 0 points to very high risk. the subcomponents are: • contract viability/expropriation • profits repatriation • payment delays icrg 19902017 note: the description of the data has been taken from the source document/site of the data without making any changes. journal of applied economics and business studies, volume. 5, issue 1 (2021) 109-130 https://doi.org/10.34260/jaebs.516 109 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x tax aggressiveness, tunneling and firm value: interactive effect of corporate governance in business groups hamid ullah1*, hamid ali shah2, sajjad ahmad khan3 1 assistant professor (finance), department of management sciences, islamia college 2 assistant professor (finance), quaid-e-azam college peshawar, university of peshawar 3 director, institute of management studies, university of peshawar abstract this study extends the literature on tax aggressiveness (ta) from agency perspectives in the rarely discussed case of group firms. more specifically, the study investigates the relationship between firms ta, tunnelling and value of firm. moreover, the study also investigates the impact of the moderating role of corporate governance (cg) in counterfeiting the conflicts of interests in group firms. for this purpose, sample data of 160 non-financial pakistani firms belonging to groups for the period from 2009 to 2018 is analyzed through two stages least square regression (2sls) models. the findings reveal that tunneling, group ownership and managerial ownership show direct association with ta. while cg agents, board and audit committee independence and external audit quality exhibits an indirect relationship with ta. moreover, the tunneling-related ta has a negative effect on the firm value. however, the estimated interaction effects show that cg mitigates this negative relationship between the tunneling-related ta and firm value. thus, good cg ameliorates the expropriations by the controlling shareholders and ta becomes a value enhancing activity in business groups. keywords tax aggressiveness, firm value, business groups, corporate governance, external audit quality jel classification h51; h75; i18 1. introduction in general governments take more than 30% of a firm’s pre-tax income and pakistan is no exception. this gigantic firms’ tax cost could influence managers to devise plans to mitigate the cost and enhance profits after tax (chen et al., 2008). looking through the lens, managers as stewards, it is not irrational to assume that managers’ aggressive * hamidullah@icp.edu.pk hamid ullah, hamid ali shah and sajjad ahmad khan 110 tax planning will contribute towards the cherished goal of shareholders wealth maximization (scholes et al., 2009). however, agency theory predicts that these plans might translate to contribute towards furthering managerial opportunism because it might reduce corporate transparency (desai & dharmapala, 2006). further, if a firm belongs to some business group then it may minimize its tax cost at the group level by shifting profits within a group (choi, sami & zhuo, 2010). to this end, jung, kim and kim (2009) reported that the marginal tax rate of group firms is lower relative to that of non-group firms of the same industry and size due to aggressive tax avoidance. similarly, klassen, lang and wolfson (1993) in the case of multinational corporations operating in countries with high and low tax rate countries found that these firms, through related party transactions (rpt), transfer their funds from economies with high tax rates to those with low tax rates (ko, 2000). existing studies propose number of diverse factors that are associated with the ta behavior of firms. these factors are size of firm, international operations, leverage, capital intensity, executive compensations and political affiliation of firms (see e.g., richardson & lanis, 2007; and richardson & lanis, 2007). however, these studies ignore the aggressive tax behaviors that have potential non-tax agency costs such as regulatory or reputational (scholes, wolfson, erickson, maydew, & shevlin, 2005). the agency perspective of corporate ta issues is interlinked with corporate governance (cg). from the agency perspective, cash saved from ta activities is transferred by the firms’ controlling shareholders having low cash flow rights to a firm in which these shareholders have high cash flow rights (desai & dharmapala, 2006). kim, quinn, and wilson (2016) documented a positive (negative) association between managerial ownership and ta (tax expenses paid in cash). an indirect association between corporate ta and family ownership is also reported (see e.g., chen et al., 2010; and landry, deslandes, & fortin, 2013). existing literature suggests that cg mechanisms mitigate agency problems and empirical studies report significant positive effect of cg on value of firm and tax avoidance (see e.g., desai & dharmapala, 2009; and yee, sapiei, & abdullah, 2018). number of studies show that the negative effect of corporate ta on value of firm is mitigated through better cg practices in placed by the firms. arif and hashim (2013) found that the value relevance of corporate ta is relatively higher for well-governed firms. large number of studies report that independent boards with sufficient financial compensations, independent audit committees, separation of the role of chairman and ceo, quality of external audits and close monitoring by creditors and regulatory authorities discipline the expropriating behaviour of controlling shareholders (cheung et al. 2006; gao & kling 2008; nekhili & cherif, 2011; and ryngaert & thomas, 2012). most of the work on this topic is done in developed countries where dispersed ownership with conflict between managers and shareholders or in china where government act as a controlling shareholder (pilos, 2017; and kim, li & zhang, 2011). journal of applied economics and business studies, volume. 5, issue 1 (2021) 1-22 https://doi.org/10.34260/jaebs.511 111 the current study takes this literature a step further and tends to investigate the implication of non-tax costs incurred due to aggressive tax avoidance on value of firm in the case of group firms. this study is conducted in an institutional setting where business groups act as controlling shareholders of firms and are reported to be aggressively involved in tunneling-related activities (see e.g., ullah & shah, 2015). the relationships among ta, cg, and firm value are largely unexplored in pakistani group firms. earlier, shabbir, waheed, and mahmood (2017) examined the relationships of firm size, growth, debts and inventory changes, roa, ta and tax avoidance. akram, iqbal, and mughal (2012) followed exploratory research design and considered macroeconomic factors that affect corporate tax. they explored that lack of enforcement, lower penalties, lack of public awareness, and high compliance cost are the factors that discourage ta. whereas, the current study is entirely different from these aforementioned studies in a way that it follows an agency perspective of corporate ta and the impact of tunneling-related ta on firm value and further analyzes the moderating role of cg in counterfeiting the agency conflict arising from ta in unique institutional settings of business groups.the unique institutional setting is considered the most suitable to conduct this study and to analyse the behaviour of the business groups in managing the tax at group level within the agency framework and determining its effect on value of firm in the presence of cg mechanisms. thus, following an agency framework, this multi-prong study contributes to the literature and broadens the understanding of tax-sheltering in family-dominated and business groups firms. the study tests that if tax sheltering activities are benefiting the overall shareholders or if the resources saved through such activities are tunneled out by the controlling shareholders. the study also assesses the contribution of the cg agents in constraining the expropriations of saved tax through tax sheltering activities. 2. literature review the extant literature highlights different terms for the ta such as tax avoidance, tax planning, tax management, and tax sheltering. according to dyreng, hanlon & maydew (2010) the management practices whereby firms pay lower tax per dollar of the pre-tax accounting income is known as ta. whereas, frank, lynch and rego (2009) stated that ta is tax planning that involve manipulating and reporting reduced accounting income and that may or may not be fraudulent tax evasion. similarly, chen et al. (2010) also stated that ta could involve legal and illegal practices. whereas, hanlon and heitzman (2010) describe ta to include activities such as investments in municipal bond, noncompliance and tax evasion. as such ta is multifaceted phenomenon that influences firms’ tax payment. 2.1 tax aggressiveness and tunneling the traditional theory of finance assumed that managers undertake ta activities to maximize shareholders wealth (scholes, et al., 2009). the tax saved may be reinvested or given to the shareholders. however, agency theory predicts that in the case of group hamid ullah, hamid ali shah and sajjad ahmad khan 112 firms, cash saved from ta activities might be transferred to other firm(s) by the controlling shareholders to protect selfinterests (desai & dharmapala, 2006; desai et al., 2009). there is an opportunity for firms belonging to business group to play down their tax liabilities through rpts (choi et al., 2010). findings of empirical studies show that firms belonging to group shift income and expropriate resources within the group (ullah & shah, 2015). other studies provide more direct evidence of ta by multinational corporations (mncs); mncs are reported to shift incomes to business units in other countries and exposed to lower tax rates (klassen, lang & wolfson, 1993; ko 2000). as revealed in the literature, in economies with concentrated or family ownership of firms the conflict between principals-principals are of major concern (see e.g., la porta et al., 1999). further, utility maximizer majority shareholders can influence the management and will expropriate wealth at the expense of minority shareholders through rpts due to their more cash flow rights (jensen & meckling, 1976). empirical studies document that controlling shareholders with excess power are involved in self-dealing such as ta, excessive compensation, personal loans, transfer pricing, corporate opportunities, outright expropriation of assets and profits (desai & dharmapala, 2006; djankov et al., 2008; shleifer & vishny, 1997). few studies documented value destroying effect of ta because cash saved through ta is expropriated by the controlling shareholders through rpts at the expense of other shareholders (chen et al., 2010; desai & dharmapala, 2006; and ryngaert & thomas, 2007). thus, rpts and tax avoidance will relate positively on the conjecture that the saved tax could be expropriated by the controlling shareholders in group firms. h1: there is positive relationship between rpts and ta in group firms. 2.2 ownership structure and tax aggressiveness firms’ structures of ownership and agency problems have association (shleifer & vishny, 1997). majority of the pakistani firms are family owned, have concentrated ownerships and belong to some groups (abdullah, shah, gohar, & iqbal, 2011). family controlled associated firms have concentrated ownership either due to the weaker enforcement of property rights or due to the absences of specialized institutions in the capital market (fan & wong, 2005; and la porta et al., 1997). the possible nexus between family members and the opportunistic could spin the ta activities in the favour of the associated firms. given the majority and minority shareholders conflict, it cannot be ruled out that the later could discount the stock prices coupled with other tax avoidance cost. however, due to lesser dependence on the capital market and lesser reputational impact in case of penalty imposed, associated family-controlled firms are expected to gain more benefits from ta actions in countries like pakistan with relatively poor or inefficient law enforcement. literature reveals that controlling shareholders with the control of the group can shift or share resources within a group, which benefits those firms operating in developing economies (khanna & yafeh, 2005; ullah & shah, 2015). moreover, the existing literature documented other benefits associated with business groups such as sharing financial resources, financing needs of newly journal of applied economics and business studies, volume. 5, issue 1 (2021) 1-22 https://doi.org/10.34260/jaebs.511 113 established firms, ta within group firms (almeida & wolfenzon, 2006; and ryngaert & thomas, 2007). therefore, it is hypothesized: h2: associated ownership has positive association with ta in group firms. extant literature document the relationship between ta and family ownership (chen et al., 2010), private equity ownership (badertscher, shroff & white, 2012) and dual-class ownership (mcguire et al., 2014). however, there is no study of managerial ownership and ta in the case of group firms. desai and dharmpala (2006) suggested that manager execute complex ta transactions that extract rents from shareholders. thus, the tax saved through ta is used by the opportunistic managers for their private benefits. kim, quinn, and wilson (2016) and rego and wilson (2012) found that an increase in managerial ownership increases ta. the possible collusion between majority shareholders and managers also provide the opportunity of sharing the saved tax pie. hence, it is hypothesised that: h3: managerial ownership and ta exhibit positive relationship in group firms. 2.3 tunneling, tax aggressiveness, and firm value theoretically, rpts can have either positive or negative effect on the wealth of minority shareholders and value of firms (gordon, henry, louwers & reed, 2007). on one hand, the “conflict of interest hypothesis” argues that rpts negatively affect firms’ performance and value due to the existence of conflicts of interests between principals and agents or principals and principals. on the other hand, the “efficient transactions hypothesis” argues that due to inherent expected efficiency, rpts could improve performance and value of group firms. wang (2011) investigated the role of corporate transparency in ta and value of firm and concluded that tax sheltering dilute value of firm in opaque firms and positive influence in case of informational transparent firms. the resulting lack of transparency due to the complex structure and complex nature of transactions suggests that that there will be negative association between ta and value of firm in case of group firms. empirical studies about the effect of ta on the value of firm have mixed results. for example, gupta and newberry (1997) found that firm profitability, liquidity and leverage are positively associated with ta. in the support of a positive effect of tax sheltering activities and firm value, wilson (2009) reported that market returns were higher for the firms involved in ta practices. others, such as arif and hashim (2013), moradi, mohammadi, and saeedi (2015) also found that ta enhances firm value. however, cloyd, mills, and weaver (2003) found no significant association between ta and firm value in the case of us firms. in contrast, hanlon and slemrod (2009) found that the market responds negatively to the firms involved in aggressive ta. similarly, desai and dharmapala (2009) found a significant and negative effect of ta on firm value. their findings are in line with the agency perspective that information asymmetry could favor controlling shareholders to engage in ta activities therefore, hamid ullah, hamid ali shah and sajjad ahmad khan 114 tax saved from tax sheltering activities may not increase shareholders wealth due to the expropriation by the major shareholders or opportunistic managers at the expense of minority shareholders. in this study, it is hypothesized that: h4: ta is inversely associated with the value of firm in group firms. 2.4 corporate governance, tax aggressiveness, and value of firm shleifer and vishny (1997) describe cg as “the system of control mechanisms, through which the suppliers of finance to corporations assure themselves of getting a return on their investment.” thus, cg is value maximizing framework that preserves interests of all stakeholders including minority shareholders and mitigates agency costs. desai and dharmapala (2009) opined that the influence of ta is value enhancing (destroying) for well-governed (poor governed) firms. similar results are reported by arif and hashim (2013) and yee et al., (2018). however, the later further added that tax saved is expropriated by the controlling shareholder and the weak cg couldn’t benefit the minority shareholders in malaysian firms. moreover, chan, mo, and tang (2016) suggested that the value relevance of ta reduces for the firms having high association between ta and tunnelling. in pakistani business groups with family dominance and expropriations by the controlling shareholders (ullah & shah, 2015), the ta is expected to be used for tunnelling. tax saved will be beneficial for the controlling shareholders but of no value to the minority shareholders. hence, ta shall negatively affect value of pakistani firms. the mitigating effect of cg here is an empirical issue. it might partially or completely control the negative impact of ta on value of firm. contrary to this view, “efficient transactions hypothesis” will predict that rpts, not meant to expropriate resource, might increase the performance of pakistani group firms with family dominance. thus, tax saved through ta might bear shareholders’ wealth maximizing effect. following the “conflict of interest hypothesis”, it is hypothesized that the relationship between ta and value of firm is moderated by the level of cg factors in firms belonging to business groups. the current study considered the following three main pillars of cg. 2.4.1 board independence, ta and value of firm an effective board of directors composed of independent (neutral) members with ample experience and repute would effectively safeguard the interests of all stakeholders (khaoula & moez, 2019). due to their effective monitoring, the boards in group firms are expected to control expropriation and rules violation-oriented ta activities and shall positively influence value of firms (adams & ferreira, 2007; and masulis & mobbs, 2011). on the other hand, it can be argued that independent members who are supposed to be more concerned about their repute shall disallow tax planning practices by the firms or controlling shareholders having political powers might negatively influence the independent role of the board members. hence, an indirect effect of the board independence on the ta and value of firm is also possible. we hypothesize that: journal of applied economics and business studies, volume. 5, issue 1 (2021) 1-22 https://doi.org/10.34260/jaebs.511 115 h5a: board independence positively influences the association between ta and value of firm in group firms. 2.4.2 independent audit committee, ta and value of firm expert, independent member directors in the audit committee, having incentives, will increase monitoring effectiveness and hence quality of the financial information of these firms by sustaining any influence of the firms’ management (beasley, carcello, hermanson & neal, 2010). in fact, the auditors are expected to have the expertise to identify the complex tax avoidance practices in the perplex management structure of group firms. a us based study by dhaliwal, naiker, and navissi (2010) reports that accounting frauds are controlled by fully independent audit committee. under the premise that ta is a tunneling activity by the controlling shareholders, an independent audit committee will ensure that the saved tax is not expropriated at the expense of the minority shareholders. therefore, an independent audit committee will positively influence the association between ta and value of firm. however, auditors also provide advices on managing tax planning risk (mcguire, omer & wang, 2012). thus, in the presence of expert and independent audit committee, shareholders might perceive ta as value enhancing practice rather than an opaque action (see dhaliwal et al., 2010). here, we hypothesize that: h5b: independent audit committee will have mitigating effect on the value destroying impact of at in group firms. 2.4.3 eexternal audit from big4 experience, expertise, repute, presumed greater independence and accompanied incentives are some of the reasons due to which big audit firms are expected to offer better monitoring of the accounting information of firms. audit by big firms shall reduce the chances of presumed controlling shareholders driven ta. moreover, it shall positively influence the ta and value of firm association because the quality of the financial information in the presence of big audit firms could be highly valued by the investors. in addition, the advising role on tax planning also suggests a positive impact of ta on value of firm. we hypothesize that: h5c: in group firms, audit by big4 is expected to positively affect the association of ta and value of firm. 3. methodology 3.1 data collection and sample size this study used a sample of 160 firms belonging to different business groups and listed in pakistan stock exchange for a period from 2009 to 2018. data of ta measures, cg, ownership, and rpts and other variables is extracted from the companies’ published annual reports available on their respective websites. the sample size is primarily determined by the availability of data on the variables of interest for the sample period. hamid ullah, hamid ali shah and sajjad ahmad khan 116 3.2 regression models and estimation the study uses panel method to estimate the models used in the study of yee et al., (2018). the study used 2 stage least square (2sls) regressions wherein the 1st stage equation the dependent variable is detr and the independent variables include cg factors and firm-specific variables. it is least expected that all firms have same level of ta e.g., intangible assets intensive firms may engage in more ta, hence endogeneity cannot be completely ignored. the 2sls is expected to control the issue. two instrumental variables, inst (intangible assets scaled by total assets) and atr (applicable tax rate) are added to eq -2. 2-stages least square regression models 2-stages least square regression models with interaction terms 3.3.1 measures of tunneling this study examines that controlling shareholders of business groups use rpts to tunnel the saved tax. therefore, tunneling is measured through related party receivables and related party loans both scaled by total asset (rpr) (jiang, lee & yue, 2010; ullah and shah, 2015). 3.3.2 measures of tax aggressiveness the study follows amiram, bauer, and frank (2013) and inger (2014) to compute ta as the difference of effective tax rate (etr) and applicable tax rate (atr). etr is current income tax expense to current pre-tax income. the other two measures of ta that are used for robustness check are detr2 and detr3; where detr2 is provisions of income tax scaled by pre-tax income (gupta & newberry, 1997; hanlon & heitzman, 2010), and detr3 is income tax divided by cash flows from operations (annuar, salihu & obid, 2013). journal of applied economics and business studies, volume. 5, issue 1 (2021) 1-22 https://doi.org/10.34260/jaebs.511 117 3.3.3 measure of firm value value of firm is measured through the frequently used tobin's q (wang, 2011; and inger, 2014). 3.3.4 measures of corporate governance and ownership structure cg elements include board independence, independent audit committee and external audit from big41. board independence is determined through the numbers of outside directors divided by the total directors (tandean & winnie, 2016). the independence of the audit committee is the ratio of non-executive members to total members of the audit committee (tandean & winnie, 2016). it is expected that an increase in the numbers of outside member will increase the independence of board and audit committee. the higher independence of these two bodies could lower the likelihood of diverting the saved tax from ta activities and are expected to positively affect value of the firm. an external audit from big4 is valued as 1 if the firm auditor is from big4 otherwise it is coded as 0 (ullah & shah, 2015). the engagement of big4 would signal that value enhancing tax sheltering decisions would be made and diversion of the saved tax by the controlling shareholders would be refrained. managerial ownership is the percentage of shares held by the directors and their family members out of total shares and associated firm ownership is the shares held by other firms or affiliated firms (abdullah et al., 2011). 3.3.5 control variables following the literature, this study includes various control variables to account for their effects on the ta and firm value. firm size and growth of firm is computed as log of total assets and changes in fixed assets respectively. leverage is represented by debts to equity ratio and roa is computed as the ratio of net income to total assets (sari, utama & rossieta, 2017; and yee et al., 2018). 4. results and discussions this section includes discussions on the results of descriptive analysis, correlation, and panel regression analysis. 4.1 descriptive statistics on the average firms have 18.9% related party receivables out of the total assets, 7.6% difference of receivables to payables scaled by total assets, and 4% related party loans out of total assets. these statistics show that a significant amount of assets are transacted through rpts by the group firms. the detr(s) represents the difference of the tax amount that group firms were required to pay and what actually these firms have paid. the positive average value of detr(s) suggests that the group firms have practiced ta. the average value of the detr1 is 0.138, detr2 is 0.183 and detr3 is 0.204. the ownership structure variables show that on average managerial ownership is 1 the big four accounting firms are pricewaterhousecoopers (pwc), deloitte and touche, kpmg and ernst & young. hamid ullah, hamid ali shah and sajjad ahmad khan 118 25.5% and associated ownership 34.5% out of total shares of firms. on the average 40% of the group firms take services from big4 audit firms. about 66% of the directors are non-executive directors where audit committee includes two non-executive directors. further, on the average firm assets show 5.7% growth, the average size of firms is 15, whereas, the debt to equity ratio has an average value of 0.65 which suggests that 0.65 out of 1 rupee is used to finance assets. the average roa is 6.9% and the average value of tobin’s ratio is 1.791 which shows that group firms have high market value relative to their book value. 4.2 pearson correlation matrix table 4.2 reports no case of high correlation among the variables and rules out the issue of multicollinearity. the measures of ta (detr1, detr2, detr3) and rpts (rpr, rploan) exhibit positive correlations between them. this association of rpts with ta suggests that the former is used to save tax. furthermore, managerial and associated ownership are positively related to the different measures of rpts and ta. as such it can be inferred that managerial ownership in pakistani group firms might be a mean to collude with the controlling shareholders. as expected, the three cg variables are inversely linked to both rpts and ta. similarly, growths, roa and debts to equity exhibit an inverse relationship with the measures of rpts and ta. finally, rpts is inversely related to tobin's q but ta show evidence of a positive correlation with the value of firm. 4.4 tax aggressiveness, tunneling and firm value table 4.3 shows results of the 2sls models with ta and tobin’s q as the explained variables. the results of the 1st stage model of ta show a positive effect of the rpra on ta. an increase in rpra motivates the group firms to aggressively engage in tax planning (chan et al., 2016; jiang et al., 2010). this indirectly implies that the saved tax could be diverted by the controlling shareholder to other group firms through rpts. in fact, the positive association between rpts and ta is indicative of the incidents of the income diversion from a firm with probably a high tax rate in the group to a firm with probably a low tax rate and hence the findings are in line with jung et al., (2009) who reported lower pre-tax return on assets for group firms with more tax burden. as expected, managerial and associated ownership has direct significant effect on ta (ryngaert & thomas, 2007). thus, firms with relatively more managerial and associated ownership are most likely to engage in ta activities (kim et al., 2016). desai and dharmapala (2006) suggested that ta facilitates managerial rent extraction and could strengthen its opportunistic behavior. the cg variables such as board independence, independence of audit committee and audit from big4 show statistically significant inverse relationship with ta in all the estimated models. these findings are similar to those documented by jiang et al., (2010) and li et al., (2016). thus, increase in the outside members in board and audit committee would improve governance and journal of applied economics and business studies, volume. 5, issue 1 (2021) 1-22 https://doi.org/10.34260/jaebs.511 119 reduce the aggressive tax opportunism that would risk the minority shareholders wealth. among the control variables growth, leverage and roa have a negative influence on the ta activities, whereas, size of firm shows a direct and significant influence on ta activities in the sample firms. thus, it is argued that to preserve reputation and image growing firms and firms with higher level of leverage would have lower level of ta practices (jiang et al. 2010). however, relatively large firms are more engaged in aggressive tax sheltering due to their large volume and sophisticated transactions. the results of the tobin's q models show that all the three measures of ta i.e. detr1, detr2, and detr3 have an inverse and significant relationship with the value of firm. as expected, the results imply that ta is conflicting in nature in the pakistani group firms and perceived as value reducing activity by investors. these results supports the view that ta increase after-tax cash flows, however its effect on firm value is diluted, particularly in group firms, due to the possible expropriation of saved tax by the controlling shareholders (desai & dharmapala, 2006, 2009). likewise, perceived as a mean of expropriation by the investors, rpra shows an indirect influence on the tobin's q (gao & kling, 2012; ullah & shah, 2015). in line with the entrenchment hypothesis, both, managerial and associated ownership are inversely related to the tobin's q. therefore, following the results of ullah and shah (2015), it is argued that increase in associated ownership or managerial ownership enhance controlling power in other group firms and the controlling shareholder, in self-interest, thus tunnel out resources through rpts that eventually decrease the firm value (gao & kling, 2012; and ullah & shah, 2015). consisting with the prior studies, firm growth, leverage and applicable tax rate have negative effect on tobin’s q, and firm size and intangible assets have positive and significant effect on tobin’s q (chan et al., 2016; yee et al., 2018). the moderating effect of the cg variables between the relationship of ta and value of firm is determined through interaction term of the governance and ta variables.2 as reported in table 4.4, the interaction of cg variables with ta has a positive and significant effect on the firm value. these findings demonstrate that in the presence of more independent members on the board, independence of audit committee, and big4 auditor, the effect of the ta on the firm value becomes positive. even in the case of group firms, these measures of cg constrain expropriation of the saved tax through tax sheltering activities, improve the overall governance, and hence improve the firm value (chan et al., 2016; jiang et al., 2010; li et al., 2016). these results are similar to the findings documented by arif and hashim (2013), yee et al., (2018), and desai and dharmapala (2009). they reported that case of well-governed firm, there is a significant positive influence of ta on firm value; the results of all other variables remain consistent with the results of the baseline model. from these findings we conclude that 2 the study also used interactions of corporate governance variables with deter in related party loan models and found consiste nt results with the baseline interaction models. the other variables results also remained the same but not discussed here. hamid ullah, hamid ali shah and sajjad ahmad khan 120 presence of good cg mechanisms is perceived by investors to reduce the probability of expropriations and shifting resources from a group firm to another by the controlling shareholders. 4.5 robustness check to check that these findings are robust and remain consistent for another type of rpts; this study also considered related party loan scaled by total assets and examines its effect on ta and firm value (rploan). table 4.5 presents the results of 2sls models. the only changed proxy is rploan in the model. the results of rploan are not different from rpra, the earlier proxy of tunneling. more specifically, an increase in rploan increases ta in the pakistani group firms because controlling shareholders might pressurize managers to aggressively engage in the tax sheltering activities (jiang et al. 2010; sari, utama & rossieta, 2017). chan et al., (2016) also reported that rploan is used to divert saved tax and further showed its detrimental effect for the minority shareholders. frank et al., (2009) suggested that information risk from ta deepens the uncertainty of firms which confers on the controlling shareholders greater flexibility in expropriating resources. therefore, following the results of sari et al., (2017) we also concluded that increase in the related party loan would increase the demand for aggressive tax sheltering activities in group firms in pakistan and this prediction is remained consistent with the baseline model. similar to the results of the base line regression, the ownership variables, managerial and associated ownership, are directly and statistically significantly influencing the various measures of ta; whereas, cg variables, board independence, independent audit committee, and audit from big4 exhibit the same negative and significant effect on ta. the results of other control variables also exhibit no inconsistency such that firm growth, leverage, and roa have negative influence and size of firm shows a direct effect on ta. 5. conclusion to the extent aggressive tax planning could reduce tax costs, the shareholders would prefer it. however, this argument overlooks the prospective non-tax costs associated with ta, say for example, agency related costs. firms owned or controlled by business group or family are exposed to conflict of interests between majority and minority shareholders. this study analyzed data of 160 pakistani group firms through 2sls estimation to find impact of rpts on ta and the later impact on the value of firm. in addition, the moderating role of measures of cg on the relationship between rpts and ta is analyzed. it is found that rpts, associated firm ownership and managerial ownership have direct influence on the ta, whereas, board and audit committee independence and audit from big4 have negative effect on the ta. moreover, the ta and rpts show indirect impact on the value of firm while controlling for ownership and cg variables. the interaction models show that cg variables positively moderate the journal of applied economics and business studies, volume. 5, issue 1 (2021) 1-22 https://doi.org/10.34260/jaebs.511 121 ta and firm value relationship. we conjecture that good cg practices are evenly important in the case of group firms. in these firms, as perceived by investors, the indirect effect of ta on the value of firm is significantly mitigated by employing good cg practices. references abdullah, f., shah, a., gohar, r., & iqbal, a. m. (2011). the effect of group and family ownership on firm performance: empirical evidence from pakistan. international review of business research papers, 7(4), 191-208. adams, r., & ferreira, d. (2009). women in the boardroom and their impact on governance and performance. journal of financial economics, 94(2), 291-309. amiram, d., bauer, a. m., & frank, m. m. (2013). manager-shareholder alignment, shareholder dividend tax policy, and corporate tax avoidance. ssrn electronic journal. doi:10.2139/ssrn. 2111467 annuar, h. a., salihu, i. a., & obid, s. n. (2014). corporate ownership, governance and tax avoidance: an interactive effects. procedia social and behavioral sciences 164: 150-160. doi:10.1016/j.sbspro.2014.11.063 badertscher, b., shroff, n., & white, h. d. (2012). externalities of public firms’ disclosures: evidence from private firms’ investment decisions. ssrn electronic journal. doi:10.2139/ssrn.2037287 beasley, m. s., carcello, j. v., hermanson, d. r., & neal, t. l. (2010). the audit committee oversight process. contemporary accounting research, 26(1), 65122. chan, k. h., mo, p. l., & tang, t. (2016). tax avoidance and tunneling: empirical analysis from an agency perspective. journal of international accounting research 153: 49-66. doi:10.2308/jiar-51345 chen, s., chen, x., shevlin, t., chen, s., chen, x., & shevlin, t. (2010). institutional knowledge at singapore management university are family firms more tax aggressive than non-family firms ? are family firms more tax aggressive than non-family firms ? * university of texas at austin university of wisconsinmadison. 41–61. chen, c., hepfer, b. f., quinn, p. j., & wilson, r. j.(2018). the effect of taxmotivated income shifting on information asymmetry. review of accounting studies 233: 958-04. doi:10.1007/s11142-018-9439-1 chen, s., chen, x., cheng, q., & shevlin, t. (2010). are family firms more tax aggressive than non-family firms? journal of financial economics 951: 41-61. doi:10.1016/j.jfineco.2009.02.003 cheung, y., rau, p., & stouraitis, a. (2006). tunneling, propping, and expropriation: evidence from connected party transactions in hong kong. journal of financial economics, 822: 343-86. doi:10.1016/j.jfineco.2004.08.012 hamid ullah, hamid ali shah and sajjad ahmad khan 122 choi, j. j., sami, h., & zhou, h. (2010). the impacts of state ownership on information asymmetry: evidence from an emerging market. china journal of accounting research 3: 13-50. doi:10.1016/s1755-30911360018-0 cloyd, c. b., mills, l. f., & weaver, c. d. (2003). firm valuation effects of the expatriation of u.s. corporations to tax‐haven countries. the journal of the american taxation association 25: 8709. doi:10.2308/jata.2003.25.s-1.87 desai, m. a., & dharmapala, d. (2006). corporate tax avoidance and high-powered incentives. journal of financial economics 791: 145-79. doi:10.1016/j.jfineco.2005.02.002 desai, m. a., & dharmapala, d. (2009). corporate tax avoidance and firm value. review of economics and statistics 913: 537-46. doi:10.1162/rest.91.3.537 desai, m., & dharmapala, d. (2004). corporate tax avoidance and high powered incentives. doi:10.3386/w10471 dhaliwal, d., naiker, v., & navissi, f. (2010). the association between accruals quality and the characteristics of accounting experts & mix of expertise on audit committees. contemporary accounting research, 27(3), 787-827. djankov, s., ganser, t., mcliesh, c., ramalho, r., & shleifer, a. (2008). the effect of corporate taxes on investment and entrepreneurship. doi:10.3386/w13756 doi:10.1016/j.jaccpubpol.2010.09.003 dyreng, s. d., hanlon, m., & maydew, e. l. (2010). the effects of executives on corporate tax avoidance. the accounting review 854: 1163-89. doi:10.2308/accr.2010.85.4.1163 fama, e. f., & jensen, m. c. (1983). separation of ownership and control. the journal of law and economics 262: 301-25. doi:10.1086/467037 fan, j. p., & wong, t. j. (2005). do external auditors perform a corporate governance role in emerging markets? evidence from east asia. journal of accounting research, 43(1), 35-72. frank, m. m., lynch, l. j., & rego, s. o. (2009). tax reporting aggressiveness and its relation to aggressive financial reporting. the accounting review 842: 467-96. doi:10.2308/accr.2009.84.2.467 gao, l., & kling, g. (2012). journal of international accounting, auditing and taxation. journal of international accounting, auditing and taxation, 21, 1731. gao, l., & kling, g. (2008). corporate governance and tunneling: empirical evidence from china. pacific-basin finance journal 165: 591-05. doi:10.1016/j.pacfin.2007.09.001 gordon, e. a., henry, e., louwers, t. j., & reed, b. j. (2007). auditing related party transactions: a literature overview and research synthesis. accounting horizons, 21: 81-102. doi:10.2308/acch.2007.21.1.81 journal of applied economics and business studies, volume. 5, issue 1 (2021) 1-22 https://doi.org/10.34260/jaebs.511 123 gupta, s., & newberry, k. (1997). determinants of the variability in corporate effective tax rates: evidence from longitudinal data. journal of accounting and public policy 161: 1-34. doi:10.1016/s0278-42549600055-5 hanlon, m., & heitzman, s. (2010). a review of tax research. journal of accounting and economics 3: 127-78. doi:10.1016/j.jacceco.2010.09.002 inger, k. k. (2014). relative valuation of alternative methods of tax avoidance. the journal of the american taxation association 361: 27-55. doi:10.2308/atax50606 jensen, m. c., & meckling, w. h. (1976). theory of the firm: managerial behavior, agency costs, and ownership structure. economic analysis of the law, 16276. doi:10.1002/9780470752135.ch17 jiang, g., lee, c. m., & yue, h. (2010). tunneling through intercorporate loans: the china experience☆. journal of financial economics 981: 1-20. doi:10.1016/j.jfineco.2010.05.002 jung, k., kim, b., & kim, b. (2009). tax motivated income shifting and korean business groups chaebol. journal of business finance & accounting 6: 55286. doi:10.1111/j.1468-5957.2009.02141.x khaoula, f., & moez, d. (2019). the moderating effect of the board of directors on firm value and tax planning: evidence from european listed firms. borsa istanbul review, 19(4), 331-343. kim, j. b., li, y., & zhang, l. (2011). corporate tax avoidance and stock price crash risk: firm-level analysis. journal of financial economics 1003: 639–62. https://doi.org/10.1016/j.jfineco.2010.07.007 kim, j., quinn, p., & wilson, r. (2016). managerial ownership and tax avoidance. working paper. klassen, k., lang, m., & wolfson, m. (1993). geographic income shifting by multinational corporations in response to tax rate changes. journal of accounting research 31: 141-82. doi:10.2307/2491168 la porta, r., lopez‐de‐silanes, f., shleifer, a., & vishny, r. (1998). law and finance. journal of political economy, 1066: 1113-55. doi:10.1086/250042 landry, s., deslandes, m., & fortin, a. (2013). tax aggressiveness, corporate social responsibility, and ownership structure. ssrn electronic journal. doi:10.2139/ssrn.2304653 masulis, r. w., & mobbs, s. (2011). are all inside directors the same? evidence from the external directorship market. journal of finance, 66, 823-872. mcguire, s. t., omer, t. c., & wang, d. (2012). tax avoidance: does tax specific industry expertise make a difference? accounting review, 87(3), 975-1003. nekhili, m., & cherif, m. (2011). related parties transactions and firm's market value: the french case. review of accounting and finance, 103: 291-15. doi:10.1108/14757701111155806 hamid ullah, hamid ali shah and sajjad ahmad khan 124 ryngaert, m. d., & thomas, s. e. (2007). related party transactions: their origins and wealth effects. ssrn electronic journal. doi:10.2139/ssrn.970689 sari, d. k., utama, s., & rossieta, h. (2017). tax avoidance, related party transactions, corporate governance and the corporate cash dividend policy. journal of indonesian economy and business 323: 190-14. doi:10.22146/jieb.28658 scholes, m. s., wolfson, m. a., erickson, m. m., hanlon, m. l., maydew, e. l., & shevlin, t. j.. (2015). taxes & business strategy, global edition. new york, ny: pearson higher ed. tandean, v. a., & winnie, w. (2016). good corporate governance dan ukuran perusahaan pengaruhnya pada tax avoidance. jurnal ilmiah akuntansi dan bisnis. doi:10.24843/jiab.2016.v11.i01.p07 ullah, h., & shah, a. (2015). related party transactions and corporate governance mechanisms: evidence from firms listed on the karachi stock exchange. pakistan business review, 17(3), 663-680. yee, c. s., sapiei, n. s., & abdullah, m. (2018). tax avoidance, corporate governance and firm value in the digital era. muhammadiyah university yogyakarta. journal of applied economics and business studies, volume. 5, issue 1 (2021) 1-22 https://doi.org/10.34260/jaebs.511 125 table 4.1 descriptive statistics variable obs mean std.dev min max rpra 1494 0.189 0.063 0.000 0.685 rploan 1494 0.040 0.026 0.000 0.364 detr1 1494 0.138 0.200 -0.499 0.898 detr2 1494 0.183 0.155 -0.420 0.636 detr3 1494 0.204 0.138 -0.246 0.636 mos 1494 0.255 0.208 0.000 0.775 assoctosp 1494 0.345 0.288 0.000 0.903 big4 1494 0.406 0.510 0.000 1.000 nxd 1494 0.666 0.294 0.300 0.800 iac 1494 2.106 0.269 1.000 4.000 fg 1494 0.057 0.155 0.000 0.832 size 1494 15.889 1.529 12.065 20.132 roa 1494 0.069 0.093 -0.205 0.460 de 1494 0.652 0.204 0.031 0.994 related party lending’s measured through related party receivables scaled by total asset and related party loan scaled by total assets. (dter1) is the difference of effective tax rate (etr) and applicable tax rate (atr), whereas, effective tax rate is the ratio of current income tax expense to current pre-tax income. detr2 is the effective tax rate is computed as provisions of income tax divided by pre-tax income, detr3 is the effective tax rate is computed as the ratio of income tax to operating cash flows. tobin's q is computed as the market value of equity plus book value of liabilities divided by total assets. board independence (ned) is computed as the ratio of outside directors to total directors. the independence of the audit committee (iac) is computed by the ratio of non-executive members to total members of the audit committee. an external audit from big4 is coded as 1 if the firm auditor is from big4 otherwise it is coded as 0. firm size is measured by the log of total assets, firm growth is measured by changes in fixed assets, de represents leverage which is measured by the ratio of debts to equity and roa is computed as net income divided by total assets. hamid ullah, hamid ali shah and sajjad ahmad khan 126 table 4.2 correlation variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 (1) rpra 1 (2) rploan 0.11 1 (3) detr1 0.24 0.11 1 (4) detr2 0.23 0.18 0.54 1 (5) detr3 0.23 0.13 0.42 0.51 1 (6) mos 0.09 0.09 0.15 0.2 0.18 1 (7) assoct 0.15 0.12 0.15 0.14 0.13 -0.5 1 (8) big4 -0.16 -0.13 -0.14 -0.16 -0.14 -0.07 0.02 1 (9) nxd -0.17 -0.15 -0.12 -0.19 -0.17 -0.09 -0.04 0 1 (10) iac -0.08 -0.11 -0.11 -0.18 -0.15 -0.18 0.1 0.03 0.02 1 (11) fg -0.01 -0.02 -0.16 -0.25 -0.27 -0.19 0.02 0.31 0.16 0.06 1 (12) size 0.31 -0.11 0.03 0.03 0.02 -0.24 0.08 0.14 0.06 0.22 -0.16 1 (13) roa -0.01 -0.04 -0.24 -0.31 -0.39 -0.07 0.02 0.17 -0.06 0.09 0.01 0.07 1 (14) de -0.11 -0.03 -0.12 -0.22 -0.27 0.05 -0.04 -0.03 -0.14 -0.01 -0.12 -0.02 -0.16 1 table 4.2 shows correlation results. journal of applied economics and business studies, volume. 5, issue 1 (2021) 1-22 https://doi.org/10.34260/jaebs.511 127 table 4.3 2sls regression results of detr and tobin's q (detr1) (tobin’s q) (detr2) (tobin’s q) (detr3) (tobin’s q) variables model 1 model 2 model 3 model 4 model 5 model 6 detr1 0.8483*** -0.339 detr2 0.6559*** -0.216 detr3 0.3591*** -0.191 rpra 0.462*** -0.455** 0.256** -1.077*** 0.0958 -0.396** -0.124 -0.18 -0.0904 -0.224 -0.0731 -0.177 mos 0.160*** 0.422*** 0.119*** 0.805*** 0.0734** 0.288*** -0.0586 -0.0772 -0.0414 -0.12 -0.0307 -0.0656 assoctosp 0.0906** -0.194*** 0.0714** -0.118** 0.0631** -0.0902* -0.0461 -0.0566 -0.0347 -0.0529 -0.0299 -0.052 big4 -0.0498** 0.0927*** -0.0442*** 0.0544** -0.00629 0.0459** -0.0224 -0.0205 -0.0164 -0.0229 -0.013 -0.0197 nxd -0.0728** 0.0679** -0.0411** 0.0653** -0.0522** 0.0625* -0.0369 -0.0383 -0.0271 -0.0318 -0.0238 -0.0375 iac -0.0976*** 0.0199 -0.0238 0.132*** -0.00844 0.00648 -0.0376 -0.0379 -0.0289 -0.0426 -0.0251 -0.0378 fg -0.206*** -0.282** -0.200*** -1.078*** -0.222*** -0.567*** -0.0609 -0.113 -0.0476 -0.198 -0.057 -0.14 size 0.0210*** 0.0700*** 0.0183*** 0.138*** 0.0126** 0.0630*** -0.00743 -0.00975 -0.0061 -0.0173 -0.00555 -0.00915 de -0.0863** -0.397*** -0.0994** -0.0782** -0.0748** -0.195** -0.0374 -0.0596 -0.0468 -0.0368 -0.036 -0.0827 ina 0.1251*** 0.1251*** 0.1251*** -0.047 -0.047 -0.047 art -0.5590*** -0.5590*** -0.559*** -0.1906 -0.1906 -0.1906 roa -0.634*** -0.24 -0.439*** -0.158 -0.169 -0.119 constant -0.198 1.562** -0.119 1.275* -0.00623 2.031*** -0.125 -0.68 -0.1 -0.682 -0.0911 -0.682 fixed-industry yes yes yes yes yes yes fixed-year yes yes yes yes yes yes observations 1464 1464 1464 1464 1464 1464 r-squared 0.192 0.468 0.195 0.468 0.178 0.468 table 4.3 shows regression results, variables definitions are given in the table 4.1. robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 hamid ullah, hamid ali shah and sajjad ahmad khan 128 table 4.4 detr and tobin’s q with interactive effect of corporate governance variables (detr1) (tobins_q) (detr2) (tobins_q) (detr3) (tobins_q) model 1 model 2 model 3 model 4 model 5 model 6 edetr 0.0512*** -0.0786*** -0.0356*** 0.0036 0.0075 0.005 detr1_big4 0.0855*** 0.0235 detr1_iac 0.0560*** 0.027 detr1_ned 0.0422*** 0.0102 detr2_big4 0.0708*** 0.012 detr2_iac 0.0574*** 0.0145 detr2_ned 0.0484** 0.0166 detr3_big4 0.0363** 0.0143 detr3_iac 0.0382** 0.0156 detr3_ned 0.0414** 0.0205 arrp 0.3961*** -0.170*** 0.1981*** -0.227*** 0.2001*** -0.00714 -0.0398 -0.0487 -0.0326 -0.0519 -0.0278 -0.0488 mos 0.137** -0.391*** 0.0925** -0.776*** 0.0661** -0.269*** -0.0593 -0.0747 -0.0421 -0.115 -0.0319 -0.0668 assoctosp 0.6631*** -0.206*** 0.2122*** -0.207*** 0.0208 -0.106** -0.0454 -0.0578 -0.0336 -0.0563 -0.0302 -0.051 big4 -0.137** 0.0434** -0.0973*** 0.0895*** -0.0485** -0.131*** -0.0422 -0.0218 -0.0161 -0.0313 -0.0128 -0.0393 ned -0.0516*** 0.0427** -0.0496** 0.0825*** -0.0118** -0.0794** -0.0171 -0.0187 -0.0218 -0.0221 -0.0234 -0.0337 iac -0.0314** 0.08121** -0.0445** -0.229*** -0.0363** -0.0699** -0.0184 -0.0387 -0.0213 -0.0534 -0.0157 -0.031 fg -0.212*** -0.301** -0.2115*** -1.428*** -0.227*** -0.569*** -0.0613 -0.118 -0.0466 -0.242 -0.0584 -0.14 size 0.0222*** 0.0683*** 0.0210*** 0.180*** 0.0134*** 0.0558*** -0.00744 -0.0106 -0.00569 -0.0229 -0.00509 -0.00966 de -0.0499** -0.460*** -0.023 -0.153* -0.0719** -0.186** -0.0278 -0.0561 -0.0465 -0.0799 -0.0353 -0.0778 ina 0.114** 0.0831** 0.0864* -0.048 -0.0412 -0.051 art -5.576*** -5.434*** -4.763** -2.068 -2.098 -2.148 roa -0.594*** -0.193 -0.430*** -0.16 -0.17 -0.122 constant -0.0773 1.789** 0.0197 2.084*** 0.0566 1.910** -0.122 -0.738 -0.101 -0.753 -0.0902 -0.776 fixed-industry yes yes yes yes yes yes fixed-year yes yes yes yes yes yes observations 1378 1378 1378 1378 1378 1378 r-squared 0.102 0.384 0.129 0.393 0.182 0.397 the definitions of the other variables remain the same as the baseline model, moreover, the interaction of journal of applied economics and business studies, volume. 5, issue 1 (2021) 1-22 https://doi.org/10.34260/jaebs.511 129 various measures of tax aggressiveness represented by detr1, detr2 and detr3 are interacted with the independence audit committee (iac), board independence (ned) and audit from big4 (big4). robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 table 4.5 2sls regression results of detr and tobin's q variables detr1 tobins_q detr2 tobins_q detr3 tobins_q model 1 model 2 model 3 model 4 model 5 model 6 detr1 0.5458*** -0.1336 detr2 0.5844*** -0.1799 detr3 0.3514*** -0.48 rploan 0.428** -0.648** -0.337 -1.634*** -0.0692 0.0889 -0.256 -0.261 -0.226 -0.336 -0.313 -0.247 mos 0.169*** 0.448*** 0.118*** 0.725*** 0.0757** 0.299*** -0.0584 -0.0799 -0.0413 -0.111 -0.0303 -0.0665 assoctosp 0.0961** -0.209*** 0.0997*** -0.127** 0.0954*** -0.101** -0.0457 -0.0569 -0.0346 -0.0524 -0.0297 -0.0514 big4 -0.0776*** 0.0641*** -0.0955*** 0.0757*** -0.0420** 0.0426** -0.0223 -0.0206 -0.0162 -0.0211 -0.0199 -0.0199 nxd -0.0761** 0.0692* 0.0593** 0.0742** -0.0626** 0.0758** -0.0372 -0.0395 -0.0271 -0.0341 -0.0299 -0.0382 iac -0.0359** 0.0743** -0.0229 0.109*** -0.0766*** 0.0815** -0.0176 -0.0378 -0.0289 -0.0412 -0.0252 -0.0377 fg -0.205*** -0.270** -0.202*** -0.944*** -0.223*** -0.548*** -0.0611 -0.112 -0.0474 -0.182 -0.0568 -0.137 size 0.0190** 0.0654*** 0.0141** 0.101*** 0.0107** 0.0562*** -0.00736 -0.00933 -0.00578 -0.013 -0.00519 -0.0086 de -0.0316** -0.409*** 0.0905** -0.0945 0.0605** -0.154* -0.0175 -0.0587 -0.0464 -0.0869 -0.0302 -0.0806 ina 0.121*** 0.121*** 0.121*** -0.0464 -0.0464 -0.0464 art -0.434*** -0.5434*** -0.5434*** -0.1897 -0.1897 -0.1897 roa -0.646*** -0.272 -0.452*** -0.161 -0.172 -0.12 constant -0.173 1.555** -0.0515 1.680** 0.022 2.058*** -0.129 -0.682 -0.0988 -0.682 -0.087 -0.684 fixedindustry yes yes yes yes yes yes fixed-year yes yes yes yes yes yes observations 1164 1164 1164 1164 1164 1164 r-squared 0.19 0.469 0.105 0.469 0.177 0.469 table 4.5 shows regression results, variables definitions are given in the table 4.1. robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 hamid ullah, hamid ali shah and sajjad ahmad khan 130 table 4.6 2sls regression results of detr, tobin's q and variables (detr1) (tobinsq) (detr2) (tobinsq) (detr3) (tobinsq) model 1 model 2 model 3 model 4 model 5 model 6 edetr -0.0411*** -0.0516*** -0.0252*** 0.0046 0.0045 0.007 detr1_big4 0.0915*** 0.0135 detr1_iac 0.6401* 0.41 detr1_ned 0.0626*** 0.01 detr2_big4 0.0923*** 0.013 detr2_iac 0.8147** 0.415 detr2_ned 0.0584*** 0.0132 detr3_big4 0.0763*** 0.0241 detr3_iac 0.6829* 0.416 detr3_ned 0.0514** 0.0219 arrp 0.3261** -0.7492** -0.7345** -0.6344*** -0.6912** 0.8893** 0.156 0.361 0.351 0.216 0.213 0.447 mos 0.1991*** 0.5482*** 0.2181*** 0.7315*** 0.0857** 0.1798*** 0.0484 0.0799 0.0313 0.211 0.0403 0.0665 assoctosp 0.0991*** -0.2894*** 0.0817*** -0.1255** 0.0854*** -0.1115** 0.0357 0.0537 0.0246 0.0123 0.0195 0.0414 big4 -0.0776** 0.0641** -0.0955** 0.0757*** -0.0420** 0.0421** 0.0353 0.0306 0.0462 0.0211 0.0199 0.0189 ned -0.0851** 0.0692** 0.0593** 0.0732*** -0.0826*** 0.0718*** 0.0312 0.0315 0.0271 0.0241 0.0299 0.0282 iac -0.0419** 0.0743** -0.0229 0.0109 -0.0766*** 0.0815 0.0176 0.0378 0.0289 0.412 0.0252 0.0677 fg -0.2053*** -0.2701** -0.2024*** -0.9443*** -0.2235*** -0.5486*** 0.0611 0.112 0.0474 0.182 0.0568 0.137 size 0.0190** 0.0654*** 0.0141** 0.101*** 0.0107** 0.0562*** 0.00736 0.00933 0.00578 0.013 0.00519 0.0086 de -0.2167 -0.409*** 0.0805* -0.0745 0.0705* -0.254** 0.275 0.0587 0.0564 0.0769 0.0412 0.0806 ina 0.1213*** 0.1213*** 0.1211*** 0.0464 0.0464 0.0464 art -0.1241*** -0.1441*** -0.1434*** 0.01897 0.01897 0.01897 roa -0.646 -0.272 -0.452 0.561 0.172 0.32 constant -0.1731 0.5551 -0.0515 0.6801** 0.022 0.0581 0.129 0.682 0.0988 0.682 0.087 0.684 fixed-industry yes yes yes yes yes yes fixed-year yes yes yes yes yes yes observations 1145 1145 1145 1145 1145 1145 r-squared 0.19 0.469 0.105 0.469 0.177 0.469 table 4.6 shows regression results, variables definitions are given in the table 4.1. robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 journal of applied economics and business studies, volume 4, issue 1 (2020) 77-96 https://doi.org/10.34260/jaebs.414 77 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online): 2663-693x cohort risk sharing in pakistan: relative wage and consumption movements over time syed hassan raza1* 1 school of social sciences and humanities, national university of science and technology, pakistan abstract this study analyzes to which extent the distribution of consumption is affected by the relative wage movement among birth cohorts and education groups. our empirical design is based on a synthetic panel constructed using repeated cross-sectional data from “household integrated economic survey of pakistan.” we limit our analysis to persons aged between 26 to 50 years at the time of survey. to see the evolution of change in income and consumption we measured growth by taking 6, 8and 10-years’ difference respectively. the findings ascertained there is limited risk-sharing across cohorteducation groups in pakistan, but the measured extent of risk-sharing increases over longer horizons. furthermore, we observe relatively higher consumption smoothing among the less educated people over the period of ten years. in the university education group, results reveal less consumption smoothing in the shorter, sixand eight-year time periods. the study concludes that the relative risk-sharing over a decade is better in pakistan than the shorter growth horizon. keywords insurance, risk, consumption smoothing jel classification o12, e21 1. introduction consumption insurance hypothesis has been a long-heated debate in the economic literature over the last two decades (attanasio and davis, 1996; attanasio and székely, 2004; zheng and lolips, 2018). the question arises whether an income shocks penetrate consumption or whether consumption is well insured against income shocks. theorists investigated this phenomenon from simpler to complex dimension, however, little evidence is available for developing countries due to non-availability of longitudinal data sets. we encounter similar issues while working on consumption smoothing and choosing ideal dataset for pakistan. therefore, as a solution to non-availability of longitudinal data we used synthetic panel formulated through repeated cross section data. this paper aims to investigate the consumption smoothing in response to relative wage movement over time. we intend to provide estimates concerning consumption and income risk sharing over different education groups and cohorts. * corresponding author: dr.hassanraza@s3h.nust.edu.pk syed hassan raza 78 during the last decade, pakistan has experienced massive economic transformation coupled with several important shocks to its economy. such shocks primarily consist of huge energy crisis, terrorism, earth quakes, wheat shortages and massive flooding. some of those shocks affected between group inequality and both consumption and income were affected. although the effects of such shocks have been long heated debate in the empirical literature, but to the best of our knowledge in recent years this is first attempt to capture the evolution of inequality in pakistan. this study incorporates the repeated cross-sectional data to investigate the evolution of consumption across the age cohorts and education groups, and relates it to the changes in the distribution of wage. in other words, we measured the effect of growth in hour wage with respect to cohorts and education groups on the distribution of household consumption. basically, we build synthetic panel which is justified in the absence of longitudinal data. since it is a cohort analysis, it is even justified on theoretical ground by blundell and preston (1998) as they argue consumption could be compared for individuals at the same point in their life cycle. previous research pertaining to risk sharing in consumption, discussed several formal and informal insurance mechanisms which protects consumption from income shocks. cochrane 1991; mace 1991; altonji, hayashi, and kotlokoff 1992; townsend 1994; kazianga and udry 2006; gertler, paul, david, levine and moretti 2009 are the main contributors to this part of literature. they have exploited the consumption theory related with allocation that under reasonable restriction of preferences, brings sturdy implications on consumption growth such as role of microfinance or asset as buffer stocks in smoothing the consumption over time. earlier research pertaining to consumption smoothing relies on publicly unobserved income shocks or the focus mainly relies on the consumption behavior of individuals or their earning capacity. however, attanasio and davis (1996) examined the power of systematic, publicly evident swings in per hour wage structure on the distribution related with household consumption. this research design, focusing on relative wage structure has quite a number of advantages in comparison to the traditional mechanism where the research is more centered on individual shocks. for instance, one of the major benefits of this study arises because relative wage movements across the group of workers are least likely to be correlated with idiosyncratic component of individual preference shift. changes in household preferences may affect the household marginal utility, thus, it may lead to false rejection of consumption insurance hypothesis. moreover, inability to notice publicly observable income shocks limits the scope of old studies, as their main focus remains on unobservable income shocks. consequently, such theories cannot be rationalized which stresses on the role of unseen income shocks in an informationally constrained journal of applied economics and business studies, volume 4, issue 1 (2020) 77-96 https://doi.org/10.34260/jaebs.414 79 optimal allocation of consumption. the mechanism of this study relies on publicly observable income shocks as attanasio and davis (1996) in their paper focused on publicly observable income shocks and how they affect the distribution of consumption as it enables us to see to which extent consumption is insured. as discussed earlier one of the key objectives of this study is to document the growth of relative wages across the birth cohorts and education groups. but unlike attanasio and davis (1996) where college graduates were uniformly better-off during 1980 to 1990. in pakistan the picture is slightly different, in fact it is more mixed. in particular, we observe for some years and cohorts, better educated people are priced similar to the people with less education. later on, we establish the nexus by comparing the changes in relative income to changes in relative consumption. this part is interpreted as test of consumption insurance hypothesis. in this paper section 2 presents detailed literature review. there we outline the evolution of literature pertaining to risk and consumption smoothing. in section 3, we provide the detailed mechanism through which we have constructed the synthetic panel and have run the regression for the desired residuals. whereas, section 4 of this paper outlines the data section in which we discuss the key properties of our dataset. descriptive analysis has been used to get better insights about the data. this helped us to see average trend and overall variation in the data across and within the cross sections. section 5 is reserved for results and discussion in which we presented the results with the help of graphs. finally, in section 6, the conclusion and future recommendations are given. 2. literature review there is an escalating amount of literature which documented the relative wage movement and consumption smoothing through various perspectives. major contributions are accredited to developed countries due to better tracking and availability of datasets. however, in empirical research it remains underdeveloped area for developing economies. cochrane (1991) was the first who tested whether the consumers are well insured compared to idiosyncratic shocks to income and wealth. considering the role of both formal and informal financial institutions and transfers, he concluded full insurance can be rejected for long illness or involuntary job loss but not for the spells of unemployment. townsend (1994) further implemented the full insurance model on three indian villages. one of the major findings of this paper was the co-movement of household consumption with village average consumption. he proved household consumption was not influenced by factors like idiosyncratic shocks, unemployment or illness, however, people with no land holdings were not well insured. syed hassan raza 80 attanasio and davis (1996) implemented the test of cochrane (1991) on synthetic panel. but they exploit the basic assumption maintained by cochrane (1991) and townsend (1994) which basically states the ability of social planner to easily transfer or move leisure to the households. they alter the need for this assumption by examining the relative wage movements amongst a cluster of males with inelastic labor supply. this enables them to develop a better consumption insurance hypothesis test when preferences are non-separable among consumption and leisure along with imperfect transfer of leisure across the households. given their empirical design, short frequency movements were detected in the cohort and education structure of pre-tax hourly wages amongst men troop large and substantial changes in the distribution of household consumption. attanasio and székely (2004) investigate the high volatility period in mexico using a similar research question. they presented evidence of relationship between relative male wage and growth in consumption. they added four different types of expenditure categories and checked how they respond to income shocks. in some years, highly educated people earned less money in comparison to less educated people in mexico. they attributed this to a peso crisis and decline in gdp per capita. blundell et al. (2008) analyzed the extent to which consumption is insured against income shocks. they combined two different panel datasets to observe the transmission of inequality from income to consumption. study concludes that the people in the us offset transitory shocks by greater insurance arrangements, whereas the degree of insurance varies across demographic groups. krueger and perri (2006) used the consumer expenditure survey and document despite of noteworthy increase in both between and within group of household income inequality in united stated, there is moderate increase in consumption inequality. krueger and perri (2006) document the noteworthy increase in both, between and within group income inequality for households having similar characteristics, for instance education and race. but they found consumption and income inequality were not coupled together as there was a moderate increase in consumption inequality. they held within group inequality, responsible for this dichotomy or divergence as inequality in income increased more than inequality in consumption. heathcote, perri and violante (2010) used four different surveys to document cross sectional inequality in united states between 1967 to 2006. they measured the different dimensions of inequality through market, choices and institutions. findings reveal reasonable increase in income inequality before 1982, whereas, taxes and transfers compress the inequality especially at the bottom of income distribution. study concludes with the evidence that changes in distribution of hours worked generates income inequality. zheng and lolips (2018) investigated the transmission of income risk and consumption in context of growing economy like china. they utilized a longitudinal panel for household income and consumption data. they concluded that consumption insurance declines as the economy progresses towards higher attainment of growth. hence journal of applied economics and business studies, volume 4, issue 1 (2020) 77-96 https://doi.org/10.34260/jaebs.414 81 the literature suggests important implications of income and consumption risk with multiple dimensions. our study brings novel insights about pakistan’s economy in context of publicly observable income and consumption shocks since we build a synthetic panel. furthermore, this study helps us to analyze the role of education in response to income shocks and consumption readjustments. 3. methodology to estimate whether the variation in consumption is affected by income shocks, we need longitudinal data. but since we have repeated cross sectional data, we can use an alternative mechanism which is through synthetic panel. we build our synthetic panel from repeated cross-sectional data after creating age cohorts and education groups. these age cohorts and education groups become our unique identifier and creates a synthetic panel over the period of time. the advantages of using a synthetic panel are coupled with one disadvantage as one focuses only on insurance across the groups. by taking averages over the members of groups, one cannot say anything about the extent of risk sharing within the group. this is the method developed and used by attanasio and davis (1996) and here are details of economic environment of his model along with equations in order to see if income shocks contribute to consumption variation. as per consumption insurance hypothesis, both direct and indirect mechanism for sharing the consumption risk matches the growth in consumption related marginal utility among various individuals and group of individuals. if we have a fixed set of pareto weights, it is important to note this sort of condition can be obtained through central planner’s first order condition. as it is clear, across individuals and time, the central planner assigns resources under uncertainty. let’s assume ξt represents the langrage multiplier related with the aggregate or overall feasibility constraint at time t. now we present the planner’s first order condition. (ѳ j) t λj uc (ct j, δt j) = ξt, j = 1,…., j, (1) now considering all the states in the world t, cj represents the consumption of concerned individual j, δj implies “arbitrary preference shocks,” ѳj is representing a “discount factor,” and uc (., .) signifies the marginal utility function. individual fixed effects are equivalent to λj which are time invariant pareto weights and can easily be eliminated through the ratios of first order conditions at two points in time: ѳ𝑗 𝑈𝑐(𝐶𝑡+1 𝑗 ,𝛿𝑡+1 𝑗 ) 𝑈𝑐(𝐶𝑡 𝑗 ,𝛿 𝑡 𝑗 ) = 𝜉𝑡+1 𝜉𝑡 , 𝑗 = 1, … , 𝐽. (2) syed hassan raza 82 as per equation (2) any variable which is in term of cross sections, not correlated with the variation in preferences and the measurement error related with consumption growth is also not correlated to cross-sectional distribution of consumption related growth. in measuring consumption, we assume iso-elastic utility through multiplicative shocks and multiplicative error which implies log linear form for equation (2). specially, let uc (ct j, δt j) = uc (ct j, btjγt j) = btj(ct j)1+γj/1+γj, and let εt+1 represents the “error in log consumption change,” thus, equation 2 becomes 𝑙𝑜𝑔 ( 𝐶𝑡+1 𝑗 𝐶 𝑡+1 𝑗 ) = 1 𝛾𝑗 [𝑙𝑜𝑔 ( 𝜉𝑡+1 𝜉𝑡 ) − 𝑙𝑜𝑔 ( 𝑏𝑡+1 𝑗 𝑏 𝑡 𝑗 ) − 𝑙𝑜𝑔(ѳ 𝑗 )] + 𝜀𝑡+1 𝑗 (3) therefore, we see now, if a variable xjt+1 is independent of variation in preferences and measurement error, in this case consumption insurance suggests x has nothing in context of explanatory power related to cross-sectional distribution of growth in consumption. about the shocks related with bequest or the individual’s endowment and earning size becomes an ideal candidate for x under interesting substitutions to consumption insurance null hypothesis. next, households are portioned into groups and indexed by i. by taking “logs in equation” (1) and averaging over the sample of group “i” households at time “t,” we get ô ≡ ∑ 𝑙𝑜𝑔[𝑈𝑐(𝐶𝑡 𝑗 ,𝛿𝑡 𝑗 )𝑗∈𝑖(𝑡) #𝑖(𝑡) (4) = 𝑙𝑜𝑔𝜉𝑡 − 𝑡 ∑ 𝑙𝑜𝑔 ѳ𝑗𝑗∈𝑖(𝑡) #𝑖(𝑡) − ∑ 𝑙𝑜𝑔 𝜆𝑗𝑗∈𝑖(𝑡) #𝑖(𝑡) here “#i(t)” signifies the “number of households” who were sampled at particular time “t”. now we can rewrite the first order condition of sample average as ô = 𝑙𝑜𝑔𝜉𝑡 − 𝑡ṝ𝑖 − ῑ𝑖 + 𝜀𝑖𝑡 , 𝑖 = 1, … , 𝐼, 𝑡 = 1, … , 𝑇, (5) in equation (5), ṝ and ῑ are the population counterparts to the sample in equation (4) and the reason we use error term in the equation is due to finite sampling from heterogeneous population. moreover, it leads to a “level” regression on the synthetic panel created from group average data, given the parameterization of preferences. to be specific, if the form of preferences becomes iso-elastic as specified above, the sample mean ô converts to ô = ∑ (𝑏𝑡 𝑗 𝑗∈𝑖(𝑡) ) #𝑖(𝑡) + ∑ 𝛾𝑗 𝑙𝑜𝑔(𝐶𝑡 𝑗 )𝑗∈𝑖(𝑡) #𝑖(𝑡) ≡ ḡ𝑖𝑡 ȗ + ∑ 𝛾𝑗 𝑙𝑜𝑔(𝐶𝑡 𝑗 )𝑗∈𝑖(𝑡) #𝑖(𝑡) + 𝜐𝑖𝑡 journal of applied economics and business studies, volume 4, issue 1 (2020) 77-96 https://doi.org/10.34260/jaebs.414 83 combining this mean with equation (5) produces ∑ 𝛾𝑗𝑙𝑜𝑔(𝐶𝑡 𝑗 )𝑗∈𝑖(𝑡) #𝑖(𝑡) = 𝑙𝑜𝑔 𝜉𝑡 − 𝑡ṝ𝑖 − ῑ𝑖 − ḡ𝑖𝑡 + 𝜐𝑖𝑡, (6) 𝑖 = 1, … , 𝐼, 𝑡 = 1, … , 𝑇. now we are able to present the regression of the sample mean of log consumption on the full set of time and group fixed effects, along with a variable “xit” which captures time variation in relative group endowment. 𝑙𝑜𝑔�̂�𝑖𝑡 = 𝛼𝑡 + ɡ𝑖 + 𝛽𝑋𝑖𝑡 + 𝑒𝑖𝑡 , 𝑖 = 1, … , 𝐼, 𝑡 = 1, … , 𝑇. (7) consumption insurance implies 𝛽=0, if we compare equation (6) with (7), if and only if “xit” satisfies the set of auxiliary assumptions. for instance, xit should be not be correlated with the error in the log consumption, group differences related to the mean of time discount rate, group differences related to distribution of preference parameter γj and variation in the mean preferences disturbances ḡit. similarly, it is easy to find that consumption insurance also suggests 𝛽=0 in first difference regression specifications of the form 𝑙𝑜𝑔�̂�𝑖𝑡 − 𝑙𝑜𝑔�̂�𝑖𝑠 = 𝛼𝑡 + 𝛽(𝑋𝑖𝑡 −𝑋𝑖𝑠) + 𝑒𝑖𝑡 −, 𝑒𝑖𝑠, 𝑖 = 1, … , 𝐼, 𝑡 = (𝑡 − 𝑠), … , 𝑇, (8) for equation (8) it is also necessary to satisfy the statistical assumption discussed above for xit. 4. data section the household integrated income survey of pakistan (hies) has been conducted with some breaks since 1963. the survey was renamed as “pakistan social and living standard measurement (pslm) survey” in 2004, but the basic module of hies remained intact. since 2004, the survey is conducted in every alternate year. it’s a repeated cross-sectional survey which covers detailed information on consumption, education, income and employment. individuals were asked about income and education status, however, consumption level data is obtained at the household level. households were asked simple questions pertaining to durable and non-durable consumption and we choose the reported total consumption on non-durable items. the universe of this survey consists on all the rural-urban areas of pakistan except capital territory islamabad and military restricted areas. every city or town has been divided into enumeration blocks consisting of 200 to 250 households identifiable through sketch map. every sketch map has been classified into three classes of income groups: low, middle and high, keeping in mind the living standard of the majority of the people. a twostage stratified sample design has been adopted in this survey. villages and enumeration blocks in rural or urban areas are chosen as the primary sampling units (psus). sample syed hassan raza 84 psus have been selected from strata/sub-strata with probability proportional to size (pps) method of sampling. households within the psus have taken as secondary sampling units ssus. in every psu of rural or urban areas 12 to 16 households are selected by using a systematic sampling technique with a random start. we restrict our sample to the male heads of the households having age between 26 to 50 years. moreover, we exclude those heads who are currently enrolled in any educational institution at any level. it is well-established economic fact that head of the household usually earns more and is usually considered as the keeper of the house (blumberg, 1988; charles and james, 2005; treas and ruijter, 2008). therefore, it is more important to observe the shocks in the income of the head, but for type of design problem appears in consumption as we don’t have consumption of the head of the household. consequently, as a solution to this problem we used oecd scales to generate adult equivalent average consumption. this enabled us to study the shocks in the income of the head, and its impact on adult equivalent average consumption per household. we categorize the data in five distinct age cohorts and three levels of educational categories. the youngest cohort in the data are those born in 1984-1988 and oldest cohort are those born in 1954-1958. the lowest education group is “junior middle” followed by “intermediate” and highest education group is “university.” since we are more focused on labor earnings, and it is difficult to monitor agriculture income, we limit ourselves to nonagricultural labor income. lastly, we narrow the focus to those heads who worked for at least ten months in a year with more than 20 days per month with paid employment. age cohorts along with education groups become the unique identifier in the data which enables us to formulate the synthetic panel from repeated cross sections. in data, we assume working day implies working eight hour a day, this assumption allows us to compute the hourly wage for heads of the households. possibility of outliers in the data have been removed by dropping the top and bottom 1% of the observations which makes the data ready to study relative wage movement and consumption distribution. except in table 1.1, all the analysis is based on real values for which we have deflated the income and consumption by using the values of the gdp deflator obtained from world bank indicators. table 1.1 summary statistics min max p10 p50 p90 mean consumption 23.6663 67.4972 25.2636 34.2498 53.8943 38.0360 wage 0.1649 0.70819 0.1973 0.3137 0.5522 0.34850 birth year 1954 1988 1958 1960 1970 1970 education 1 3 1 2 3 2 source: authors calculations journal of applied economics and business studies, volume 4, issue 1 (2020) 77-96 https://doi.org/10.34260/jaebs.414 85 in table 1.1, we present the summary statistics of the dataset. consumption and wages are measured in terms of pak rupees. we have reported minimum, maximum, mean, 10th, 50th and 90th percentile in the data. the distribution of our key variables shows that the difference between minimum and maximum value of monthly non-durable consumption and hour wage is more than double. moreover, at the 90th percentile, most people have a university education, so they earn higher hourly wage a and consume more. table 1.2 average real hourly wage for head of household by cohort and education group five years birth cohort year 1984-1988 1979-1983 1974-1978 1969-1973 1964-1968 1959-1963 1954-1958 junior middle 2004 ----0.2444 0.2768 0.2808 0.31196 0.2999 2006 --0.1789 0.1934 0.2122 0.2296 0.2390 0.2442 2008 --0.1531 0.1703 0.2256 0.2143 0.1822 0.1790 2010 0.1751 0.2015 0.2218 0.2423 0.2310 0.2336 -- 2012 0.1845 0.1953 0.2250 0.2414 0.2449 0.2401 -- 2014 0.2054 0.2102 0.2450 0.2492 0.2640 ---- intermediate 2004 ----0.3048 0.3471 0.3822 0.4246 0.4213 2006 --0.1885 0.2403 0.2736 0.3201 0.3572 0.3472 2008 --0.2206 0.2376 0.2611 0.3334 0.3103 0.3218 2010 0.1883 0.2711 0.3241 0.3291 0.3603 0.3587 -- 2012 02119 0.2534 0.2947 0.3292 0.3798 0.3659 -- 2014 0.2378 0.2682 0.3156 0.3674 0.3946 ---- university 2004 ----0.4412 0.4829 0.5497 0.6787 0.7081 2006 --0.2941 0.4041 0.5014 0.5224 0.6479 0.7130 2008 --0.2987 0.3817 0.4126 0.3957 0.4999 0.4388 2010 0.1061 0.4962 0.6173 0.6095 0.6320 0.7632 -- 2012 0.3502 0.4399 0.4998 0.5626 0.6351 0.7189 -- 2014 0.3910 0.4679 0.549 0.5374 0.7031 ---- source: authors calculations furthermore, we observe that the youngest birth cohort in the data is 1988 and the oldest one is 1954, whereas, the median man was born in 1970. in education, the median man obtained intermediate education as we have classified education into three categories such as junior middle, intermediate and university education. in table 1.2, we document the average pre-tax real wage on a yearly basis because it helps us to observe the evolution of average hourly wage on yearly basis by cohort and education group. it is evident in general that, on average, people with university education or old cohorts earns more relative to other categories. however, we observe in some of the years younger cohorts earn more than the older ones. for instance, in year 2012, within the intermediate education group, for cohort 1964-1968 average hourly wage was 0.3798 pkr, syed hassan raza 86 whereas, for cohort 1959-1963 average hourly wage was 0.3659 pkr. but this sort of situation is limited to a couple of cohorts in the less educated group, thus, we see men with university education earn more over the period of time. this was not the case in attanasio and davis (1996), but attanasio & székely (2004) encountered a similar trend in their data. in addition, we observe an overall decline in wage for year 2006 across the education groups. however, from 2008 onwards we see a noteworthy increase in wages. it is evident from table 1.2 that people with university education are relatively better off. intuitively the economic slow-down in pakistan which started in year 2006 could be one of the reasons for the sudden decline in real wages as it began with an energy crisis and food inflation. we see average real monthly consumption of non-durable items in table 1.3 crossed with cohorts and education groups. non-durable items in our data mainly consist of food, textile and miscellaneous expenditures. in food expenditures we have items like staple crops, eatables, pulses, oil, tea, and other bakery items, in addition to this, textile expenditures consist of expenses on clothing, embroidery and sewing. lastly, expenses on religious or festive occasion and house rent fall into the miscellaneous category. as we can see for cohort 1974-1978, people with junior middle education, on an average basis spent 30.7638 pkr in 2004, which fell to 23.8473 pkr in 2006. during the same period in table 1, there is a decline in real wage as well. in all the cohorts with junior middle education there is a sudden decline in real consumption for year 2012. there is an overall decline in consumption for year 2006, just like we see in table 1.1 for real wages. there is an upward trend in consumption for later years especially for people with intermediate and university education. furthermore, this table reveals that older cohorts consumed more in comparison to the younger cohorts, and higher educated men consumed more than lower educated men. table 1.3 average real consumption for head of household by cohort and education group five years birth cohort year 1984-1988 1979-1983 1974-1978 1969-1973 1964-1968 1959-1963 1954-1958 junior middle 2004 ----30.7638 30.928 28.8463 30.5317 29.9717 2006 --23.6663 23.8473 23.9125 24.7859 25.0600 25.0108 2008 --26.7830 24.5410 29.0237 30.9255 24.3115 28.8305 2010 25.9362 28.2912 29.6621 29.5170 31.9105 32.7330 -- 2012 24.5443 23.9703 24.9468 25.1834 27.1800 27.5366 -- 2014 26.0949 26.3300 27.6141 27.7845 30.8667 ---- intermediate 2004 ----35.9491 38.9695 39.1120 41.7763 39.0028 2006 --27.557 29.1672 30.0009 31.7300 35.9817 36.3239 2008 --32.7482 34.5104 35.5961 36.5188 45.2793 38.3048 2010 30.9797 35.5648 40.5093 39.7500 39.2518 42.8678 -- 2012 30.0662 30.6529 32.3272 32.7494 35.9109 33.2192 -- 2014 31.9842 33.4639 35.3683 37.3314 39.9968 ---- journal of applied economics and business studies, volume 4, issue 1 (2020) 77-96 https://doi.org/10.34260/jaebs.414 87 university 2004 ----49.5925 49.9373 53.7658 60.04717 60.0334 2006 --30.0324 45.6930 48.5953 47.3523 55.0256 57.2592 2008 --45.2432 45.9092 51.3861 49.2820 62.9737 56.5886 2010 44.6402 53.8513 69.6926 66.6436 63.0003 78.3772 -- 2012 34.7618 47.6504 51.8958 49.8831 49.9002 53.7790 -- 2014 48.5662 51.7343 50.7679 49.8437 57.5022 ---- source: authors calculations this implies education and age are key determinants of consumption here in this dataset, which is a well-established economic result. before we proceed to main results it is important to understand how average wage and consumption evolve over time as shown in figure 1.1. the lower panel shows increasing pattern of average real hourly wage from 2008 to 2012, which can be further break down in such a way, for instance, from 2008 to 2010 it increased at increasing rate, but from 2010 to 2012 it increased at decreasing rate. production of staple food increased remarkably in 2006-07 with surplus supplies to export, but in 2008 a wheat crisis in afghanistan affected pakistan resulting in spiked food inflation in “regional food prices” (world bank, 2009). figure 1.1 trend of average real consumption and real wage figure 1.2 presents the result of variance in real consumption and real wages. it reveals a sharp decline in the variance of hour wage in year 2006 but moderate increase from 2008 till 2010. as discussed before, this was the period of great uncertainty in pakistan due to economic and political crises. thus, we notice similar fluctuations in consumption and wage, but there is greater variation in wage in comparison to consumption. table 1.1 syed hassan raza 88 and table 1.2 showed a sudden decline in both real wage and consumption for year 2006, hence, we see the similar trend here in this figure. in last two years variance of wage declined sharply but there is not much difference in the variance of consumption. figure 1.2 trend of variance in real consumption and hour wage source: authors calculations 5. results and discussion this paper examines whether the shocks to income drill through the consumption or not. evidence from all the growth specifications except ten years growth shows consumption is susceptible to income shocks and we observe that a smaller variation in income results in a larger variation in consumption. this finding implies that, in pakistan relative risk sharing is better over longer time horizon than a shorter one. zheng and lolips (2018) also argued that over a longer time horizon, risk sharing in developing countries is better in comparison to developed countries. therefore, we conclude people in pakistan usually share the income risk and smooth consumption over a longer time horizon which could be argued to be a consequence of the existence of strong social insurance and family bonding. one of the biggest advantages of our empirical design is that it enabled us to capture the evolution of the slope of the growth in different growth specifications for 6, 8 and 10 years. for instance, in the majority of our six-year growth specifications we observed that slope varies around 1, but when we increased the time period from six to ten years, we see the slope declines to 0.254 which is in accordance with economic theory. attanasio and davis (1996) concluded failure of consumption insurance in the long run, but we conclude that people tend to smooth more over the longer period of time; hence, our empirical findings substantially differ in the context of pakistan. in figure 1.3 we visualize the change in log disposable income and change in log consumption over the period of six years in the similar way as it was visualized by attanasio and davis (1996). we observe equal variation journal of applied economics and business studies, volume 4, issue 1 (2020) 77-96 https://doi.org/10.34260/jaebs.414 89 in income and consumption for people with university or junior middle education. but for men with intermediate education there is not much variation in income for six years growth horizon. figure 1.3 change in income and consumption by education groups and cohorts “adult equivalent household consumption” vs. “man’s wage (head of the household), 2004-2010 log change residuals.” groups are defined by three different types of education crossed with 5 years birth cohorts. plotted values are residuals from regression on cubic age. income variation significantly passes through to consumption as we see in the slope and standard error, therefore, there is not much consumption smoothing in this figure. in terms of age cohorts, we see the youngest cohort in the right-hand panel of figure 1.3, is 36 to 40 years of age. we notice closer movement of change in income and change in consumption for people with 41-45 and 36 to 40 years of age. therefore, people belonging to these particular to age groups tend to smooth less. it is important to understand as lagakos et al., (2018) argued that earning peaks are usually realized in the later part of one’s career. since these age groups are right in the middle of their working life and they share a lot of family responsibilities including financial burden, they are more susceptible with limited income and higher consumption risk. syed hassan raza 90 figure 1.4 change in income and consumption by education groups and cohorts “adult equivalent household consumption” vs. “man’s wage (head of the household), 2006-2012 log change residuals.” groups are defined by three different types of education crossed with 5 years birth cohorts. plotted values are residuals from regression on cubic age. in figure 1.4 we observe the similar slope but majority of the observations are clustered together. however, for one of the observations pertaining to university education we see higher growth in consumption than income. there is a negative growth in income for people with junior middle education but there is not equal variation in income. thus, there is a greater consumption smoothing observed for this education group during the mentioned time period. similarly, in the right-hand panel of this figure, cohorts are clustered together but we observe similar variations in in income and consumption. journal of applied economics and business studies, volume 4, issue 1 (2020) 77-96 https://doi.org/10.34260/jaebs.414 91 figure 1.5 change in income and consumption by education groups and cohorts “adult equivalent household consumption” vs. “man’s wage (head of the household), 2008-2014 log change residuals.” groups are defined by three different types of education crossed with 5 years birth cohorts. plotted values are residuals from regression on cubic age. in figure 1.5 we document the last plot of six years difference for change in income and consumption with respect to education group age-cohorts. it suggests higher growth in income and consumption remains low for least educated people during this period, however, university and intermediate education groups realized significant fluctuations in income and consumption growth. there is not much change in consumption even in the right-hand panel where we present the growth with regard to cohorts. slope of the line is much lower in contrast to figure 1.3 and figure 1.4, as we have seen in the figure 1.1 and figure 1.2, there are a lot of variation in 2008 due to economic crisis, and here we see much variation at wage side instead of consumption. during this period somehow consumption is better insured against income shocks as we see much lower slope but still it is significant. syed hassan raza 92 figure 1.6 change in income and consumption by education groups and cohorts “adult equivalent household consumption” vs. “man’s wage (head of the household), 2004-2012 log change residuals.” groups are defined by three different types of education crossed with 5 years birth cohorts. plotted values are residuals from regression on cubic age. in figure 1.6 we see 8 years growth horizon for change in consumption and change in change in income from 2004 to 2012. we observe reasonable departure from consumption insurance hypothesis as slope is significant and in accordance with economic theory, we see there a decline in slope relative to majority of six years specifications. since we increase the time period so the slope comes down, thus people do more smoothing in longer time period. but we see change in consumption for people with university education is much smoother in comparison to the rest of observations. in terms of cohort it has been ascertained that people between of 36 to 40 years realize the greater change in consumption given the change in income. journal of applied economics and business studies, volume 4, issue 1 (2020) 77-96 https://doi.org/10.34260/jaebs.414 93 figure 1.7 change in income and consumption by education groups and cohorts “adult equivalent household consumption” vs. “man’s wage (head of the household), 2006-2014 log change residuals.” groups are defined by three different types of education crossed with 5 years birth cohorts. plotted values are residuals from regression on cubic age. figure 1.7 presents the last result of our eight years difference design which shows the slope is still significant but much higher relative to the previous figure. for some of the observations we see greater change in consumption is realized through the change in income for people with university and junior middle education. besides this, we see greater consumption smoothing for people with intermediate education. in the right-hand panel of this figure we see greater change in consumption than income for one of the observations of 31 to 35 years cohort. syed hassan raza 94 figure 1.8 change in income and consumption by education groups and cohorts “adult equivalent household consumption” vs. “man’s wage (head of the household), 2004-2014 log change residuals.” groups are defined by three different types of education crossed with 5 years birth cohorts. plotted values are residuals from regression on cubic age. figure 1.8 shows ten years difference plot from 2004 to 2014 and it shows consumption is more insured over a decade with respect to relative wage movements. unlike 6and 8years difference growth, here we see different trend as the slope is insignificant, implying the existence of insurance mechanism for consumption. it is evident here bigger variations in wage drawing very small variations in wage for instance we observe regardless of negative growth in real wage, people with junior middle education tends to maintain the consumption or there is a lesser decline in consumption growth than wage. similarly, in the right-hand panel we observe, for the cohort from 46 to 50 years of age, consumption does not move with wages, implying there is consumption smoothing. this contrast we see, highlights the benefit of our empirical strategy. by drawing on cross-sectional datasets to build a pseudo or long synthetic panel, we can detect the persistent apparatuses of relative wage movements causing relative consumption movements. journal of applied economics and business studies, volume 4, issue 1 (2020) 77-96 https://doi.org/10.34260/jaebs.414 95 6. conclusion we started researching on relative risk sharing in consumption issue as we observed pakistan’s economy experienced definite and persistent movements in the structure of relative wages since 2004. we outline that relative wage movements across birth cohorts and education groups bring noteworthy changes in distribution of adult equalized household consumption. we document the greater variation in relative wage in comparison to consumption for men with lower educational attainment. in six years difference specification we observe, for all the education groups there is a clear evidence that adult equalized real consumption moves with hourly real wage which supports the rejection of consumption insurance hypothesis. in addition to this, under eight years difference design, we observe, for some of the observations, higher growth in relative wages are for more educated people in comparison to people with junior middle education. for them, it is evident that consumption was relatively smoother which suggests the existence of insurance arrangements. for young cohorts, we see there is greater variation in income in comparison to older cohorts. overall results suggest that the university education group and junior middle education group are susceptible to transmission of relative wage shocks into adult equalized household consumption. our ten years difference design suggests the existence of consumption insurance, which means food security, clothing and housing are insured against income variations which differs from earlier research by attansio and davis (1996). but these findings are largely in line with the argument that in developing countries people tend smooth more over a longer time horizon (zheng and lolips, 2018). however, in lesser duration, our result documents the significant departure from consumption insurance hypothesis. since the magnitude of covariance between consumption growth and income growth is similar in 6 years and 8 years growth design. therefore, it is right to say both move together in short periods but over a longer duration, people tend to smooth their consumption. the authors do acknowledge the current limitations of this primary effort. we are unable to identify whether the consumption insurance in particular to longer time horizon exist due to some counter cyclical fiscal policies or due to some other social factors. in future, this research could be extended to check for the roles of fiscal policies along with other social factors in determining the consumption insurance in longer and shorter time horizons. furthermore, the research could be extended to check if these results hold for developing economies especially in south asia. it is ideal to see risk and insurance from different spectra as in the developing economies repeated cross section data is easily available than longitudinal panel due to higher tracking cost of individuals across different waves of the surveys. moreover, inclusion of endowment shocks could be more useful for syed hassan raza 96 getting better insights. another interesting expansion of this research could be to break down consumption into different bundles. in order to see which type of consumption is more affected from income shocks such as comparison of non-durable consumption bundles with durable ones. this will be important addition to this paper to observe if results vary among different consumption bundles. references altonji, j. g., hayashi, f., & kotlikoff, l. j. (1992). is the extended family altruistically linked? direct tests using micro data. the american economic review, 1177-1198. attanasio, o., & davis, s. j. (1996). relative wage movements and the distribution of consumption. journal of political economy, 104(6), 1227-1262. attanasio, o. p., & székely, m. (2004). wage shocks and consumption variability in mexico during the 1990s. journal of development economics, 73(1), 1-25. blumberg, r. l. (1988). income under female versus male control: hypotheses from a theory of gender stratification and data from the third world. journal of family issues, 9(1), 51-84. blundell, r., & preston, i. (1998). consumption inequality and income uncertainty. the quarterly journal of economics, 113(2), 603-640. blundell, r., pistaferri, l., & preston, i. (2008). consumption inequality and partial insurance. the american economic review, 98(5), 1887-1921. charles, n., & james, e. (2005). ‘he earns the bread and butter and i earn the cream’ job insecurity and the male breadwinner family in south wales. work, employment and society, 19(3), 481-502. cochrane, j. h. (1991). a simple test of consumption insurance. journal of political economy, 99(5), 957-976. gertler, p., levine, d. i., & moretti, e. (2009). do microfinance programs help families insure consumption against illness? health economics, 18(3), 257-273. mace, b. j. (1991). full insurance in the presence of aggregate uncertainty. journal of political economy, 99(5), 928-956. kazianga, h., & udry, c. (2006). consumption smoothing? livestock, insurance and drought in rural burkina faso. journal of development economics, 79(2), 413-446. santaeulalia-llopis, r., & zheng, y. (2018). the price of growth: consumption insurance in china 1989–2009. american economic journal: macroeconomics, 10(4), 1-35. townsend, r. m. (1994). risk and insurance in village india. econometrica: journal of the econometric society, 539-591. treas, j., & de ruijter, e. (2008). earnings and expenditures on household services in married and cohabiting unions. journal of marriage and family, 70(3), 796-805. krueger, d., & perri, f. (2006). does income inequality lead to consumption inequality? evidence and theory. the review of economic studies, 73(1), 163-193. heathcote, j., perri, f., & violante, g. l. (2010). unequal we stand: an empirical analysis of economic inequality in the united states, 1967–2006. review of economic dynamics, 13(1), 15-51. journal of applied economics and business studies, volume. 4, issue 2 (2020) 101-118 https://doi.org/10.34260/jaebs.425 101 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x mobilising social capital in healthcare industry of pakistan mahreen hussain1, mahwish zeeshan2& abid ghafoor chaudhry3 1senior assistant director, national electric power regulatory authority, islamabad 2lecturer, department of anthropology, pmas-auur 3 chairman/associate professor, department of anthropology, pmas-aaur abstract this study intended to learn the mobilization of social capital for the access to better health care facilities emphasizing the role of religiosity culture and social identity. social capital was limited to religiosity culture and the social identity using the bonding and bridging concepts for this study. the three variables gauged in the study were social capital, religiosity culture and social identity which are barely used for the health care industry. the target population was public health care practitioners of the federal capital with sample size of 215 doctors and nurses over a period of six months. the data was quantitative and analyzed through co-relational tests. questionnaire was developed for the study using the validity and reliability statistics. however, the results from the study reflected the significant impact of religiosity culture and social identity. thus, it was concluded that positive and negative externalities affect the social identity in the creation of social capital. the findings of this study can provide a framework for future reference and to the policy makers in enhancing the social responsibility through mobilization of social capital of healthcare professionals in the industry. keywords social capital; bridging; bonding; individualistic approach; collective approach; religiosity culture; social identity; jel classification i11, i10, n03, z12 1. introduction in healthcare industry, the interest of social capital is increasing at a very rapid rate(pitkin derose and varda, 2009).scientists analyzed the importance of social capital in different domains for different countries including pakistan(gupta et al., 2017). this study aims to learn the mobilization of social capital for the access to the better healthcare facilities. all the major details of the study including research problem, background, conceptual definitions, and aim of this study are discussed in this section. 1.1 research problem the inspiration for this research problem arises from the neo-tocquevillean philosophy mahreen hussain, mahwish zeeshan & abid ghafoor chaudhry 102 that performance of public life and social institutes are affected by the social norms and civic engagements (sander and putnam, 2010). therefore, social capital and religiosity culture with the dimensions of norms and values will affect the social responsibility of the healthcare practitioners. also the theory lacks certain dimensions of social capital when studied in relation with social responsibility i.e. structural and cognitive. so, there is a need to study the social capital with the social responsibility and religiosity culture and social identity for the performance of social institutes and improved public life. 1.2 background of the study in pakistan last decade brought reforms in the health sector and the consequences as well. these reforms addressed the ongoing process of devolution-decentralization since the annulment of 18th amendment. decentralization is allocating authority where the decision makers are and those who have the knowledge but the decision making comes with the responsibility. in the case of health sector, the responsibility is towards the society. according to undp (1997), decentralization is re-formatting and re-organizing authority with the intention of creating co-responsibility at multiple levels bestowing good governance. the responsibility does not only lie with the authority nonetheless with the health practitioners. the model that was adopted required the state to develop a monitoring and appraisal system for the provinces regarding the authority and responsibility. the fulfillment of the duty is not only the responsibility assigned but also the moral obligation of the health practitioners as well. the constitutional provision implies upholding social justice and minimizing inequities in the society. inequities exist in the health sector at provincial level and repercussions for the well-being of society and endorsing health facilities(nishtar et al., 2010). it was suggested that the health policy should have high level norms, values, principles a standard. members of the society are mobilizing their referrals and finding access points to better health care services. the social capital in this context is created by the social groups of similar religious and identity groups. the health sector of federal is selected for the study because of the ease of access to the information and for the reason that national health subjects and their responsibilities are delegated to the federal institution (nishtar et al., 2010). the situation provides the opportunity of studying responsibility of the health practitioners towards society and social actors. 1.3 conceptual definitions when a term is created it leads to the expansion of a concept however, the mushroom growth in the field of social capital in 1970’s, when researchers from diverse backgrounds borrowed concepts from their respective fields to place it below t he “umbrella-concept” of social capital; it was not sure “whether more was achieved or lost” (farr, 2004; hirsch and levin, 1999).it is based on the relationships quality build by an organization with various stakeholders (servaes and tamayo, 2017). putnam’s definition on social capital suggests that it is an association constituted voluntarily, or the so-called social networks, along with social or interpersonal journal of applied economics and business studies, volume. 4, issue 2 (2020) 101-118 https://doi.org/10.34260/jaebs.425 103 trust. his central argument revolved around the notion that citizens organizing into social networks and organizations on a voluntary basis develops and generates (generalized) trust among people(putnam, 2000). according to(portes and landolt, 1996)social capital resides in family units, communities and developing cities hence every positive thing in the social life was associated with the social capital. this uncertainty in the usage of the term still exists in the literature because of the evolution of concept from individual, collective level to national level. this barred the expansion of the concept on the individual level which was based on the network ties that an individual builds relations especially for the material gains and benefits (hagan et al., 1996; mclanahan and sandefur, 1994). it is believed that in absence of optimal investment in social capital by the firms, there is a lower tendency in investment returns pertaining to intellectual, physical and even human capital. it is defined that the social capital incorporates networks based on social bonds, their reciprocities and value gained out of these businesses (sen and cowley, 2013).(farooq et al., 2019), suggests that this commonality of religions serve as the basis of strengthening these relationships. also, the understanding of social capital is based on environmental, cultural and political ties (hao et al., 2018). identity-bridging social capital accentuates the debate of diversity as it focuses on the need for cooperation and mutual respect. this pooling in the values and lifestyles of the diversified groups require the cooperation and mutual respect (sander and putnam, 2010). identity is also defined through its function of mediating power through prestige, and can consist of economic, social or cultural capital which is a building block of the developed state. perceived identity also plays role in social responsibility i.e., socially responsible acts may create identity or are in themselves source of identity or can be symbolized in a way that increases and signals prestige. according to the actor if the action is aligned with his preferences then his perceptions of value and honor will be augmented (wuthnow, 2000). the perceived identity of an individual in a social unit is the source of social capital arising from his collective action on the part of the group or having a responsibility towards it. also, in this point of view, companies are increasing the extent of their corporate social responsibility for three major reasons: to achieve competitive advantage, to enhance the employees’ productivity, and to erase the stigmatized image (if any) in the minds of employees that causes hindrances to reach the job satisfaction at an optimal level (flammer and luo, 2017). the immature development of the concept excluded all the other possible causes ofaltruistic behavior and effectiveness of the community (portes, 2000). the two major dimensions of social capital in bonding and bridging is based on the diversity and the number of ties are nevertheless one of the most pertinent factors of workforce diversity is religion in global enterprises(hitt et al., 2002).thus the research aimed to link other possible causes of altruistic behavior with the social capital such as culture and identity. themain objective of this research was to define social capital using the bridging and bonding mahreen hussain, mahwish zeeshan & abid ghafoor chaudhry 104 concept using the internal and external ties approach. further this research study utilizes the religiosity culture and social identity for the bridging and bonding concepts in the healthcare industry of pakistan. 1.4 aims and significance of the research the proposition for this study was that social capital will affect social responsibility with religiosity culture and social capital moderating the relationship. the aim of the research was to study the relation of social capital and social responsibility for the health care practitioners with culture and identity being the agent in between. it aimed to fill gaps in literature by (i) integrating the process of creation, formulation and operation; (ii) providing social responsibility as the motive for the formulation of social capital; (iii) employing bridging and bonding concepts for the creation of social capital; and, (iv) utilizing the three dimensions of social capital i.e., relational, structural and cognitive, with both the approaches. thus, the ultimate motive of social capital for instigating the social responsibility in the health care practitioners is achieved. 2. literature review the term social capital is quite frequently used in the literary circles now-a-days in different contexts and holding diversified meanings. business community relationships operating in developing countries require the advancement of endogenous theories (jamali and karam, 2018). the discussion ofwhat isactually social capital is still going on as it is a concept with roots deep in many domains. the debate of social capital was already geared up when the term social capital was first cited in the works of (hanifan, 1916) who defined it as neighbor coming into contact with other neighbors and they coming in contact with others for the accumulation of social capital (maak, 2007). the concept of social capital dates back toearly nineteenth century capitalist theory when social economists like marx, hume and smith studied the society. the use of the term in nineteenth and eighteenth century suggested that term social capital is geared up for the conceptdevelopment. social capital has been defined from the individual’s point, collective aspect and later on the literature established it is the combination of both individual and collective efforts. earlier on, coleman’s (1988) definition ofsocial capital suggests that it is a collection of multiple entities which hastwo common facets among them: they facilitate certain actions in collaborationofperformers and uphold a certain social structure. (quigley, 1996)definedsocial capital as having the characteristicsoftrust and shared values which enable collaboration among the social units for mutual benefit. similarly, (cohen and prusak, 2002)defined social capital as an active stock of links among people sharing values, mutual understandings, trust and behaviors which bind the network members and facilitate cooperation. (bourdieu and wacquant, 1992)identifysocial capital as the actual or virtual resources in sum, inherent in a group or an individual viamutual acquaintances and recognition that structures a network. (portes, 1998)defined the social capital as securing benefits by being the member of a journal of applied economics and business studies, volume. 4, issue 2 (2020) 101-118 https://doi.org/10.34260/jaebs.425 105 social network. social capital when defined from the perspective of both the individualistic and collectivist approaches is based on the potentiala n d a c t u a l resources rootedwhich are rooted derived from the relationships which are linked and possessed by asocial unit or an individualmaking it accessible through them (nahapiet and ghoshal, 1998). this definitionis now usedto define the social capital as it follows both the approaches. (woolcock and narayan, 2000)termed thatsocial capital is an attribute which facilitates collective acting. it engulfs all the characteristics that servein the formationof social capital. whereas, (adler and kwon, 2002)entail the attributes of social capital as network, relationships, norms, trust and goodwill. these authors focused on the three different yet important aspects of the social capital i.e., how it is formulated, purpose of formulation and how it willoperate. religiosity is a subjective concept and varies in terms of personal and public religiosity since the effects of individual religiosity are theorized to be relative to the norms of one’s social environment(hayward and kemmelmeier, 2011).according to (wuthnow, 2002) the bonding refers to the interpersonal cohesion in the members of small groups, local communities and etc., on the long term and it occurs in the identical groups more easily endowing emotional support and friendship. the identification could be based on the race, gender, religion, culture or both the culture and religion simultaneously, forming a religiosity culture. in this context culture refers to extent to which norms and behaviors direct relations. it is a set of standard norms and values directing individual’s behavior of a group or social unit. while these are sometimes spelled out in formal contracts, often they are simply understandings that evolve within the dyad and the network”(gulati et al., 2000). however, the culturally obliged involvement in the religious acts comes under the domain of culture. religious attachments are also a source of networks and may be one source of such networks where worshippers get together might be a place individual’s meet influential people to form ties. the bonding conept is focused on the relations internal to the society and organization. the religiosity culture was first acknowledged by (durkheim, 1947) as ultimate superiority of society as accepted religious beliefs over the inferior individual. since the religion determines people’s behavior (sadler, 1970) directly through the taboos and rules stimulated (macdonald, 1986) and indirectly through the institutionalizing code of conduct and behaviors acceptable and prioritizing them (sood and nasu, 1995). (wuthnow, 2000) used religious involment as a status-bridging social capital. nevertheless, religiousity culture has its implications on the individual level within homogeneous group although it lacks the empirical analysis as an individual is a part of as different groups simulataneously having diverse goals. thus, the religiosity culture will be used to measure the socail capital internal to the group or social unit with which they identify on the basis of shared culture and goals only. in this sense culture is a fixed locale that outlines the identity and will affect it. the bridging concept encompasses the formation of ties across the social units of citizens who donot essentially share the same social identity but works for the harmony and mutual respect (poortinga, 2006). it is also defines as the external resources inherent in network ties mahreen hussain, mahwish zeeshan & abid ghafoor chaudhry 106 and relations of the social network (bourdieu and wacquant, 1992). the scope of the bridging is to develops ties with the hetrogeneous groups and linking them to form a large society to fortify it(paxton, 1999). thus, the bridging concept is related to the external relations on the individual or collective level and will help in maintaining the access to resources. the bridging concept promotes civic responsibilty, increases tolerances and boosts coopertion to address the larger societal issues(portes and landolt, 1996; skocpol and fiorina, 1999). bridging is not easy to create and sustain because the actors are required to interact outside their social circle and in communities helping in managing the diversified groups and cultivate collaboration. thus the identity-bridging social capital includes classification of ethnicity, religious traditions, origion and race embedded in the culture and sub-culture differentiating them on their value system and preferences. this classification is based on what people think of “us” and what is their perception of themselves. consequently, the identity of an individual becomes the source of the capital creation (wuthnow, 2002). investment in the internal ties enhances the collective identity and the collective action. (tajfel and turner, 2004) states that the individual’s knowledge that he belongs to certain social groups having emotional and value significance to him for that certain group membership is social identity. according to (mael and ashforth, 1992) it is individual’s perception of oneness or belongingness to an organization, where he defines one’s self in terms of the organization in which he holds a membership. table 1: summary of review of literature on social capital group author approaches key terms i (coleman, 1988) (quigley, 1996) (cohen and prusak, 2002) inindividualisticapproach socialstructureand actors. networks, norms and trust. trust, values, understanding and behaviors. ii (bourdieu and wacquant, 1992) (portes, 1998) collective approach acquaintances and recognition for networkstructure. securing benefits bybeing a member. iii (nahapiet and ghoshal, 1998) (woolcock and narayan, 2000) (adler and kwon, 2002) individualistic and collective approach sum of actual and potential resources and accessible through social unit. attributesthat facilitates acting collectively. attributes: networks, relationships, norms, trust and goodwill. in the table 1 above, three patterns emerged based on the literature above individualistic, collectivist and individualistic-collectivist approaches with regards to the development of term social capital over the period. journal of applied economics and business studies, volume. 4, issue 2 (2020) 101-118 https://doi.org/10.34260/jaebs.425 107 further to this social capital was classified on the basis of bridging and bonding concepts (gittell and vidal, 1998; wuthnow, 2002). bonding focuses on the “actual or potential” social capital of a group or society inherent in their structures, shared norms and values (coleman, 1988). bonding pertains to the ties that humans build for the social cohesion with the particular group with which they identify (harpham et al., 2002). to build the network and ties the individuals tap into cultural, ethnic or religious social repositories. according to (wuthnow, 2002)the bonding refers to the interpersonal cohesion in the members of small groups, local communities and etc., for a longer period among the members of identical groups establishing emotional support and friendship. the individuals identify with each other based on the race, gender, religion, culture or the culture & religion simultaneously, known as religiosity culture. in this context culture refers norms and behaviors impacting relations; a set of standard norms and values directing individual’s behavior of a group or social unit. further to this religious attachment provide a source of networking, a place where worshippers get together and meet people of means to form ties. the religiosity culture was first acknowledged by (durkheim, 1947) as the society’s religious beliefs takes precedence over the individual’s inferior belief system. since the religion determines people’s behavior (sadler, 1970)directly through the taboos and rules stimulated (macdonald, 1986)and indirectly through the institutionalizing code of conduct and behaviors acceptable and prioritizing them (sood and nasu, 1995). nevertheless, religiosity culture has implications within homogeneous (same) groups as an individual can be a part of various social groups simultaneously having diverse goals. this study will assess the impact of religiosity culture on the social capital internal to asocial unit with which they identify on the basis of shared culture and goals. bridging focuses to develop ties with the heterogeneous groups and linking them to form a large society to fortify it (paxton, 1999). however, social units of citizens might not essentially share the same social identity but works for harmony and mutual respect (poortinga, 2006). the social identity can be derived from any association with a social unit other than that of ethnicity, race or culture. (granovetter, 1977)insisted for the strong mutual ties however he emphasized the importance of weaker ties that helps an individual to connect with social groups and units outside his comfort zone for individual growth and creating large diverse groups. bridging social capital is not easy to create and sustain because the actors are required to interact outside their social circle and in communities helping in managing the diversified groups and cultivate collaboration along with their differences. thus the identity-bridging social capital’s classification is based on what people think of “us” and what is their perception of themselves; the identity of an individual becomes the source of the capital creation. (tajfel and turner, 2004)suggest the knowledge of an individual on his affiliation to a certain social group may serve in reflecting emotional significance to him for that certain group membership is social identity. according to (mael and ashforth, 1992)it is individual’s awareness of belongingness with an organization or oneness with that group, where he defines one’s self in terms of the social unit of which he holds a membership. identity is also defined through its mediating power and functions through the prestige based on social, economic or cultural capital which is a building block of the developed state. mahreen hussain, mahwish zeeshan & abid ghafoor chaudhry 108 3. conceptual framework the review of literature above indicates some deficiencies that authors have defined over the years in terms of approaches and basis of creation i.e., bridging & bonding. as per the table 1.1group i (cohen and prusak, 2002; coleman, 1988; quigley, 1996)used individualistic approach; group ii (bourdieu and wacquant, 1992; portes, 1998)defined social capital on the basis of the collectivist approach; and, group iii (adler and kwon, 2002; nahapiet and ghoshal, 1998; woolcock and narayan, 2000)employed both the approaches i.e., individualistic and collectivist. scientists discussed the comparison between both approaches in detail in different research works (beilmann et al., 2018; song, 2020). yet the creation of social capital and a concise model using both the approaches in view the culture and identity were still deficient in the literature. the perspectives of social capital for pakistan have also been discussed by the authors in detail (bhatti et al., 2020; hafeez et al., n.d.; sana et al., 2020). the studies over years have contradicted that the purpose of forming ties are either mutual benefit or altruistic behavior. nevertheless, suggested that individuals establish networks and units for the sense of unity or sharing same identity; as individuals prefer the actions that enhance their honor and value (wuthnow, 2002). the aim of this study is to investigate the relationship of religiosity culture and social identity for the creation of social capital. in pakistan the major religion is islam having altruistic behavior as a basic tenant which institutionalizes kinship based on the religious ties. whereas, the people apart from religion, identify with each other on the basis of ethnicity & caste for example people who speak hindko (native language of hazara division) identify themselves as hazara. therefore, the shared values and social norms inherent in the language, ethnicity and castes facilitates in forming large homogenous groups. the research intended to study the interaction of public sector healthcare practitioners while creating social capital through religiosity culture and social identity reflecting upon the neo-tocquevillean philosophy that performance of public life and social institutes are affected by the social norms and civic engagements. figure 1: conceptual framework journal of applied economics and business studies, volume. 4, issue 2 (2020) 101-118 https://doi.org/10.34260/jaebs.425 109 3.1 proposed hypothesis the hypothesis of the study was grounded in the literature and aimed at determining the factor responsible for social capital in the healthcare industry. the proposed hypotheses of this study are given below as follows: • h1: religiosity culture predicts social capital. • h2: social identity predicts social capital. whereas the formulas used in the research for religiosity culture and social identity are given as shown in eq.1 and eq.2 respectively. 4. research methodology the methodology section defines the constructs of the religiosity culture, social identity and social capital. further, explaining the sample selection, tools used for measurement of the data and techniques used for the data analysis. the validity and reliability tests were carried out because the questionnaire was self-constructed. the religiosity culture was derived from theanthropologyto the field of organizational behavior, since its inception the term has defied all the rules of scientific consensus with regards to empirical testing. according to (durkheim, 1947) religiosity culture can be better explained through its characteristics that could be found in the religion. however, (koenig et al., 2001) defined it as an organized system of beliefs, symbols and practices that extend one’s understandingof their relationships and responsibilities towards the society they are part of. the institutions of the religiosity are based on the belief system; whereas the interactions are based on the value system. for this study both the values and belief systems were used to measure the religiosity culture as an extrinsic measure. whereas, the social identity comprises of various characteristics that helps an individual to consider him/herself as a part of a group with whom they identify (ashforth and mael, 1989). social identity requires an individual to be a part of the socialgroup where he performs according to the norms of the society to feel accepted (simon, 1992). this variable was measured through social commitments and interactions. social capital was measured on the trust and trustworthiness of the actors enabling them to achieve goals working collectively (baron and hannan, 1994).the social capital also refers to the capital hidden in the relations. (granovetter, 1992) defines it as social network of the individuals where he mobilizes his contacts to access the resources which would have been possible in the absence of the network. mahreen hussain, mahwish zeeshan & abid ghafoor chaudhry 110 the population selected for this study comprised of the public healthcare practitioners of federal capital of pakistan comprisingof doctors and nurses. the sample was purposively selected from the 5000 healthcare practitioners of islamabad only because of the geographical delimitation. convenience sampling was used because the sampling frame did not exist; sampling of a population has the same generalizability as of the random sampling technique employed on them retaining its uniqueness,if selection biasness is avoided (hultsch et al., 2002) according to (viswesvaran, 1998)subject-variable ratio requires fifteen subjects per variables but in this case there were only three variables requiring the minimum count of 60 subjects but a sample size of 215 was collected forth is study. a survey questionnaire was designed comprising of two sections: demographics and questionnaire designed on likert scale having thirty-five questions. since, the questionnaire was developed by the researcher; it was tested for the validity and reliability using the factor analysis, correlations and alpha reliability respectively. insignificantly correlated or reverse coded questions were excluded. the reliability of the sub-scales and components extracted were all greater than 0.5, whereas the reliability of the total scale was 0.697. the primary tool for the data collection was questionnaire. the questionnaire was designed for the said variables i.e., religiosity culture, social identity, social responsibility and social capital, specific to the healthcare industry of pakistan. consequently, this questionnaire enquires about the belief system, values and a moral obligation of an individual. this questionnaire is divided into two sections. the first section pertains to the demographic variable whereas; the second section measures the four variables through self-ranking. the purpose of this questionnaire was to gauge the causation of social capital by the social identity and religiosity culture. it also measures the effect of social capital on the social responsibility of the healthcare individuals. table 2: summary of data collection questionnaire respondents population size sample size number of responses received percentage of responses received variables measured section-i: demographics section-ii: main questionnaire 139doctors 76nurses 5000 500 215 43% i. religiosity culture, ii. social identity, iii. social responsibility, iv. social capital. busha and harter (1980) states that to conduct a survey the questionnaire should be administered and fairly designed to improve the results. this study developed the questionnaire fair enough and the questionnaires were self-administered to increase the journal of applied economics and business studies, volume. 4, issue 2 (2020) 101-118 https://doi.org/10.34260/jaebs.425 111 response rate and result reliability. the main questionnaire was measured using the likert scale with five options instead of the three and seven. the five-point likert scale is the widely used because of easy to code and label. this scale rated 1 for strongly disagree, 2 for disagree, 3 for neutral, 4 for agree and 5 for strongly agree. prior appointments were made for the survey and were contacted in the premises of the hospital. the questionnaires were self-administered and were returned during the stay. the time duration required to fill a questionnaire was 10-15 minutes with the informal interview regarding the sensitive questions of religiosity and ethnicity. the informal interviews were used to analyze the data that was received through the survey and helped in defining the anomalies occurring. the respondents feel at ease with oral conversations than verbal statements. the data collected through survey questionnaire was in a self-administered manner such that response error and bias could be avoided, and the researcher is there to answer the questions. for the dispensary and mch the questionnaires had to be translated in urdu for the convenience of the respondents as their qualification level was matriculation and intermediate. the statistical techniques applied for the questionnaire development and for the hypotheses testing include respondents’ profile, validity and reliability statistics, data description using mean and standard deviation and comparison of demographic variables using anova and t-test. 5. results and discussions the statistical method used in different disciplines to determine the strength and characteristics of the relationship between dependent and other variables is called regression. regression is widely used method to find the importance of dependent variables. scientists used regression analysis for different problem solving related to social capital (benbow and lee, 2019; carrillo‐álvarez et al., 2019; saptutyningsih et al., 2020).the findings were based on the testing of the hypothesisusingthelinear regression. following results were obtained for the hypothesis 1 i.e., religiosity culture predicts social capital. thus, results from the table 2.1 revealed that the r value in the model summary expresses the simplecorrelationcoefficientthatwas 59.6 forthis equation table 3: summary of regression analysis for hypothesis 1 social capital variables coefficient confidence interval constant 1.109 (0.6341.585) religiosity culture 0.731 (0.5960.866) r2 0.356 f 113.16 whereas n=206, *=p<0.05, ** p<0.10. mahreen hussain, mahwish zeeshan & abid ghafoor chaudhry 112 however, r2 explains the amount of variance in the outcome as a ratio of how much variation there was to be explained, which was 162 in totaland this regressionmodel defined 57 percent of the total. however, the r2for thehypothesis1 was 35.6% and f value was significant at 95% confidence interval for 113.631. this f value explainshow accurately religiosity predicted social capital even with the presence of errors. nevertheless, the t-test next to the β value was significant at p<0.05 at 10 explains predictor contributes significantly to the outcome variable. thus, it was concluded that religiosity culture contributes 36 percent in explaining social capital and h1 was not rejected. table 4: summary of regression analysis for hypothesis 2 social identity as predictor variables coefficient confidence interval constant 1.559 (1.02-2.09) social capital 0.67 (0.49-0.84) r2 0.226 f 60.04 whereas, n=206, *=p<0.05, ** p<0.10. the table 4 reflects the results for the hypothesis h2 i.e., socialidentity predicts social capital. the results show that the correlation for the social identity and social capital was estimated to be 47.5%. the total variance to be explained was 164 and social identity explained 37 percent of that. r2 was estimated to be around 22percent with the f-value of 60.47 at the significance level of 95 percent. however, the beta (β)was estimated to be 67 percent corresponding with the t 7.7 at p<0.05. thisexplainsthat the contribution of social identity to the social capital was significant for this model in the presence of the errors. thus, the hypothesis h2 was not rejected implyingthat social identity causes socialcapital. the results were in accordance with the (wuthnow, 2002)that social capital uses culture bonding and identity bridging. share values, culture, and social networks to which the individuals belong, in turn influence behavior, the economy’s functioning and the legal and institutional basis (hofstede, 2001).(putnam, 2000)states that religious attachments are also a contributor to the social capital. religiosity affects the number and diversity of ties in enterprises in a higher proportion (ramasamy et al., 2010). however, the previous studies used religious involvements only as the creation of the social capital, but this study used involvement and belief system. it is believed that all big religions of the world such as christianity, islam, buddhism, hinduism, confucianism, etc. preach social responsibility and support (ramasamy et al., 2010). the variable of the religiosity culture does notuse empirical studies often either because of the sensitivity or improper tools. but this provided the future researchers with the tool and constructs that could be used for the measurement. journal of applied economics and business studies, volume. 4, issue 2 (2020) 101-118 https://doi.org/10.34260/jaebs.425 113 the results for the social identity are in accordance with the (wuthnow, 2002)that individuals look outside their circles and institutes to enhance the social capital. however, the contribution to the social capital was lesser than religiosity culture as putnam (2000) suggests that it is created by individuals from middle and upper middle class mostly. nevertheless, the results were in accordance with the (woolcock and narayan, 2000)that developing countries should be moving towards the bridging instead of the religious bonding.(servaes and tamayo, 2017), suggest that social capital can be put together through activities of csr. in adding up to the investments in human, physical and intellectual capital, firms are required to extend the priorities related to social capital based on the relationship qualityestablished by a firm with various stakeholders (farooq et al., 2019). it is believed that by paying attention to the wants and unattended social needs of the community (hao et al., 2018).this study shows that the focus of the social capital is religious groups and beliefs instead of social identity. also, the decision makers have to take into account the religious affiliations of the employees as well as their culture in order to mobilize the social capital to enhance the effectiveness of healthcare industry. 6. conclusion this study revealed that hypothesis 1 and 2 were not rejected. so, the study concludes that religiosity culture and social identity predicts social capital. these results are in congruence with putnam (2000) stating that religiosity culture is abetter opportunity for the creation of social capital using bonding concept as compared tothe social identity using bridging concept. the network ties for the religiosity culture are strong for the under developed countries. the managerial implications for this study lies in this fact as the mangers have to create and cultivate the values vital for the identity creation such as reward management, recognition, work environment, interpersonal relations and motivating factors using herzberg’s two factor theory. for the policy makers this would help them in utilizing what they have and cultivating identity among the healthcare professionals. the health care professionals require social responsibility which tends to be on the downfall, thus measures are needed to enhance it. there are no research implications for doctors and other health professionals because the participants who belong to health sector such as doctors and nurses have been consulted for filling of the questionnaire. the data anonymization techniques were applied so that privacy of the participants can be ensured. the future recommendations for this study include religiosity culture with only two constructs beliefs and groups whereas there are number of other items that could lead to religiosity culture. these items include religious affiliations and attendance of the religious activities which this study didnottake intoaccount. another gapfor future research lies in testing for the factors that made social identity less significant for this society. thus, religiosity culture and social identity predicts the social capital for the healthcare professionals of pakistan working in public sector. mahreen hussain, mahwish zeeshan & abid ghafoor chaudhry 114 references adler, p.s., kwon, s.-w., 2002. social capital: prospects for a new concept. academy of management review 27, 17–40. ashforth, b.e., mael, f., 1989. social identity theory and the organization. academy of management review 14, 20–39. baron, j.n., hannan, m.t., 1994. the impact of economics on contemporary sociology. journal of economic literature 32, 1111–1146. beilmann, m., kööts-ausmees, l., realo, a., 2018. the relationship between social capital and individualism–collectivism in europe. soc indic res 137, 641–664. https://doi.org/10.1007/s11205-017-1614-4 benbow, r.j., lee, c., 2019. teaching-focused social networks among college faculty: exploring conditions for the development of social capital. high educ 78, 67–89. https://doi.org/10.1007/s10734-018-0331-5 bhatti, s.h., vorobyev, d., zakariya, r., christofi, m., 2020. social capital, knowledge sharing, work meaningfulness and creativity: evidence from the pakistani pharmaceutical industry. journal of intellectual capital. bourdieu, p., wacquant, l.j., 1992. an invitation to reflexive sociology. university of chicago press. busha, c.h., harter, s.p., 1980. research methods in librarianship: techniques and interpretation. academic press. carrillo‐álvarez, e., kawachi, i., riera‐romaní, j., 2019. neighbourhood social capital and obesity: a systematic review of the literature. obesity reviews 20, 119–141. https://doi.org/10.1111/obr.12760 cohen, d., prusak, l., 2002. in good company: how social capital makes organizations work. harvard business review 80, 107–113. coleman, j.s., 1988. social capital in the creation of human capital. american journal of sociology 94, s95–s120. durkheim, é., 1947. the elementary forms of the religious life: a study in religious sociology. free press, glencoe, ill. farooq, q., hao, y., liu, x., 2019. understanding corporate social responsibility with crosscultural differences: a deeper look at religiosity. corporate social responsibility and environmental management 26, 965–971. https://doi.org/10.1002/csr.1736 farr, j., 2004. social capital: a conceptual history. political theory 32, 6–33. journal of applied economics and business studies, volume. 4, issue 2 (2020) 101-118 https://doi.org/10.34260/jaebs.425 115 flammer, c., luo, j., 2017. corporate social responsibility as an employee governance tool: evidence from a quasi-experiment. strategic management journal 38, 163–183. gittell, r., vidal, a., 1998. community organizing: building social capital as a development strategy. sage publications. granovetter, m., 1992. problems of explanation in economic sociology. networks and organizations: structure, form, and action 25–56. granovetter, m.s., 1977. the strength of weak ties, in: social networks. elsevier, pp. 347– 367. gulati, r., nohria, n., zaheer, a., 2000. strategic networks. strategic management journal 21, 203–215. https://doi.org/10.1002/(sici)1097-0266(200003)21:3<203: aidsmj102>3.0.co;2-k gupta, r., kaur, m., islam, s., mohan, v., mony, p., kumar, r., kutty, v.r., iqbal, r., rahman, o., deepa, m., 2017. association of household wealth index, educational status, and social capital with hypertension awareness, treatment, and control in south asia. american journal of hypertension 30, 373–381. hafeez, h., abdullah, m.i., riaz, a., shafique, i., n.d. prevention of occupational injuries and accidents: a social capital perspective. nursing inquiry n/a, e12354. https://doi.org/10.1111/nin.12354 hagan, j., macmillan, r., wheaton, b., 1996. new kid in town: social capital and the life course effects of family migration on children. american sociological review 368– 385. hanifan, l.j., 1916. the rural school community center. the annals of the american academy of political and social science 67, 130–138. hao, y., farooq, q., zhang, y., 2018. unattended social wants and corporate social responsibility of leading firms: r elationship of intrinsic motivation of volunteering in proposed welfare programs and employee attributes. corporate social responsibility and environmental management 25, 1029–1038. harpham, t., grant, e., thomas, e., 2002. measuring social capital within health surveys: key issues. health policy and planning 17, 106–111. hayward, r.d., kemmelmeier, m., 2011. weber revisited: a cross-national analysis of religiosity, religious culture, and economic attitudes. journal of cross-cultural psychology 42, 1406–1420. hirsch, p.m., levin, d.z., 1999. umbrella advocates versus validity police: a life-cycle model. organization science 10, 199–212. mahreen hussain, mahwish zeeshan & abid ghafoor chaudhry 116 hitt, m.a., lee, h.-u., yucel, e., 2002. the importance of social capital to the management of multinational enterprises: relational networks among asian and western firms. asia pacific journal of management 19, 353–372. hofstede, g., 2001. culture’s consequences: comparing values, behaviors, institutions and organizations across nations. sage publications. hultsch, d.f., macdonald, s.w., hunter, m.a., maitland, s.b., dixon, r.a., 2002. sampling and generalisability in developmental research: comparison of random and convenience samples of older adults. international journal of behavioral development 26, 345–359. jamali, d., karam, c., 2018. corporate social responsibility in developing countries as an emerging field of study. international journal of management reviews 20, 32–61. https://doi.org/10.1111/ijmr.12112 koenig, h.g., mccollough, m., larson, d.b., 2001. handbook of religion and health, new york: oxford univ. press. maak, t., 2007. responsible leadership, stakeholder engagement, and the emergence of social capital. j bus ethics 74, 329–343. https://doi.org/10.1007/s10551-007-9510-5 macdonald, d., 1986. imposing aid: emergency assistance to refugees. refuge: canada’s journal on refugees 12–14. https://doi.org/10.25071/1920-7336.21138 mael, f., ashforth, b.e., 1992. alumni and their alma mater: a partial test of the reformulated model of organizational identification. journal of organizational behavior 13, 103–123. https://doi.org/10.1002/job.4030130202 mclanahan, s., sandefur, g., 1994. growing up with a single parent. what hurts, what helps. harvard university press, 79 garden street, cambridge, ma 02138 ($19. nahapiet, j., ghoshal, s., 1998. social capital, intellectual capital, and the organizational advantage. amr 23, 242–266. https://doi.org/10.5465/amr.1998.533225 nishtar, s., khalid, f., ikram, a., kazi, a., abbas, y., mirza, khattak, i. ul h., javad, a., jaffer, h., karl, brown, badsha, t., 2010. protecting the poor against health impoverishment in pakistan: proof of concept of the potential within innovative web and mobile phone technologies [www document]. url /paper/protecting-the-pooragainst-health-impoverishment-%3a-nishtarkhalid/adb1b64d4e792701c242f25a34d75850309be506 (accessed 9.5.20). paxton, p., 1999. is social capital declining in the united states? a multiple indicator assessment. american journal of sociology 105, 88–127. https://doi.org/10.1086/210268 pitkin derose, k., varda, d.m., 2009. social capital and health care access: a systematic journal of applied economics and business studies, volume. 4, issue 2 (2020) 101-118 https://doi.org/10.34260/jaebs.425 117 review. medical care research and review 66, 272–306. poortinga, w., 2006. perceptions of the environment, physical activity, and obesity. social science & medicine 63, 2835–2846. https://doi.org/10.1016/j.socscimed.2006.07.018 portes, a., 2000. the two meanings of social capital. sociological forum 15, 1–12. https://doi.org/10.1023/a:1007537902813 portes, a., 1998. social capital: its origins and applications in modern sociology. annual review of sociology 24, 1–24. https://doi.org/10.1146/annurev.soc.24.1.1 portes, a., landolt, p., 1996. the downside of social capital. putnam, r.d., 2000. bowling alone: america’s declining social capital, in: crothers, l., lockhart, c. (eds.), culture and politics: a reader. palgrave macmillan us, new york, pp. 223–234. https://doi.org/10.1007/978-1-349-62965-7_12 quigley, k.f.f., 1996. making democracy work: civic traditions in modern italy: by robert d. putnam. (princeton, n.j.: princeton university press, 1993.258 pp. $24.95.). orbis 40, 333–341. https://doi.org/10.1016/s0030-4387(96)90074-5 ramasamy, b., yeung, m.c.h., au, a.k.m., 2010. consumer support for corporate social responsibility (csr): the role of religion and values. j bus ethics 91, 61–72. https://doi.org/10.1007/s10551-010-0568-0 sadler, w.a., 1970. personality and religion: the role of religion in personality development, harper forum books. harper & row. sana, a., rida, f., tayyaba, i., masooma, m., ayesha, z., kalsoom, a., 2020. willingness to pay for community-based healthcare insurance schemes in developing countries: a case of lahore, pakistan. ethiopian journal of health sciences 30, 135. sander, t.h., putnam, r.d., 2010. democracy’s past and future: still bowling alone? the post-9/11 split. journal of democracy 21, 9–16. https://doi.org/10.1353/jod.0.0153 saptutyningsih, e., diswandi, d., jaung, w., 2020. does social capital matter in climate change adaptation? a lesson from agricultural sector in yogyakarta, indonesia. land use policy 95, 104189. sen, s., cowley, j., 2013. the relevance of stakeholder theory and social capital theory in the context of csr in smes: an australian perspective. j bus ethics 118, 413–427. https://doi.org/10.1007/s10551-012-1598-6 servaes, h., tamayo, a., 2017. the role of social capital in corporations: a review. oxf rev econ policy 33, 201–220. https://doi.org/10.1093/oxrep/grx026 skocpol, t., fiorina, m., 1999. civic engagement in american democracy. the brookings institution press and the russell sage foundation, washington dc. mahreen hussain, mahwish zeeshan & abid ghafoor chaudhry 118 song, l., 2020. social capital, social cost, and relational culture in three societies. soc psychol q 0190272520939880. https://doi.org/10.1177/0190272520939880 sood, j., nasu, y., 1995. religiosity and nationality: an exploratory study of their effect on consumer behavior in japan and the united states. journal of business research 34, 1–9. https://doi.org/10.1016/0148-2963(94)00015-7 tajfel, h., turner, j.c., 2004. the social identity theory of intergroup behavior, political psychology: key readings. psychology press, new york, ny, us. https://doi.org/10.4324/9780203505984-16 viswesvaran, c., 1998. multiple regression in behavioral research: explanation and prediction. personnel psychology 51, 223. woolcock, m., narayan, d., 2000. social capital: implications for development theory, research, and policy. the world bank research observer 15, 225–249. https://doi.org/10.1093/wbro/15.2.225 wuthnow, r., 2002. religious involvement and status-bridging social capital. journal for the scientific study of religion 41, 669–684. https://doi.org/10.1111/14685906.00153 wuthnow, r., 2000. how religious groups promote forgiving: a national study. journal for the scientific study of religion 39, 125–139. https://doi.org/10.1111/00218294.00011 journal of applied economics and business studies, volume. 4, issue 2 (2020) 75-100 https://doi.org/10.34260/jaebs.424 75 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x a descriptive analysis of research culture in pakistan with contextual reference to management sciences imran naseem1* saima imran2, muhammad tahir3 & bilal bin saeed4 1 honorary associate, university of liverpool school of management sciences, liverpool, uk. assistant professor, department of management sciences, comsats university islamabad, abbottabad campus, pakistan. 2 ms scholar, department of development studies, comsats university islamabad, abbottabad, pakistan 3 assistant professor, department of management sciences, comsats university islamabad, abbottabad campus, pakistan. 4 assistant professor, department of management sciences, comsats university islamabad, abbottabad campus, pakistan. abstract the study takes on to gauge evolution of research culture in pakistan with special focus to management sciences. the study is descriptive in script and exploratory in nature. its primary analysis evolves around number of universities, phds produced, and their different contextual ratios. the study finds relatively recent emergence of certain strengths in terms of a research culture and anticipates an opportunity to build upon these strengths in future. public universities with better experience are in the leading role in almost all domains and private and relevantly new universities lag behind a bit too far. if these could also start contributing, the near future could muster the fruit of a rich research culture. keywords research, evolution; management sciences; pakistan jel classification a20; a23; c44 1. introduction developing countries, unlike developed ones, do not have strong research culture. these countries have strong roots in conventional teaching with remote inclination for knowledgebased economy (salazar-clemeña & almonte-acosta, 2007). apparently, pakistan also shares this trait. there is a huge difference between the top countries with strong research culture and the developing world in terms of gdp spending on r&d (research and development) expenditure. whereas israel has been spending on an average 3.9% of its gdp on r&d in the last 23 years and stands on top of the list, germany spends 2.5% of its gdp and stands 10th on the list. none of the south asian countries touches even 1% of gdp on r&d spending. pakistan spends a meagre 0.3% of gdp on r&d. india stands above while sri lanka and nepal stand below this percentage whereas data are missing for other south asian countries. see graph 1. likewise, world bank provides data for 31 out of 47 countries with lower middle *imran.naseem@liverpool.ac.uk, drimran@cuiatd.edu.pk mailto:imran.naseem@liverpool.ac.uk mailto:drimran@cuiatd.edu.pk imran naseem, saima imran, muhammad tahir & bilal bin saeed 76 income. thirteen countries stand above pakistan while 17 countries stand below on the list. similar trend is seen in average researchers in r&d per million people. israel again stands on top with 7771 researchers in r&d per million people. pakistan on the other hand, though maximum in the region, has only 157 researchers in r&d per million people. see graph 2. similarly, it performs better than 17 and 10 countries perform better than pakistan on the list of lower middle income where data are available for 28 countries. this development indicates that pakistan is doing well in its respective region as well as among countries of relevant income group. this development can be easily traced back to the establishment of higher education commission (hec) of pakistan that was established in 2002 on the recommendation of a task force for the improvement of higher education. the task force blamed ineffectiveness of the administration, absence of required infrastructure, and incompetent academic researchers for decaying standards of public university education (akbari & naqvi, 2008). consequently, a world record budget jumped 7 fold from 2002 to 2008. this increase in budget resulted in the increase of universities that grew from 74 to 145 and enrolled students grew from 276,000 to over ten lac from 2001 to 2012 respectively (shaukat, 2012). universities further grew to 195 in 2019 (anon., 2019). a visible improvement in publication number can also be seen in the country which only published 68 if articles in 1975 jumped to 1138 articles in 2005 and kept on leaping higher in 2010, 2015, and 2018 with 4452, 7833, and 12413 publications respectively. see graph 3. such great leaps put pakistan on top of the countries that exhibited biggest rise in research output from 2017 to 2018. see graph 4. however, this increase does not result from an even base as many universities still face the staff shortage, especially the remote ones. to mock it further, any government change brings volatility in hec budget and performance (shaukat, 2012), a clear indicator of the leadership importance. in us 15 out of top 20 universities are reported to be private (bridgestock, 2013). it might be useful to have a comparison of public and private universities’ performance. similarly, women with phd have been lesser in numbers historically but have been increasing over the time (studio & foundation, 2009). this would also be interesting to see if the same pattern prevails in pakistani context. since literature is predominantly from developed countries, it is hard to apply its finding on developing countries like pakistan. this is why we assume it appropriate to explore research culture in pakistan. 2. literature review different authors have given different definitions and explanation of research culture over the years. schein, (1985) explained it in 6 different angles such as 1) regulations observed in behavior, 2) norms of the research, 3) the values dominating the research, 4) research philosophy of the organization, 5) organizational culture, and 6) the research climate. hauter explains it in the context of an organization that “how each individual should think, act and make decisions about research” (hauter, 1993) while to hill, (1999) it is an environment in which research grows and multiplies. perhaps owing to these different definitions evans, journal of applied economics and business studies, volume. 4, issue 2 (2020) 75-100 https://doi.org/10.34260/jaebs.424 77 (2009) terms research culture as vague concept hence rosas, (2013) explains it as “a blind date with the unknown”. however mapa, (2017) puts it in routine work “the way we do research round here”. the following sections present geographic review of literature in chronological order. 3. international studies literature reflects inability of the research to give a complete set of variables to explain ‘pan cultural’ aspects to theorize it (bhagat & mcquaid, 1982) (shalom, h. schwartz; maria, 1995). jan-benedict, e.;steenkamp, (2001) are also convinced that a limited set cannot properly explore all cultural forces at work. hence, dilworth-anderson, williams, & gibson, (2002) and farley & deshpandé, (2004) revealed through review of international studies significant cultural differences in different countries whereas craig & douglas, (2006) found that these differences are mixing up with other cultures causing plurality, contamination, and hybridization. culture, owing to its definitional diversity and measurement, also poses challenge to research, be it at group, organizational, or national level (leidner & kayworth, 2006). deshpande & webster, (1989) do not consider culture as endogenous variable to firm when studying comparative management approach rather identify it as exogenous. bland & ruffin, (1992) highlight the significance of robust leadership among common features to develop a research culture. bland & ruffin, (1992) and slade et al., (2018) advocate to give equal important in synergizing the consolidation and sustainability of employees as well as institutions rather than dealing the two aspects separately. leaders are equally pivotal at individual and institution base for not only setting clear research objectives but also to meritoriously communicate them. management investment is another common framework (slade, philip, & morris, 2018) for developing a research culture. for a given motivated and collaborative faculty, reasonable resource allocation for training and support, with the flexibility to give space to the faculty’s area of interest, is central to any effort aiming at cultivating a robust research culture (hanover, 2014) and we need this synergy and commitment for developing national research culture at a higher level. a research culture may take reasonably long time to evolve, even in the presence of all prerequisites (hanover, 2014). researchers’ compliance with the internationally accepted norms to transparency, openness, and productivity is a major challenge (vandenbos et al., 2015). these international studies suggest absence of a common set of variables for a rich and healthy global research culture. however, leadership, considering employees and institute cohesively, provision of resources and training to the faculty along with flexibility of research in areas of interest, and a long timeframe are some potential variables that may help foster a research culture. imran naseem, saima imran, muhammad tahir & bilal bin saeed 78 3.1 uk studies ebbutt & ebbutt, (2006) on the basis of studies conducted in the uk are contend about having a proper national mechanism that directly extends funds to the individuals or groups of teachers (even) at the school level. sheffield university highlights 8 main variables of research culture in the uk i.e. a) quality research, b) recruitment, c) human resource development, d) research discussion, e) departmental structure, f) culture, g) management, and h) doctoral researchers (“fostering an effective research environment,” 2017). however, without compromising the role of competition and differential assessment, team work and collaborative efforts are still missing in the research culture (sarah, 2018). it also fails to encourage the activities, from researcher’s point of view, necessary for quality research. hence, it needs to ensure efforts aiming at cultivating the culture of research support for the quality research practice (notes, 2015). these studies show that uk has a strong research culture with focus on a) quality research, b) recruitment, c) human resource development, d) research discussion, e) departmental structure, f) culture, g) management, and h) doctoral researchers but still needs to encourage activities from researcher’s point of view and research support for the quality research as envisioned in the international studies as well. 3.2 scandinavian studies (denmark, finland, netherland) surfing the scandinavian countries presents the vitality of action research in blossoming a research culture (kjerholt & hølge-hazelton, 2018). kessels & keursten, (2002) emphasize the supporting skills to guarantee smooth knowledge construction in a knowledge economy, anticipating blessings to the future elite of knowledge workforces. the digital humanities in the area are in search of identity and creating new space for the research culture with reasonable skill set to look after swarm of problems. this can be attained only through national collaboration and synergizing the strengths of the existing digital humanities hubs (matres, oiva, & tolonen, 2018). these studies indicate that action research, supporting skills, and digital humanities are important actors in the scandinavian region. 3.3 asia pacific studies (australia, new zealand) studies in this region, like international studies (bland & ruffin, 1992) (slade et al., 2018), (hanover, 2014), scandinavia (kjerholt & hølge-hazelton, 2018), and uk (“fostering an effective research environment,” 2017), southeast asia (rosas, 2013), schein (1985), also stress on the sturdy leadership for developing a vibrant research culture along with decentralization of the university management structure (pratt, margaritis, & coy, 2007) (marchant, 2009). a resilient leadership through optimal use of resources can change customary teaching loaded model into a more research oriented dynamic, facilitating creation of new knowledge which is essential to the universities’ mission (marchant, 2009). in a pursuit of evolving and maintaining research culture mentors must be available for the young journal of applied economics and business studies, volume. 4, issue 2 (2020) 75-100 https://doi.org/10.34260/jaebs.424 79 researchers, mentoring network can significantly contribute to research culture (studman, 2003). like scandinavian countries, asia pacific studies also consider action research as an established way for developing the research skills (ferguson, 1999). research culture is nurtured at institutional and individual level. the former requires consistency and comfort in terms of knowledge sharing, research direction, research support, and resource availability while the latter requires inspiration, research skills development, the parallels between the study of research culture and the organizational culture (hill, 2002). studies in the region highlight the importance of leadership, decentralization of university management system, optimal resource utility, mentoring network, action research, knowledge sharing, research direction, research support, resource availability, inspiration, and research skill development. 3.4 southeast asian studies (malaysia, philippines, vietnam) while a study imitates a stated objective for research (rosas, 2013), we also find stony reaction of the officials to acknowledge policy-practice gap and a less cooperative research culture, though, with a gradually developing appreciation for moving from orthodox quantitative research to qualitative one (scott, miller, & lloyd, 2006). a research culture in its evolution may pass through phases like creating, growing, developing, and nurturing research (anuar & abdul, 2013). mapa, (2017) classifies some significant causative features to research culture such as research policies, budget, benefits and incentives, research committees, culture and working environment, infrastructure, and inter-institute collaboration. similarly, dacles et al., (2016) identify that institutional support motivates individual faculty to get into the organizational culture. their causative factors are research unit, incentive, expertise, research programs, and institutional policies where research output stands the weakest factor. however, individually targeted peripheral incentives are not sufficient to nurture a research culture, it needs an all-inclusive approach on behalf of the administrators to implant research in their culture (teehankee, n.d.). however, whereas a sturdy research culture would cause in high research output, the same is not essentially correct vice versa (anuar & abdul, 2013) dissimilar to the many who believe in this bidirectional relation. these studies highlight the grey areas that need to improve such as acknowledging the policy-practice gap, less cooperative research culture, institutional support, and all-inclusive approach. the same problem is true in india, outside southeast asia, where an inclination succeeds to treat research and publication alike causing failure of the organizations to have an honest research drive. since publication are the result of individual requirements for its continued existence or promotion rather than culminating from a shared passion (chakaraborty, 2017), there is a dire need to aim research culture rather than research publication. developing countries have been recently seen appreciating the need for research orientation such as mexico that is seeing a prevalence of research in the academic setting lately (mendez & cruz, 2014). imran naseem, saima imran, muhammad tahir & bilal bin saeed 80 similarly, china is seeing a research passion where the government funds for research have been increasing annually at the rate of more than 20% which is shocking even to the most excited scientists (shi & rao, 2010). 3.5 pakistan studies not much literature is available with contextual reference to pakistan. however, with the establishing of higher education commission in september 2002 (akbari & naqvi, 2008) pakistan has been seeing an upward inclination in if publications since 2002 and has seen amazing growth in producing phds (lodhi, 2012) with a growing number of students registering for masters’ and doctoral degrees. however, this does not bring the research culture any closer to being inspirational and the shortage of third party supervision and quality academic professionals adversely affect it (agha, 2015). besides referring to this trend as an indicator of approval of the research culture in the country, lodhi, (2012) also warned for more hard work to get balance between prevailing teaching customs and a robust research culture in pakistani universities. agha, (2015) also approves it and points out adverse effects of absence of third party supervision and quality academic professionals ensuing in compromised caliber of pakistani students. the study also classifies culture in three domains i.e. 1) an individual’s ability to embark on research activities, whereas this ability can be built, improved, and refined through training, 2) human development through unstipulated medium, and 3) a set of shared concepts, customs, skills etc. held by some people that are transferred to their successors. these are diverse but valuable research culture features (lodhi, 2012). according to thomson reuters, “in the last decade, pakistan’s scientific research productivity has increased by more than four times, from approximately 2,000 articles per year in 2006 to more than 9,000 articles in 2015. during this time, the number of highly cited papers (hcps) featuring pakistan-based authors increased tenfold from 9 articles in 2006 to 98 in 2015” (herciu, 2016). but hoodbhoy, (2016) is unconvinced to this progress as he does not see on ground many of the fundamentals for such growth and terms it ‘playing the ranking game’. the studies suggest that research culture in pakistan has seen improvement in the recent decades but this is far from ideal as the mere increase in number of research publication may not bring the desired improvement in solving the societal problems. the literature above indicates several aspects helping foster a rich and healthy research culture. these include leadership (bland & ruffin, 1992), (slade et al., 2018), (hanover, 2014), (kjerholt & hølge-hazelton, 2018), (“fostering an effective research environment,” 2017), (rosas, 2013), (schein, 1985), & (marchant, 2009), culture (farley & deshpandé, 2004),(dilworth-anderson, williams, & gibson, 2002), (craig & douglas, 2006), (deshpande & webster, 1989), (jan-benedict, e.;steenkamp, 2001), (bhagat & mcquaid, 1982),(shalom, h. schwartz; maria, 1995), (leidner & kayworth, 2006), (“fostering an effective research environment,” 2017), (mapa, 2017), (dacles et al., 2016), (anuar & abdul, 2013),(agha, 2015), & (lodhi, 2012), resources (hanover, 2014), (ebbutt & ebbutt, 2006), (marchant, journal of applied economics and business studies, volume. 4, issue 2 (2020) 75-100 https://doi.org/10.34260/jaebs.424 81 2009), & (hill, 2002), doctoral research (“fostering an effective research environment,” 2017), (lodhi, 2012), & (agha, 2015), management (deshpande & webster, 1989), (slade, philip, & morris, 2018), (“fostering an effective research environment,” 2017), (pratt, margaritis, & coy, 2007), & (marchant, 2009), collaboration (mapa, 2017), (hanover, 2014), (“fostering an effective research environment,” 2017), (sarah, 2018), & (matres, oiva, & tolonen, 2018), development (“fostering an effective research environment,” 2017), (ferguson, 1999), (hill, 2002), (scott, miller, & lloyd, 2006), (anuar & abdul, 2013), (lodhi, 2012), & (agha, 2015), and quality (“fostering an effective research environment,” 2017), (sarah, 2018), (notes, 2015), 2002 (akbari & naqvi, 2008), (agha, 2015), & (lodhi, 2012) etc. 4. methods since the definitions of a research culture verity do not provide a standard set of variables, we shall take a few internationally used variables in our research such as leadership, resources, doctoral research, and culture to ascertain evolution of research culture in pakistan. to gauge the impact of leadership, we have taken four bench marks for leadership change i.e. 1) till 1947 when pakistan came into being, 2) 1948-1971 when the countries split into two separating eat pakistan as bangladesh, 3) 1972-1998 when pervez musharraf the then army chief toppled the democratic government and enforced emergency, and 4) 1999-2016 the era that began after the topple. the same bench marks are used for gauging the impact of resources in terms of increased number of universities and phds where number of doctoral research (phds) also serves as proxy for change in culture. in addition, our study takes into account different demographic indicators such as geographic, gender, and sector diversity along with university and supervisor ratio with respective number of phds produced. although in our wider scope we used mixed methodologies using both quantitative (descriptive analysis) as well qualitative (swot analysis) techniques but this paper would only elaborate descriptive approach to see through various sequential stages of the research evolution in pakistan. whereas our first version of the study dealt with the pakistani research culture in general (naseem, et al., 2019), second version of the study deals with qualitative aspect (naseem, et al., 2020), this version has special focus on management sciences. hence, this study is primarily quantitative in in nature. for this purpose, it takes number of universities and phds output across the country, gender diversity, and contribution of private and public sectors as indicators to demonstrate evolution of research culture with contextual reference to management sciences. it is a broad and general picture derived from the descriptive analysis and does not deal with the in-depth qualitative analysis. time series data start from country’s inception in 1947 and comes down to 2016. whereas gender, public, and private sectors are obvious, geographic division needs a bit explanation as follows. it is primarily divided into 7 administrative units i.e. 4 provinces i.e balochistan, khyber pakhtukhwa, punjab, and sind along with two semiautonomous regions i.e. azad jammu and kashmir (ajk) and gilgit baltistan. islamabad is the federal unit comprising of capital city and adjacent areas. the data imran naseem, saima imran, muhammad tahir & bilal bin saeed 82 source is primarily hec and respective universities’ websites with a few exceptions where websites were not available or updated, phone calls were made to avail requisite data. the graphs are taken from other sources and cited accordingly while tables are based on the data retrieved from hec and developed during the study. 5. findings we find that with each change in leadership, resources have increased and resultantly number of universities and doctoral researches have also increased. however, this change is immensely noted in the last change of leadership i.e. 1998. similarly, these changes in leadership and resource allocation have also been impacting research culture of the country in desirable direction. at the time we fetched data, hec website had 180 universities on the list. however, there seems a disparity in the demographic stretch as islamabad capital territory despite being a single city boasts of 21 universities in sharp contrast to quite a vast spread geography of three regions i.e. azad jammu kashmir, balochistan, and gilgit baltistan, where we only find a meagre number of 16 universities across the area. khyber pakhtunkhwa, sind, and punjab have 33, 53, and 57 universities respectively which again shows that though punjab has a greater number but it is far less than its share in population which is 30, 47.8, and 110 million respectively for khyber pakhtunkhwa, sind, and punjab respectively (“population of pakistan census 2017 information report,” 2017). except sind where private universities are in greater number than the public universities, in every other region, the latter surpass the former. see graph 5. considering management phd producing universities, majority of universities come from public sector except sindh that has equal share. ajk and gilgit baltistan have no contribution and balochistan has only 2 public sector universities contributing management phds. see graph 6. a bird eye view over era based breakup, we find an encouraging trend in the recent years. this correlates with changes in leadership and allocation of resources for higher education, comparing with the first 5 decades, the last 2 decades have witnessed a much increased number of universities. see graph 7. era-based breakup of management phds producing universities reveals that the majority of older universities are producing management phds while newer universities are still settling down and may take a while to gear up. out of 116 new universities, only 17 are on the contributing list. see graph 8. only 76 out of 180 are producing phds. only about 1 in 3 private and 1 in 2 public universities are producing phds. this trend can be linked with their respective establishment dates as majority of the private universities are only recently established. see graph 9. similarly, all phd producing universities are not producing management phds as well. about 2/3rd of these make on management phd producing university list. see graph 10. data from hec website indicates that there are 13259 indigenous phds, produced by pakistani universities. this number does not include the pending list. the picture reveals not much difference except that islamabad has produced more than twice phds than kp, though it journal of applied economics and business studies, volume. 4, issue 2 (2020) 75-100 https://doi.org/10.34260/jaebs.424 83 hosts universities in much lesser number. similarly, while punjab and sind have almost same in number there is a huge difference in favour of punjab in terms of phds produced. quaid e azam university in islamabad and relatively older universities in punjab could be the decisive factor that have produced a larger number of phds in comparison to their counterparts. see graph 11. in terms of management phd production, islamabad stands taller than all followed by punjab, sindh, and khyber pakhtunkhwa respectively. whereas balochistan has a very meagre contribution, ajk and gilgit baltistan have no contribution at all. see graph 12. phds breakup in chronology indicates that the last 2 decades have been more productive than the first 7 seven decades as university of the punjab produced first phd in 1930. see graph 30. the breakup in female phds shows more than 6 times increase in number of phds in the last 2 decades than all the previous phds that is very encouraging. see graph 14. phds growth in males also boasts an encouraging trend with more than 4 times production is recorded in last 2 decades than the all previous decades. see graph 15. growth trend of management phds reflects a huge leap of more than 20 times in the last two decades. at the time of independence, we only had two phds that also in economics which is an associated subject of management sciences. first ever pure management phd appeared in the mid 1990s. hence, management sciences appears to be a relatively recent phenomenon, in pakistan context at least. see graph 16. first female phd appeared in 1990. however, it was also in economics. first ever female phd in core management subject was produced in 2007. we have produced 136 female management phds so far. see graph 17. similarly, first ever male management phd was produced in 1994 and 542 altogether till date. see graph 18. twenty six (26) out of 76 phd producing universities are private and 50 are public sector universities. seventy four (74) phds per university is the ratio considering all 180 universities. segregation reveals that this ratio sharply declines to 8 phds per university in private sector universities and goes up to 102 phds per university in public sector. if we exclude non-phd producing universities, ratio for overall phd producing universities goes up from 74 to 174 phds per university. similarly, ratio for private and public sector universities also goes up from 8 and 102 to 23 and 214 respectively. see graph 19. exploring management phd producing universities reveals that about 1/3rd of private and about half of the public sector universities also produce management phds. their overall ratio 4 management phds per university, 2 and 5 management phds per private and public sector university respectively. these figures go up for 9, 5, and 11 respectively if we exclude non-phd producing universities. see graph 20. university of karachi and university of the punjab have produced almost 2500 and 2200 phds respectively that make them stand on the top and second positions respectively. university of agriculture faisalabad and quaid e azam university are next in ranks that have produced more than 1000 phds each followed by university of peshawar and university of sind producing more than 500 phds. out of 13259 indigenous phds a huge chunk of 55 universities have produced only 1350 phds with an average of 24.5 phds per university imran naseem, saima imran, muhammad tahir & bilal bin saeed 84 whereas rest of the 11909 phds are produced by the top 21 universities with an average of 567 phds each. all of these 21 universities hail from public sector and none from private sector. see graph 21. hamdard university, among private universities, stands on top with 96 phds. next in number are ghulam ishaq khan institute of technology with 54, muhammad ali jinnah university with 51, and qurtuba university of science and information technology with 44 phds. see graph 22. a major share of 7361 phds come from 4 universities which is more than half of the total. by including 2 other universities producing more than 500 phds, this ratio goes to more than 2/3rd of the total phds. fifteen universities have produced more than 100 phds. as discussed earlier, none of these 21 universities come from private sector. private universities appear on the scene with less than 100 universities where hamdard university tops the list with 96 phds. there are 2 other private and 7 public universities that have produced more than 50 phds. similarly, 11 private and 13 public universities have produced 10 or more phds with 21 universities producing less than 10 phds notwithstanding the 104 universities with no phd output. see graph 23. university of karachi again tops the list for producing the most number with 72 management phds. however, numl takes over university of the punjab with 42 management phds and university of peshawar takes over university of agriculture faisalabad with 35 management phds. all top 5 universities are public sector universities again. top 22 universities have produced 576 management phds while remaining 25 universities have produced 102 management phds. see graph 24. similarly, majority of management phds come from public sector universities. see graph 25. national college of business administration and economics stands on top among private universities with 31 management phds. this is followed by foundation university, muhammad ali jinnah university, and qurtuba university of science and information technology with 28, 23, and 19 management phds respectively. see graph 26. out of 47 management phd producing universities 7 have produced 276 management phds. this makes more than 1/3rd of the total. next 7 universities have produced 182 management phds that is almost 1/4th of the total. collectively these 14 universities have contributed about 3/4th management phds. public sector has a clear dominant role in the first four major segments while private sector comes closer to public sector in the last segment. see graph 27. further segregating these 678 management phds we realise that 217 are from core disciplines whereas most of the phds i.e. 335 belong to the subject that are closely associated with management sciences but are not primarily core subjects of management sciences. economics stands taller in these subjects with more than 260 phds. since management sciences is also taught in quite a few other subjects, 126 management phds come from such disciplines. see graph 28. further scrutinising core management phds, we realise that it is a very recent emergence as the first two management phds appeared in mid 1990s. and all 215 phds from core management sciences are produced in recent years. further private sector has journal of applied economics and business studies, volume. 4, issue 2 (2020) 75-100 https://doi.org/10.34260/jaebs.424 85 shown a strong contribution along with public sector. see graph 29. out of 217 core management phds, 34 are females and all have been produced in the last two decades. see graph 30. similarly, 183 management phds are males in this total of 217 core management phds and apart from 2, rest of all have been produced in the last two decades. see graph 31. similarly, exploring the supervisor to phds ratio, we find that there are 100 supervisors that have produced 10 or more phds each. dr. atta ur rehman tops the list with over 50 phds. further, 360 supervisors have contributed with 5 or more phds each. furthermore, 223 supervisors stand with 4 phds under their name while 448 and 1030 supervisors are tagged with 3 and 2 phds respectively. a huge number of 4060 supervisors could only contribute 1 phd each. see graph 32. dr. abuzar wajidi from university of karachi is the only supervisor who has produced 10 management phds. it is obvious that majority of the supervisors have produced less than 5 phds, and only a few supervisors stand with 5 or more phds. see graph 33. this chart does not count a reasonable majority of the faculty who have produced no phd far. 6. conclusion one prominent conclusion that we can draw from the study is that last major change in the leadership about two decades back has been a turning point where at least tremendous numerical improvement has been seen in almost all aspects of the research whether it was increase in resources, number of universities or number of indigenous phds. leadership change in 1999 in terms of government and in 2002 the transformation of university grants commission into higher education commission can clearly take this credit and the result confirms literature (bland & ruffin, 1992), (pratt, margaritis, & coy, 2007), (marchant, 2009) that leadership is an important ingredient in cultivating research culture. however, this improvement has resulted primarily from a few older and public universities and is not evenly distributed over geographic or sectorial stretch. hence, room for enormous improvement is still there and much needed contribution from universities with recent establishment dates and private sector would certainly add to the research cultural depth. similarly, there is much room for many public sector universities that need to become at par with the leading universities. this would surely add to the evolution of the research culture. similarly, there is a huge geographical stretch that has very limited set of educational resources and facilities. a policy initiative based on equity may bring benefits to the whole society for sure. benefits for gender equity have also well recorded. stronger contribution from private universities would definitely play a catalyst role in this evolution in the years to come. last but not the least, more efficient supervision is expected to add to the improvement. however, in nutshell, the recent trends are encouraging and we anticipate their continuity and more improvement. imran naseem, saima imran, muhammad tahir & bilal bin saeed 86 6.1 limitations and future course of action current study only focused on number of universities and phds as indicators of evolution of the research culture based on the leadership and resource allocation benchmarks. the study does not explore characteristics of the respective leaderships. further, the study also does not cater for political atmosphere of the country in terms of its priorities regarding education. hence future research could incorporate such indicators as well. similarly, any future study could do a comparative study of indigenous and foreign phds as hec has awarded a huge number of foreign scholarships for phd programs and many of them have returned and joined different institutes. there are also many jobless phds and we need to explore the possible reasons for this bleak scenario. future research, among others, can also study phds produced and their respective publications and citations. we also need to explore the quality of pakistani journals and their comparison with the foreign journals in respective fields. though this paper presents quantitative outlook, qualitative aspects need also exploration. this also remains unanswered whether this improvement results from change in leadership or resources alone or it results from combination of the both. references agha, n. (2015, november 15). research culture in pakistan_ the ‘undisclosed’ moral decadence dawn. the dawn. retrieved from https://www.dawn.com/news/1218751/ research-culture-in-pakistan-the-undisclosed-moral-decadence anuar, m., & abdul, b. (2013). understanding of research culture levels : a review of literature, (i), 120–125. bhagat, r. s., & mcquaid, s. j. (1982). role of subjective culture in organizations: a review and directions for future research. journal of applied psychology. https://doi.org/10.1037/0021-9010.67.5.653 bland, c. j., & ruffin, m. t. (1992). characteristics of a productive research environment: literature review. academic medicine : journal of the association of american medical colleges. retrieved from http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?dbfrom= pubmed&id=1596337&retmode=ref&cmd=prlinks%0apapers3://publication/uuid/14de 9a14-395a-4d26-ad32-b5e6d64e3e6b chakaraborty, s. (2017). creating a culture of research in india the hindu business line. craig, c. s., & douglas, s. p. (2006). beyond national culture: implications of cultural dynamics for consumer research. international marketing review, 23(3), 322–342. https://doi.org/10.1108/02651330610670479 dacles, d. d. m., valtoribio, d. c., yolanda, f., rosario, g. del, matias, c. a., & saludarez, m. u. (2016). cultivating research culture : an analysis of contributing factors , the institution ’ s research initiatives , and collaboration among the hei ’ s trifocal functions, 4(6), 439–449. https://doi.org/10.12691/education-4-6-2 journal of applied economics and business studies, volume. 4, issue 2 (2020) 75-100 https://doi.org/10.34260/jaebs.424 87 deshpande, r., & webster, f. e. (1989). organizational culture and marketing: defining the research agenda. journal of marketing, 53(1), 3. https://doi.org/10.2307/1251521 dilworth-anderson, p., williams, i. c., & gibson, b. e. (2002). issues of race, ethnicity, and culture in caregiving research. the gerontologist, 42(2), 237–272. https://doi.org/10.1093/geront/42.2.237 ebbutt, d., & ebbutt, d. (2006). the development of a research culture in secondary schools a research culture in secondary schools, 0792. https://doi.org/10.1080/09650790200200171 evans, l. (2009). developing research capacity in the social sciences: a professionality based model. international journal for researcher development, 1(2), 134–149. https://doi.org/10.1108/1759751x201100010 farley, j. u., & deshpandé, r. (2004). organizational culture, market orientation, innovativeness, and firm performance: an international research odyssey. international journal of research in marketing. https://doi.org/10.1016/j.ijresmar.2003.04.002 ferguson, p. b. (1999). developing a research culture in a polytechnic : fostering an effective research environment. (2017). the university of sheffield. retrieved from https://www.sheffield.ac.uk/rs/ethicsandintegrity/research-environment/research _environment hanover. (2014). building a culture of research: recommended practices. hanover research, (may), 1–33. retrieved from http://www.hanoverresearch.com/media/building-aculture-of-research-recommended-practices.pdf herciu, l. (2016). pakistan: another bric in the wall. retrieved from http://ipscience.interest.thomsonreuters.com/incites-pakistan/ hill, r. (1999). revisiting the term “research culture.” higher education research and development (herd), (july), 12–15. hill, r. (2002). establishing and sustaining a research culture. https://doi.org/10.1073/pnas.1112965109 hoodbhoy, p. (2016). playing the rankings game in pakistan _ the world view. jan-benedict, e.;steenkamp, m. (2001). the role of national culture in international marketing. international marketing review, 18(1), 30–44. kessels, j. w. m., & keursten, p. (2002). creating a knowledge productive work environment. lline, lifelong learning in europe, vii(2), 104–112. kjerholt, m., & hølge-hazelton, b. (2018). cultivating a culture of research in nursing through a journal club for leaders: a pilot study. journal of nursing management, 26(1), 42–49. https://doi.org/10.1111/jonm.12518 leidner, & kayworth. (2006). review: a review of culture in information systems research: imran naseem, saima imran, muhammad tahir & bilal bin saeed 88 toward a theory of information technology culture conflict. mis quarterly. https://doi.org/10.2307/25148735 lodhi, a. s. (2012). a pilot study of researching the research culture in pakistani public universities: the academics’ perspective. procedia social and behavioral sciences, 31(2011), 473–479. https://doi.org/10.1016/j.sbspro.2011.12.089 mapa, d. s. (2017). research culture in the philippines, 1–4. marchant, t. (2009). developing research culture overcoming regional and historical obstacles. professional doctorate research in australia: commentary and case studies from business, education and indigenous studies, 55–64. https://doi.org/10.1080/00220620.2012.659948 matres, i., oiva, m., & tolonen, m. (2018). in between research cultures – the state of digital humanities in finland. informaatiotutkimus, 37(2). https://doi.org/10.23978/inf.71160 mendez, e. h., & cruz, m. del r. r. (2014). research culture in higher education: the case of a foreign language department in mexico. profile: issues in teachers’ professional development, 16(2), 135. https://doi.org/10.15446/profile.v16n2.40819 notes, l. (2015). the culture of research in the uk, (november), 1–9. pratt, m., margaritis, d., & coy, d. (2007). developing a research culture in a university faculty. journal of higher education policy and management, 21(1), 43–55. https://doi.org/10.1080/1360080990210104 rosas, n. l. (2013). strengthening research culture in higher education. sgs research journal, 5–22. salazar-clemeña, r. m., & almonte-acosta, s. a. (2007). competition, cooperation and change in the academic profession: shaping higher education’s contribution to knowledge and research; developing research culture in philippine higher education institutions: perspectives of university faculty; 2007, (september). sarah, c. (2018). transforming research culture – policy postings_ ucl public policy blog – medium. university college london. schein, e. h. (1997). schein’s organizational culture & leadership. scott, s., miller, f., & lloyd, k. (2006). doing fieldwork in development geography: research culture and research spaces in vietnam. geographical research, 44(1), 28–40. https://doi.org/10.1111/j.1745-5871.2006.00358.x shalom, h. schwartz; maria, r. (1995). values in the west: a theoretical and empirical challenge to the individualism-collectivism cultural dimension. world psycology, 91– 122. shaukat, a. (2012, december 9). 10 years of hec_ a revolution in education or misplaced journal of applied economics and business studies, volume. 4, issue 2 (2020) 75-100 https://doi.org/10.34260/jaebs.424 89 priorities_ _ the express tribune. the tribune. retrieved from https://tribune.com.pk/ story/477192/10-years-of-hec-a-revolution-in-education-or-misplaced-priorities/ shi, y., & rao, y. (2010). china’s research culture. science, 329(5996), 1128. https://doi.org/ 10.1126/science.1196916 slade, s. c., philip, k., & morris, m. e. (2018). frameworks for embedding a research culture in allied health practice : a rapid review, 1–15. studman, c. (2003). growing a research culture. journal of research administration, 34(1), 19. retrieved from http://search.ebscohost.com/login.aspx?direct=true&db=a9h&an= 9609684&site=ehost-live teehankee, b. (n.d.). philippine higher education : a systems action research approach. vandenbos, g., humphreys, m., karlan, d., yarkoni, t., dafoe, a., alter, g., … contestabile, m. (2015). promoting an open research culture. science, 348(6242), 1422–1425. https://doi.org/10.1126/science.aab2374 imran naseem, saima imran, muhammad tahir & bilal bin saeed 90 graph 1: average of 23 years' r&d expenditure (% of gdp) 1996-2018, graph 2: average of 23 years' researchers in r&d/million people:1996-18, (wdi 2019) graph 3: if articles published from pakistan over the years, (wdi 2019) 3.9 3.4 3.2 3.1 3.0 2.8 2.6 2.6 2.6 2.5 0.8 0.3 0.2 0.1 0.0 1.0 2.0 3.0 4.0 5.0 7771 7377 6572 5649 5616 5126 5086 5005 4611 4281 157 148 121 61 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 68 178 228 355 524 579 1138 4452 7833 9050 10863 12413 0 2000 4000 6000 8000 10000 12000 14000 1975 1985 1995 2005 2015 2017 journal of applied economics and business studies, volume. 4, issue 2 (2020) 75-100 https://doi.org/10.34260/jaebs.424 91 graph 4, source: web of science analysis: isi, clarivate analytics graph 5: breakup of hec recognized universities in pakistan graph 6: breakup of management sciences phd producing universities 180 7 8 1 21 33 57 53 75 3 1 0 6 10 25 31 105 4 7 1 15 23 32 22 0 50 100 150 200 total universities azad jammu and kashmir balochistan gilgit baltistan islamabad capital territory khyber pakhtoonkhwa punjab sindh total private public 47 0 2 0 10 6 19 10 14 0 0 0 1 2 6 5 33 0 2 0 9 4 13 5 0 10 20 30 40 50 total azad jammu and kashmir balochistan gilgit baltistan islamabad capital territory khyber pakhtoonkhwa punjba sindh total private punjab imran naseem, saima imran, muhammad tahir & bilal bin saeed 92 graph 7: era-wise breakup of 180 universities graph 8: era-wise breakup of 47 management sciences phds producing universities graph 9: breakup of universities with and without phd output graph 10: breakup of universities with and without producing management sciences phds 1 5 58 116 0 0 28 47 1 5 30 69 0 50 100 150 till 1947 1948-1971 1972-98 1999-2016 total private public 1 4 25 17 0 0 7 7 1 4 18 10 0 10 20 30 till 1947 1948-71 1972-98 1999-2016 total private public 180 75 105 76 26 50 104 49 55 0 50 100 150 200 total private public total universities universities producing phds universities not producing phds 180 75 105 47 14 33 133 61 72 0 50 100 150 200 total private public total universities producing mgt phds not producing mgt phds journal of applied economics and business studies, volume. 4, issue 2 (2020) 75-100 https://doi.org/10.34260/jaebs.424 93 graph 11: region wise breakup of total phds graph 12: region-wise breakup of management sciences phds graph 13: era-wise distribution of 13259 phds graph 14: era-wise distribution of 3685 female phds 13259 27 92 0 2559 1267 5737 3577 495 0 0 0 51 120 43 281 12764 27 92 0 2508 1147 5694 3296 0 5000 10000 15000 total phd azad jammu & kashmir baluchistan gilgit baltistan islamabad capital territory khyber pakhtunkhwa punjab sindh total private public 678 0 6 0 216 89 212 155 143 0 0 0 51 24 38 30 535 0 6 0 165 65 174 125 0 200 400 600 800 total azad jammu and kashmir balochistan gilgit islamabad capital territory khyber pakhtoonkhwa punjba sindh total private public 29 282 2258 10690 0 0 2 49329 282 2256 10197 0 5000 10000 15000 till 1947 1948-71 1972-98 1999-2016 total phds private public 1 20 412 3252 0 0 1 1081 20 411 3144 0 1000 2000 3000 4000 till 1947 1948-71 1972-99 1999-2016 total private public imran naseem, saima imran, muhammad tahir & bilal bin saeed 94 graph 15: era-wise distribution of 9574 male phds graph 16: era-wise distribution of 678 management sciences phds graph 17: era-wise distribution of 136 female management sciences phds graph 18: era-wise distribution of 542 male management sciences phds 28 262 1846 7438 0 0 1 38528 262 1845 7053 0 2000 4000 6000 8000 till 1947 1948-71 1972-99 1999-2016 total private public 2 7 29 640 0 0 0 143 497 0 200 400 600 800 till 1947 1948-71 1972-98 1999-2016 total private public 0 0 8 128 0 0 0 21 0 0 8 107 0 50 100 150 till 1947 1948-71 1972-98 1999-2016 total private public 2 7 21 512 0 0 0 122 2 7 21 390 0 200 400 600 till 1947 1948-71 1972-98 1999-2016 total private public journal of applied economics and business studies, volume. 4, issue 2 (2020) 75-100 https://doi.org/10.34260/jaebs.424 95 graph 19: public private ratio of university phds graph 20: ratio of university to management sciences phds graph 21: 13259 phds produced by 76 universities 180 76 74 174 75 26 8 23 105 50 102 214 0 50 100 150 200 250 total universities phd producing universities ratio of over all universities ratio of phd producing universities total private public 180 76 4 9 75 26 2 5 105 50 5 11 0 50 100 150 200 total universities mgt phd producing universities ratio of total universities ratio of mgt phd producing universities total private public 2443 2188 1392 1338 778 655 495 342 314 278 247 189 165 164 147 142 137 132 126 123 114 1350 0 500 1000 1500 2000 2500 3000 imran naseem, saima imran, muhammad tahir & bilal bin saeed 96 graph 22: 495 phds produced by 26 private universities graph 23: major segments of universities with number of phds 96 54 51 44 41 34 26 25 22 19 12 11 10 10 8 7 6 4 4 3 2 2 1 1 1 1 0 20 40 60 80 100 120 4 2 15 10 24 21 7361 1433 3115 708 581 61 0 0 0 3 11 124 2 15 7 13 9 0 1000 2000 3000 4000 5000 6000 7000 8000 >= 1000 >=500&<1000 >=100&<500 >=50&100 >=10 & <50 <10 no. of universities phds private public journal of applied economics and business studies, volume. 4, issue 2 (2020) 75-100 https://doi.org/10.34260/jaebs.424 97 graph 24: distribution of 678 management sciences phds by 47 universities graph 25: distribution of 535 management sciences phds by 33 public universities 72 42 35 34 32 31 30 28 28 28 28 27 23 20 19 17 17 15 15 14 12 10 102 0 20 40 60 80 100 120 72 42 35 34 32 30 28 28 28 27 20 17 16 15 15 14 12 9 9 7 7 6 6 5 5 5 3 2 2 1 1 1 1 0 10 20 30 40 50 60 70 80 u n iv e rs it y o f k a ra ch i, … n a ti o n a l u n iv e rs it y o f… u n iv e rs it y o f p e sh a w a r, … u n iv e rs it y o f a g ri cu lt u re ,… b a h a u d d in z a ka ri ya … p a k is ta n i n st it u te o f… u n iv e rs it y o f th e p u n ja b , … q u a id -i -a za m u n iv e rs it y, … p m a s -a ri d a g ri cu lt u re … u n iv e rs it y o f s in d h ,… in te rn a ti o n a l is la m ic … s h a h a b d u l la ti f u n iv e rs it y, … t h e u n iv e rs it y o f… u n iv e rs it y o f e n g in e e ri n g & … f e d e ra l u rd u u n iv e rs it y o f… g o m a l u n iv e rs it y, d .i .k h a n . a lla m a i q b a l o p e n … c o m sa t s i n st it u te o f… u n iv e rs it y o f s a rg o d h a ,… g o ve rn m e n t c o ll e g e … n a ti o n a l u n iv e rs it y o f… g o ve rn m e n t c o ll e g e … s in d h a g ri cu lt u re … is la m ia u n iv e rs it y, … u n iv e rs it y o f b a lo ch is ta n ,… u n iv e rs it y o f v e te ri n a ry … m e h ra n u n iv re si ty o f… u n iv e rs it y o f e d u ca ti o n ,… b a h ri a u n iv e rs it y , i sl a m a b a d u n iv e rs it y o f e n g in e e ri n g & … f a ti m a j in n a h w o m e n … la h o re c o ll e g e f o r w o m e n … b a lo ch is ta n u n iv e rs it y o f… imran naseem, saima imran, muhammad tahir & bilal bin saeed 98 graph 26: distribution of 143 management sciences phds by 14 private universities graph 27: major segment of management sciences phd producing universities graph 28: distribution of 678 management sciences phds in subject relevancy 31 28 23 19 10 9 9 5 2 2 2 1 1 1 0 5 10 15 20 25 30 35 7 7 8 12 13 276 182 118 82 20 1 2 2 3 66 5 6 9 7 0 50 100 150 200 250 300 >= 30 phds >=20&<30 >=10&<20 >=5&<10 <5 universities phds private public 217 335 126 0 50 100 150 200 250 300 350 400 core subjects associated subjects mgt in other discilpines journal of applied economics and business studies, volume. 4, issue 2 (2020) 75-100 https://doi.org/10.34260/jaebs.424 99 graph 29: distribution of core management sciences phds graph 30: distribution of 34 female core management sciences phds graph 31: distribution of 183 male core management sciences phds 0 0 2 215 0 0 0 106 0 0 2 109 0 50 100 150 200 250 till 1947 1948-71 1972-98 1999-2016 total private public 0 0 0 34 0 0 0 16 0 0 0 18 0 10 20 30 40 till 1947 1948-71 1972-98 1999-2016 total private public 0 0 2 181 0 0 0 90 0 0 2 91 0 20 40 60 80 100 120 140 160 180 200 till 1947 1948-71 1972-98 1999-2016 total private public imran naseem, saima imran, muhammad tahir & bilal bin saeed 100 graph 32: number of supervisors with number of phds graph 33: number of supervisors with number of management sciences phds 4060 1030 448 223 360 100 4060 2060 1344 892 2241 1610 0 500 1000 1500 2000 2500 3000 3500 4000 4500 one two three four =>five =>ten supervisors phds 311 69 18 17 6 5 1 2 1 311 138 54 68 30 30 7 16 10 0 50 100 150 200 250 300 350 one two three four five six seven eight ten supervisors phds journal of applied economics and business studies, volume. 4, issue 2 (2020) 119-140 https://doi.org/10.34260/jaebs.426 119 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x enterprise resource planning systems and user performance in higher education institutions of pakistan abrar ullah 1,2*, rohaizat bin baharun1, muhammad yasir3 & khalil md nor 1 1 azman hashim international business school, universiti teknologi malaysia, johor, 81310, malaysia 2 department of management sciences, university of swabi, swabi, pakistan 3 department of management sciences, bacha khan university, charsadda, pakistan abstract the study is designed to assess the impact of enterprise resource planning (erp) systems on the performance perceived by users in higher education institutions (heis) of pakistan. this study sought to evaluate the effect of erps quality factors including information quality (iq), system quality (sq) and service quality (srq) on user performance (up) towards system usage through the role of perceived usefulness (pu), perceived ease of use (peou) and user satisfaction (us). consequently, a framework is proposed by integration of delone & mclean (d&m) information systems (is) success model and technology acceptance model (tam) to address the research questions. the study used quantitative research methodology and data were collected from 317 employees from eight universities in pakistan. structural equation modelling (sem) was performed using smartpls. the results indicated that sq, iq and srq has direct and positive effect on up, pu, peou and us. additionally, pu, peou and us are found to have influence on up. theoretically, the study contributes by integrating factors from d&m is success model and tam to investigate the effect of erp systems on up through pu, peou and us. practically, the study implied that practitioners needs to put efforts to provide a system which users perceive as useful and free of efforts. keywords information systems, erp systems, user performance, management information systems, higher education jel classification c32, c38, c52, c58 1. introduction the term enterprise resource planning (erp) was introduced by gartner group in 1990s (arif et al., 2004), comprising of “computer software systems that integrate all related processes within enterprise and provides users with services to manage all functions” (swartz et al., 2001). erp systems implementation brings great advantages (ullah et al., 2018; umar et al., 2016), but due to the uncertainties of technological complexities, the completion of these projects remains a challenge (xu et al., 2010). millions of dollars are invested on erp systems (beheshti et al., 2010), and despite of enormous growth, it is claimed that these systems failed * abrar.ullah@uoswabi.edu.pk abrar ullah, rohaizat bin baharun, & muhammad yasir, khalil md nor 120 with a higher rate of about to be 60 -90 percent (ahmad et al., 2014; alshamlan et al., 2011; gill et al., 2020). similarly, it is reported that the organization’s expenditure on erp systems implementation is about $6.1 million, of which 58% are cost overrun, 65% experienced schedule overrun, and in the post implementation stage 53% achieved less than 50-percent of anticipated measurable benefits (solutions, 2015). thus, making this a thorny problem deserving further exploration that how to achieve related outcome from implemented systems (sun et al., 2015). in the is related research, assessing its impact is a crucial concern (petter et al., 2008) and it is repeatedly reported as the main problem by organizational administrators (gable et al., 2008). practitioners as well as researchers yet to find answers for how to measure is successfully (rabaa'i et al., 2009). in this quest, several models and theories have emerged to measure and recognize technology and specifically is such as, the social cognitive theory (bandura, 1986), d&m is success model (delone et al., 1992, 2003), theory of reasoned action (tra), diffusion of innovation (doi) theory (rogers, 1995), theory of planned behavior, the unified theory of acceptance and use of technology (utaut) (venkatesh et al., 2003), tam (davis, 1989). however, with the continues innovation in technologies, the focus is shifted towards the outcome in terms of performance while evaluating system success (a. h. aldholay et al., 2018; isaac et al., 2017). within the erp environment, previous studies mentioned performance evaluation of these systems and only some of the reported articles discussed its impact on the productivity and performance (shen et al., 2016). reviewing the literature related to erp system performance shows that majority of prior studies measure the erp systems in terms of performance at organizational level rather than individual and the outcome variables discussed are product quality (banker et al., 2006), benefits (zhu et al., 2010), financial performance (ayman et al., 2015), organizational service enhancement (gorla et al., 2010), market value (ranganathan et al., 2006), process efficiency (chou et al., 2008), shareholder return (galy et al., 2014), competitive advantage (ram et al., 2014), organisational benefits (almahamid et al., 2015). thus, providing enough evidence to investigate from users perspective at individual level rather organizational level, that is investigating the actual impacts of erp systems on users (hsu et al., 2015), as users itself are consumers while receiving services from it department (alsaleh et al., 2016) and user performance studies are given less attention (ullah, baharun, nor, siddique, & sami, 2018). although, erp systems have implemented in large manufacturing organizations, but research in higher education is still limited (albarghouthi et al., 2020), especially in pakistan (ahmer et al., 2018; nizamani et al., 2014). in summary, the preceding background presents the motivation for the research, and it shows that studies tested these models in different areas with different stakeholders and systems. some studies adopted these models partially and tested specific constructs, while others tried to extend the models by adding additional constructs. keeping in view the above, journal of applied economics and business studies, volume. 4, issue 2 (2020) 119-140 https://doi.org/10.34260/jaebs.426 121 d & m model and tam integration is another route to evaluate antecedents that determine the erp systems success in higher education of pakistan. accordingly, objective of the study to propose a model based on d & m is success model representing quality factors and tam to examine whether quality factors as antecedents to acceptance and user satisfaction effect user performance while using erp system. 2. literature review and development of hypotheses 2.1 system quality in related literature to is, quality is somewhat “ill-defined” (nelson et al., 2005) and debate exists among authors that the quality factors are either self-defined or derived empirically (ullah, baharun, nor, siddique, & bhatti, 2018). rai et al. (2002) defined, system quality as “the degree to which a system is user friendly”. system quality is often found relevant construct and always found in support, while evaluating the matters of is (urbach et al., 2010) and is studied widely as explanatory construct in erp research (gorla et al., 2010; rajan et al., 2015; sternad et al., 2013). in studies, positive impact of sq on perceived usefulness has been found. among them lin (2010) found the influence of sq on pu and the study also indicated that sq can increase the belief about erp system usefulness. zhou (2014) investigated system quality influence on perceived usefulness to measure trust in mobile payment and found the effect significant. this relationship is also found positive in other related studies of the erp system environment (abugabah et al., 2015; ali et al., 2013; cheng, 2019; gorla et al., 2014; lin, 2010; tseng et al., 2018; yang et al., 2017; zhou, 2011). in their study on erp satisfaction and user adoption costa et al. (2016) found that sq have positive effect on perceived ease of use. other related studies also tested and confirmed the effect of sq on peou (abugabah et al., 2015; ali et al., 2013; lin, 2010; rana et al., 2015; yang et al., 2017). satisfied erp employees may be more efficient, when system usage is mandatory (hsu et al., 2015) as us refers to “the degree to which users believe that the is meets their requirements” (delone et al., 2003). according to tsai et al. (2012) sq is the key antecedent of user satisfaction in erp environment. therefore, we expect that the sq has a significant influence on us. prior studies backed this claim and showed significant effect of sq on us such as (cheng, 2019; costa et al., 2016; landrum et al., 2010; lin, 2010; masrek et al., 2016; noorman masrek et al., 2010; rana et al., 2015; shahibi et al., 2016; tam et al., 2016). based on these pieces of evidence, it is postulated that: h1: system quality positively influence perceived usefulness h2: system quality positively influence perceived ease of use h3: system quality positively influence user satisfaction abrar ullah, rohaizat bin baharun, & muhammad yasir, khalil md nor 122 2.2 information quality information quality (iq) is “the degree to which information generated possess content, accuracy, and format” (rai et al., 2002). higher information quality enhances performance of users with system usage, reinforcing the perceptions of usefulness (chen et al., 2015). researchers supported the effect of information quality on perceived usefulness like (abugabah et al., 2015); floropoulos et al. (2010). in similar environment zhou (2011) found the positive effect of iq on pu in the study of mobile web sites adoption. other researchers (alfarraj et al., 2017; chen et al., 2015; cheng, 2019; lin, 2010; tseng et al., 2018; e. s.-t. wang, 2016; zhou, 2014) also confirmed this effect. the effect of information quality on perceived ease of use was found by the study of ali et al. (2013) while evaluating is impact on user performance. their result confirmed that user performance is improved when system is useful and easy to use. this relationship is also explored in studies of (abugabah et al., 2015; rana et al., 2015; zhou, 2011) in different settings. information quality plays a vital role to satisfy users to achieve goals and it depends on the purpose of the users (y. s. wang, 2008). mohammadi (2015a) tested the influence of iq on user satisfaction to investigate user’s perceptions about e-learning system in universities and found this relationship significant. moreover, previous researchers found that iq is a key predictor of user satisfaction (chen et al., 2015; cheng, 2019; floropoulos et al., 2010; hsu et al., 2015; landrum et al., 2010; lin, 2010; masrek et al., 2016; mohammadi, 2015b; noorman masrek et al., 2010; shahibi et al., 2016; tam et al., 2016). these studies showed that satisfaction of users with the system depends on higher quality of information. hence, it is assumed that: h4: information quality positively influence perceived usefulness h5: information quality positively influence perceived ease of use h6: information quality positively influence user satisfaction 2.3 service quality petter et al. (2009) described service quality as “the quality of support received by users when interacting with the system”. srq is treated as peripheral to sq and iq, but recent research claimed that it can be seen as substantial construct due to the improved development in the service role of is (mohammadi, 2015a). in previous literature, the influence of service quality on perceived usefulness is confirmed as positive (ahn et al., 2007; chen et al., 2015; floropoulos et al., 2010; landrum et al., 2010; yang et al., 2017). in a study by zhou (2011), it is found that srq is a predictor to perceived ease of use and this relationship is also confirmed by ahn et al. (2007). the relationship of service quality with user satisfaction came from delone et al. (2003). the same relationship was tested by masrek et al. (2016) to investigate the determinant of us. journal of applied economics and business studies, volume. 4, issue 2 (2020) 119-140 https://doi.org/10.34260/jaebs.426 123 in erp related studies, several authors (alshibly, 2014; hsu et al., 2015; landrum et al., 2010; mohammadi, 2015a; noorman masrek et al., 2010; rana et al., 2015; shahibi et al., 2016; tam et al., 2016) confirmed the effect of srq on us. thus, srq is assumed to have the same effect: h7: service quality positively influence perceived usefulness h8: service quality positively influence perceived ease of use h9: service quality positively influence user satisfaction 2.4 relationship between pu, peou, us to user performance the constructs perceived usefulness and perceived ease of use came from tam and this part represents the acceptance of technology. pu and peou of tam are important elements to decide the acceptance of technology of the system (hsieh et al., 2013). the success of the system also relies on satisfaction level towards the system (y.-s. wang et al., 2008). several studies supported these relationships, for example, abugabah et al. (2015) confirmed that pu influences the performance of users. the same relationship have been found positive and significant in studies by (ali et al., 2013; chen et al., 2015). similarly, the effect of peou on user performance in an erp environment was also found in studies (abugabah et al., 2015; ali et al., 2013). lastly, the influence of user satisfaction on the up is confirmed positive and significant in previous studies (chen et al., 2015; hsu et al., 2015; noorman masrek et al., 2010; tam et al., 2016). based on the preceding paragraph, the same outcome is expected. hence the following is postulated. h10: perceived usefulness positively influence user performance h11: perceived ease of use positively influence user performance h12: user satisfaction positively influence user performance 2.5 user performance scholars used organizational impact (gorla et al., 2014), individual impact (ifinedo et al., 2010), system usage (lin, 2010), satisfaction (floropoulos et al., 2010; landrum et al., 2010), organizational performance (choi et al., 2013), perceived net benefits (chen et al., 2015) as dependable variable to measure erp systems. however, with growth in technology, the focus is diverted to the outcome as performance to measure the success (abugabah et al., 2015; a. aldholay et al., 2018; montesdioca et al., 2015). user performance is referred to the outcome of doing a set of tasks (alfarraj et al., 2017; ali et al., 2013; ullah et al., 2018). in this study, user performance is measured with questions related to efficiency and effectiveness. according to abugabah et al. (2015) efficiency is “the extent to which the method of performance minimizes the efforts that are required to perform the task”, and the effectiveness is “the degree of the objective accomplishment showing how well a set of results could be accomplished”. in this study context, users are individuals who use erp system for daily job, have some abrar ullah, rohaizat bin baharun, & muhammad yasir, khalil md nor 124 knowledge about the functionality of the system, and are familiar with other users (liu et al., 2011). 3 research methods 3.1 overview of proposed framework in the current study, the constructs and the hypothesized relationships among them are derived from the literature as preceding sections. the framework is shown in figure 1, where it is mentioned that the erp quality factors (system quality, information quality, service quality) influence perceived usefulness, perceived ease of use and user satisfaction, which further predict user performance. the quality factors are derived from delone et al. (2003), perceived usefulness and perceived ease of use from davis (1989) and user performance is taken from abugabah et al. (2015). the proposed framework set to test 12 hypotheses. figure 1: study framework 3.2 development of instrument in line with existing literature, a total of 51 items questionnaire was developed to capture the constructs. a 5-point likert scale is employed for each item of the constructs, with 1 as strongly disagree, 5 strongly agree, and 3 as not applicable (na). all the items in the questionnaire are adapted from related studies as shown in table 1 and appendix b. journal of applied economics and business studies, volume. 4, issue 2 (2020) 119-140 https://doi.org/10.34260/jaebs.426 125 table 1: construct measurement 4 data collection as discussed, the instruments for this study are adapted from established sources but pretesting is still needed to ensure that each question fits well with a different set of respondents (kumar et al., 2013). the process of pretesting was conducted by a panel comprising of renowned academicians in the field. each questionnaire item was examined, and cases of ambiguous wording are rectified and rephrased. this process fine-tunes the research instrument based on their feedback and recommendations such as tangibility dimension of service quality was removed as most of the questions remained similar to system quality. hereafter, the data was collected through google form. the questionnaire was sent by email to all employees in 8 universities in pakistan. the respondents are erp system users with the name of campus management solutions (cms) system. in total, we collected 356 responses from 8 universities. after the completion of initial data screening including missing values and outlier a total of 317 usable sample were retained. the respondents’ demography is presented in table 2. table 2: respondents demographics demographic feature frequency percentage gender male 247 77.9 female 70 22.1 age above 21 and below 25 years 18 5.7 above 25 and below 30 years 38 12.0 above 30 and below 35 years 107 33.8 above 35 and below 40 years 61 19.2 40 and above 93 29.3 education higher secondary school certificate (hssc) 3 0.9 graduation (14 years) 10 3.2 master (16 years) 45 14.2 mphil/ms (18 years) 131 41.3 construct items source system quality sq1 – sq9 (hsu et al., 2015); (gable et al., 2008); (nelson et al., 2005) information quality iq1 – iq6 (hsu et al., 2015); (gable et al., 2008); (nelson et al., 2005) service quality reliability srq1 – srq4 (hsu et al., 2015); (pitt et al., 1995); (parasuraman et al., 1988) responsiveness srq5 – srq7 assurance srq8 – srq11 empathy srq12 – srq15 perceived usefulness pu1 – pu4 (abugabah et al., 2015) perceived ease of use peou1 – peou4 (abugabah et al., 2015) user satisfaction us1 – us3 (lin, 2010) user performance up1 – up10 (abugabah et al., 2015) abrar ullah, rohaizat bin baharun, & muhammad yasir, khalil md nor 126 doctorate 128 40.4 experience less than 3 years 61 19.2 3 to 6 years 78 24.6 7 to 10 years 78 24.6 more than 10 years 100 31.5 usage once a week 74 23.3 once a day 69 21.8 several times a day 51 16.1 regular use, many times a day 123 38.8 5 data analysis and results structure equation model (sem) technique was used for hypotheses testing. two-stage approach was followed as recommended by hair et al. (2017) to examine the measurement model and structural model using smartpls 3.0 software (c. ringle et al., 2015; c. m. ringle et al., 2020). 5.1 measurement model assessment the measurement model stage determines the reliability and validity of the constructs. this includes convergent, discriminant validity and reliability. convergent validity is “the degree to which a measure correlates positively with alternative measures of the same construct” (hair et al., 2017). for factors loading the recommended value is ≥ 0.708 but loading greater than 0.6 or 0.5 is adequate (ramayah et al., 2018). on the basis of analysis 7 items found to be short of the required level are deleted. the values of composite reliability (cr) shown in table 3 were used for construct reliability. as shown cr ranging between 0.796 and 0.929 are over the recommended value of 0.70 by gefen et al. (2000); kline (2010). average variance extracted (ave) criterion was employed to determine convergent validity. ave ranging between 0.602 and 0.711, meeting the recommended threshold above 0.50 (hair et al., 2010). table 3: reliability and loading construct item loading (>0.5) cr ave system quality (sq) sq2 0.686 0.896 0.612 sq3 0.777 sq4 0.791 sq5 0.781 sq6 0.650 sq8 0.727 sq9 0.785 information quality (iq) iq1 0.776 0.929 0.685 iq2 0.819 iq3 0.849 iq4 0.874 iq5 0.851 iq6 0.794 service quality (srq) reliability (rel) srq1 0.662 0.796 0.610 journal of applied economics and business studies, volume. 4, issue 2 (2020) 119-140 https://doi.org/10.34260/jaebs.426 127 discriminant validity is “the extent to which a construct differs from other constructs in the research model, that is the construct measures what is intended to measure” (hair et al., 2017). discriminant validity was determined using three criteria namely fornell and lacker’s criterion (fornell et al., 1981), items cross loading and recently developed heterotrait-monotrait ratio (htmt) (henseler et al., 2015). fornell-larcker’s criterion is the comparison of square root of ave values with construct correlations (hair et al., 2017), in other words, the construct share more variance with its own block of items compare to other construct. table 4 presents discriminant validity using this criterion is established as the values of each construct’s square root (values presented in bold) are larger than the correlation value between construct. srq2 0.750 srq3 0.837 responsive (resp) srq5 0.813 0.872 0.694 srq6 0.864 srq7 0.821 assurance (assu) srq9 0.693 0.810 0.602 srq10 0.802 srq11 0.802 empathy (emp) srq12 0.740 0.832 0.624 srq13 0.779 srq14 0.847 perceived usefulness (pu) pu1 0.813 0.908 0.711 pu2 0.841 pu3 0.880 pu4 0.836 perceived ease of use (peou) peou1 0.686 0.838 0.620 peou2 0.621 peou3 0.860 peou4 0.820 user satisfaction (us) us1 0.843 0.829 0.618 us2 0.721 us3 0.789 user performance (up) up1 0.723 0.874 0.650 up2 0.612 up3 0.704 up4 0.580 up5 0.798 up6 0.676 up7 0.724 up8 0.655 all loadings are statistically significant (p<0.01) abrar ullah, rohaizat bin baharun, & muhammad yasir, khalil md nor 128 table 4: fornell-larcker’s criterion assu emp iq peou pu rel resp sq up us assu 0.776 emp 0.512 0.790 iq 0.495 0.519 0.828 peou 0.518 0.542 0.642 0.787 pu 0.462 0.312 0.504 0.542 0.843 rel 0.592 0.496 0.509 0.303 0.461 0.781 resp 0.517 0.527 0.407 0.430 0.447 0.578 0.833 sq 0.584 0.557 0.434 0.417 0.428 0.524 0.530 0.782 up 0.557 0.435 0.420 0.532 0.320 0.506 0.429 0.334 0.806 us 0.582 0.550 0.462 0.320 0.416 0.538 0.544 0.320 0.548 0.786 the second approach is cross loading, “that is outer loading of each indicator on the associated construct should be higher than the cross loadings on the other construct” (hair et al., 2017). as shown in appendix a the values of the indicators have loaded highly on their respected construct. thus, providing adequate evidence of fulfilling the convergent validity requirement. 5.2 structural model assessment as per hair et al. (2017) the structure model assessment includes: assessment of collinearity by evaluating the predictor constructs, assessing the significance and relevance of the path coefficients representing hypotheses among constructs, coefficient of determination (r2), effect size (f2), and assess the predictive relevance (q2). for this study, srq construct is higher order construct (hoc) with 4 sub-dimensions. hoc assessment is necessary to determine whether their first order (lower order) constructs load onto their respective second order (higher order) construct. to achieve hoc assessment, the researcher employed the two stage higher order construct modeling approach by (becker et al., 2012; c. m. ringle et al., 2012; wilson, 2010). the first stage is repeated indicator approach. in this approach hoc’s measurement model is represented by assigning all the indicators of the first order construct to the hoc. the second stage is using latent variable scores as indicator representing the first order construct in the final structural model (hair et al., 2017). the pls-sem produced latent variable scores for each construct and saved for further analysis. 5.2.1 hypotheses tests table 5 gives the hypothesized direct relationships results. specifically, sq (β = 0.474, t = 7.797, p< 0.01) has significance influence on pu. similarly, sq has significance relationship with peou (β = 0.381, t = 6.388, p< 0.01) and the results further confirmed the positive relationship between sq and us (β = 0.396, t = 6.740, p< 0.01). thus, providing support for h1, h2 and h3. furthermore, iq has significance influence on pu (β = 0.131, t = 3.124, p< 0.01), peou (β = 0.227, t = 4.441, p< 0.01) and us (β = 0.219, t = 4.302, p< 0.01). therefore, h4, h5 and h6 are supported. figure 2 shows structural model assessment. journal of applied economics and business studies, volume. 4, issue 2 (2020) 119-140 https://doi.org/10.34260/jaebs.426 129 figure 2: pls algorithm of structural model likewise, srq has positive and significance relationship with pu (β = 0.192, t = 3.017, p< 0.01), peou (β = 0.208, t = 3.546, p< 0.01) and us (β = 0.237, t = 3.933, p< 0.01). hence, providing enough evidence to support hypotheses h7, h8, and h9. the results provide further indication that pu (β = 0.304, t = 9.760, p< 0.01) exhibits positive and significance influence on up. in addition, peou (β = 0.227, t = 3.555, p< 0.01) has positive and significance effect on up. lastly, us (β = 0.439, t = 7.566, p< 0.01) shows significance effect on up. therefore, hypotheses h10, h11, and h12 are supported. based on the table 5, r2 value (0.826) for user performance imply that the combination of constructs: system quality, information quality, service quality, perceived usefulness, perceived ease of use and user satisfaction explain 82% of the variance in user performance. similarly, system quality, information quality and service quality jointly contribute 55% variance in perceived usefulness. the same set of predictors i.e. sq, iq, srq account for 56% and 60% variance in perceived ease of use and user satisfaction respectively. in conclusion, as per the recommendation of chin (1998) the results shows that the predictive values for user performance can be considered as substantial, while perceived usefulness, perceived ease of use and user satisfaction qualify as moderate. abrar ullah, rohaizat bin baharun, & muhammad yasir, khalil md nor 130 table 5: summary of results of hypotheses hyp. path () se t value p value decision r2 f2 q2 h1 sq -> pu 0.476 0.061 7.797 0.000 supported 0.553 0.133 0.360 h2 sq -> peou 0.381 0.060 6.388 0.000 supported 0.557 0.086 0.289 h3 sq -> us 0.396 0.059 6.740 0.000 supported 0.607 0.105 0.347 h4 iq -> pu 0.131 0.042 3.124 0.001 supported 0.017 h5 iq -> peou 0.227 0.051 4.441 0.000 supported 0.053 h6 iq -> us 0.219 0.051 4.302 0.000 supported 0.055 h7 srq -> pu 0.192 0.064 3.017 0.001 supported 0.028 h8 srq -> peou 0.208 0.059 3.546 0.000 supported 0.033 h9 srq -> us 0.237 0.060 3.933 0.000 supported 0.048 h10 pu -> up 0.304 0.031 9.760 0.000 supported 0.826 0.208 0.345 h11 peou -> up 0.227 0.064 3.555 0.000 supported 0.032 h12 us -> up 0.439 0.058 7.566 0.000 supported 0.140 note: p<0.01 the effect size (f2) for the study is also assessed. the effect size (f2) is “a measure used to assess the relative impact of a predictor construct on an endogenous construct” (cohen, 1988). the results in table 5 indicate that sq has medium effects on pu (0.133), small effect on peou (0.086), and us (0.102). information quality has small effects on pu (0.017) peou (0.053), us (0.055). similarly, service quality exhibited small effect on pu (0.028), peou (0.033), and us (0.048). the result also indicated that the effect of pu on up (0.208) is medium. lastly, peou (0.032) and us (0.140) has small effect on up. in terms of predictive relevance, the blindfolding procedure was employed to measure predictive relevance. predictive relevance (q2) represents “how well the data collected empirically can be reconstructed with the help of the model and the pls parameters” (fornell et al., 1994). as per ramayah et al. (2018) and hair et al. (2017) the recommended q2 value should be larger than zero for dependent construct in structural model. the recommended q2 values of 0.02 (small), 0.15 (medium) and 0.35 (large) indicate the level of predictive relevance for endogenous construct (hair et al., 2017). table 5 presents that the q2 values for all endogenous constructs are above zero, clearly indicate that the model has predictive relevance. thus, the result shows that one endogenous variable has a large effect and the remaining have medium predictive relevance. 6 discussions the aim of the study is to evaluate the effect of quality factors on user performance. on integration of d&m is success model and tam, the study proposes and tests a framework. the result revealed that system quality positively effect perceived usefulness, perceived ease of use and user satisfaction. the effect of sq on perceived usefulness (h1) is found significant and in line with studies of (alfarraj et al., 2017; ali et al., 2013; cheng, 2019; gorla et al., 2010; lin, 2010; tseng et al., 2018; zhou, 2011, 2014). h2 is found supported, postulated as sq positively effect peou. this relationship provides further support to the previous studies (costa et al., 2016; rana et al., 2015; yang et al., 2017; zhou, 2011). another hypothesis as the effect on sq on us (h3), was found supported and consistent with prior research in similar journal of applied economics and business studies, volume. 4, issue 2 (2020) 119-140 https://doi.org/10.34260/jaebs.426 131 context (cheng, 2019; masrek et al., 2016; ojo, 2017; tam et al., 2016). this imply that sq is important antecedent to the perception of usefulness, ease of use and satisfaction. in this study context, system accuracy, system easiness, easy to learn are important considerations for users to perceive the system as useful, free of efforts and be satisfied. with regards to information quality relationships with pu, peou and us, three hypotheses h4, h5, h6 were tested. consistent with previous research the influence of information quality on perceived usefulness (alfarraj et al., 2017; chen et al., 2015), perceived ease of use (abugabah et al., 2015; rana et al., 2015; zhou, 2011) and user satisfaction (chatterjee et al., 2018; cheng, 2019; hsu et al., 2015; lin, 2010; masrek et al., 2016; sharma et al., 2019; tam et al., 2016; urbach et al., 2010) is confirmed. in this study context in higher education, users find the system as useful, ease of use when the obtained information from the system are clear, readable, well formatted and concise. higher quality of information increases the satisfaction with the system by balancing the system output and user requirements. the relationship of service quality with perceived usefulness, perceived ease of use and user satisfaction were postulated as hypotheses h7, h8, h9 and found supported. the result supported the effect of srq on pu and confirmed the findings of (ahn et al., 2007; chen et al., 2015; floropoulos et al., 2010; landrum et al., 2010; yang et al., 2017; zhou, 2011). this imply that srq is key antecedent to pu. when users get high quality service from support staff, then the system is perceived as more useful. similarly, in line with the studies of (ahn et al., 2007; lin, 2015) the result shows that srq has positive influence on peou. the system is perceived to be free of efforts when it staff provide timely service, shows sincere efforts to solve their problems, and the services are reliable and error free. moreover, the users will find the system as ease of use when it staff boost their confidence, feels safe while dealing with it staff, find the it staff knowledgeable, and it staff provide them individual attention. the result also reveals that srq has significant influence on user satisfaction. as expected the result supported previous studies (floropoulos et al., 2010; hsu et al., 2015; landrum et al., 2010; masrek et al., 2016; mohammadi, 2015a; noorman masrek et al., 2010; rana et al., 2015; sharma et al., 2019; tam et al., 2016). the findings imply that service quality strongly contributes to the satisfaction level of users upon providing services to them. in this study context, when the services provided by it staff are reliable, timely and boost confidence to them, then users feel satisfaction with the system. this study also tested the effect of perceived usefulness, perceived ease of use and user satisfaction as antecedents to user performance and postulated as h10, h11, and h12. consistent with the previous work (abugabah et al., 2015; ali et al., 2013; chen et al., 2015), the effect of pu on user performance is confirmed. the result shows that pu is strong proxy for user performance, that is the perception of the system’s usefulness in terms of job performance. in the same vein the effect of perceived ease of use on up is confirmed and abrar ullah, rohaizat bin baharun, & muhammad yasir, khalil md nor 132 implies that the system perceived as free of efforts have impact on the performance on users. lastly, the findings confirmed the significant influence of user satisfaction on user performance. previously, positive effect of us on user performance was found in literature (a. h. aldholay et al., 2018; choi et al., 2013; noorman masrek et al., 2010). this implies that user satisfaction is a key factor of user performance, as satisfied users with the system leads to better performance. 7 theoretical implications from the theoretical angle, this work provides several contributions in erp system area. the first contribution is to bring together the quality factors of d&m is success model and confirmed that each have effect on perceived usefulness, perceived ease of use and user satisfaction. although, service quality has been part of updated d&m is success model since 2003 but most of the research focuses on the relationships of sq and iq with different constructs to measure system success. in current study service quality is used to captures user’s assessment of services delivered by it staff as they collect holistic view of service quality while interacting with them. second, the study incorporates perceived usefulness, perceived ease of use and user satisfaction. in erp systems evaluation studies these effects are tested singularly or in combination involving other antecedents. thus, according to researcher knowledge, this is the first empirical work to measure the impact of the erp systems on through pu, peou and us in a framework involving sq, iq, and srq as their antecedents. this confirms the importance of tam as pu and peou facilitates user performance in the presence of quality factors in this study context. hence, incorporating erp quality factors of d&m is success model with pu and peou of tam can capture complete picture to evaluate erp system. contextually, the work generally contributes to the body of knowledge on user performance and particularly to higher education from the perspective of users. mostly, in higher education the focus remained on technical aspects of the systems. thus, the study contributes to literature on higher education, especially higher education of pakistan. 8 practical implications in terms of practical contribution, one of the implications is the establishment that perceived usefulness and perceived ease of use were shown vital impact on user performance. this shows that users put emphasis on usefulness and easy to use. this provides practitioners the opportunity to help users to realize the benefits of the system with regards to usefulness and ease of use. therefore, the practitioners need to put efforts to provide a system which user perceive as useful and free of efforts, that is to produce a system that make their job easy, to enhance their productivity, easy to use, easy to learn, understandable. journal of applied economics and business studies, volume. 4, issue 2 (2020) 119-140 https://doi.org/10.34260/jaebs.426 133 in case of user satisfaction, it can be concluded that us plays the role between the relationship of system quality and information quality and service quality with user performance. hence, the practitioners need to ensure that users are satisfied with the functional features, output of the system and services provided to them. another major conclusion is that among all quality predictors the results suggest that system quality is having great impact. the possible conclusion is that users place more emphasis on system quality which represent functional features of the system. the practitioners need to put efforts on the functional features of the system. 9 limitations and future recommendations the first the limitation to be considered is that the study proposes and tests a new framework in erp environment. the data for this scholarly work was collected and analyzed from a subsect of universities of pakistan. regardless of the significant relationships among constructs, these results may not be generalized to other sectors. the constructs used in this study and their relationships can be used for further investigation in different industries. second, the framework of the current study is based on the constructs from d & m is success model and tam. therefore, other factors related to users such as user resistance, user characteristics influencing user performance may be explored by extended this framework. third, the target population were the users regardless of their designations. hence, researchers may target users in different layers such as academic and non-academic staff. 10 conclusions upon the integration of the d&m is success model and tam, the study proposes and tested a framework to test system’s effect on users. the framework clarifies how erp quality factors (system quality, information quality, service quality) influence perceived usefulness, perceived ease of use and user satisfaction from user’s perspective. the study goes further to investigate the effect of perceived usefulness, perceive ease of use and user satisfaction on user performance. the results showed that significance influence exists from exogenous constructs on user performance in erp system environment. the findings also established the importance of perceived usefulness and perceived ease of use representing tam’s portion in erp environment in higher education’s institutions of pakistan. references abrar ullah, rohaizat bin baharun, & muhammad yasir, khalil md nor 134 abugabah, a., sanzogni, l., & alfarraj, o. (2015). evaluating the impact of erp systems in higher education. the international journal of information and learning technology, 32(1), 45-64. ahmad, n., haleem, a., & ali syed, a. (2014). study of reasons for enterprise systems adoption among indian organizations. journal of enterprise information management, 27(6), 696-718. ahmer, z., demir, e., tofallis, c., & asad, h. (2018). usage of enterprise resource planning systems in higher education institutions in pakistan. doi:https://uhra.herts.ac.uk/handle/2299/19625 ahn, t., ryu, s., & han, i. (2007). the impact of web quality and playfulness on user acceptance of online retailing. information & management, 44(3), 263-275. albarghouthi, m., qi, b., wang, t. c., & abbad, m. (2020). erp adoption and acceptance in saudi arabia higher education: a conceptual model development. international journal of emerging technologies in learning(15). aldholay, a., isaac, o., abdullah, z., abdulsalam, r., & al-shibami, a. h. (2018). an extension of delone and mclean is success model with self-efficacy: online learning usage in yemen. the international journal of information and learning technology, 35(4), 285-304. aldholay, a. h., isaac, o., abdullah, z., & ramayah, t. (2018). the role of transformational leadership as a mediating variable in delone and mclean information system success model: the context of online learning usage in yemen. telematics and informatics, 35(5), 1421-1437. alfarraj, o., & abugabah, a. (2017). extending information system models to the health care context: an empirical study and experience from developing countries. international arab journal of information technology, 14(2), 159-167. ali, b. m., & younes, b. (2013). the impact of information systems on user performance: an exploratory study. journal of knowledge management, economics and information technology, 3(2), 128-154. almahamid, s., & awsi, o. (2015). perceived organizational erp benefits for smes: middle eastern perspective. interdisciplinary journal of information, knowledge, and management, 10. alsaleh, i., & bageel, m. (2016). measuring user satisfaction with service quality of it department support as perceived by the users: case study of service industry sector in jeddah, saudi arabia. international journal of liberal arts and social science, 4(1), 65-82. alshamlan, h. m., & almudimigh, a. s. (2011). the chang management strategies and processes for successful erp implementation: a case study of madar. international journal of computer science, 8(2), 399-407. alshibly, h. h. (2014). evaluating e-hrm success: a validation of the information systems success model. international journal of human resource studies, 4(3), 107. arif, m., kulonda, d. j., proctor, m., & williams, k. (2004). before you invest: an illustrated framework to compare conceptual designs for an enterprise information system. information knowledge systems management, 4(2), 119-135. ayman, b., & kamaljeet, s. (2015). factors influencing the acceptance of enterprise resource planning system (erp) and financial performance of saudi arabia listed companies: multivariate data analysis using structural equation modeling (sem). international journal of business and management review, 3(4), 93-118. bandura, a. (1986). social foundations of thought and action: a social cognitive theory: englewood cliffs, nj, us: prentice-hall, inc. banker, r. d., bardhan, i. r., chang, h., & lin, s. (2006). plant information systems, manufacturing capabilities, and plant performance. mis quarterly, 315-337. becker, j.-m., klein, k., & wetzels, m. (2012). hierarchical latent variable models in pls-sem: guidelines for using reflective-formative type models. long range planning, 45(5-6), 359-394. beheshti, h. m., & beheshti, c. m. (2010). improving productivity and firm performance with enterprise resource planning. enterprise information systems, 4(4), 445-472. https://uhra.herts.ac.uk/handle/2299/19625 journal of applied economics and business studies, volume. 4, issue 2 (2020) 119-140 https://doi.org/10.34260/jaebs.426 135 chatterjee, s., kar, a. k., & gupta, m. (2018). success of iot in smart cities of india: an empirical analysis. government information quarterly. chen, j. v., jubilado, r. j. m., capistrano, e. p. s., & yen, d. c. (2015). factors affecting online tax filing–an application of the is success model and trust theory. computers in human behavior, 43, 251-262. cheng, y.-m. (2019). a hybrid model for exploring the antecedents of cloud erp continuance: roles of quality determinants and task-technology fit. international journal of web information systems, 15(2), 215-235. chin, w. w. (1998). issues and opinion on structural equation modelling. management information systems quarterly, 22(1), 1-8. choi, w., rho, m. j., park, j., kim, k. j., kwon, y. d., & choi, i. y. (2013). information system success model for customer relationship management system in health promotion centers. healthcare informatics research, 19(2), 110-120. chou, s.-w., & chang, y.-c. (2008). the implementation factors that influence the erp (enterprise resource planning) benefits. decision support systems, 46(1), 149-157. cohen, j. (1988). statistical power analysis for the behavioral sciencesnew jerseylawrence erlbaum associates. inc. publishers. costa, c. j., ferreira, e., bento, f., & aparicio, m. (2016). enterprise resource planning adoption and satisfaction determinants. computers in human behavior, 63, 659-671. davis, f. d. (1989). perceived usefulness, perceived ease of use, and user acceptance of information technology. mis quarterly, 319-340. delone, w. h., & mclean, e. r. (1992). information systems success: the quest for the dependent variable. information systems research, 3(1), 60-95. delone, w. h., & mclean, e. r. (2003). the delone and mclean model of information systems success: a ten-year update. journal of management information systems, 19(4), 9-30. floropoulos, j., spathis, c., halvatzis, d., & tsipouridou, m. (2010). measuring the success of the greek taxation information system. international journal of information management, 30(1), 47-56. fornell, c., & cha, j. (1994). partial least squares. advanced methods of marketing research, 407(3), 52-78. fornell, c., & larcker, d. f. (1981). structural equation models with unobservable variables and measurement error: algebra and statistics. journal of marketing research, 382-388. gable, g. g., sedera, d., & chan, t. (2008). re-conceptualizing information system success: the isimpact measurement model. journal of the association for information systems, 9(7), 377408. galy, e., & sauceda, m. j. (2014). post-implementation practices of erp systems and their relationship to financial performance. information & management, 51(3), 310-319. gefen, d., straub, d., & boudreau, m.-c. (2000). structural equation modeling and regression: guidelines for research practice. communications of the association for information systems, 4(1), 7. gill, a. a., amin, s., & saleem, a. (2020). investigation of critical factors for successful erp implementation: an exploratory study. journal of business and social review in emerging economies, 6(2), 565-575. gorla, n., & somers, t. m. (2014). the impact of it outsourcing on information systems success. information & management, 51(3), 320-335. gorla, n., somers, t. m., & wong, b. (2010). organizational impact of system quality, information quality, and service quality. the journal of strategic information systems, 19(3), 207-228. abrar ullah, rohaizat bin baharun, & muhammad yasir, khalil md nor 136 hair, j. f., anderson, r. e., babin, b. j., & black, w. c. (2010). multivariate data analysis: a global perspective (vol. 7): pearson upper saddle river, nj. hair, j. f., hult, g. t. m., ringle, c., & sarstedt, m. (2017). a primer on partial least squares structural equation modeling (pls-sem) (2nd edition ed.): sage publications. henseler, j., ringle, c. m., & sarstedt, m. (2015). a new criterion for assessing discriminant validity in variance-based structural equation modeling. journal of the academy of marketing science, 43(1), 115-135. hsieh, p., huang, c., & yen, d. c. (2013). assessing web services of emerging economies in an eastern country taiwan's e-government. government information quarterly, 30(3), 267-276. hsu, p.-f., yen, h. r., & chung, j.-c. (2015). assessing erp post-implementation success at the individual level: revisiting the role of service quality. information & management, 52(8), 925942. ifinedo, p., rapp, b., ifinedo, a., & sundberg, k. (2010). relationships among erp postimplementation success constructs: an analysis at the organizational level. computers in human behavior, 26(5), 1136-1148. isaac, o., abdullah, z., ramayah, t., & mutahar, a. m. (2017). internet usage, user satisfaction, tasktechnology fit, and performance impact among public sector employees in yemen. the international journal of information and learning technology, 34(3), 210-241. kline, r. b. (2010). principles and practice of structural equation modeling: guilford press. kumar, m., talib, s. a., & ramayah, t. (2013). business research methods. oxford: oxford university press. landrum, h., prybutok, v. r., & zhang, x. (2010). the moderating effect of occupation on the perception of information services quality and success. computers & industrial engineering, 58(1), 133-142. lin, h. f. (2010). an investigation into the effects of is quality and top management support on erp system usage. total quality management, 21(3), 335-349. lin, h. f. (2015). the impact of company-dependent and company-independent information sources on organizational attractiveness perceptions. journal of management development, 34(8), 941959. liu, l., feng, y., hu, q., & huang, x. (2011). from transactional user to vip: how organizational and cognitive factors affect erp assimilation at individual level. european journal of information systems, 20(2), 186-200. masrek, m. n., & gaskin, j. e. (2016). assessing users satisfaction with web digital library: the case of universiti teknologi mara. the international journal of information and learning technology, 33(1), 36-56. mohammadi, h. (2015a). factors affecting the e-learning outcomes: an integration of tam and is success model. telematics and informatics, 32(4), 701-719. mohammadi, h. (2015b). investigating users’ perspectives on e-learning: an integration of tam and is success model. computers in human behavior, 45, 359-374. montesdioca, g. p. z., & maçada, a. c. g. (2015). measuring user satisfaction with information security practices. computers & security, 48, 267-280. nelson, r. r., todd, p. a., & wixom, b. h. (2005). antecedents of information and system quality: an empirical examination within the context of data warehousing. journal of management information systems, 21(4), 199-235. nizamani, s., khoumbati, k., ismaili, i. a., & nizamani, s. (2014). a conceptual framework for erp evaluation in universities of pakistan. sindh university research journal, 45(3), 467-475. noorman masrek, m., jamaludin, a., & awang mukhtar, s. (2010). evaluating academic library portal effectiveness: a malaysian case study. library review, 59(3), 198-212. journal of applied economics and business studies, volume. 4, issue 2 (2020) 119-140 https://doi.org/10.34260/jaebs.426 137 ojo, a. i. (2017). validation of the delone and mclean information systems success model. healthcare informatics research, 23(1), 60-66. parasuraman, a., zeithaml, v. a., & berry, l. l. (1988). servqual: a multiple-item scale for measuring consumer perc. journal of retailing, 64(1), 12. petter, s., delone, w., & mclean, e. (2008). measuring information systems success: models, dimensions, measures, and interrelationships. european journal of information systems, 17(3), 236-263. petter, s., & mclean, e. r. (2009). a meta-analytic assessment of the delone and mclean is success model: an examination of is success at the individual level. information & management, 46(3), 159-166. pitt, l. f., watson, r. t., & kavan, c. b. (1995). service quality: a measure of information systems effectiveness. mis quarterly, 173-187. rabaa'i, a. a., bandara, w., & gable, g. (2009). erp systems in the higher education sector: a descriptive study. paper presented at the 20th australasian conference on information systems. rai, a., lang, s. s., & welker, r. b. (2002). assessing the validity of is success models: an empirical test and theoretical analysis. information systems research, 13(1), 50-69. rajan, c. a., & baral, r. (2015). adoption of erp system: an empirical study of factors influencing the usage of erp and its impact on end user. iimb management review, 27(2), 105-117. ram, j., wu, m.-l., & tagg, r. (2014). competitive advantage from erp projects: examining the role of key implementation drivers. international journal of project management, 32(4), 663-675. ramayah, t., cheah, j., chuah, f., ting, h., & mumtaz, a. m. (2018). partial least squares structural equation modeling (pls-sem) using smartpls 3.0: an updated and practical guide to statistical analysis. (2nd edition ed.). pearson, malaysia. rana, n. p., dwivedi, y. k., williams, m. d., & weerakkody, v. (2015). investigating success of an e-government initiative: validation of an integrated is success model. information systems frontiers, 17(1), 127-142. ranganathan, c., & brown, c. v. (2006). erp investments and the market value of firms: toward an understanding of influential erp project variables. information systems research, 17(2), 145161. ringle, c., wende, s., & becker, j.-m. (2015). b. smartpls gmbh. ringle, c. m., sarstedt, m., mitchell, r., & gudergan, s. p. (2020). partial least squares structural equation modeling in hrm research. the international journal of human resource management, 31(12), 1617-1643. ringle, c. m., sarstedt, m., & straub, d. (2012). a critical look at the use of pls-sem in mis quarterly. rogers, e. m. (1995). diffusion of innovations. new york: simon and schuster. shahibi, m. s., saidin, a., & izhar, t. a. t. (2016). evaluating user satisfaction on human resource management information system (hrmis): a case of kuala lumpur city hall, malaysia. international journal of academic research in business and social sciences, 6(10), 95-116. sharma, s. k., & sharma, m. (2019). examining the role of trust and quality dimensions in the actual usage of mobile banking services: an empirical investigation. international journal of information management, 44, 65-75. shen, y.-c., chen, p.-s., & wang, c.-h. (2016). a study of enterprise resource planning (erp) system performance measurement using the quantitative balanced scorecard approach. computers in industry, 75, 127-139. solutions, p. c. (2015). erp report: a panorama consulting solutions research report. retrieved from sternad, s., & bobek, s. (2013). impacts of tam-based external factors on erp acceptance. procedia technology, 9, 33-42. abrar ullah, rohaizat bin baharun, & muhammad yasir, khalil md nor 138 sun, h., ni, w., & lam, r. (2015). a step-by-step performance assessment and improvement method for erp implementation: action case studies in chinese companies. computers in industry, 68, 40-52. swartz, d., & orgill, k. (2001). higher education erp: lessons learned. educause quarterly, 24(2), 20-27. tam, c., & oliveira, t. (2016). understanding the impact of m-banking on individual performance: delone & mclean and ttf perspective. computers in human behavior, 61, 233-244. tsai, w.-h., lee, p.-l., shen, y.-s., & lin, h.-l. (2012). a comprehensive study of the relationship between enterprise resource planning selection criteria and enterprise resource planning system success. information & management, 49(1), 36-46. tseng, t. h., & lee, c. t. (2018). facilitation of consumer loyalty toward branded applications: the dual-route perspective. telematics and informatics, 35(5), 1297-1309. ullah, a., baharun, r. b., nor, k., siddique, m., & bhatti, m. n. (2018). enterprise resource planning (erp) systems and erp quality factors: a literature review. journal of managerial sciences, 11(3), 297-322. ullah, a., baharun, r. b., nor, k., siddique, m., & sami, a. (2018). enterprise resource planning (erp) systems and user performance (up). international journal of applied decision sciences, 377-390. umar, m., khan, n., agha, m., & abbas, m. (2016). exploring the factors affecting enterprise resource planning (erp) implementation quality. journal of quality and technology management, 7(1), 137-155. urbach, n., smolnik, s., & riempp, g. (2010). an empirical investigation of employee portal success. the journal of strategic information systems, 19(3), 184-206. venkatesh, v., morris, m. g., davis, g. b., & davis, f. d. (2003). user acceptance of information technology: toward a unified view. mis quarterly, 425-478. wang, e. s.-t. (2016). the moderating role of consumer characteristics in the relationship between website quality and perceived usefulness. international journal of retail & distribution management, 44(6), 627-639. wang, y.-s., & liao, y.-w. (2008). assessing egovernment systems success: a validation of the delone and mclean model of information systems success. government information quarterly, 25(4), 717-733. wang, y. s. (2008). assessing e‐commerce systems success: a respecification and validation of the delone and mclean model of is success. information systems journal, 18(5), 529-557. wilson, b. (2010). using pls to investigate interaction effects between higher order branding constructs handbook of partial least squares (pp. 621-652): springer. xu, x., zhang, w., & barkhi, r. (2010). it infrastructure capabilities and it project success: a development team perspective. information technology and management, 11(3), 123-142. yang, m., shao, z., liu, q., & liu, c. (2017). understanding the quality factors that influence the continuance intention of students toward participation in moocs. educational technology research and development, 65(5), 1195-1214. zhou, t. (2011). examining the critical success factors of mobile website adoption. online information review, 35(4), 636-652. zhou, t. (2014). an empirical examination of initial trust in mobile payment. wireless personal communications, 77(2), 1519-1531. zhu, y., li, y., wang, w., & chen, j. (2010). what leads to post-implementation success of erp? an empirical study of the chinese retail industry. international journal of information management, 30(3), 265-276. appendix b questionnaire items journal of applied economics and business studies, volume. 4, issue 2 (2020) 119-140 https://doi.org/10.34260/jaebs.426 139 system quality (hsu et al., 2015); (gable et al., 2008); (nelson et al., 2005) sq1 the cms system is easy to use sq2 the cms system is easy to learn sq3 the cms system always processes data accurately sq4 the cms system requires only a minimum number of fields and screens to complete a task sq5 the cms system meets my requirements. sq6 the cms system includes necessary features and functions for my job. sq7 the cms system user interface can be easily adapted to my personal approach. sq8 all the data that i use within the cms system are fully integrated and consistent sq9 the cms system can be easily modified or improved according to my needs. information quality (hsu et al., 2015); (gable et al., 2008); (nelson et al., 2005) iq1 the cms system provides output that is exactly what i want iq2 information needed from the cms system is always available iq3 information from the cms system is in a form that is readily usable iq4 information from the cms system is easy to understand iq5 information from the cms system appears readable, clear, and well formatted iq6 information from cms system is concise service quality (hsu et al., 2015); (pitt et al., 1995); (parasuraman et al., 1988) reliability srq1 the it department provides its services at the time it promises to do srq2 when users have a problem, the it staff shows a sincere interest in solving it srq3 the it department is reliable srq4 the it department insists on error-free records responsiveness srq5 the it staff informs users exactly when services will be performed srq6 the it staff gives prompt service to users. srq7 the staff of the it department is never too busy to respond to user’s requests. assurance srq8 the behaviour of the staff in it department boosts the confidence of users. srq9 i feel safe in my dealings with the it staff srq10 it staff is consistently courteous with users. srq11 the it staff have the knowledge to do their job well empathy srq12 the it department has operating hours convenient to all users. srq13 the it department gives users individual attention srq14 the it department has the users’ best interests at heart srq15 the staff of the it department understand the specific needs of the users perceived usefulness (abugabah et al., 2015) pu1 using the cms system improves my performance in my job pu2 using the cms system in my job increases my productivity pu3 using the cms system enhances my effectiveness in my job pu4 i find the system to be useful in my job perceived ease of use (abugabah et al., 2015) peou1 i find the cms system easy to use peou2 i find it easy to get the cms system to do what i want it to do peou3 my interaction with the cms system is clear and understandable. peou4 interacting with the system does not require a lot of my mental efforts user satisfaction (lin, 2010) us1 the information i get from cms system is very satisfying us2 my interaction with cms system is very satisfying us3 overall, i am very satisfied with cms system user performance (abugabah et al., 2015) up1 the quality of the cms system enables me to accomplish my work up2 our cms system has a positive impact on my productivity up3 using cms system in my job enables me to accomplish multiple tasks more quickly up4 overall, our cms system improves my efficiency in my job up5 our cms system helps me solve my job problems up6 our cms system reduces performance errors in my job up7 our cms system enhances my effectiveness in my job abrar ullah, rohaizat bin baharun, & muhammad yasir, khalil md nor 140 up8 our cms system helps me create new ideas in my job up9 our cms system enhances my creativity up10 overall, our erp system helps me achieve my job goals appendix c cross loadings sq iq rel resp assu emp pu peou us up sq2 0.686 0.489 0.452 0.505 0.486 0.442 0.480 0.476 0.490 0.401 sq3 0.777 0.471 0.482 0.548 0.527 0.468 0.406 0.531 0.541 0.451 sq4 0.791 0.527 0.538 0.605 0.584 0.492 0.402 0.530 0.547 0.468 sq5 0.781 0.558 0.464 0.527 0.529 0.511 0.582 0.533 0.492 0.471 sq6 0.650 0.532 0.424 0.511 0.472 0.489 0.458 0.498 0.534 0.531 sq8 0.727 0.586 0.425 0.516 0.447 0.492 0.465 0.492 0.500 0.574 sq9 0.785 0.473 0.463 0.588 0.510 0.533 0.578 0.546 0.570 0.589 iq1 0.412 0.776 0.454 0.578 0.487 0.516 0.581 0.590 0.409 0.445 iq2 0.421 0.819 0.455 0.527 0.443 0.371 0.459 0.526 0.544 0.595 iq3 0.561 0.849 0.371 0.477 0.311 0.382 0.491 0.544 0.540 0.565 iq4 0.401 0.874 0.451 0.481 0.419 0.436 0.470 0.504 0.544 0.590 iq5 0.582 0.851 0.417 0.471 0.407 0.413 0.456 0.509 0.544 0.509 iq6 0.553 0.794 0.370 0.461 0.372 0.442 0.528 0.493 0.489 0.570 srq1 0.422 0.296 0.662 0.375 0.468 0.392 0.344 0.403 0.429 0.456 srq2 0.417 0.365 0.750 0.351 0.363 0.305 0.286 0.299 0.329 0.365 srq3 0.553 0.472 0.837 0.553 0.493 0.413 0.399 0.421 0.446 0.524 srq5 0.522 0.418 0.534 0.813 0.485 0.479 0.426 0.478 0.490 0.557 srq6 0.471 0.569 0.501 0.864 0.539 0.518 0.315 0.496 0.408 0.478 srq7 0.427 0.526 0.411 0.821 0.517 0.571 0.571 0.499 0.512 0.496 srq9 0.372 0.204 0.512 0.362 0.693 0.352 0.141 0.228 0.314 0.346 srq10 0.408 0.407 0.449 0.578 0.802 0.520 0.492 0.492 0.519 0.581 srq11 0.569 0.502 0.417 0.461 0.802 0.521 0.390 0.441 0.486 0.552 srq12 0.369 0.254 0.356 0.375 0.500 0.740 0.287 0.245 0.269 0.352 srq13 0.492 0.498 0.371 0.496 0.424 0.779 0.514 0.472 0.486 0.519 srq14 0.369 0.463 0.443 0.595 0.527 0.847 0.317 0.542 0.528 0.321 pu1 0.590 0.476 0.350 0.498 0.358 0.504 0.813 0.541 0.530 0.410 pu2 0.580 0.472 0.384 0.537 0.368 0.491 0.841 0.553 0.532 0.511 pu3 0.561 0.525 0.386 0.575 0.398 0.553 0.880 0.437 0.593 0.474 pu4 0.520 0.556 0.429 0.564 0.426 0.512 0.836 0.560 0.389 0.498 peou1 0.493 0.429 0.350 0.454 0.380 0.396 0.421 0.686 0.463 0.455 peou2 0.524 0.468 0.396 0.395 0.442 0.300 0.329 0.621 0.421 0.494 peou3 0.420 0.556 0.429 0.564 0.426 0.512 0.436 0.860 0.489 0.598 peou4 0.516 0.471 0.338 0.468 0.319 0.401 0.685 0.820 0.543 0.548 us1 0.613 0.530 0.442 0.539 0.508 0.456 0.496 0.397 0.843 0.524 us2 0.524 0.468 0.396 0.395 0.442 0.300 0.329 0.321 0.721 0.494 us3 0.520 0.556 0.429 0.564 0.426 0.512 0.533 0.586 0.789 0.440 up1 0.562 0.513 0.574 0.500 0.546 0.532 0.559 0.543 0.539 0.723 up2 0.536 0.510 0.332 0.443 0.400 0.375 0.493 0.404 0.377 0.612 up3 0.594 0.524 0.419 0.475 0.456 0.380 0.540 0.482 0.481 0.704 up4 0.448 0.436 0.395 0.397 0.459 0.383 0.377 0.436 0.451 0.580 up5 0.520 0.556 0.429 0.564 0.426 0.512 0.436 0.350 0.489 0.798 up6 0.567 0.478 0.387 0.547 0.457 0.468 0.587 0.501 0.456 0.676 up7 0.313 0.530 0.442 0.539 0.508 0.456 0.496 0.397 0.443 0.724 up8 0.493 0.429 0.350 0.454 0.380 0.396 0.421 0.486 0.463 0.655 journal of applied economics and business studies, volume. 4, issue 3 (2020) 159-184 https://doi.org/10.34260/jaebs.437 159 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x financing behavior of small and medium sized enterprises of khyber pakhtunkhwa at life cycle stages shahid ali*1, maryam jabeen2, laila taskeen qazi3, aisha jabeen4 1associate professor at institute of management sciences, peshawar, pakistan. 2assistant professor at institute of management sciences, peshawar, pakistan. 3assistant professor at institute of management sciences, peshawar, pakistan. 4lecturer at institute of management sciences, peshawar, pakistan. abstract the present study explores the financing patterns of small and medium enterprises (smes) across different age categories in khyber pakhtunkhwa. the study aims to indicate whether these enterprises are congruent with or contradict to the predictions of existing theories that explains the financing decisions in smes. the survey covered 341 smes operating across the seven districts of khyber pakhtunkhwa (kp): peshawar, mardan, karak, chitral, manshera, bannu and d.i. khan. the sample is represented by three main industries: agriculture, industrial and service. chi-square test is used to examine the objective of the study, followed by a post-hoc analysis using standardized residuals to distinguish the inter-dependence among the variables more effectively. the findings show that the owner-managers’ personal savings are preferred by enterprises that are relatively younger (age 0-10 years). smes in operation for 11 to 20 years seem to take up more debt than entrepreneurs’ capital. the industry breakdown confirms the robustness of our results. smes in the service sector prefer internal funds and government backed loans at the start-up stage. however, enterprises across all the three business sectors, agriculture, industrial and service, with 11-20 years of existence prefer asset-based lending and debt financing. the findings from the survey study confirm that smes across different districts and business sectors of kp follow the assumptions predicted by the trade-off theory. several policy implications have been recommended not only for regulatory bodies and lending institutions but also for the sme sector based on the results from empirical investigation and survey findings. keywords small and medium-sized enterprises, firms’ age, pecking order theory, tradeoff theory, debt financing, internal sources of financing jel classification i21;i25; i26; o11, q10 *shahid.ali@imsciences.edu.pk mailto:shahid.ali@imsciences.edu.pk shahid ali, maryam jabeen, laila taskeen qazi, aisha jabeen 160 1. introduction the wealth of an economy apart from agricultural output is generated through entrepreneurial ventures. when general public, policy makers, and economists talk about economic growth, they mostly relate it to the expansion of output by enterprises. although there is substantial contribution of large firms to wealth creation and economic development but in some countries the aggregate contribution of smes is greater in comparison to large organizations (hallberg, 2000). besides output, smes make investments, create jobs, and export their products to international markets. a significant contribution is made by smes to asia’s economic growth. the sheer number is overwhelming as in most of the countries they account for 98 percent or more of the total number of businesses. over the period from 2007 to 2012, smes employed approximately 66 percent of national labor force and have on average 38 percent contribution to gross domestic product (asian development bank, 2014). progressive and dynamic smes expand the industries technological frontiers through innovation of product and process technologies. the cumulative clout of smes is impressive; however, their diversity is even more. a large number of smes are service providers and retailers. they run business operations as mundane as hair and beauty salon, auto repair and service shops, or small fleets of three wheeled vehicles that carry passengers from suburban regions to metropolitan areas. the heterogeneity that exists among these enterprises makes it difficult to address them as a single group when trying to pin down the market failures they encounter or in designing effective policies (vandenberg, yoshino, goto, intarakumnerd, & miyamoto, 2016). development of smes is considered a panacea for poverty alleviation, employment generation and economic empowerment especially in the third world countries. there is considerable literature that goes in favor of the argument that economies can be built using potential of smes (ndiaye, razak, nagayev & ng, 2018). smes cater as change agents in an economy as they not only bring innovation in obsolete business practices but change the stereotype thinking of a society. smes are a catalyst to a broader social change as they open windows to gender equity in the economic pool. the only course to resolve the burning issues of poverty, unemployment and areal disparities lies in the hands of smes. small enterprises often utilize domestic resources more proficiently than their larger counterparts when imperfection in the marketplace prevent them to maximize the benefits of increasing efficiency in operating activities. smes are major contributors to pro-poor growth that cannot be matched by large scale corporations (jones & lee, 2018; hoang, pham & ulubaşoğlu, 2019). the management of smes journal of applied economics and business studies, volume. 4, issue 3 (2020) 159-184 https://doi.org/10.34260/jaebs.437 161 depend on owners and for that reason they have more incentives to control everything in an efficient manner with greater flexibility. also, the competition in the labor market is expected to push up the wage rate due to growing number of smes. the poor class benefit from the improvement in the availability of products and services. in terms of job creation, medium sized businesses are in a better position to grow if business support services and credit are easily accessible and readily available. the initiatives of government in promoting smes for development are hence justified. smes form the backbone of industry, trade and the economy. smes in pakistan seem not to have benefited from technological progress which is evident from the absence of mass production facilities, low prevalence of quality and efficiency, and a lack of access to formal financing facilities (nenova & niang, 2009; afraz, hussain & khan, 2014). generally, the problems being faced by the sme sector are well known, but a lack of updated and quantifiable research makes it an arduous job for policy makers and the stakeholders to develop an effective framework for facilitating the growth of smes. from the sixth sme conference titled ‘engaging smes to stabilize the economy’ arranged by state bank of pakistan some worrisome figures came out like bank lending to smes dropped from rs. 437 billion in fy07 to rs. 248 billion in fy12. there was an academic consensus in the conference that banks exhibit a risk-averse posture in financing smes in pakistan. they were deemed to customize their financing products and make them more attractive to cater to the dire finance needs of this sector. the declining bank lending to smes needs investigation as it has implications on the survival and growth of smes. banks offer a variety of financial products to smes in pakistan including term financing, lease financing, overdraft, revolving credit, letter of credit, bills of exchange purchased, unsecured or clean financing, and demand finance etc. these financing products also need to be examined as their structure may not suit the requirements of smes especially under stringent prudential regulations of state bank of pakistan (sbp). the varying financial needs of smes are a key for banks/dfis to design attractive financing solutions. the financing preferences of firms’ change depending on the age of the enterprises. the literature examines the financial life cycle theories of smes to understand the varying levels of financing needs at different levels of business life cycle. however, none of the identified theories on smes life cycle stages best explicate the capital structure practices. the literature on the topic present varying dimensions, however, understanding the local dynamics is both important and essential in providing solutions to this important sector of our economy. shahid ali, maryam jabeen, laila taskeen qazi, aisha jabeen 162 the present survey study investigates whether financing structure of enterprises exhibit any pattern along its life cycle. the financing of smes is examined through a financial growth life cycle paradigm. the samples of different ages provide us with a basis of analyzing the financial behavior of smes in different districts of kp along the business life cycle. empirical investigation of the financing patterns of firms across different age categories indicate whether these enterprises are congruent with or contradict to the predictions of existing theories that explains the financing decisions in sme firms. the research further explores the factors that might underlie firms’ choice of mix between debt financing and equity financing. this diagnosis of smes intends to guide regulatory agencies and policy makers on how to sustain and accelerate the on-going transition towards a more entrepreneurial culture. based on the findings from empirical investigation and survey results, several implications for policy have been recommended not only for regulatory bodies and lending institutions but also for sme sector firms. 2. literature review and hypotheses development ample literature is available around the world that emphasizes the appreciating the role of smes in economic development. irwin & scott (2010) present pre-requisites for effective sme support in an economy; this includes conducive public policy, understanding the contribution of smes, and business development support. this paper clearly suggests that financial institutions and banks need to understand the financial needs of smes at different stages of growth. this would call for designing any sme financing product after carefully assessing the need for financing. it is an understood fact that businesses survive on positive npv projects, for which one has to see that how such projects are financed. according to the pioneering theory of modigliani and miller (1958) there is no dependence between capital structure and its cost of capital. this theory is the reason for subsequent work on agency theory and signaling theory. while working on smes it becomes important to look into the debt-equity structure that how projects are financed before we come to a financing need solution. asymmetric information is one of the most common imperfections in the credit market between borrowers and lenders. this is in particular true for smaller enterprises (usai, scuotto, murray, fiano & dezi, 2018; miglo & miglo, 2019), as an end result, the lending party does not know the actual worth of investment plans that are projected by small businesses (adverse selection) and cannot be certain how the proposed funds will be applied (moral hazard). this is primarily caused by the low standard of financial reporting information generated by small organizations. the lack of self-assurance on journal of applied economics and business studies, volume. 4, issue 3 (2020) 159-184 https://doi.org/10.34260/jaebs.437 163 part of credit lending institutions results in high risk premium, which in turn attract relatively riskier projects. this undermines the absolute profitability for credit institutions of lending to small firms. credit institutions frequently request guarantees as a compensation for taking risk, even though they are often greater than the existing assets. therefore, funds available by the lenders are often rationed at given levels of interests, and only provide finances for those projects that offer sufficient guarantees (lopez-gracia & aybar-arias, 2000). offering static financial solutions to smes especially where their financing needs vary with their age demand due attention. the life cycle theory suggests a business may have differing financing needs at inception, middle age, and old age. the life cycle theory of a business actually originates from economics but was first given as an analogy in biology (penrose, 1952) and later on was developed in a formal economic theory (penrose, 1959). some of the earlier texts like (weston & brigham, 1970) present a description of firm at different stages with sources of finance for its business needs and the potential problems it faces at varying stages. researchers like (berger & udell, 1998) have developed a financial growth life cycle model for smes which need to be tested for smes around the world. there exist myriad theories about the different lifecycle stages that small businesses typically go through. the variety of models that have been proposed to describe the life cycle stages in small enterprises differ significantly in range and setup from a mere three stages to models that depict no less than ten stages. however, from the extant literature it can be concluded that in vast majority of firms some general form of a life cycle can undeniably be identified, and most organizations follow roughly the same trajectory (lewis & churchill, 1983; miller & friesen 1984). agency theory explains conflict of interests among different stakeholders in an organization. on one hand, the theory considers conflict of interest between owners and managers, and on the other, between owners and debt holders. smes are not likely to suffer from the first problem as ownership and control are generally merged in these enterprises. adverse selection and moral hazard are significant problems that happen from the agency conflicts in smes (ang, 1992). smes face severe adverse selection problems, as majority of these enterprises are not listed on the stock exchanges, resulting in high degree of uncertainty, regarding the information that a firm makes publicly available. the shortage of sufficient funds, via access to formal debt, is noticeably pertinent in sme, an outcome of problems caused by asymmetric information (serrasqueiro & caetano, 2015). shahid ali, maryam jabeen, laila taskeen qazi, aisha jabeen 164 pecking order theory, based on the idea of information asymmetry, attempts to explain the financial structure of a firm. it deals with the asymmetry of information between the internal stakeholders and the external stakeholders; and which can lead the market to undervalue a firm. this theory has been applied particularly to small organizations. small firms are often run and managed by a single individual. the owners of small enterprises function without targeting an optimal capital structure, and thus show an obvious preference for those modes of financing that can reduce interference into their business. therefore, financing is based on internally generated resources and personal savings. also, it depends on short-term and long-term debt financing and finally, the least preferred is the issuance of new shares that dilute control over the firm (lopez-gracia & aybar-arias, 2000). trade-off theory (tot) is based on the assumption that firms desire to allocate their resources by comparing the tax benefits of debt with the bankruptcy costs thereof targeting an optimal debt ratio. empirical studies predict a positive association between the age of a firm and use of debt financing in comparison to younger firms. in the later stages in view of life cycle, organizations can obtain debt on sufficiently favorable terms (serrasqueiro & caetano, 2015). la rocca, la rocca and cariola (2009) tested the explanatory power of different financing theories on the basis of informational opacity at various stages of business life cycle of smes. the empirical findings of their study show that specific financing strategies are adopted by these organizations in economies that are bank-oriented through different stages of their life cycle. the role of financial institutions and information asymmetry considerations tend to influence the financing strategy of smes. their findings further show that an increasing amount of debt is needed by young and start-up organizations to retain their growth though they progressively rebalance their capital structure after the business consolidation. as a result, the pecking order theory is not applicable to younger firms. young enterprises have insufficient internal liquidity to sustain growth and therefore do not remain capable to seek additional equity funds (bhama, jain & yadav, 2018). mateev, poutziouris and ivanov (2013) suggest that smes prefer internal rather than external sources of finances. firms that are highly profitable resort to retained earnings to finance their investment projects rather than employing debt. this means of financing can certainly be applied to smes based on the following rationale: managers in smes who remain owners, do not prefer to lose their business assets and remain in-charge of their own firms. they have an insignificant acceptance for new shareholders, as a result they rely on internal financing sources than external finances. if external funds are required, smes desire short-term loans, which is less likely to contain restrictive journal of applied economics and business studies, volume. 4, issue 3 (2020) 159-184 https://doi.org/10.34260/jaebs.437 165 covenants and does not even lessen management flexibility (holmes & kent, 1991). . the above discussion leads to the formulation of following hypotheses: h1: small and medium-sized enterprises follow the predictions of trade-off theory at different stages of business life cycle h2: small and medium-sized enterprises follow the predictions of pecking order theory at different stages of business life cycle 3. research objective the objective of the present survey study is to examine and understand the financing of sme businesses through a financial growth life cycle paradigm. observing financing patterns of firms across different age categories indicate whether these enterprises are congruent with or contradict to the predictions of existing theories that explains the financing decisions in smes. 4. research methodology: 4.1. sample size and data sources deciding about a representative sample for this study seemed a tricky proposition as organizations in pakistan differ not only on the definition but also on the exact number of smes. let alone the population size, the following table gives a glimpse of the definition of smes in pakistan; the state bank of pakistan defines smes as; institution annual sales turnover (upto) annual assets at cost (upto) number of employees (upto) sbp manufacturing rs. 300 m rs. 100 m 250 service 〃 rs. 50 m 250 trading 〃 rs. 50 m 50 small and medium enterprise development authority of pakistan defines smes as; institution annual sales turnover (upto) paid up capital (upto) number of employees (upto) smeda rs. 250 milliom rs. 25 million 250 the number of smes in kp differ as pakistan bureau of statistics, sarhad development authority, and smeda report different numbers about it. smeda has a walk-in facility for troubled smes; it is deemed as a most relevant sampling frame for this study. their latest set of kp troubled smes comprises of 2504 walk-ins which would be used as a most suitable population in getting a random sample for this study as it is believed that the core problem of a troubled sme is financial hardships. the study randomly samples seven districts from 35 districts of kp initially. next, for sampling the smes, industrial units of the sampled 7 districts are considered. as shahid ali, maryam jabeen, laila taskeen qazi, aisha jabeen 166 2504 financially troubled smes is considered as the population. the standard statistical objective procedure to determine the sample size by taking n= 2504 at 95 percent confidence interval with 5% margin of error is computed using the following formula; 𝑋 = 𝑍(𝑐 100⁄ ) 2 𝑟(100 − 𝑟);𝑛 = 𝑁𝑥 ((𝑁 − 1)𝐸2 + 𝑥) ; 𝐸 = √[ (𝑁 − 𝑛)𝑥 𝑛(𝑁 − 1)⁄ ] 2 n shows the population size, z(c/100) is the critical value, and the fraction of responses are denoted by r. the calculated sample size following this method of sample size determination is 377 (n). the sampled firms are deemed to be investigated from firms operating in three major business sectors including industrial, agriculture, and services. keeping in view the greater concentration of small industrial business, the detail of the survey respondents suggested for sampling the smes is mentioned in table 1. since the total industrial estates for these 7 districts is 30, the same is used for sampling smes proportionally from these major districts. table 1 district no of industrial estates total number of smes stratified by districts ni = (ni/n0)*n peshawar 7 88 mardan 3 38 karak 3 38 chitral 4 51 mansehra 5 62 bannu 5 62 d.i.khan 3 38 30 377 where, n0 =30; n = 377; ni = number of industrial estates in a district a fully planned survey was conducted using trained manpower and questionnaires were distributed through personal visits among the smes owner-manager operating in different sectors across seven districts of kp. the questionnaire was developed using input from literature and considering several local perspectives. surveying smes from these major districts are deemed to unfold important regional and social factors towards percieving the local dimension of availing or ignoring formal financing opportunities. 4.2. study variables and measures the present study has performed the statistical analysis namely on three variables: source of capital for smes, age of enterprise and business sector. the sources of capital for smes are measured on ten (10) items. the ten items are evaluated on a five-point likert scale ranging from not preferred at all to extremely preferred (the detail is mentioned in section 5.3). age is computed by taking the difference between year 2018 journal of applied economics and business studies, volume. 4, issue 3 (2020) 159-184 https://doi.org/10.34260/jaebs.437 167 and the year in which the business was established. the average age of the enterprises is approximately 16 years. firm age varies from 0 to 30 years. the age of surveyed enterprises is divided into three (3) groups (the detail is given in sub-section 5.1). business sectors are classified in three (3) broad categories: agriculture, industrial and service. 5. results and discussions 5.1. descriptive analysis the study was aiming to survey respondents from the desired sample size (377) and therefore it targeted more firms from each district. smes from karak, bannu, and d.i.khan generated more survey responses compared to other districts. table.1 shows that the survey actually covered 341 smes operating across seven districts of kp. this left a nonresponse of 9.5% as some of the randomly sampled firms were either found shut or were not willing to participate in the survey. table 2: districts frequency percent peshawar 75 22 mardan 37 10.9 karak 40 11.7 chitral 47 13.8 mansehra 36 10.6 bannu 63 18.5 d.i. khan 43 12.6 total 341 100 table. 2 shows majority of the respondent of smes belonged to service sector, 36.4 percent, while industrial and agriculture sector accounted for 35.2 percent and 27.3 percent respectively. the service sector includes sub-sectors such as wholesale and retail trade, transport, storage and communication, finance and insurance. table 3: sectors frequency percent valid percent agriculture 93 27.3 27.6 industrial 120 35.2 35.6 service 124 36.4 36.8 total 337 98.8 100 missing 4 1.2 total 341 100 industrial sector includes mining & quarrying; manufacturing; electricity generation & distribution & gas distribution and construction. while, service sector include wholesale & retail trade; transport, storage and communication; finance & insurance. firm age varies from 0 to 30 years. the age of surveyed enterprises is divided into three (3) groups: 0 to 10 years, from 11 to 20 years and 21 to 30 years. majority of shahid ali, maryam jabeen, laila taskeen qazi, aisha jabeen 168 smes falls between the age group of 11-20 years (see the detail given in table 5). to investigate the research question of the study based on the financial life cycle theory model of pakistani smes, age of the enterprises is computed by taking the difference between year 2018 and the year in which the business was established. the average age of the enterprises is approximately 16 years. firm age has a minimum value of 0 (years) and maximum value is 30 (years). firm’s age takes the value of zero (years) for those businesses that were established in year 2018. table 4: age of smes valid 338 missing 3 mean 15.6243 standard deviation 6.3413 variance 40.212 minimum 0 maximum 30 from the output it is evident that majority of smes (59.8 percent) falls between the age group of 11-20 years. it is followed by 20.2 percent of the respondents aged between 21-30 years and the third category has the lowest proportion (17.6 percent) of survey respondents that ranges between 1 to 10 years (table 5). table 5: age categories frequency percent valid percent 1-10 60 17.6 18 11-20 204 59.8 61.3 21-30 69 20.2 20.7 total 333 97.7 100 missing 8 2.3 total 341 100 5.2 validity and reliability tests – smes questionnaire survey on the basis of convenient accessibility, the survey research pilot study is carried out on a group of 50 respondents from district peshawar. a general accepted rule is that a cronbach alpha coefficient of 0.8 and greater indicates a good level of reliability, and 0.6 to 0.7 shows an acceptable level of reliability (hulin, netemeyer & cudeck, 2001; ursachi, horodnic & zait, 2015). the result for cronbach’s alpha reliability coefficient for the questionnaire related to smes owner-manager is 0.639, indicating an overall consistency and accuracy of the instrument that has been developed. however, the content validity of the instrument is established through the calculation of content validity ratio (cvr). according to lawshe (1975), with a panel of 10 experts, the minimum required cvr for each item is 0.62. the rationale to apply journal of applied economics and business studies, volume. 4, issue 3 (2020) 159-184 https://doi.org/10.34260/jaebs.437 169 lawshe method is to quantify the assessment of the content validity. to evaluate the content validity and internal consistency of the questionnaire, at least eight (8) members must agree in order for the items to be included in the final form of a questionnaire. in this study, the cvr for each item ranged between 0.8 and 1, pointing towards the significance and the need to add relevant items in the scale. table 6: lawshe’s content validity ratio (cvr) sme questionnaire survey n ne cvr gender of the entrepreneur q #1 10 10 1 in which year was the business established? (age of business enterprise) q #2 10 9 0.8 location of the business q #3 10 9 0.8 classifying the business under the different sectors? q #4 10 10 1 how will you rank the following sources of capital in terms of preference? owner’s saving, retained earnings, family and friends, angel financing, venture capital, trade credit, leasing, hire purchase, loan from bank, and government scheme q #5 10 10 1 n is the total number of panel members; ne is the number of panelists that point towards ‘essential’ cvr= (ne – n/2) / (n/2) 5.3 smes financial growth life cycle model to investigate the financial behavior of smes through its lifecycle, the sources of capital for entrepreneurs are classified in ten (10) financing options: owner‘s saving, retained earnings, family and friends, angel financing, venture capital, trade credit, leasing, hire purchase, loan from bank, and government scheme (the detail is given in q #5-sme questionnaire survey). a chi-square test is employed to measure if there is a significant association between sme's sources of capital and the age of enterprises. the study further investigates the relationship between these two variables across major sectors of kp, i.e. agriculture, industrial, and service. also known as pearson’s chisquare test, it is the most frequently used nonparametric statistical test. the results of the chi-square tests are estimated by reference to the chi-square distribution. the chisquare statistic is used to test the hypothesis of no association between two or more categorical variables or to evaluate how a sample data fits the distribution of a finite population (franke, ho & christie, 2012; rana & singhal, 2015). according to benhamou & melot (2018), the biggest advantages of using pearson's chi-squared test is its robustness and simplicity as it only depends on two key assumptions: independence of observations and large sample size. the present survey study has relied on a sufficiently large sample size. the mutually exclusive variables are chosen for the shahid ali, maryam jabeen, laila taskeen qazi, aisha jabeen 170 purpose of analysis by drawing random samples from each stratum (districts). chisquare test is used to test the hypothesis of no association between two or more categorical variables or to evaluate in the manner in which a sample fits the distribution of a known population (franke, ho, & christie, 2012; rana & singhal, 2015). ,. the analysis of chi-square test is followed by a post-hoc analysis using standardized residuals to distinguish the inter-dependence among the variables more effectively. smes sources of capital in terms of preference (q# 5) are cross tabulated across three age groups. the preferences for financing by smes are categorized into ten (10) distinct sources of capital. however, the statistical significance of the results, as shown by the p-value of chi-square test, is confirmed only for owners saving (0.004), leasing (0.018), hire purchase (0.070) and loans from banks (0.030). table 7 reports the survey findings: table 7 chi-square test for association (sources of capital in terms of preference * age categories) pearson chi-square likelihood ratio n of valid cases value df asymp. sig (2-sided) value df asymp. sig (2sided) owner’s saving 22.516 8 0.004* 29.304 8 0.000 328 retained earnings 7.360 8 0.498 6.980 8 0.539 322 family and friends 5.993 8 0.648 5.744 8 0.676 319 angel financing 2.062 8 0.979 1.944 8 0.983 313 venture capital 13.335 8 0.101 11.845 8 0.158 319 trade credit 8.195 8 0.415 8.034 8 0.430 324 leasing 18.433 8 0.018** 18.418 8 0.018 313 hire purchase 14.475 8 0.070*** 14.399 8 0.072 319 loans from banks 17.044 8 0.030** 14.636 8 0.067 312 government scheme 6.013 8 0.646 5.772 8 0.673 323 *p<0.01; **p<0.05; ***p<0.1 chi-square test does not identify the combination of factors affecting the desired level of significance. if a difference is statistically significant, the study moves forward to conduct a post hoc analysis using the standardized residual method to determine which combination of categories contributes to statistical significance. if the absolute value of the residual in the detailed cross-tabulation analysis is greater than 2 (miller, 2017), it can be concluded that the categories (cells) are having major contribution to the chi-square test results. it is important to mention here that a detailed analysis has been performed only for those variables that have a statistically significant association. the findings and the discussion of the results is given below: journal of applied economics and business studies, volume. 4, issue 3 (2020) 159-184 https://doi.org/10.34260/jaebs.437 171 5.3.1. internal sources of financing (owner’s saving) the findings in table 8 shows that the owner-managers’ personal savings (residual= 9.1) are preferred by enterprises that are aged between 0-10 years. young firms are often informationally opaque as a consequence of failing to maintain an established track record, which may lead to the reluctance of financial institutions and banks to lend to these enterprises. klapper, sarria-allende and sulla (2002) observed that businesses that are established for less than four years rely more heavily on informal financing and less on financing from banks. at the start-up stage, smes mostly raise money internally. the major reason is that the external sources of finance are limited at this stage. the first-generation business owners do not favor external debt due to difficulties and discrimination in access to intermediate external financing. as the firm grows, they then look for external financing such as equity capital or debt finance when the capital requirements become higher (ampenberger, schmid, achleitner & kaserer, 2013). table 8: owner’s saving * age categories age categories total 1-10 11-20 21-30 not preferred at all count 16 40 16 72 residual 3.0 -3.9 0.9 slightly preferred count 26 55 13 94 residual 9.1 -2.3 -6.8 moderately preferred count 10 60 18 88 residual -5.8 6.3 -0.5 highly preferred count 0 28 12 40 residual -7.2 3.6 3.6 extremely preferred count 7 17 10 34 residual 0.9 -3.7 2.8 total count 59 200 69 328 5.3.2. asset-based lending and debt financing (leasing, hire purchase and loan from bank): in kp, smes with 11 to 20 years of existence seem to take up more debt than entrepreneurs’ capital. around the world, sme business possesses experience of more than ten (10) years (abbasi, wang & abbasi, 2017). the findings show that the preferred options for smes are leasing (residual= 7.7), hire-purchase (residual= 7.6) and loan from banks (residual= 6.4). research has shown that leasing as a source of finance is availed by smes on a regular basis. lease transaction is not only recognized as a viable alternative to outright secure or purchase assets; it also allows for better shahid ali, maryam jabeen, laila taskeen qazi, aisha jabeen 172 management of cash flows in particular for those enterprises that lack collateral to apply for bank loans. another reason leasing is an attractive option for smes is that it facilitates growth, offers flexibility, and grants tax related incentives (lasfer & levis, 1998; kraemer-eis & lang, 2012). banks are the primary source of external finance to smes both in developed and developing economies (carey & flynn, 2005; ono & uesugi, 2009; vera & onji, 2010). credit from banks is utilized by small businesses on regular basis, despite the fact that traditional bank finance poses challenges to smes, particularly to innovative, start-ups and fast growing businesses, with a high risk and high return profile and this may not be well suited at different phases of firms’ life cycle. moreover, lenders often require personal guarantee from the founders and executives of sme businesses, which implies that the owner or manager need to risk his own personal assets with a particular end goal to subsidize the organization (abbasi, wang & abbasi, 2017). bank loan is considered to be more expensive in comparison to different sources of finance available to businesses but it generates massive returns for smes. businesses relying on banks for credit perform better as they are being monitored by and are answerable for the decisions taken by the entrepreneurs to banks (keasey & mcguinness, 1990). in the context of developing economies, abor and biekpe (2009), li, griffin, yue & zhao (2011) and saarani & shahadan (2013) suggest a positive correlation between age of the firm and access to bank loans. older smes tend to have good association and better track records with their lenders in comparison to young enterprises, which ultimately tilt the preferences of these enterprises towards borrowing. another critical issue related to the age of the firm is information asymmetry and principal-agent problem. older businesses typically have a habit of hoarding financial documents, which in some way lowers the agency problem and addresses the issue of information asymmetry and consequently organizations enjoy better access to debt markets. alternatively, young and small firms’ face many problems in accessing external finance because of high information costs, that may indirectly discourage the use of external debt (mac an bhaird, 2010; saarani & shahadan, 2013). table 9 leasing * age categories table 10 hire purchase * age categories table 11 loans from banks * age categories age categories total age categories total age categories total 110 1120 2130 110 1120 2130 1-10 11-20 21-30 not preferred at all count 6 12 9 7 10 16 8 34 12 10 6 28 residual 1.2 4.6 3.4 4.1 -5.0 0.9 7.1 -6.9 -0.2 slightly preferred count 16 27 14 57 14 31 11 56 11 33 14 58 residual 5.8 8.0 2.2 4.3 -3.6 -0.8 0.8 -1.9 1.2 journal of applied economics and business studies, volume. 4, issue 3 (2020) 159-184 https://doi.org/10.34260/jaebs.437 173 moderately preferred count 15 67 16 98 9 65 19 93 14 63 21 98 residual 2.5 6.9 4.4 -7.0 7.6 -0.5 -3.3 3.9 -0.7 highly preferred count 9 66 20 95 12 63 18 93 13 63 18 94 residual 8.0 7.7 0.3 -4.0 5.6 -1.5 -3.6 6.4 -2.8 extremely preferred count 10 20 6 36 10 22 11 43 5 19 10 34 residual 3.6 2.1 1.5 2.6 -4.6 2.0 -1.0 -1.5 2.5 total count 56 192 65 313 55 197 67 319 55 188 69 312 the findings from the survey study confirm that smes across different districts of kp follow the assumptions predicted by trade-off theory. the theory assumes a positive relationship between age of a firm and use of debt financing. the financial information of smes is more uncertain in comparison to large organizations, and unlike big businesses, with more choices to raise money from the debt market; smes primarily rely on lending from banks. several schemes have been initiated by the government of pakistan to increase the smes loans in bank lending as part of financial sector reforms. age is a significant determinant of capital structure choices. organizations at the later stages of business life-cycle have several advantages in acquiring loans than young firms. older firms have better reputation, which may imply an easy access to debt and reducing the severity of agency problems. 5.4 smes financial growth life cycle model sector-wise analysis the survey study further investigates sector-wise financing patterns of smes across their business life cycle. a sectoral evaluation reveals a better understanding of how the distinct sources of financing are used by smes during distinct stages of an organization’s life cycle. smes sources of capital in terms of preference (q# 5) are cross tabulated across three age groups and business sectors (q# 4). smes have been classified into three (3) major sectors: agriculture, industrial and service while the preferences for financing are categorized into ten (10) different sources of capital. the statistical significance of the results, as shown by the p-value of chi-square test, is confirmed only for owners’s saving in industrial sector (0.027) and service sector (0.005), trade credit in industrial sector (0.058), leasing in agriculture sector (0.080), loans from banks in agriculture sector (0.053) and service sector (0.029) and government scheme in service sector (0.082). table 12 reports the sector-wise financial behavior of smes across the business life cycle stages. shahid ali, maryam jabeen, laila taskeen qazi, aisha jabeen 174 table 12 chi-square test for association (sources of capital in terms of preference * age categories*business sectors) agriculture industrial service pearson chi-square value df asymp. sig (2-sided) value df asymp. sig (2-sided) value df asymp. sig (2-sided) owner’s saving 8.988 8 0.343 17.336 8 0.027** 22.158 8 0.005* retained earnings 7.061 8 0.530 3.267 8 0.917 10.726 8 0.218 family and friends 6.823 8 0.556 6.178 8 0.627 7.332 8 0.501 angel financing 4.037 8 0.854 4.457 8 0.814 8.369 8 0.398 venture capital 12.786 8 0.119 9.407 8 0.309 6.330 8 0.610 trade credit 11.520 8 0.174 15.036 8 0.058*** 7.456 8 0.488 leasing 19.082 8 0.080*** 5.732 8 0.677 8.734 8 0.365 hire purchase 9.159 8 0.329 11.035 8 0.200 9.740 8 0.284 loans from banks 15.328 8 0.053*** 10.493 8 0.232 17.060 8 0.029** government scheme 11.602 8 0.170 2.011 8 0.981 13.984 8 0.082*** *p<0.01; **p<0.05; ***p<0.1 an effective analysis of association among the variables is provided by the residual analysis. the value of chi-square only provides a general diagnosis of the relationship; however, the accurate correlation can be found through residuals. the statistical significance due to correlation can be proved by using chi-square, but it does not predict the strength of the association for the variables. a residual having absolute value greater than 2 offers a strong evidence of association (miller, 2017). a detailed analysis has been performed only for those variables that have a statistically significant association. the findings and the discussion of the results is mentioned below: 5.4.1 internal sources of financing (owner’s saving): the findings in table 13 shows that the owner-managers’ personal savings are preferred by enterprises that are aged between 1-10 years (residual = 7.8) and 11-20 years (residual = 3.5) in the service sector and the industrial sector respectively. due to complex and high cost procedure, in our country, credit from the bank is ranked by the smes after the funds from ngos and family and friends. when smes face lack of funding, the leading option is the investment by the owner himself from the existing sources and to seek financial assistance from friends and family. this type of financing is known as the source of trust funds (abbasi, wang & abbasi, 2017). smes financing needs vary substantially as they move along the stages of their business life cycle (fort, haltiwanger, jarmin & miranda, 2013). it is imperative that smes should be provided journal of applied economics and business studies, volume. 4, issue 3 (2020) 159-184 https://doi.org/10.34260/jaebs.437 175 with sufficient funds to finance their operations at early stages of business with government credit guarantees, as in the initial stages there is reluctance from the private sector to offer finance. at the start-up business life cycle stage, smes have informational opacity, low likelihood of survival, higher chances of credit risk failure, lack a proven track record, history of trading activities is limited and lack of collateralized assets. thus, smes at the start-up phase relies heavily on the internal sources of financing. in the successive phases of growth, the business seems to endure, grow, and become mature, henceforth, then they attain information transparency, track important details of their business, offer more assets for use as a collateral, and develop greater access to financial markets (de la torre, soledad martinez peria, & schmukler, 2008; steijvers, voordeckers & vanhoof, 2010; shaikh & khoso, 2019). table 13: owner’s saving * age categories * business sectors industrial service age categories total age categories total 1-10 11-20 21-30 1-10 11-20 21-30 not preferred at all count 8 12 9 29 1 15 5 21 residual 2.7 -5.9 3.2 -3.1 2.3 0.8 slightly preferred count 7 22 1 30 15 17 5 37 residual 1.5 3.5 -5.0 7.8 -5.4 -2.5 moderately preferred count 4 24 6 34 3 23 6 32 residual -2.2 3.0 -0.8 -3.2 3.6 -0.5 highly preferred count 0 6 5 11 0 11 3 14 residual -2.0 -0.8 2.8 -2.7 2.5 0.2 extremely preferred count 2 7 2 11 4 6 5 15 residual 0.0 0.2 -0.2 1.1 -3.1 2.0 total count 21 71 23 115 23 72 24 119 5.4.2 asset-based lending and debt financing (trade credit, leasing and loan from bank): smes aged between 11-20 years prefer asset-based lending and debt financing to other sources of financing. the findings show that leasing and loan from banks are the preferred options for smes that are aged between 11-20 years in the agriculture sector (residual = 3.1 for leasing; residual = 4.6 for loan from banks) and service sector (residual = 4.2 for loan from banks) of kp. through asset-backed financing, businesses can obtain funds on the basis of the value of a particular asset, that consists of inventory, real estate and equipment, machinery, accounts receivables etc., rather than on its own credit worthiness. in this manner, it can serve the requirements of small and young firms that face hurdles in accessing traditional loans. asset-backed loans provide flexible financing terms and in recent years has become increasingly popular specifically in countries with advanced financial expertise and support and efficient judicial and legal systems (cusmano & koreen, 2015). according to pakistan economic survey (pes) shahid ali, maryam jabeen, laila taskeen qazi, aisha jabeen 176 for the fiscal year 2018-19, agricultural credit advisory committee (acac) has set the indicative disbursement of loan for agriculture sector development targets to rs. 1,250 billion to 50 financial institutions in agriculture sector including 5 islamic banks, 19 commercial banks, 11 microfinance banks, 2 specialized banks and 13 rural support programs and microfinance institutions. to accommodate the needs of small farmers, rs. 35 billion has been assigned to 13 rural support programs and microfinance institutions, rs. 156 billion to 11 microfinance banks and rs. 13 billion to punjab provincial cooperative bank limited. table 14 leasing * age categories * business sectors table 15 loans from banks * age categories * business sectors table 16 loans from banks * age categories * business sectors agriculture agriculture service age categories total age categories total age categories total 110 1120 2130 110 1120 2130 110 1120 2130 not preferred at all count 2 3 4 9 2 3 1 6 5 2 4 11 residual 0.7 2.7 2.0 1.2 -0.8 -0.4 2.8 -4.5 1.7 slightly preferred count 3 6 1 10 1 8 3 12 7 12 6 25 residual 1.6 0.3 1.3 -0.6 0.5 0.1 2.0 -2.9 0.8 moderately preferred count 1 17 4 22 7 18 6 31 1 22 7 30 residual 2.1 3.1 1.0 2.9 -1.4 -1.5 -4.9 4.2 0.8 highly preferred count 3 24 9 36 0 19 4 23 7 27 6 40 residual 2.1 1.3 0.9 -3.0 4.6 -1.5 -0.9 3.2 -2.3 extremely preferred count 3 3 1 7 1 4 6 11 3 6 1 10 residual 2.0 1.4 0.6 -0.5 -2.9 3.3 1.0 0.1 -1.1 total count 12 53 19 84 11 52 20 83 23 69 24 116 the findings further show a high preference for trade credit (residual = 7.0) for smes operating in the industrial sector of kp. industrial sector includes smes from manufacturing, mining & quarrying, construction, electricity generation & distribution & gas distribution sectors. a report published by sbp compared financing to smes at a macro-level among the three subsectors: services, manufacturing and trading for the year 2009 to 2014. after december 2012, a rising trend has been recorded in growth in flow of credit to both the sme sector and manufacturing smes. the corresponding amount financed during 2009-2014 is rs. 121 billion and rs. 301 billion. at the end of december 2014, services, trading, and manufacturing accounted for 17.1 percent, 40.3 percent, and 42.5 percent, respectively, of the overall financing to the sme industry. financing to subsectors in service industry showed a similar trajectory, nonetheless the growth is more projecting for smes operating in the manufacturing sector. journal of applied economics and business studies, volume. 4, issue 3 (2020) 159-184 https://doi.org/10.34260/jaebs.437 177 table 17 trade credit * age categories * business sectors industrial age categories total 1-10 11-20 21-30 not preferred at all count 3 2 3 8 residual 1.6 -3.0 1.4 slightly preferred count 4 10 4 18 residual 0.8 -1.3 0.5 moderately preferred count 4 21 7 32 residual -1.7 0.9 0.8 highly preferred count 3 29 3 35 residual -3.2 7.0 -3.8 extremely preferred count 6 9 5 20 residual 2.5 -3.6 1.1 total count 20 71 22 113 5.4.3 government scheme: the findings in table 18 show a preference for government backed loans and schemes for smes aged between 1-10 years (residual = 3.6) in the service sector of kp. there are myriad of demand side and supply side issues that restraint credit to smes. lack of incentives, poor management of cash flow, lack of documentation and unawareness of financing products etc. are some of the reasons for low demand to bank finance from the smes. the biggest challenges, though, are the collateral requirements and information asymmetry that exists between the smes and the financial institutions. these issues are being addressed to an extent through initiatives taken by the government such as the credit guarantee schemes and through the financial institutions (secured transactions) act, 2016, for the registry of moveable and immovable assets. government and regulatory authorities should strengthen the laws and introduce reforms that reduce the credit risk in smes loans portfolios, lessen information asymmetry, offer remedies for collateral requirement and harness trust between smes and lending institutions. although the national assembly approved the financial institutions (secured transactions) act 2016, it needs to be implemented now by fast tracking the setup of an electronic registry for collateral that would bring in record moveable asset that can be used as collateral. the credit information bureau (cib) comprises all the loan transactions that has taken place, however, more work is required to provide detailed smes credit reports and gauge their credit worthiness. this is where an effective role can be played by the bureaus operating in the private sector. by expanding the work being done under the credit guarantee schemes, the program should be modified into an on-going concern basis, the department for international development (dfid) and the state bank of pakistan are presently working on an initiative that will result in the establishment of a corporation for loan guarantees. shahid ali, maryam jabeen, laila taskeen qazi, aisha jabeen 178 localized loan guarantee funds and private institutions with credit guarantee schemes should also be encouraged (aslam & sattar, 2017). table 18 government scheme * age categories * business sectors service age categories total 1-10 11-20 21-30 not preferred at all count 4 7 6 17 residual 0.7 -3.3 2.6 slightly preferred count 9 15 4 28 residual 3.6 -1.9 -1.6 moderately preferred count 3 23 7 33 residual -3.4 3.0 3 highly preferred count 6 19 2 27 residual 0.8 2.7 -3.4 extremely preferred count 1 8 5 14 residual -1.7 -0.5 2.2 total count 23 72 24 119 the sector-wise findings from the survey study confirms that smes across different sectors of kp follow the assumptions predicted by trade-off theory. the sector-wise results are also in the conformity of a positive association between the age of a firm and use of debt financing. in the context of developing economies, a large number of empirical studies report a positive relationship between age of a firm and access to financing from lending institutions. in comparison to young enterprises, older smes have good relationships and better track records with their lenders (abor, & biekpe, 2009). the positive relationship between age of a firm and leverage in smes has been hypothesized for the following reasons: firstly, older firms have reputable track record that enabled them to have improved access to external sources of financing. secondly, entrepreneurs who have been running their businesses over a long period have good relationships with bank officials, politicians and other firms enabling them to have improve access to debt finance (saarani, & shahadan, 2013; mat nawi, 2015). 6. conclusions the study has the motivation to comprehend the financing behavior of smes of kp at different life cycle stages. for this, a thorough literature review suggested designing a survey methodology to find answers to the posed research hypotheses. the study identified financially troubled smes from seven districts of kp and sampled firms from industrial, agricultural, and services sectors. the survey was designed by putting emphasis on every detail and 377 firms were aimed to be sampled. the survey actually covered 341 firms by generating a 9.5% of nonresponse. the findings show that the journal of applied economics and business studies, volume. 4, issue 3 (2020) 159-184 https://doi.org/10.34260/jaebs.437 179 owner-managers’ personal savings are preferred by enterprises that are aged between 010 years. however, smes with 11 to 20 years of existence seem to take up more debt than entrepreneurs’ capital. the industry breakdown confirms the robustness of our results. smes in the service sector prefer internal funds and government backed loans at the start-up stage. however, the enterprises across all the sectors of the economy, agriculture, industrial and service, aged between 11-20 years prefer asset-based lending and debt financing. the findings from the survey study confirm that smes across different districts and business sectors of kp follow the assumptions predicted by tradeoff theory. as reflected by survey findings, smes in kp have their own inadequacies and limitations such as the organizational structure and decision making styles, small-scale operations, high level of dependence on owner, poor management and lack of brand identity, unfavorable balance of power and shortage of surplus funds. the survey findings imply that smes operating in kp can join hands to become a network of member firms to nurture exports and make a common offer to large set of clients as retail houses. to increase outreach, they should diversify their product range for the population of low-income groups. future studies should consider in-depth qualitative approaches to gain better insights on exploring the smes financing gap that can be bridged by an interplay of lending institutions and other stakeholders. detailed interviews from smes or credit lending institutions can provide a sound theoretical base to identify the demand side problems in obtaining funds from the lending institutions by catering to business pressures and taking into account the external environment. this, in view of supply side responses can lead to a formal financing model development for the smes of kp. acknowledgement this paper is an extraction from a nrpu funded project of higher education commission of pakistan. shahid ali, maryam jabeen, laila taskeen qazi, aisha jabeen 180 references abbasi, w. a., wang, z., & abbasi, d. a. (2017). potential sources of financing for small and medium enterprises (smes) and role of government in supporting smes journal of small business and entrepreneurship development, 5(2), 3947. abor, j., & biekpe, n. (2009). how do we explain the capital structure of smes in subsaharan africa? evidence from ghana. journal of economic studies, 36(1), 83-97. afraz, n., hussain, s. t., & khan, u. (2014). barriers to the growth of small firms in pakistan: a qualitative assessment of selected light engineering industries. the lahore journal of economics, 19, 135. ampenberger, m., schmid, t., achleitner, a. k., & kaserer, c. (2013). capital structure decisions in family firms: empirical evidence from a bank-based economy. review of managerial science, 7(3), 247-275 ang, j. s. (1992). on the theory of finance for privately held firms. the journal of entrepreneurial finance, 1(3), 185-203. aslam, a. k., & sattar, h. (2017). nature and characteristics of sme financing and npls in pakistan. benhamou, e., & melot, v. (2018). seven proofs of the pearson chi-squared independence test and its graphical interpretation. arxiv preprint arxiv:1808.09171. berger, a. n., & udell, g. f. (1998). the economics of small business finance: the roles of private equity and debt markets in the financial growth cycle. journal of banking & finance, 22(6-8), 613-673 bhama, v., jain, p. k., & yadav, s. s. (2018). relationship between the pecking order theory and firm's age: empirical evidences from india. iimb management review, 30(1), 104-114. carey, d., & flynn, a. (2005). is bank finance the achilles' heel of irish smes?. journal of european industrial training. cusmano, l., & koreen, m. (2015). new approaches to sme and entrepreneurship financing: broadening the range of instruments. pariz: oecd centre for entrepreneurship, smes and local development. journal of applied economics and business studies, volume. 4, issue 3 (2020) 159-184 https://doi.org/10.34260/jaebs.437 181 de la torre, a., soledad martinez peria, m., & schmukler, s. l. (2008). bank involvement with smes: beyond relationship lending. the world bank fort, t. c., haltiwanger, j., jarmin, r. s., & miranda, j. (2013). how firms respond to business cycles: the role of firm age and firm size. imf economic review, 61(3), 520-559. franke, t. m., ho, t., & christie, c. a. (2012). the chi-square test: often used and more often misinterpreted. american journal of evaluation, 33(3), 448-458. hallberg, k. (2000). a market-oriented strategy for small and medium scale enterprises. the world bank. hoang, t. x., pham, c. s., & ulubaşoğlu, m. a. (2019). institutions for private sector development and pro‐poor growth: evidence from vietnam. economics of transition and institutional change, 27(3), 699-728. holmes, s., & kent, p. (1991). an empirical analysis of the financial structure of small and large australian manufacturing enterprises. journal of small business finance, 1(2), 141-154. hulin, c., netemeyer, r., & cudeck, r. (2001). can a reliability coefficient be too high?. journal of consumer psychology, 10(1/2), 55-58. irwin, d., & scott, j. m. (2010). barriers faced by smes in raising bank finance. international journal of entrepreneurial behavior & research. jones, r. s., & lee, j. w. (2018). enhancing dynamism in smes and entrepreneurship in korea. keasey, k., & mcguinness, p. (1990). small new firms and the return to alternative sources of finance. small business economics, 2(3), 213-222 klapper, l. f., sarria-allende, v., & sulla, v. (2002). small-and medium-size enterprise financing in eastern europe. the world bank. kraemer-eis, h., & lang, f. (2012). the importance of leasing for sme finance (no. 2012/15). eif working paper la rocca, m., la rocca, t., & cariola, a. (2009). small business financing. financial preferences throughout the life cycle of a firm. univercity of calabria (italy) lasfer, m. a., & levis, m. (1998). the determinants of the leasing decision of small and large companies. european financial management, 4(2), 159-184 shahid ali, maryam jabeen, laila taskeen qazi, aisha jabeen 182 lewis, v. l., & churchill, n. c. (1983). the five stages of small business growth. harvard business review, 61(3), 30-50. li, k., griffin, d., yue, h., & zhao, l. (2011). national culture and capital structure decisions: evidence from foreign joint ventures in china. journal of international business studies, 42(4), 477-503. lopez-gracia, j., & aybar-arias, c. (2000). an empirical approach to the financial behaviour of small and medium sized companies. small business economics, 14(1), 55-63. mac an bhaird, c. (2010). the modigliani–miller proposition after fifty years and its relation to entrepreneurial finance. strategic change, 19(1‐2), 9-28. mat nawi, h. (2015). determinants of capital structure in small and medium sized enterprises in malaysia (doctoral dissertation, brunel university london). miglo, a., & miglo, v. (2019). market imperfections and crowdfunding. small business economics, 53(1), 51-79. miller, r. l. (2017). spss for social scientists. macmillan international higher education. miller, d., & friesen, p. h. (1984). a longitudinal study of the corporate life cycle. management science, 30(10), 1161-1183. modigliani, f., & miller, m. h. (1958). the cost of capital, corporation finance and the theory of investment, in “american economic review”, n. 48. ndiaye, n., razak, l. a., nagayev, r., & ng, a. (2018). demystifying small and medium enterprises’(smes) performance in emerging and developing economies. borsa istanbul review, 18(4), 269-281. nenova, t., niang, c. t., & ahmad, a. (2009). bringing finance to pakistan's poor: a study on access to finance for the underserved and small enterprises. world bank. ono, a., & uesugi, i. (2009). role of collateral and personal guarantees in relationship lending: evidence from japan's sme loan market. journal of money, credit and banking, 41(5), 935-960 penrose, e. t. (1952). biological analogies in the theory of the firm. the american economic review, 42(5), 804-819. penrose, e. (1959). 1959 the theory of the growth of the firm oxford: blackwell. journal of applied economics and business studies, volume. 4, issue 3 (2020) 159-184 https://doi.org/10.34260/jaebs.437 183 rana, r., & singhal, r. (2015). chi-square test and its application in hypothesis testing. journal of the practice of cardiovascular sciences, 1(1), 69. saarani, a. n., & shahadan, f. (2013). the determinant of capital structure of smes in malaysia: evidence from enterprise 50 (e50) smes. asian social science, 9(6), 64. serrasqueiro, z., & caetano, a. (2015). trade-off theory versus pecking order theory: capital structure decisions in a peripheral region of portugal. journal of business economics and management, 16(2), 445-466. shaikh, s., & khoso, i. u. (2019). impact of financial constraints on the performance of smes a study of sindh province. journal of independent studies & research: management & social sciences & economics, 17(1). steijvers, t., voordeckers, w., & vanhoof, k. (2010). collateral, relationship lending and family firms. small business economics, 34(3), 243-259. ursachi, g., horodnic, i. a., & zait, a. (2015). how reliable are measurement scales? external factors with indirect influence on reliability estimators. procedia economics and finance, 20, 679-686. usai, a., scuotto, v., murray, a., fiano, f., & dezi, l. (2018). do entrepreneurial knowledge and innovative attitude overcome “imperfections” in the innovation process? insights from smes in the uk and italy. journal of knowledge management. vandenberg, p., yoshino, n., goto, a., intarakumnerd, p., & miyamoto60, j. (2016). policies to enhance sme internationalization. integrating smes into global value chains, 117. vera, d., & onji, k. (2010). changes in the banking system and small business lending. small business economics, 34(3), 293-308. shahid ali, maryam jabeen, laila taskeen qazi, aisha jabeen 184 appendix 1 sme’s questionnaire survey the purpose of this study is to explore the factors that might underlie firms’ choice of mix between debt financing and equity financing of smes owner-manager in khyber pakhtunkhwa in order to examine the financing of through a financial growth life cycle paradigm. section 1: general information & business information: 1) gender of the entrepreneur • male (---) • female (---) 2) in which year was the business established? ----------------------------------------------------------------- (age of business enterprise) 3) location of the business: ----------------------------------------------------------------- 4) please classify your main business under the following sectors? • agriculture • industrial (mining & quarrying; manufacturing; electricity generation & distribution & gas distribution and construction) • service (wholesale & retail trade; transport, storage and communication; finance & insurance) 5) how will you rank the following sources of capital in terms of preference? not preferred at all slightly preferred moderately preferred highly preferred extremely preferred owner‘s saving 1 2 3 4 5 retained earnings 1 2 3 4 5 family and friends 1 2 3 4 5 angel financing 1 2 3 4 5 venture capital 1 2 3 4 5 trade credit 1 2 3 4 5 leasing 1 2 3 4 5 hire purchase 1 2 3 4 5 loan from bank 1 2 3 4 5 government scheme 1 2 3 4 5 source: world bank (2014) journal of applied economics and business studies, volume. 4, issue 4 (2020) 215-236 https://doi.org/10.34260/jaebs.4410 215 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x an appraisal of the role of islamic banking development and economic growth talat hussain1, noman arshed2* and rukhsana kalim3 1assistant professor, department of banking and finance, university of management and technology, lahore 2lecturer, department of economics, university of management and technology, lahore 3professor, department of economics, university of management and technology, lahore abstract literature is well-versed with the contribution of financial inclusion from the deposit and financing size and its role in economic growth. these contributions include a boost in economic transactions and efficient resource mobilization. islamic financial system is different from conventional banking as it distributes the risk equitably and promotes fairness in dealings. it helps in the integration of business gains as a borrower of islamic capital with the earnings of savers as depositors. this study has proposed two channels via which islamic financial development may incur growth. first is bank financing penetration, and second is depositor financial inclusion. based on the data of 41 full-fledged islamic banks between 2012 and 2017, the results show that both increases in bank and depositor returns have a growth-promoting effect. this prompts the policymakers with new insights. policymakers should increase islamic banking penetration to different sectors and regulate for increased extraction of the depositor contribution from the banking financing activity. keywords capital structure; financing penetration; financial inclusion; income integration 1. introduction economies enjoyed rapid growth with the introduction of currency as a common medium of exchange created ease in long-distance trading and storing value (mishkin, 2007). this fiat currency allowed economies to specialize in production rather than focus on the production of necessities to avoid meet coincidence of wants (kiyotaki & wright, 1989; ritter, 1995). sollow (1956) illustrated that a certain level of investment is required to counter the depreciation and needs of capital. such that without it, the current infrastructure will eventually fail to support the population needs. here the * noman.arshed@umt.edu.pk talat hussain, noman arshed and rukhsana kalim 216 investment is the source of injection of the resources into productive sectors by businesses and households, which were leaked out of the economy’s cash flow in the form of household saving (daraban, 2010). the invention of a common medium of exchange means that people now have two forms of wealth, first is in physical goods and services form. second is in terms of the medium of exchange, which people and households can use to save their wealth for more extended periods (mishkin, 2007). however, the real advantage is that people can access their future wealth today (by borrowing), or they can send their present wealth into the future (by lending)1. the circular flow model of 3 sector economy, in figure 1 depicts that, households need the assistance of the financial market in order to mobilize the saving in the economy. which can be used in the form of the firm, government and foreign borrowing (parkin, 2016). since economies have adapted and specialized, it led to the division of people into households (buyers) and firms (producers), and with the introduction of medium of exchange, the 3rd sector of financial market can become arbitrageur between the fund provider (savers/investors) and fund demanders (firms/government). in this case, every household cannot be an entrepreneur/producer; this leads to the inability of the households to mobilize the savings themselves efficiently (abel, 2014). this resorts them to depend on the saving mobilization mechanism provided by the financial markets in the economy. the use of idle money, which was initially kept in temples for safekeeping, led to the evolution of the banking system. priests in temples used to lend the money to the people who needed for some compensation. this compensation was later shared with the original owner to wait for the lent money to be returned. this arrangement later evolved into banks, which were known as financial institutes that accept deposits and make loans (mishkin, 2007). so, banks use the idle resources of the households and lend it to the firms who are willing to pay for it; this becomes earning of the bank, which it shares with the depositors. fig. 1 – 3 sector circular flow model (parkin, 2016). 1 irving fisher theory of intertemporal choice. (mankiw, 2014) journal of applied economics and business studies, volume. 4, issue 4 (2020) 215-236 https://doi.org/10.34260/jaebs.4410 217 under such situations, financial intermediaries/institutions play an important role in injecting the surplus income in the circular flow. it facilitates the people with excess resources (lenders/savers) with returns at reduced risk of default by diversifying and sharing of risk and addressing the asymmetric information problems. while it helps the people with the shortage of resources (borrowers) with low-cost liquidity services, furthermore, it facilitates specialization and technological innovations in businesses (levine, 1997). the more the financial institutions are integrated within the economy and across the world, the better portfolio they will be able to make, which will reduce the systemic risk (tarashev, borio, & tsatsaronis, 2009). demirgüç-kunt and maksimovic (1999) provided evidence that economies having more robust financial and banking sector tend to have a stable supply of finances to firms. this enables them to access long-term debt. better financial institutions also help the government to mobilize resources quickly and at a lower cost (greenbaum & thakor, 1995; mishkin, 2007). the study by (gurley & shaw, 1955) first discussed the contribution of the financial sector in economic development which is then followed by (mckinnon, 1973; santomero, 1984). with time, the financial system developed as per the need. as the financial system developed under the patronage of the western world, it lacked the underlying spirit and guidelines of the islamic system. these conventional banks became accustomed to using interest-based means, which ensured their returns irrespective of the outcomes of investments. this system was ideal for the fund suppliers (savers) and harmful for the fund demanders (firms). the foundation of the islamic capital market ensures the creation of difference in lives of people by providing valuebased business support leading to a reduction in inflation, job creation and social prosperity (wief, 2014). commercial banks are the most crucial component of financial intermediaries. they ensure that the savings are efficiently channelled to the most efficient investment options to induce growth (patrick, 1966). the uniqueness of commercial banks is that they specialize in short-term financing, while the other non-banking financial institutions like investment houses specialize in long term financing (ebrahim & joo, 2001). this conventional financial system has been working for more than 600 years. this conventional financial system has contributed to producing 358 billionaires while keeping 1.3 billion people below the poverty line. this signifies that the conventional financial system is mature, but it is not equitable (yakcop, 2002). nowadays, islamic banks present themselves as a responsible banking variant of commercial banks. the evolution of islamic banks is transcribed to follow a completely different path rather than a coincidence for the case of conventional banks. the early foundations of islamic banking are derived from al qur’an and sunnah under the subject of fiqh al-muamalat. the early era of islamic banking is observable under the rule of caliphs. the initiative of a practical version of islamic banking in the modern era started in 1963 with the establishment of an islamic bank in egypt. further organization of islamic conference (oic) worked in the development of the financial talat hussain, noman arshed and rukhsana kalim 218 system in 1973 (ahmed, 2014; hassan & lewis, 2009). other pioneers in islamic banking is seen in philippine in 1973 (patrick & moreno, 1982), dubai in 1975 (ariss & sarieddine, 2007), in sudan and egypt (alharbi, 2015; fada & weabekwa, 2012), bahrain in 1979 (abdulla, 2016) and pakistan in 2002 (ansari & rehman, 2011). currently, countries like bahrain, qatar, indonesia, saudi arabia, malaysia, uae, turkey, kuwait and pakistan (qismut +3) are leaders in islamic banking with industry assets of us$ 920 billion in 2015 (ey, 2016). islamic banks are the type of commercial banks whose objectives are set by the moral code provided by islamic law (shari’ah). this banking system is not free in terms of defining new transaction mechanism that seems fit. here islamic banking system follows the boundaries set by islamic law which defines all the rules of faith (aqedah), prayers (ibadah), social dealings (muamalat) and habits (aadat). the rulings for islamic banks come under the umbrella of social dealings and the study that interprets the rulings is called fiqh al-muamalat. the foundation of islamic moral code for finance is put forward by al qur’an (2:261, 2:265, 2:270, 2:274). these mentioned verses of al qur’an depict that there is a code which needs to be followed in financial transactions, and allah rewards those who follow it. islamic banks are structured in such a way that removes all the aspects which spoil the relationship between humans. like in riba, where lender exploits the borrower for his needs without any concern to the usage of funds and sharing of risks (ebrahim & joo, 2001). this entitlement is impermissible as per islamic legal maxim narrated by holy prophet ملسو هيلع هللا ىلص “entitlement of profit depends upon the liability of loss” (sunan abi-dawud, 3508). compared to this, the lending mechanism of conventional banks transfers the total risk of loss in investment on the borrower while claiming unjustified rights to return (mansoori, 2012). as per al qur’an (2:264), there is no reward for spending other than what allah has prescribed. while studying the code set by islam, it will be noticeable that islam promotes trade and entrepreneurship (al qur’an, 2:275). here financiers are partners instead of having a relationship of creditor and debtor. islam provides directions to humans through maqasid al-shari’ah, where major categories of directions are provided for ibadah and muamalat. the principles of muamalat are designed to guide how muslims can earn their provisions (rizq) which are guaranteed by allah, as mentioned in al qur’an (11:6, 17:29, 25:67, 65:7). the economic implication of promoting muslims to participate in fair trade and business is that it leads to the integration of the muslim world and generation of economic surplus, which is curtailed in the prescribed rules in shari’ah. (wief, 2014, 2016) iterated that if the documentation of trading is standardized, the inclusion of smes and integration of halal business with the islamic finance industry, then the growth of oic countries can be boosted from inter and intra-region trade. prophet muhammad ملسو هيلع هللا ىلص recorded several instances in his life, which indicate the importance of working of sustenance, also indicated that the sustenance (rizq) from trade is nine times journal of applied economics and business studies, volume. 4, issue 4 (2020) 215-236 https://doi.org/10.34260/jaebs.4410 219 higher than working for others. narrated by abu huraira, allah's messenger (ملسو هيلع هللا ىلص) said, "the prophet daud a.s. used not to eat except from the earnings of his manual labour." (sahih al bukhari, 2073) in all forms of halal income sources, trade and business are considered best as discussed in (al qur’an, 4:29). usmani (1954) indicates that the purpose of only discussing trade in this ayat is to assert that in all means of sustenance, trade is superior. the development of the islamic banks is, therefore, a step to provide the facility of trade and investment under compliance with the canons of islam. here the banking and finance needs are fulfilled by ensuring permissible means which are proved to be less risky and are robust to financial shocks (abduh, omar, & duasa, 2011; how, abdul karim, & verhoeven, 2004). because of such features, the acceptance of islamic banks is growing in western economies, with the growth rate of about 10-15% per annum. this islamic system has now been introduced in more than 51 countries (solé, 2007). the idea of islamic banks is based on the principle of equitable sharing the surplus of business investment, rather than a fixed portion of surplus for a lender under conventional finance2. this equitable sharing in islamic finance is expected to increase the welfare of the lenders. narrated abu huraira: the prophet (ملسو هيلع هللا ىلص) said, "a time will come when one will not care how one gains one's money, legally or illegally." (sahih al-bukhari, 2059) 1.1. present scenario of islamic banking currently, islamic banks have experienced 16% growth in their assets from us$490 billion to us%882 billion, in which gulf cooperation council (gcc) countries enjoyed 34% growth3. according to the report of ey (2016), islamic banking has contributed 69% to gcc growth in 2014. figure 3 depicts the stable growth in market share of bahrain, while saudi arabia, kuwait and qatar are returning after a slump in 2013. as compared to asset growth of conventional banking, islamic banking has outgrown its counterpart in saudi arabia, malaysia, kuwait, qatar, indonesia, bahrain and pakistan for the year 2014. banks from saudi arabia, uae and qatar are market leaders with equity more than us$2 billion. similarly, banks from saudi arabia, malaysia, qatar, turkey and uae have experienced more than 10% returns on equity between 2010 and 2014. other than turkey and uae, where convention financing growth is higher than islamic financing growth shown in figure 4, islamic financing growth is towering in all major markets. this indicates that every year islamic financing is claiming a higher share in the market. even though islamic financing is showing promising growth, but still there are challenges which are bottlenecks to realize its true potential. islamic banks were initially 2 since lending interest rate is fixed and predetermined, lender cannot earn more surplus if the profit of the investment is higher than expectations. this borrower end up enjoying higher than equitable share of surplus. 3 ernst and young denote islamic banks as participatory banks in order to avoid differentiation between the participatory banks of turkey and islamic banks of rest of muslim countries. talat hussain, noman arshed and rukhsana kalim 220 designed to cater to the partnership based modes of finance (musharaka and mudarabah). however, it later developed trade based modes of finance (ijarah, salam, murabaha etc.) to address the issues and special needs of the economy (usmani, 2002). the partnership-based modes of finance were designed to transfer the surplus from the big firms to the individuals, ensuring welfare generating a distribution of income, as advised in al qur’an (59:7)4. rosly and zaini (2008) stated that because of the fixed returns to depositors in conventional banks, their return to deposit is lower than the return to equity. in contrast, the higher risk in investments like mudarabah has the potential to increase the return to deposits. while islamic banks have imitated some products (trade-based) for special needs and conditions, mass utilization of these products depicts over lenient treatment to the businessmen. fig. 3 – trends in the market share growth of islamic banking. source (ey, 2016) 4 “… so that it (the wealth) may not circulate among those who are rich among you” al qur’an (59:7) journal of applied economics and business studies, volume. 4, issue 4 (2020) 215-236 https://doi.org/10.34260/jaebs.4410 221 fig. 4 – country wise financing growth comparison in 2014. source: ey (2016) 1.2. modes of islamic banking deposits and financing there are few forms of contracts available which are used to accept deposits such as mudaraba5 for deposit in saving account, wakalah (agency)6 for portfolio investment accounts and qard e hassan as a current account (ahmed, 2011). other than attracting deposits, banks can engage clients from the financing side shown in table 1. the specificity of the contracts in islamic finance which differs from conventional banking and finance (arshed & kalim, 2020). all of these contracts have different implications to the depositors and the bank returns they receive from the investment pool. hence, changes in these social (return to deposit) and private (return to the bank) indicators of banking sector development are needed to be explored using global data set (narayan & phan, 2019). table 1 – types of contracts unilateral contracts uqud al-tabarru’at (gratuitous contract) hiba (gift), ibra (off-set of debt), wasiyyah (bequest), waqf (endowment) and qard (loan) bilateral contracts uqud al-tijjariah (trading contract) ijarah (rental leasing), salam (agri financing), istisna’a (manufacturing contact), istijrar, murabaha (cost plus sale) uqud al-musharakah (partnership contract) musharaka (partnership), mudaraba (trust financing), musaqat and muzarah’ah uqud al-khadmaat (services contract) kafalah (guarantee), wakalah (agency), rahn (mortgage) and hawalah (transfer) source: (aziz, 2013; mansoori, 2009) 5 in the mudarabah contract one partner (rabb ul mal) invests capital and other partner invests his entrepreneurial or management skills (mudarib) (usmani, 2002). 6 wakalah contract makes bank agent of the depositor to manage the investment fund on his behalf. client charges a nominal fee for services on the returns generated on the investment (ahmed, 2011). -20% -10% 0% 10% 20% 30% bahrain indonesia kuwait malaysia pakistan qatar saudi arabia turkey uae financing growth in 2014 islamic financing conventional financing talat hussain, noman arshed and rukhsana kalim 222 1.3. objectives the objective of the study is to determine, whether the islamic banking sector development leads to the economic growth of the economy while controlling for the endogenous growth model (i.e. labor, capital and human capital) and governance. this study has proposed indicators like the deposit returns and bank returns as an instrument of islamic banking sector development. this study proposes a new indicator to assess the islamic banking sector development, which represents the private and public objective maximization. the assessment will ascertain whether adherence to the objectives of the islamic business leads to economic progress. 2. literature review development of islamic banking sector as a part of developing the financial sector will influence growth in many ways. it will increase the capacity of the firms to acquire an investment and incentivize the household to save (goldsmith, 1955). the relationship between economic growth and financial sector development can be explained by two hypotheses. first is the supply leading theory proposed by (schumpeter, 1911), where development in the financial sector leads to economic development. and second is the demand following theory (robinson, 1952) where economic development leads towards financial development. following these two studies, several studies built on the relationship between financial development and growth, like (apergis, filippidis, & economidou, 2007; levine, 1997; mckinnon, 1973; shaw, 1973; shen & lee, 2006). the most recent studies by (bongini, iwanicz-drozdowska, smaga, & witkowski, 2017; durusu-ciftci, ispir, & yetkiner, 2017; ghosh, 2017; hassan & kalim, 2017; webb, grace, & skipper, 2017) coined that, the level of development of the economy defines how banking and stock market development (jointly known as financial sector development) affects economic development. the rules set for the countenance of allah in islamic finance guarantee the growth in wealth while avoiding the interest (al qur’an, 2:276, 3:130, 30:39). al qur’an (2:261) gives the example of the growth of wealth if it is spent in the way of allah, that it is like a grain seed which grows into seven branches and each branch yielding 100 grains. the spending mechanism ensures that there is moderation (al qur’an 17:29) and equal distribution of income (al qur’an, 59:7) in society. there are certain advantages of islamic financial system which makes them growthpromoting. these advantages are, the promotion of justice and objectivity (al qur’an 4:135)7, liquidity ease in hardship (al qur’an 2:280)8, transparency standards (al qur’an 2:282)9, sharing of risk (ibn-qudamah, 1983; siddiqui, 2001), stabilize inflation 7 “o you who have believed, be persistently standing firm in justice, witnesses for allah, even if it be against yourselves or parents and relatives. whether one is rich or poor, allah is more worthy of both. so follow not [personal] inclination, lest you not be just. and if you distort [your testimony] or refuse [to give it], then indeed allah is ever, with what you do, acquainted.” al qur’an (4:135) 8 “and if someone is in hardship, then [let there be] postponement until [a time of] ease. but if you give [from your right as] charity, then it is better for you, if you only knew.” al qur’an (2:280) 9 “o you who have believed, when you contract a debt for a specified term, write it down. and let a scribe write [it] between you in justice. let no scribe refuse to write as allah has taught him. so let him write and let the one who has the obligation dictate. and let him fear allah, journal of applied economics and business studies, volume. 4, issue 4 (2020) 215-236 https://doi.org/10.34260/jaebs.4410 223 (paramanik, 1993) and avoidance of interest-based exploitative tools (sayedi, 2017; zamil, 2014). while comparing the islamic financial or banking sector development and growth, there are a limited number of studies which explored this relationship. comparing two studies for the case of malaysia proposed opposite findings. a study by (furqani & mulyany, 2009) compared islamic bank financing and real gdp per capita. it concluded that demand following phenomenon is prominent. while a study by (majid & kassim, 2010) favoured the supply leading view. opposing to uni-directional or bi-directional causality, goaied and sassi (2010) selected 16 countries in the middle east and north africa (mena) region and using gmm estimation technique this study provides insignificant evidence for any contribution of the islamic banking system on economic growth. few studies like (abduh & omar, 2012; furqani & mulyany, 2009; majid & kassim, 2010; anggraini, 2019; naz & gulzar, 2020), used error correction model for the case of several countries and concluded that there is bi-directional causality between the total financing of the islamic banking and economic growth. similarly, kalim, mushtaq, and arshed (2016) explored the effects of islamic financing like ijarah and murabaha on economic growth using ardl cointegrating bounds approach for the case of pakistan. they concluded that increase in islamic financing to influence growth in long run, especially financing in ijarah. tabash and anagreh (2017) studied the case of uae and showed that islamic banking causes an increase in growth, fdi and investment. a similar outcome was witnessed in the case of indonesia (mansur, 2019). hafnida et al. (2015) compared 5 popular islamic financing products and their effect on economic growth in order to assess the gains of different forms of deposit mobilization. kalim and arshed (2018) compared 4 popular models like musharaka, mudarabah, ijarah and murabaha in terms of efficiency to achieve social goals. the results showed that the deposit mobilization via musharaka and mudarabah is effective in increasing the efficiency in the process of achieving public benefit from the islamic bank. arshed et al. (2020) extended this model to compared 7 islamic financing modes against economic growth in a panel data setup of 9 economies. this study provided the profile of popular financing modes and their connection with economic progress. the outcome of the study showed that salam, wakala and musharaka financing are the top more effective financing models for economic growth. for the case of controlling variables, solow (1956) pioneered in determining the long run growth model using the production function approach. he proposed physical his lord, and not leave anything out of it. but if the one who has the obligation is of limited understanding or weak or unable to dictate himself, then let his guardian dictate in justice. and bring to witness two witnesses from among your men. and if there are not two men [available], then a man and two women from those whom you accept as witnesses so that if one of the women errs, then the other can remind her. and let not the witnesses refuse when they are called upon. and do not be [too] weary to write it, whether it is small or large, for its [specified] term. that is more just in the sight of allah and stronger as evidence and more likely to prevent doubt between you, except when it is an immediate transaction which you conduct among yourselves. for [then] there is no blame upon you if you do not write it. and take witnesses when you conclude a contract. let no scribe be harmed or any witness. for if you do so, indeed, it is [grave] disobedience in you. and fear allah. and allah teaches you. and allah is knowing of all things.” al qur’an (2:282) talat hussain, noman arshed and rukhsana kalim 224 labor and physical capital as major inputs, in which increasing one by keeping other constant leads to diminishing returns. mankiw, romer, and weil (1992) extended this model by proposing the role of knowledge and intellect, this model is known as human capital augmented solow growth model and later on named as endogenous growth theory (aghion, caroli, & garica-penalosa, 1999; aghion & howitt, 1992). empirically, this human capital has been instrumented using health, education or sanitation related indicators (babatunde, 2012; barro, 1996; bloom, canning, & sevilla, 2004; hanif & arshed, 2016; inabo & arshed, 2019). almost all of the empirical studies which compared islamic banking and economic growth used quarterly data and aggregated time series analysis because of the limitation of data. this study will use bank-wise multicounty data of islamic banks. this allows us to incorporate the dynamics of internationalization and financial integration which time-series studies often ignore. 3. model this study proposed two channels (in figure 5) which islamic banking development may lead to economic growth. these two indicators are inspired by the objective of the islamic firm (i.e. maximization of private and social benefits) (amin & yusof, 2003; yusof & amin, 2007). here an increase in social / depositor benefits will increase saving returns of small income group leading to growth. while an increase in the banking returns, it allows the bank to sustain higher risk and returns based capital structure. this will allow banks to penetrate new markets for financing. the previous model of supply of ijarah and supply of musharaka shows that bank profitability encourages the bank to increase supply of ijarah and musharaka. consequently, banks financing is increased. further, there is evidence that an increase in musharaka and ijarah lead to economic growth (kalim et al., 2016). fig. 5 – model of effects of financial sector development this study has used the banking returns, and depositor returns are in a relative form, which indicates rules out the role of volume of banking returns and depositor returns to avoid endogeneity. rather this study talks about the maturity of the banking sector, which increases would increase the proportion of returns, leading to growth. under the premise of assessing the effects of private and public objective maximization, this study has not compared the islamic banking sector with conventional banking returns increase financial sector developme depositor returns increase increase in growth increase in bank financing increase in financial inclusion journal of applied economics and business studies, volume. 4, issue 4 (2020) 215-236 https://doi.org/10.34260/jaebs.4410 225 banks because of few stylized facts. the earning structure of the islamic and conventional banks are different, a comparative model will not be able to assess the effectiveness of islamic banks. the earning structure of conventional banks are based on riba for which there is a decree that it decreases wealth (al qur’an 2:276) and increases the righteous earning (al qur’an 2:261). hence theoretically, there is no need to compare both modes of earning and its effect on growth. 4. methods 4.1. sample in order to achieve the set objectives, this study has access to the financial statements of 41 full-fledged islamic banks around the globe for the time period of 2012-2017. the list of countries and banks are mentioned in the appendix. 4.2. estimation equation based on the literature regarding banking development and economic growth. following stochastic equation is proposed by the study. 𝐆𝐃𝐏𝐜𝐭 = 𝛂𝟎 + 𝛂𝟏𝐑𝐎𝐁𝐢𝐭 + 𝛂𝟐𝐑𝐎𝐃𝐢𝐭 + 𝛂𝟑𝐋𝐋𝐜𝐭 + 𝛂𝟒𝐋𝐊𝐜𝐭 + 𝛂𝟓𝐇𝐊𝐜𝐭 + 𝛂𝟔𝐆𝐎𝐕𝒄𝒕 + ɛ𝐢𝐭 here, in the model of economic growth, the dependent variable used in the study is economic growth estimated using the natural log of gdp per capita. here the variables used are deposit returns (rod), bank returns (rob), labor (ll), capital (lk), human capital (hk), government quality (gov), i is representing banks, c is representing countries and t is representing years. while ɛ𝑖𝑡 and 𝜇𝑖𝑡 denotes the randomly distributed factors which could determine the dependent variable, which are not included in this study and which are assumed to have a negligible effect on the dependent variable within the domain of the constructed model. the data is collected from world development indicators (wdi, 2019), worldwide governance indicators (wgi, 2019) and financial statements of banks extracted from the respective bank website. 4.3. estimation method since in the models based on panel data, which this study has built, the data has spatial-temporal variation. estimating pooled ols with over-restrictive assumptions will lead to heteroskedasticity in estimation results. this study will opt for panel feasible generalized least square (fgls) approach, which allows the bank level differences to change standard errors of each coefficient (greene, 2012; hassan et al., 2019). further to control from the cross-sectional heteroskedasticity and time series heteroskedasticity, this study will use country and time dummies (jaccard & turisi, 2003; jaccard, wan, & turisi, 1990). 5. estimation results 5.1. descriptive statistics while exploring the variables, table 2 provides descriptive statistics. for all the variables, only government quality has a standard deviation more than the mean value talat hussain, noman arshed and rukhsana kalim 226 showing that this variable is over-dispersed across the banks globally. while all other variables are under dispersed owing to mean value greater than the standard deviation. this study used the panel variant of (jarque & bera, 1987) normality test proposed by (alejo, galvao, montes-rojas, & sosa-escudero, 2015) to test the normality of the variables across cross-sections and time periods. here for the case of cross-section normality, rob, rod and ll are normal at 5%, and gdp is normal at 1%, this means that these are not skewed, and there are no outliers when assessed across banks and countries. while for the case of time-series normality, gdp, lk and hk are normally distributed at 5%, this means that these variables are not skewed, and there are no outliers when assessed across time. all other variables are non-normal, which may have caused because of the presence of outliers in the data, and there is a mixture of under and over-dispersed variable, the pooled ols model seems unfeasible10. table 2 – descriptive statistics of islamic banking development obs. mean std. dev. cross-sectional normality time normality gdp 932 8.84 1.42 7.21 (0.03)* 2.80 (0.24) rob 387 12.39 2.24 4.94 (0.08) 22.47 (0.00)* rod 431 12.93 2.42 2.74 (0.25) 12.30 (0.00)* ll 990 16.07 1.52 5.44 (0.07) 31.18 (0.00)* lk 920 3.08 0.23 16.09 (0.00)* 3.03 (0.22) gov 990 -0.18 0.66 14.57 (0.00)* 51.40 (0.00)* hk 737 1.32 0.43 6.76 (0.03)* 4.68 (0.10) * significant at 5% figure 6 provides the graphical association between the gdp per capita and return to deposit and return to bank using time-averaged data. here, it is evident that an increase in returns to deposit and return to the bank are positively associated with the growth. further, the correlation of return to bank and growth are stronger, it may be because of direct link with the economic activity. figure 6 – returns and economic growth 10 while the variables can be used in estimation as they are asymptotically normal (lind, marchal, & wathen, 2006). journal of applied economics and business studies, volume. 4, issue 4 (2020) 215-236 https://doi.org/10.34260/jaebs.4410 227 5.2.multicollinearity test to detect the present of multicollinearity, gujarati, porter, and gunasekar (2012) illustrated the vif test. presence of correlated independent variables (multicollinearity) would make regression estimates biased. the decision criterion is that if vif < 10, then there is no evidence that independent variables are collinear to each other. table 3 provides the vif values for each independent variable pairs. since none of them is above 10, it confirms the absence of multicollinearity and un-biasness of estimations. table 3 – vif of islamic banking development model vif of growth model ll lk hk gov rod lk 1.10 hk 1.13 1.00 gov 1.17 1.03 2.55 rod 1.24 1.00 1.34 1.24 rob 1.11 1.01 1.29 1.32 3.54 5.3.regression estimates in this section, the estimation of two way fixed effect panel feasible generalized least squares approach and results are reported in table 4. based on the valid sample of 239 bank years, the significant value of wald test confirms that all the proposed variables in the model are jointly significantly effecting the economy growth. and the positive value of intercept depicts that the other factors causing a positive trend in economic growth. table 4 – estimates of islamic banking development model panel fgls (two way) – economic growth (dep. var.) coef. std. err. z prob. ll -0.46 0.02 -21.43 0.00*** lk 0.01 0.004 1.47 0.14 hk 0.10 0.02 5.15 0.00*** gov 0.82 0.06 13.35 0.00*** rod 0.10 0.05 2.08 0.03** rob 0.10 0.05 1.96 0.04** rod*rob -0.01 0.00 -1.85 0.06* cons 14.31 0.65 21.96 0.00** post estimation statistics obs. 239 no of banks 41 average obs. per bank 5.82 wald chi2 6739 prob. 0.00*** wald test on time effects 29.00 0.00*** wald test on bank effects 2600 0.00*** ***significant at 1%, ** significant at 5% and * significant at 10% while exploring the effect of independent variables, it can be seen that if there is 1% increase in labor force, it decreases economic growth by 0.46%. this is complying with the demand and supply framework, whereby an increase in labor leads to a decrease in talat hussain, noman arshed and rukhsana kalim 228 the market wages as defined by equilibrium shifts and a decrease in marginal productivity. further, the results indicate that, capital has no significant effect on the economic growth for the selected countries. this is because the bank itself is moderating the supply of capital to the economy to maximize their returns; hence effects of the capital is already optimized to economic outcomes. a 1% increase in human capital helps to increase economic growth by 0.10% on average. here if the skill level of the existing labor is increased, it makes them more productive, which increases production and wages in the economy. increase in the skilled human resource will increase the overall productivity of businesses which leads to growth. this result is complying to (hadia & arshed, 2020) which confirmed this outcome for the case of mudarabah companies of pakistan. also, if a 1% increase in institutional quality, there is 0.82% increase in economic growth. hereby better governance improves the systems and procedures and reduces the leakages like corruption which helps to maximize economic gains. a study by grant et al. (2019) confirmed that improvement in institutional quality increases the development related effectiveness of entrepreneurs. lastly, while discussing the concerned variables of banking sector development, it can be seen that economic growth increases by 0.10% for both cases when there is 1% increase in returns to deposit and 1% increase in return to the bank. hereby profitable banks and the income disbursement to the savers increase economic activity in terms of investment and consumption, respectively. this confirms the path discussed in the theoretical model. while the cross product of the returns to deposit and returns to bank shows that if both of them jointly increase, it slows down economic growth by 0.01% on average. we can coin it to the fact that when banks and depositors are earning high from the financing, it may consequently make the debt expensive for the firms leading to a decrease in demand for lending. table 5 – autocorrelation test first difference residuals of model coefficient of lag of residuals (prob.) decision economic growth -0.03 (0.00) no autocorrelation table 5 provides the autocorrelation test on the model. the coefficient of the lag value of residuals against the first difference of residuals showed that there is no hint of autocorrelation in the model. 5. conclusion the objective was to explore whether pursuing the shari’ah based objective of the islamic banks are favorable for the economy or not. the shari’ah based objectives include pursing for higher bank returns and deposit returns which are labelled as private returns and social returns of a bank. this study used the endogenous growth model with institutions as a base model and analyzed the role of the increase in bank returns and journal of applied economics and business studies, volume. 4, issue 4 (2020) 215-236 https://doi.org/10.34260/jaebs.4410 229 deposit returns on the economic growth of the selected economies which have fullfledged islamic banks. it is hypothesized that an increase in bank profit is private returns of bank while an increase in deposit returns are public returns of the bank. in an islamic financial system increase in banking sector development is coined as an increase in both private and public returns which are expected to increase growth. this is because of the increase in financing base and financial inclusion in the economy. the estimates were calculated using panel feasible generalized least square model on the data of all islamic banks whose financial statements are available between 2012 and 2017. the results show that both bank returns and deposit returns have a positive and similar effect on economic growth. the outcome of bank returns is similar to (furqani & mulyany, 2009; kalim et al., 2016; tabash & anagreh, 2017) while it opposes the outcome (goaied & sassi, 2010). while institutional quality has the highest impact on economic growth as proposed earlier by (dollar & kraay, 2003; hassan, bukhari, & arshed, 2019; knack & keefer, 1995; zaman, mehmood, aftab, siddique, & ameen, 2017). lastly, the study indicates that simultaneous increase in bank’s public and private returns will decrease growth, this indicates the increase in the cost of lending. hence, this study proposes that central banks and governments should postulate policies which can assist the bank’s efficiency to earn profits in investments without a major increase in the cost of lending (ahmad, 2016; kalim & arshed, 2018). this higher profits can then be translated into higher bank returns and deposit returns. this study aligns with the directive that the cash should not be held in a few hands (al qur’an 59:7). references abduh, m., & omar, m. a. (2012). islamic banking and economic growth: the indonesian experience. international journal of islamic and middle eastern finance and management, 5(1), 35-47. abduh, m., omar, m. a., & duasa, j. (2011). the impact of crisis and macroeconomic variables towards islamic banking deposits. american journal of applied science, 8(12), 1413-1418. abdulla, y. (2016). evaluation of bahrain’s islamic finance industry. journal of islamic finance studies, 2(1), 1-12. abel, s. (2014). importance of financial intermediation, the herald. retrieved from http://www.herald.co.zw/importance-of-financial-intermediation/ aghion, p., caroli, e., & garica-penalosa, c. (1999). inequality and economic growth: the perspective of the new growth theories. journal of economic literature, 37(4), 1615. aghion, p., & howitt, p. (1992). a model of growth through creative destruction. econometrica, 60, 323-351. ahmad, a. y. (2016). does islamic banking help in economic development of muslim countries? journal of islamic business management, 6(2) http://www.herald.co.zw/importance-of-financial-intermediation/ talat hussain, noman arshed and rukhsana kalim 230 ahmed, h. (2011). product development in islamic banks. uk: edinburgh university press. ahmed, h. (2014). islamic banking and shari'ah compliance: a product development perspective. journal of islamic finance, 3(2), 15-29. alejo, j., galvao, a., montes-rojas, g., & sosa-escudero, w. (2015). tests for normality in linear pane data models (l. y. s. centro de estudios distributivos, trans.). la palta. alharbi, a. (2015). development of the islamic banking system. journal of islamic banking and finance, 3(1), 12-25. amin, r. m., & yusof, s. a. (2003). allocative efficiency of profit maximization: an islamic perspective. review of islamic economics, 13, 5-21. anggraini, m. (2019). islamic banking development and economic growth: a case of indonesia. asian journal of islamic management, 1(1), 51-65 ansari, s., & rehman, a. (2011). financial performance of islamic and conventional banks in pakistan: a comparative stud. paper presented at the 8th international conference on islamic economics and finance. apergis, n., filippidis, i., & economidou, c. (2007). financial deepening and economic growth linkages: a panel data analysis. review of world economics, 143(1), 179-198. ariss, r. t., & sarieddine, y. (2007). challenges in implementing capital adequacy guidlines to islamic banks. journal of banking regulation, 9(1), 46-59. arshed, n., yasmin, s., & gulzar, m. (2020). islamic financing portfolio and its comparative growth potential. islamic banking and finance review, 7(1). aziz, u. r. (2013). islamic contracts. https://www.youtube.com/watch?v=duhxhpjkdro. babatunde, m. a. (2012). an analysis of the growth-health relationship in nigeria. paper presented at the center for the study of african economies. barro, r. (1996). health and economic growth. cambridge: harvard university press. bloom, d. e., canning, d., & sevilla, j. (2004). the effect of health on economic growth: a production function approach. world development, 32(1), 1 13. bongini, p., iwanicz-drozdowska, m., smaga, p., & witkowski, b. (2017). financial development and economic growth: the role of foreign-owned banks in cesee countries. sustainability, 9(3), 335-360. daraban, b. (2010). introducing the circular flow diagram to business students. journal of education for business, 85(5), 274-279. demirgüç-kunt, a., & maksimovic, v. (1999). institutions, financial markets, and firm debt maturity. journal of financial economics, 54(3), 295-336. journal of applied economics and business studies, volume. 4, issue 4 (2020) 215-236 https://doi.org/10.34260/jaebs.4410 231 dollar, d., & kraay, a. (2003). institutions, trade and growth. journal of monetary economics, 50(1), 133-162. durusu-ciftci, d., ispir, m. s., & yetkiner, h. (2017). financial development and economic growth: some theory and more evidence. journal of policy modeling, 39(2), 290-306. ebrahim, m. s., & joo, t. k. (2001). islamic banking in brunei darussalam. international journal of social economics, 28(4), 314-337. ey. (2016). world islamic banking competitiveness report 2016. fada, k. a., & weabekwa, b. (2012). people’s perception towards islamic banking: a field work study in gombe local government area, nigeria. international journal of business, humanities and technology, 2(7), 121-131. furqani, h., & mulyany, r. (2009). islamic banking and economic growth: empirical evidence from malaysia. journal of economic cooperation and development, 30(2), 59-74. ghosh, a. (2017). how does banking sector globalizaton affect economic growth? international review of economics & finance, 48, 83-97. goaied, m., & sassi, s. (2010). financial development and economic growth in the mena region: what about islamic banking development. institut des hautes etudes commerciales, carthage, 1-23. goldsmith, r. w. (1955). financial structure and economic growth in advanced countries: an experiment in comparative financial morphology capital formation and economic growth (pp. 112-167): princeton university press. grant, k. a., aziz, o., & arshed, n. (2019). the impact of entrepreneurial activity on poverty alleviation. economic and social development: book of proceedings, 215-224. greenbaum, s., & thakor, a. (1995). contemporary financial intermediation. new york: dryden press. greene, w. h. (2012). econometric analysis (6th ed.). nj: prentice hall. gujarati, d. n., porter, d. c., & gunasekar, s. (2012). basic econometrics (5th ed.). new york: tata mcgraw hill. gurley, g. j., & shaw, e. s. (1955). financial aspects of economic development. the american economic review, 45(4), 515-538. hafnida, h., maamor, s., & abdullah, h. (2015). an empirical study of the relationship between islamic modes of finance and financial intermediation. international journal of business and management, 10(7), 93-98. hanif, n., & arshed, n. (2016). relationship between school education and economic growth: saarc countries. international journal of economics and financial issues, 6(1), 294-300. talat hussain, noman arshed and rukhsana kalim 232 hassan, m. k., & lewis, m. (2009). handbook of islamic banking: edward elgar publishing. hassan, m. s., bukhari, s., & arshed, n. (2019). competitiveness, governance and globalization: what matters for poverty alleviation? environment, development and sustainability(1-28). doi: 10.1007/s10668-019-00355-y hassan, m. s., & kalim, r. (2017). stock market and banking sector: are they complement for economic growth in low human developed economy? pakistan economic and social review, 55(1), 1-30. how, j. c. y., abdul karim, m., & verhoeven, p. (2004). islamic financing and bank risks: the case of malaysia. thunderbird international business review, 47(1), 75-94. ibn-qudamah. (1983). al-mughni. beruit dar al-kitab al-arabi. inabo, a. o., & arshed, n. (2019). impact of health, water and sanitation as key drivers of economic progress in nigeria. african journal of science, technology, innovation and development, 11(2), 235-242. jaccard, j., & turisi, r. (2003). interaction effects in multiple regression. california sage. jaccard, j., wan, c. k., & turisi, r. (1990). the detection and interpretation of interaction effects between continuous variables in multiple regression. multivariate behavioral research, 25(4), 467-478. jarque, c. m., & bera, a. k. (1987). a test for normality of observations and regression residuals. international statistical review / revue internationale de statistique, 55(2), 163-172. kalim, r., & arshed, n. (2018). what determines the social efficiency of islamic banking investment portfolio? journal of islamic business and management, 8(2), 386-406 kalim, r., mushtaq, a., & arshed, n. (2016). islamic banking and economic growth: case of pakistan. islamic banking and finance review, 3(1), 29-43. khan, i. (2007). islamic finance: relevance and growth in the modern financial age. paper presented at the islamic finance seminar by harvard islamic finance project, london. kiyotaki, n., & wright, r. (1989). on money as a medium of exchange. journal of political economy, 97(4), 927-954. knack, s., & keefer, p. (1995). institutions and economic performance: cross country testing using alternative institutional measures. economics & politics, 7(3). lange, o. (1943). the theory of the multiplier. econometrica, 227-245. levine, r. (1997). financial development and economic growth: views and agenda. journal of economic literature, 35(2), 688-726. journal of applied economics and business studies, volume. 4, issue 4 (2020) 215-236 https://doi.org/10.34260/jaebs.4410 233 lind, d. a., marchal, w. g., & wathen, s. a. (2006). basic statistics for business & economics. boston mcgraw-hill / irwin. majid, s. a., & kassim, s. (2010). islamic finance and economic growth: the malaysian experience. paper presented at the kuala lumpur islamic finance forum, kuala lumpur. mankiw, n. g. (2014). principles of macroeconomics (7th ed.). usa: cengage learning. mankiw, n. g., romer, d., & weil, d. n. (1992). a contribution to the empirics of economic growth. the quarterly journal of economics, 107(2), 407-437. mansoori, d. m. t. (2009). islamic law of contracts and business transactions (5ed.). islamabad shari'ah academy. mansoori, d. m. t. (2012). shari'ah maxims; modern applications in islamic finance (2 ed.). islamabad, pakistan: shari'ah academy. mansur, a. (2019). sharia banking dynamics and the macroeconomic responses: evidence from indonesia. jurnal ekonomi malaysia, 2(53), 1-15. mckinnon, r. i. (1973). money and capital in economic development. washington d.c.: the brookings institutions press. mishkin, f. s. (2007). the economics of money, banking and financial markets: pearson education. narayan, p. k., & phan, d. h. b. (2019). a survey of islamic banking and finance literature: issues, challenges and future directions. pacific-basin finance journal, 53, 484-496. naz, s. a., & gulzar, s. (2020). impact of islamic finance on economic growth: an empirical analysis of muslim countries. the singapore economic review. paramanik, a. h. (1993). development and distribution in islam: pelanduk publications. parkin, m. (2016). macroeconomics (12th ed.): pearson education. patrick, h. t. (1966). financial development and economic growth in underdeveloped countries. economic development and cultural change, 14(2), 174-189. patrick, h. t., & moreno, h. a. (1982). philippine private domestic commercial banking, 1946-1980, in light of japanese historical experience. center discussion paper no 427. economic growth center, yale university. ritter, j. a. (1995). the transition from bater to fiat money. the american economic review, 85(1), 134-149. robinson, j. (1952). the generalisation of the general theory the rate of interest and other essays: macmillan. talat hussain, noman arshed and rukhsana kalim 234 rosly, s. a., & zaini, m. a. m. (2008). risk-return analysis of islamic banks' investment deposits and shareholders fund. managerial finance, 34(10), 695707. santomero, a. m. (1984). modeling the banking firm: a survey. journal of money, credit and banking, 16(4 (ii)), 576-602. sayedi, s. q. (2017). impact of islamic banking on economic growth. (bba), tolo-eaftab institute of higher education, afghanistan. schumpeter, j. a. (1911). the theory of economic development: harvard university press. shaw, e. (1973). financial deepening in economic development: oxford university press. shen, c. h., & lee, c. c. (2006). same financial development yet different economic growth: why? journal of money, credit and banking, 38(7), 1907-1944. siddiqui, s. h. (2001). islamic banking: true modes of financing. newhorizon, 109(2), 15-20. sohail, h., & arshed, n. (2020). cost-efficiency analysis of mudarabah companies. in handbook of research on theory and practice of global islamic finance (pp. 568-592). igi global. solé, j. a. (2007). introducing islamic banks into conventional banking system. imf working paper, wp/07/175. sollow, r. m. (1956). a contribution to the theory of economic growth. the quarterly journal of economics, 70(1), 65-94. tabash, m. i., & anagreh, s. (2017). do islamic banks contribute to growth to the economy? evidence from united arab emirates. banks and bank systems, 12(1), 113-118. tarashev, n. a., borio, c. e. v., & tsatsaronis, k. (2009). the systemic importance of financial institutions. bis quarterly review, 75-87. usmani, m. m. s. (1954). maariful quran (vol. 2): maktaba maariful quran. usmani, m. t. (2002). an introduction to islamic finance (vol. 20): brill. webb, i. p., grace, m. f., & skipper, h. (2017). the effect of banking and insurance on the growth of captital and output. revista de temas financieros. wdi. (2019). world development indicators. world bank. available at: https://databank.worldbank.org/source/world-development-indicators wgi. (2019) worldwide governance indicators. world bank. available at https://databank.worldbank.org/source/worldwide-governance-indicators wief. (2014). 10th world islamic economic forum. uae. wief. (2016). 12th world islamic economic forum. indonesia. https://databank.worldbank.org/source/world-development-indicators https://databank.worldbank.org/source/worldwide-governance-indicators journal of applied economics and business studies, volume. 4, issue 4 (2020) 215-236 https://doi.org/10.34260/jaebs.4410 235 yakcop, t. s. d. n. m. (2002). islamic financial market development: the malaysian strategy. paper presented at the kuala lumpur international islamic capital market conference, kuala lumpur. https://www.sc.com.my/post_archive/introduction-and-the-early-phases-ofdevelopment-of-the-islamic-financial-system-bytan-sri-dato-nor-mohamedyakcop-special-economics-adviser-to-the-prime-minister-of-malaysia-at-thekuala-lumpur-in/ yusof, s. a., & amin, r. m. (2007). a survey of the objective of the firm and models of producer behavior in the islamic framework. jkau: islamic economics, 20(2), 3-16. zaman, z., mehmood, b., aftab, r., siddique, m. s., & ameen, y. (2017). role of islamic financial literacy in the adoption of islamic banking services: an empirical evidence from lahore, pakistan. journal of islamic business and management, 7(2), 230-247. zamil, n. a. m. (2014). an empirical investigation into the problems and challenges facing islamic banking in malaysia. (phd), cardiff university, cardiff. http://www.sc.com.my/post_archive/introduction-and-the-early-phases-of-development-of-the-islamic-financial-system-bytan-sri-dato-nor-mohamed-yakcop-special-economics-adviser-to-the-prime-minister-of-malaysia-at-the-kuala-lumpur-in/ http://www.sc.com.my/post_archive/introduction-and-the-early-phases-of-development-of-the-islamic-financial-system-bytan-sri-dato-nor-mohamed-yakcop-special-economics-adviser-to-the-prime-minister-of-malaysia-at-the-kuala-lumpur-in/ http://www.sc.com.my/post_archive/introduction-and-the-early-phases-of-development-of-the-islamic-financial-system-bytan-sri-dato-nor-mohamed-yakcop-special-economics-adviser-to-the-prime-minister-of-malaysia-at-the-kuala-lumpur-in/ http://www.sc.com.my/post_archive/introduction-and-the-early-phases-of-development-of-the-islamic-financial-system-bytan-sri-dato-nor-mohamed-yakcop-special-economics-adviser-to-the-prime-minister-of-malaysia-at-the-kuala-lumpur-in/ talat hussain, noman arshed and rukhsana kalim 236 appendix table a1 – sample banks countries islamic banks bahrain al baraka islamic bank bahrain islamic bank citi islamic bank ithmaar bank bangladesh islamic bank of bangladesh shahjalal islamic bank exim bank al arafa islamic bank brunei darussalam bank islam brunei darussalam egypt abu dhabi islamic bank indonesia bank muamalat indonesia maybank shariah iran bank maskan bank pasargad bank tejarat jordan islamic international arab bank jordan islamic bank kenya gulf african bank kuwait kuwait finance house lebanon al baraka bank malaysia affin islamic bank al rajhi bank alliance islamic bank am-islamic bank asian finance bank bank islam malaysia berhad bank muamalat malaysia berhad bank rakyat cimb islamic bank hong leong islamic bank koperasi bank persatuan public islamic bank rhb islamic bank ocbc al amin bank nigeria jaiz islamic bank oman alizz islamic bank pakistan al baraka bank bank islami burj bank dubai islamic bank meezan bank mcb islamic philippines al amanah bank qatar qatar international islamic bank qatar islamic bank saudi arabia al bilad bank al ranjhi bank alnima bank south africa al baraka bank sri lanka amanah bank sudan al baraka bank sudan al shamal bank faisal islamic bank syria bank sham thailand islamic bank of thailand tunisia al baraka bank tunisia turkey al baraka bank turkey uae dubai islamic bank noor islamic bank sharjah islamic bank abu dhabi islamic bank al hilal bank ajman bank uk al rayan islamic bank yemen tadhamon islamic bank journal of applied economics and business studies, volume. 4, issue 1 (2020) 97-122 https://doi.org/10.34260/jaebs.415 97 journal of applied economics and business studies (jaebs) journal homepage: https://jaebs.com issn (print): 2523-2614 issn (online) 2663-693x entrepreneurial behavior, institutional trust and national innovation: a macro-level empirical study alvina sabah idrees1 & saima sarwar21 1 lecturer, department of economics, gc university lahore, pakistan 2 associate professor, department of economics, gc university lahore, pakistan abstract schumpeterian fundamentalism supports the argument that innovation is a dynamic process and novelties are initiated through economic agents namely the entrepreneurs; vis-à-vis a strong institutional environment is required to facilitate the innovation process. therefore, the present study undertakes the macro-level empirical analysis on determining the impact of entrepreneurial behavior, property rights and state effectiveness on country’s innovation. the data is of panel nature consisting of 55 countries and a time period from 2010 to 2016. the empirical analysis is done using system gmm (generalized method of moments) estimation technique. the study shows that the fear of failure rate and total early stage entrepreneurs reduces innovation in a country whereas there is a significant positive relationship between established business entrepreneurs and innovation. however, perceived opportunities have an insignificant impact. this means that it is not inevitable that opportunities necessary trigger innovation. in addition, the study shows that property rights play an integral role in developing institutional trust which boosts entrepreneurialism to undertake innovative venture. on the other hand, state effectiveness is negatively related to innovation i.e. institutional trust is brought down in fragile countries which retard country’s innovation. keywords innovation, entrepreneurship, property rights jel classification o31, l26, p48 1. introduction the legacy of innovation driven economic growth can be traced back to schumpeter (1934) who describes entrepreneur as the one who carry out new combinations. supporting this argument, drucker (1985) realizes innovation as an important attribute of an entrepreneur. economic prosperity is highly linked with entrepreneurial development of a country. united states, european union and asia pacific region are the practical illustrations who shifted from managed economies towards the entrepreneurial economy whose productivity gains is attributed to innovative entrepreneurship. based on solow 1 corresponding author: saimasarwar@gcu.edu.pk alvina sabah idrees & saima sarwar 98 model (solow, 1956), the early strategies of economic prosperity put great emphasis on investment in new capital. later, romer (1986) replaced physical capital with knowledge capital and the policy goals focused on investment in r&d and human capital. however, it was observed under swedish paradox (edquist & mckelvey, 1998) that the country produced lesser r&d intensive output despite spending heavily on r&d. later, it was witnessed under european paradox (donatiello & ramella, 2017) that southern europe performed well with respect to innovation despite weaknesses in their national innovation system. such paradoxes compel one to rethink the model of national innovation. this gap is filled by knowledge spillover theory of entrepreneurship (acs, braunerhjelm, audretsch & carlsson, 2009) which does not consider entrepreneurship to be exogenously determined. hirschman (1958) has also suggested that it is not the capital shortage that acts as a hindrance to development, but the real culprit is lack of entrepreneurial abilities. literature often narrates self-employment as a proxy to measure country’s entrepreneurship level (acs et al., 2009; stam & nooteboom, 2011; maltby, 2013; faggio & silva, 2014). but this measure has its own caveats. it is too narrow as well as too broad. since not all self-employed becomes entrepreneurs and not all entrepreneurs are selfemployed vis-à-vis not all entrepreneurs are innovative. in this regard, audretsch, keilbach and lehmann (2006) introduce us with the term of entrepreneurship capital. entrepreneurship capital deals with entrepreneurial behavior which reflects not only the risk bearing and risk sharing attitudes but also the capacity to undertake innovation activities. so, along with the traditional factors of physical capital as suggested by solow’s neoclassical growth model (solow, 1956) and knowledge capital of romer’s endogenous growth model (romer, 1986), entrepreneurship capital (audretsch, keilbach & lehmann, 2006) is a necessary ingredient for economic prosperity which brings productivity gains through innovation. defining innovation in macroeconomic perspective is troublesome as it is difficult to confine it within a particular domain. nevertheless, a greater difficulty arises when it comes down to the measurement of innovation at macro-level. past literature have used patent counts and r&d expenditures as proxies for innovation but these measures have their own limitations (mansfield, 1986; amesse et al., 1991; prodan, 2005; graham, et al., 2009; acemoglu., bimpikis, & ozdaglar, 2011; shaffer, 2011; acs & sanders, 2012; llobet, & suarez, 2013; moser, 2013; barasa et al., 2014; savrul & incekara, 2015; fischer & gamarra, 2017). alternatively, furman et al., (2002) examine multiple dimensions of national innovation framework to determine economy’s potential towards innovation. in the same spectrum, some other innovation-oriented indices include global innovation index (gii), global competitiveness index, bloomberg innovation index, knowledge economy index and european innovation scoreboard (murray and budden, 2017). global innovation index (gii) is the most elaborative measure of innovation capacity for a nation which was journal of applied economics and business studies, volume. 4, issue 1 (2020) 97-122 https://doi.org/10.34260/jaebs.415 99 initiated by insead in 2007.2 it includes multiple dimensions of innovation using 80 indicators and seems more plausible to undertake international comparability of countries which are at the various stages of development in terms of national innovation. the present study has used gii as measure of national innovation capacity. entrepreneurship is never short of its supply, but its productive contribution differs substantially (baumol, 1996). strong institutional environment provides an incentive framework for innovative entrepreneurship which can be achieved through strong property rights and state legitimacy. gerth and mills (1946) explain weberian definition of state as a legitimate use of force to exercise control over people and maintain order. the legitimacy will be lost if a country faces severe problems of security, governance and development. the property right institutions measure the extent to which the private property is secured form state’s expropriation as well as protection from corrupt government officials who demand bribes in exchange for personal favors (acemoglu and johnson, 2005). according to williamson and kerekes (2011), trust is one of the important determinants of secured property rights i.e. effectiveness of property rights is a build-in phenomenon in societies that have strong cultural values and one of them is trust. under weak state legitimacy and fragile property rights entrepreneurship becomes evasive which is unproductive for the society. entrepreneurial activities involve uncertainties (von mises, 1998; choi and shepherd, 2004; mcmullen and shepherd, 2006; knight, 2012) and trust building helps to realize potential gains by reducing uncertainties. therefore, one cannot undermine the legitimate role of institutions in an innovation-driven economy. aim should be to tailor the institutional environment in such a way that evasive entrepreneurial activities be made more costly so that the resources could be shifted towards more productive opportunities. this can be done by promoting institutional trust through well-defined property rights and provision of strong legal system. institutional trust helps in building entrepreneur’s confidence in exploiting innovative opportunities. hall and jones (1999), north (1991), rodrik, subramanian and trebbi (2004) and acemoglu and johnson (2005) explain that well-developed institutions directly affect the property rights. similarly, rodrik (2004) points out in growth perspective that rich countries are the ones where investors feel protected in terms of property rights. institutional trust is the utility that individuals get from satisfactory performance of institutions and deliverance of public services (mishler & rose, 2001; christensen & laegreid, 2005). oecd report (oecd, 2013) defines trust as governments’ ability to 2 insead is an acronym for institut européen d'administration des affaires european which is french for european institute of business administration. since then this construction is published annually in collaboration with cornell university and world intellectual property organization (wipo). gii captures the richness of a country in innovation from multiple dimensions and ranks the country on a scale from 0 to 100. alvina sabah idrees & saima sarwar 100 manage social and economic issues. in fragile states, weak institutional structure reduces the legitimacy of state policies which lower down entrepreneur’s trust on the system. until the rule of game is not visibly predominant, the legitimacy of polices cannot be established and entrepreneurs cannot gain trust in the institutions. if a country is not effective in implementing polices in the political, social and economic context then institutional trust is lost. states’ inability in effective deliverance of public services reduces entrepreneurs’ confidence on institutions and innovation activities are discouraged. the present study uses property rights and state effectiveness score as proxies for institutional trust. extant literature examines role of entrepreneurs towards economic growth by concentrating mainly on the development of real economy (gort & klepper, 1982; acs & audretsch, 1988; baumol, 1996; audretsch & thurik, 2001; grebel, pyka & hanusch, 2003; wong et al., 2005; hanusch & pyka, 2006; anokhin & schulze; 2009). literature have also settled on the fact that an entrepreneur is an inherent innovator (schumpeter, 1934; miller, 1983; drucker, 1985; lumpkin & dess, 1996; shane & venkataraman, 2000; de mel et al., 2009; garcía, jin & salomon, 2013). these studies have examined the firm level activities. nevertheless, limited literature is available that provides direct link between entrepreneurs and innovation at macro-level (draghici & albulescu, 2014). many studies have associated innovation with patents counts (mansfield, 1986; amesse et al., 1991; graham, et al., 2009; acemoglu, bimpikis, & ozdaglar, 2011; shaffer, 2011; acs & sanders, 2012; llobet, & suarez, 2013; moser, 2013) which is an output indicator and r&d expenditure (prodan, 2005; barasa et al., 2014; savrul & incekara, 201; fischer & gamarra, 2017) which is an input indicator. the major drawback of using r&d expenditure is that it includes efforts for both innovation as well as imitation activities (moreno et al., 2005). on the other hand, there are activities that contribute to existing stock of knowledge but do not get patented (moreno et al., 2005; tebaldi & elmslie, 2008). similarly, there are certain patents that do not develop into innovations (moreno et al., 2005). in addition, the relevance of such indicators is questionable when it is evaluated for developing countries which hardly invest in formal r&d and also do not apply for patents. therefore, innovation cannot be confined to just r&d and patents. to fill this gap, the present study considers global innovation index (gii) to measure innovative capacity of a nation. in the institutional context, d’agostino and scarlato (2016), anokhin and schulze (2009), dincer (2019), fischer and gamarra (2017) and tebaldi and elmslie, 2013) have examined the role of institutional quality on innovative output but none undertook the analysis from the perspective of institutional trust which is a necessary ingredient for determining credibility of institutions. the trust building hypothesis is tested by tedika and agbor (2016) but this study has limited scope from the point of data analysis. the present study aims to fill the literature gap by bringing new empirical evidence on the relationship of entrepreneurial behavior and institutional trust with innovation in a macro-economic perspective. journal of applied economics and business studies, volume. 4, issue 1 (2020) 97-122 https://doi.org/10.34260/jaebs.415 101 2. literature review the key to innovation is through entrepreneurship development. entrepreneurs are the main source of bringing change in the economic system through innovation. therefore, economic innovation is largely attributed to entrepreneurial activities. audretsch and thurik (2001) discuss fundamental elements necessary for the development of entrepreneurial economy. knowledge has emerged as the vital ingredient of a production process and the comparative advantage is shifting on the basis of innovation activates. moreover, knowledge spillover helps to flourish entrepreneurial activities. shane and venkataraman (2000) explains entrepreneurial activities as the discovery of new ways of producing goods and services and exploiting the opportunities for better organizing the economic activities, processes and the markets. thus, trust is required to facilitate the process of entrepreneurial activities which acts as a cornerstone for innovation and technological advancement. according to sautet (2005) “entrepreneurship is not dependent on the resources in an economy. rather, the key is the quality of institutions that permit the exploitation of resources and opportunities.” this is because prevailing institutional arrangements determine the uncertainties in profits and risks that determine not only the probability but also the nature of entrepreneurial activities. therefore, a robust institutional environment is required for shaping innovation activities by an entrepreneur. sarbah and xiao (2013) examine the role of trust for the entrepreneurship development and its growth. the study discusses that strong institutions need to be developed that incorporate the practices upon which the trust could be built. the study has limited scope as it has only targeted few selected entrepreneurs at micro level in ghana region. the qualitative analysis is undertaken while the quantitative side has been ignored. similar hypothesis is also tested by kodilatedika and agbor (2016) who suggest that half of the entrepreneurial spirit is compelled through trust considerations. therefore, regions like sub-saharan african countries can expand their entrepreneurial activities by formalizing the trust building institutions. for the measurement of trust, the world value survey (wvs) has been utilized. in addition, the study has also considered reverse causality where greedy entrepreneurs can create mistrust through exploitation of institutional weaknesses. aidis (2017) argues that supportive institutional environment helps to nurture entrepreneurship which is more innovative and productive. such conducive conditions are provided through property rights and good governance that helps institutional trust building. the study provides insights for post-soviet countries in the context of institutional reforms and concludes that fragile rule of law, high corruption and excessive regulation can impede entrepreneurial development. acs, desai and klapper (2008) point out that institutional environment play an important role in determining the cross-country differences in entrepreneurial activities. the impact of various country level characteristics are examined alvina sabah idrees & saima sarwar 102 on entrepreneurial activities which include political risk, economic risk, law and order, financial development, gdp per capita and regulatory barriers. acs and amorós (2008) analyzed the impact of gross domestic product and global competitiveness index (gci) on entrepreneurial dynamics by using dataset from gem’s adult population survey. it is suggested that competitiveness brings production efficiency but fails to achieve positive impact on entrepreneurship and innovation in developing countries. similarly, levie and autio (2008) examined the impact of finance, government policies, government regulations and programs, education and training, r&d, infrastructure, internal market openness, and social and cultural norms on entrepreneurship. the fixed effect gls methodology was applied. the results show a strong and positive impact of education and training on new business activity but no impact on growth expectations of entrepreneurial activities. this relationship is strongly influenced by a country’s level of economic development. henrekson and sanandaji (2011) determine the bilateral relationship between institutions and entrepreneurial development. innovative activities could be productive, unproductive or destructive depending upon the incentive structure faced by entrepreneurs. policy makers must consider such interactions while designing the rules of game. therefore, institutional change must be evaluated on the basis of type of entrepreneurship that it promotes. elert and henrekson (2017) also conclude a bidirectional relationship between institutions and entrepreneurship. entrepreneurial activities are heavily influenced by institutional environment. institutions, typically the regulatory framework, lag behind the technology driven entrepreneurship and innovation. under such circumstances, when this gap widens, evasive entrepreneurship is the usual response of entrepreneurs. on the other hand, entrepreneurs also influence the institutional framework through its productive or evasive activities. similarly, samadi (2019) also confirms the existence of bi-directional causality between entrepreneurship and institutions but the long-run relationship holds for innovation-driven economies and not for factor-driven and efficiency-driven economies. depending upon the institutional environment, entrepreneurial activities can be productive or unproductive that may reinforce or weaken the institutional quality. the empirical examination on the relationship between innovation and institutions is undertaken by tebaldi and elmslie (2013). the study applies iv technique on the crosscountry data and finds that creation of technical knowledge is not possible in the absence of good institutional quality. this study shows a significant positive impact of control of corruption, protection of property rights, judiciary effectiveness and market-friendly policies on patent production. this relationship is noteworthy not only in the technology frontier countries but also among countries that are far from technology frontier. although geography affects the innovation levels but that only takes place through institutions. anokhin and schulze (2009) examine that control of corruption will enhance innovation and entrepreneurship in a country. since, it increases the trust on state and institutional journal of applied economics and business studies, volume. 4, issue 1 (2020) 97-122 https://doi.org/10.34260/jaebs.415 103 ability to enforce rules and laws effectively. the rise of transaction costs and other implicit costs associated with corruption cause hindrance to economic activities. as a result, the prospective entrepreneurs fail to get benefit from innovation investment. d’agostino and scarlato (2016) develop an empirical link between government institutions and economic growth through intermediating role of innovation. the study confirms that inclusive institutions magnify the effects of technology shocks on economic growth. therefore, the formulation of incentive system for technological innovation must be accompanied by improved governance structure and strong political systems. the micro-level analysis on the mediating role of institutional quality on innovative output is undertaken by barasa et al. (2014). it is observed that regional institutional quality strengthens the positive impact of firm’s r&d activities on innovative output. better institutional environment enables firm to extract more value from its resources. firms that operate under poor institutional environment do not conduct r&d and fails to benefit from innovative output. therefore, policy makers must focus on good governance at national as well as regional level. dincer (2019) studies long-run relationship between corruption and innovation. the quantity and quality of patents is taken as a measure of innovation. it is observed that corruption slows down innovation. however, the study provides no details on the channels through which this negative relationship exists. fischer and gamarra (2017) also study the relationship between institutional quality and innovation. the impact of democracy, corruption and political rights is determined on gross r&d expenditures, patent activity and productivity. it is found that institutional backwardness hampers the potential benefits from productive technology. therefore, institutional framework in developing countries needs to be improved to guarantee the success of r&d investment. donges, meier and silva (2019) highlight that institutions affect economic growth through innovation. this study is carried out in an empirical setting of timing and geography of french occupation after the french revolution in germany. it is examined that inclusive institutions and financial development complements the innovative output. those regions that underwent institutional reforms performed better in innovative output compared to those that were not driven by these reforms. the efficient legal systems foster business creation and incentivize innovation. hanusch and pyka (2006) identify innovation as the driving force for economic development. innovation occurs as a result of entrepreneurial decisions regarding novelties created at micro-level. this leads to emergence of new industries through the transformation of industrial organization and structure. due to high competition, old manufacturing firms are compelled to engage themselves into innovation and some might seek collaboration with new entrepreneurs. such co-existence contributes heavily towards technological progress. amaghouss and ibourk (2013) examine the impact of entrepreneurial activities and innovation on economic growth. the total entrepreneurial participation rate, rate of owned business and patents counts had a positive impact on alvina sabah idrees & saima sarwar 104 economic growth, but nascent entrepreneurship showed insignificant result. braunerhjelm (2010) develops the relationship of knowledge, entrepreneurship and innovation with economic growth. the knowledge driven innovation is an outcome of r&d activities. despite this the role of entrepreneurial initiatives cannot be ignored which may appear with a lag. entrepreneurs help to bridge the gap between pace of knowledge development and fast pacing opportunities by speeding up knowledge creation and exploitation. therefore, entrepreneurs play an important role in knowledge based economy through its contribution towards innovation. 3. the model endogenous growth model (romer, 1986; lucas, 1988) puts great emphasis on human capital and knowledge-based inputs as determinants of economic growth. on the other hand, ehrlich, li and liu (2017) developed a framework of endogenous growth through accumulation of entrepreneurial human capital. the present study develops a model by borrowing from the analytical framework of romer (1986), lucas (1988) and ehrlich, li and liu (2017) to determine the factors that influence country level innovation through entrepreneurial efforts as an employer of innovation-based inputs. casson and wadeson (2007) treats entrepreneur as an employer who does not take up a one-time managerial decision to exploit profitable opportunities. in fact, an entrepreneur is in continuous efforts to sustain those profits by overcoming the diminishing returns through innovation. therefore, a country’s innovation, both in terms of demand for innovation-based inputs as well as supply of innovation outputs, is strongly correlated with the entrepreneurial attitudes and entrepreneurial activities. in addition, hirschman (2005) proposes that decision-making of a potential entrepreneur is highly associated with institutional factors. the potential entrepreneurs are discouraged to commercialize new knowledge capital if it is subjected to rent seeking (hoff & stiglitz, 2001). therefore, such institutional constraints act as barriers to innovation. lack of trust in institutions results in coordination failure among economic agents that brings a worse-off situation for all, a concept similar to coordination failure (hoff, 2000). the present study develops a macroeconomic model of innovation as follows: inn = z f(e, t)………….. (1) where inn = innovation, e = entrepreneurial human capital, t = institutional trust, z= exogenous factors to innovation. φ = inn(t)−inn(t−1) inn(t−1) ……………….. (2) inn(t) = inn(t − 1)(φ + 1)………………… (3) journal of applied economics and business studies, volume. 4, issue 1 (2020) 97-122 https://doi.org/10.34260/jaebs.415 105 where, φ is the rate of change of innovation. excluding the factor z and differentiating equation (1) w.r.t. time (t): dinn dt = δf δe . de dt + δf δt . dt dt … … … … … … . (4) divide equation (4) by inn dinn dt /inn = { δf δe . de dt + δf δt . dt dt } /inn … … … … … … . (5) dividing and multiplying right-hand-side of equation (5) with e and t: dinn dt /inn = { δf δe . e/inn . de dt /e + δf δt . t/inn . dt dt /t } … … … … … … . (6) where dinn dt /inn = φ which represents the rate of change in innovation, de dt /e = change in entrepreneurial human capital, dt dt /t = institutional trust building over time, δf δe . e/inn= share of entrepreneurial human capital in total innovation, δf δt . t/inn = share of institutional trust building in total innovation. thus, changes in national innovation depend upon changes in entrepreneurial human capital and institutional trust building in the economy as well as their contributions in total innovation. the entrepreneurial human capital in time t depends upon entrepreneurial behavior (eb) which consists of entrepreneurial attitudes (eatt) and entrepreneurial activities (eact) i.e. e(t) = e b(t) = eatt(t) + eact(t). the entrepreneurial attitudes are measured through perceived opportunities (po) and fear of failure rate (ffr). entrepreneurial activities are measured by total early-stage entrepreneurial activities (tea) and percentage of established business entrepreneurs (ebo). 3 the institutional trust building is captured through enactment of property rights (pr) and state effectiveness score (ses). according to mishler & rose (2001) and christensen and laegreid (2005), state’s ability to effectively deliver its services triggers institutional trust in the society. whereas, williamson and kerekes (2011) conclude that trust is one of the important determinants of secured property rights. the 3 the definition of entrepreneurial behavior is taken from gem website. alvina sabah idrees & saima sarwar 106 present study takes foreign direct investment (fdi) as the z factor. past literature has developed a strong association between fdi and technology diffusion which might contribute towards building national innovation capacity (walz, 1997; blyde, 2003; loukil, 2016; osano & koine, 2016). 4. data the present study is based on a panel data analysis, consisting of 55 countries and a time period of seven years from 2010 to 2016. the selection of countries and time period is on the basis of data availability. 4 the national innovation capacity is measured by global innovation index (gii). the entrepreneurial behavior is measured through various indicators compiled by adult population survey (aps) carried out under global entrepreneurship monitor (gem).5 these indicators include perceived opportunities (po), fear of failure (ffr), established business ownership (ebo) and total early-stage entrepreneurial activities (tea). institutional trust is measured by two indicators i.e. property rights (pr) and state effectiveness score (ses). lastly, foreign direct investment (fdi) is taken as a control variable. the detailed description of variables and data sources is given in table 1. table 1: description of variables and data sources variables description data sources global innovation index (inn) the index ranks the innovation performance of a country by using 80 indicators of innovation that includes political environment, infrastructure, education and business sophistication. the index takes the value between 0 to 100. a higher value indicates greater innovation. insead, cornell university and world intellectual property organization (wipo) perceived opportunities (po) percentage of adult population who perceive good opportunities to start a firm in the area where they live gem adult population survey (aps) fear of failure rate (ffr) percentage of adult population who perceive good opportunities to start a business, but the fear of failure prevents them from setting up a business gem adult population survey (aps) established business ownership (ebo) percentage of adult population who are currently an owner/manager of an established business. they are owning/ managing or running a business that has paid salaries, wages, or any other payments to the owners for more than 42 months gem adult population survey (aps) 4 the list of countries are provided in the appendix (table 4). 5 global national level data is retrieved from gem website (https://www.gemconsortium.org/data) https://www.gemconsortium.org/data journal of applied economics and business studies, volume. 4, issue 1 (2020) 97-122 https://doi.org/10.34260/jaebs.415 107 total early-stage entrepreneurial activity (tea) percentage of adult population who are either the owner/manager of a new business or are nascent entrepreneur gem adult population survey (aps) foreign direct investment (fdi) it reports the net inflows (new investment inflows less disinvestment) in the reporting economy from foreign investors, as a percentage of gdp word development indicators (wdi)6 property rights (pr) it is the assessment of ability to accumulate private property and wealth that are safe from unfair expropriation. the indicator belongs to the subcategory of the index of economic freedom i.e. rule of law. the index is measured on a scale of 0 to 100. the heritage foundation7 state effectiveness score (ses) effectiveness score is a subindicator of state fragility index. the components included in effectiveness score are security, political, economic and social effectiveness. the score is between 0 to 13. the high value indicates greater fragility. polity iv8 5. econometric methodology the functional form of the model is as follows: innovation = f (entrepreneurial behavior, property rights, state effectiveness, foreign direct investment)……….. (7) the baseline model with its econometric specification is given below: inni,t = α + β inni,t-1 + δ1 poi,t + δ2 frri,t + δ3 teai,t + δ4 eboi,t + δ5 pri,t + δ6 sesi,t + δ7 fdii,t + ɛi,t …………..(8) where inn represents innovation measured through global innovation index, po is perceived opportunities i.e. the percentage of adult population who perceive good opportunities to start a business, and those who perceive good opportunities but fear of failure prevents them from setting up a business is represented by ffr i.e. fear of failure rate, tea is total early-stage entrepreneurial activities representing nascent entrepreneurs and ebo are the established business entrepreneurs. these variables are used to measure 6 https://databank.worldbank.org/source/world-development-indicators 7 https://www.heritage.org/index/explore?view=by-region-country-year&u=637256645917098083 8 http://www.systemicpeace.org/inscrdata.html https://databank.worldbank.org/source/world-development-indicators https://www.heritage.org/index/explore?view=by-region-country-year&u=637256645917098083 http://www.systemicpeace.org/inscrdata.html alvina sabah idrees & saima sarwar 108 entrepreneurial behavior. po and ffr reflects the entrepreneurial attitudes whereas entrepreneurial activities are measured by the variables tea and ebo. pr represents property rights and ses is the state effectiveness score. these two variables are used to measure institutional trust. fdi (foreign direct investment) is used as a control variable. literature (walz, 1997; blyde, 2003; loukil, 2016; osano & koine, 2016) has taken fdi as an important source of technology diffusion, which can be a source of innovation. δ's are the slope coefficient of each explanatory variable. the subscript ‘i’ and ‘t’ denotes cross sections and time period. α is the intercept and ɛi,t is the overall error term. innt-1 is the lagged value of dependent variable to indicate the dynamic nature of the model and β is its slope coefficient. innovation activities are cumulative in nature i.e. the current level of innovation is dependent on its past levels (lee, 2013). as robinson (2009) states that the success of innovation depends upon the way it evolves. hence, it is necessary to incorporate the dynamic model of innovation by incorporating its lagged impact. koçak (2017) also estimates the dynamic model of innovation by examining the impact of institutional quality and measures innovation through patent applications. the dynamic model also helps to reduce panel bias when t is sufficiently smaller than n (baum et al., 2003; baltagi, 2008; sarafidis, yamagata and robertson, 2009). there are multiple econometric approaches to estimate equation 8. table 5 (see appendix) provides some preliminary results of the model. the dynamic panel data model is estimated, where lagged dependent variable is taken as explanatory variable along with a list of regressors. the model is first estimated through pooled ols which shows misleading parameter estimates in the presence of heteroskedasticity. in a multiple linear regression modeling, the assumption of exogeneity is violated if one of the regressors is correlated with the residual. the dynamic pooled ols produce biased and inconsistent estimates as explored by anderson and hsiao (1982). the estimated pooled ols also suffers from model misspecification (see table 5, in appendix). to capture individual specific effects, the panel fixed effect regression is estimated. but the diagnostics suggest inefficient estimates due to the presence of heteroskedasticity, serial correlation and cross sectional dependence. in addition, there might exist specification bias in the model which gives inconsistent estimations of the parameters i.e. the possibility of endogenous regressors. the instrumental variable (iv) or 2sls regression technique provides a way to overcome the issue of endogeneity (baltagi, 2008). it is a better econometric approach to obtain consistent parameter while dealing with the impact of moderating variable i.e. when some instrumental variables are uncorrelated with error term but highly correlated with one of the regressors (wooldridge, 2002). under homoscedasticity, the 2sls estimates are efficient (roodman, 2009). however, in the presence of heteroskedasticity, the 2sls estimates remain consistent but the standard journal of applied economics and business studies, volume. 4, issue 1 (2020) 97-122 https://doi.org/10.34260/jaebs.415 109 errors becomes inconsistent that prevents valid inferences (baum, schaffer and stillman, 2002). gmm (generalized method of moments) estimations is the usual approach that allows to cater the problem of endogeneity and provides better estimates whilst the presence of heteroskedasticity in iv/2sls regression. gmm uses orthogonality conditions to allow efficient estimates when heteroskedasticity is present (baum, schaffer and stillman, 2002). in addition, gmm estimator allows for over-identification of parameters as more moment conditions can be added than the number of parameters to be estimated (wooldridge, 2001). although literature provides limited guidance on how many are too many instruments but this number should not be greater than the number of cross sections (roodman, 2009). the present study employs both difference gmm (arellano and bond, 1991) and system gmm (blundell and bond, 1998) estimations. both estimators are designed for panel data when t is small and n is large. the estimated value of β coefficient for pooled ols is considered the upper bound while the corresponding fixed effect estimate is considered a lower bound. if difference gmm estimate, for β, is closer to fixed effect estimate then finite sample bias is suspected due to weak instruments and system gmm is preferred (bond, 2002). the parameters under system gmm is better determined than difference gmm when β (coefficient of lagged dependent variable) is higher than fixed effect but below the ols estimates. in system gmm, variables are instrumented by using first differences in level equation and level instruments in first differenced equation. the monte carlo simulations also suggest that system gmm is more efficient (blundell and bond, 1998). the standard errors under two-step system gmm is reduced considerably compared to one-step system gmm. 6. results and discussions the estimated results of difference gmm and system gmm are presented in table 2. the estimates of short-run coefficients are provided and the value of β shows the speed of adjustment which also indicates the persistency of dependent variable. the results from system gmm estimation indicated that the lagged impact of innovation (inn) is statistically significant i.e. the present innovation depend upon past innovation levels. this is the reason of persistence dominance of technological advanced countries over the world. the result is consistent with lee (2013) who determined that current technological innovation depends upon its past levels. alvina sabah idrees & saima sarwar 110 table 2: estimated result of dynamic panel model using difference and system gmm approach (dependent variable: inn) variable one-step difference gmm two-step difference gmm one-step system gmm two-step system gmm coefficient std. error coefficient std. error coefficient std. error coefficient std. error innt-1 -0.066 0.059 -0.069 0.070 0.318* 0.174 0.318*** 0.0155 po 0.228** 0.089 0.241*** 0.088 0.002 0.037 -0.003 0.011 ffr -0.226** 0.095 -0.278** 0.111 -0.104** 0.116 -0.105*** 0.044 tea -0.912*** 0.279 -0.872*** 0.298 -0.673*** 0.175 -0.663*** 0.034 ebo 0.931*** 0.265 0.940*** 0.301 0.450** 0.200 0.444*** 0.044 fdi -0.216 0.241 -0.189 0.283 0.048 0.030 0.048*** 0.009 pr 0.063 0.205 0.094 0.319 0.1823** 0.086 0.182*** 0.009 ses -0.777 0.520 -0.875 0.642 -0.495 0.359 -0.472*** 0.104 constant 27.828*** 4.701 28.082*** 0.824 diagnostics tests: ar (1) prob. = 0.001 prob. = 0.007 prob. = 0.005 prob. = 0.000 ar (2) prob. = 0.415 prob. = 0.626 prob. = 0.351 prob. = 0.127 hansen (j-test) prob. = 0.214 prob. = 0.165 prob. = 0.185 prob. = 0.185 c-test (diff-inhansen) gmm prob.= 0.989 prob. = 0.984 prob. = 1.000 prob. = 1.000 iv prob. = 0.981 prob. = 0.960 prob. = 0.994 prob. = 0.994 no. of instruments 49 48 54 54 cross sections 54 55 55 55 total observations 323 329 384 384 ***, ** and * indicates the significance level at 1%, 5% and 10% level, respectively. year dummies are included in the model estimation. the probability of hansen (j-test) suggests the joint validity of instruments. ar (1) indicates first order autocorrelation. whereas, ar(2) test validates the hypothesis of no second order serial correlation. the difference-in-hansen test which is also known as c-test (baum, schaffer and stillman, 2003) is valid for the subsets of instruments. the number of instruments is also less than the total crosssections. the long-run coefficients are estimated using the stata command nlcom, after applying the two-step system gmm. the results are provided in table 3. the long-run coefficients measure the sustained impact of independent variables on dependent variable. table 3: the estimated long-run coefficients variable coefficient std. error ffr -0.154*** 0.012 tea -0.972*** 0.049 ebo 0.651*** 0.064 fdi 0.070*** 0.013 pr 0.266*** 0.010 ses -0.691*** 0.154 constant 41.155*** 0.868 ***, ** and * indicates the significance level at 1%, 5% and 10% level, respectively. the long –run coefficient of each variable, measured as (δi/1β), explains the sustained impact on dependent variable. journal of applied economics and business studies, volume. 4, issue 1 (2020) 97-122 https://doi.org/10.34260/jaebs.415 111 in the long-run, the established business ownership (ebo) has a significant positive impact on innovation capacity. the result indicates that a 1% increase in ebo will increase the global innovation index (gii) by 0.65 points. ebo consists of those firms that usually observe economies of scale and larger firms are in a better position to undertake innovation as these tend to have access to sufficient resources to build-up new technologies as explained by de mel, mckenzie and woodruff (2009). they analyzed that there is a greater chance of process innovation with increase in firm size. hence, established business entrepreneurs are inclined to innovate more. innovation is risky in terms of costs, uncertainty and peril of failure. the prolonged presence of established entrepreneurs in the market gives them a competitive edge against the newly established early stage entrepreneurs who are usually the imitator in their early stage production. hence it is wiser for early entrepreneurs to let the innovation be done by others and then jump into the pool later (lumpkin & dess, 1996). therefore, entrepreneurs who have established business ownership contribute positively towards innovation as they have greater capacity to bear the risk of innovation. kirchhoff (1994) also suggests that entrepreneurs of the established firms have the ability to scale up production towards innovation. the coefficient of total early-stage entrepreneurial activity (tea), which is highly significant, shows that it has a negative relationship with innovation capacity i.e. 1% increase in tea reduce the innovation index by 0.97 points. tea includes that proportion of the adult population that own new businesses and are the nascent entrepreneurs. it has been observed by picot et al., (1989) and aldrich (1990) that nascent entrepreneurs are imitators and bring little or no additional knowledge to the society. innovation involves novelty whereas the routines and abilities of nascent entrepreneurs differ significantly from already established firms (koellinger, 2008). therefore, it can be stated that early stage entrepreneurs are usually imitators and hence do not have a positive impact on innovation, thereby reducing the national innovation level. so, a country would innovate less that has a greater proportion of early-stage entrepreneurs that are not innovators, rather imitators, compared to those who have greater concentration of entrepreneurs with established business ownership. peris, ferreira and fernandes (2018) also state that there is a significant difference in the relationship between innovation practices and tea depending upon the level of economic development of a country. moreover, the business models of early-stage entrepreneurs have greater uncertainty and high risk of failure which makes it difficult to attract highly competent and motivated employees who are necessary for such early-stage businesses to be innovative (block, fisch & van praag, 2017). at macro level, the preferences and decision making of entrepreneurs depend highly upon the circumstances and the environment they operate in. many studies have pointed out that the expected returns and uncertainty greatly influence the entrepreneurial behavior (amit et al., 1995; van stel et al., 2006; acs, audretsch &lehmann, 2013). therefore, the alvina sabah idrees & saima sarwar 112 distribution of imitative and innovative business opportunities vary across countries and the potential failure is also relevant (koellinger, 2008). the current study also supports this argument, as depicted in table 3, that there is a significant negative relationship between entrepreneurs’ fear of failure rate (ffr) and country’s innovation. 1% increase in ffr reduces the innovation index by 0.15 points. this means that a higher rate of fear of failure will be a deterrence for entrepreneurs to take new initiatives for innovation. the estimates of one-step and two-step system gmm indicate that perceived opportunities in a nation have insignificant impact on its innovation capacity. therefore, its long-run estimate is not calculated. the higher percentage of perceived opportunities for entrepreneurial activities could contribute significantly towards innovation but high risk of perceived failure discourages entrepreneurs to undertake innovative ventures even though there seem to exist good opportunities. as koellinger (2008) describes that opportunities exists objectively, and it is not inevitable that those opportunities would certainly trigger innovation practices. innovation activities are risky in nature. therefore, to thrive innovation, a strong and supportive institutional environment is required. the result in table 3 shows a significant positive impact of property rights (pr) on innovation i.e. a 1 point increase in property rights index increases the value of gii by 0.26 points. property rights play an important role in building up institutional trust that helps to nurture entrepreneurship which is more innovative and productive. tebaldi and elmslie (2013) also conclude the same impact of property right protection on patent counts. when private property and wealth will be safe from unfair expropriation, then there will be greater incentive for entrepreneurs to seek profits through innovation. the coefficient of ses (state effectiveness score) is significant and negative. the high value of ses indicates greater inability of state to deliver its services and respond effectively under crisis. hence, there is trust deficit in fragile states. the high value of ses will deteriorate institutional trust that reduces the innovation capacity of a nation. the result indicates that in the long-run, 1 point increase in ses will lower down the gii value by 0.69 points. a higher value of ses reduces peoples’ confidence on reliability of state to carry out its activities. this uncertainty compels firms not to take innovative initiatives that are prone to risks. the result supports the argument provided by speakman and rysova (2015) who point out that risky ventures and innovation activities are highly suppressed among countries that are more fragile. kaasa, kaldaru, and parts (2007) also states that investors are less risk averse when there is higher trust in the society and it enables firms to spend their finance and time in innovative activities. similarly, audretsch, seitz and rouch (2018) find that institutional trust play a vital role in stimulating innovation performance. however, the finding of present study contradicts with that of kashi and afsari (2014) who conclude no significant relationship between innovation and institutional trust in asian countries. journal of applied economics and business studies, volume. 4, issue 1 (2020) 97-122 https://doi.org/10.34260/jaebs.415 113 lastly, there is a significant positive impact of foreign direct investment on innovation. similar results are also found by walz (1997), blyde (2003), chang, chen and mcaleer (2013), loukil (2016) and osano and koine (2016) i.e. fdi can be a source of innovation through the process of technology diffusion. fdi brings spillover effects on the economy by introducing new technologies which spur innovation. 7. conclusion the present study focuses on determining the impact of entrepreneurial behavior, property rights and state effectiveness on country level innovation. past literature considers trust building as an important aspect of social capital. while examining the role of social networking on entrepreneurial activity, casson and giusta (2007) determine that government plays a significant role as trust-broker. portela, vázquez-rozas, neira, and viera (2012) also conclude that public policies must focus on institutional and interpersonal trust to improve social environment which is necessary for developing an entrepreneurial society. similar results are also found by doh & zolnik (2011). the present study contributes to the existing literature by undertaking a country-level analysis. using suitable proxies to measure institutional trust, the impact is measured on national innovation capacity which is found positive. many studies have examined the role of entrepreneurship pertaining to innovation activities (iyigun & owen, 1999; wennekers et al., 2005; van praag & versloot, 2007; burke & fraser, 2011). entrepreneurs affect innovation either through exiting established business entrepreneurs which enjoy economies of scale and are intensive in r&d activities or though increasing competition from emerging nascent firms which force existing firms towards innovation (iyigun & owen, 1999). the findings of this study also confirms the positive relationship between entrepreneurial behavior and innovation. out of the four indicators of entrepreneurial behavior, fear of failure rate and total early stage entrepreneurs do not play a role in innovation. rather, these two factors suppress innovation in a country. however, there is a significant positive relationship between established business entrepreneurs and innovation. whereas perceived opportunities have an insignificant impact. this means that it is not inevitable that opportunities necessary trigger innovation. in addition, the study shows that property rights play an integral role in developing institutional trust which boosts entrepreneurialism to exploit innovative opportunities. on the other hand, institutional trust is lost in fragile countries which retard country’s innovation. innovative entrepreneur loses confidence on legal system when a state fails to implement its policies effectively. alvina sabah idrees & saima sarwar 114 these results have some strong implications for countries in formulating policies to increase national innovation levels. firstly, countries need to develop strong property rights protection to incentivize innovative entrepreneurs. secondly, the study observes that perceived entrepreneurial opportunities are not realized if fear of failure rate is high among potential entrepreneurs. therefore, a favorable entrepreneurial ecosystem must be developed so that fear of failure rate can be reduced. thirdly, such policies must be devised that could facilitate the early-stage entrepreneurs to grow into established businesses. since established business ownerships are more competitive and contributes positively towards innovation. lastly, institutional and legal framework must be strengthened to increase state legitimacy and enactment of property rights will cultivate institutional trust to help promote entrepreneurship towards innovation. references acemoglu, d., & johnson, s. (2005). unbundling institutions. journal of political economy, 113(5), 949-995. acemoglu, d., bimpikis, k., & ozdaglar, a. (2011). experimentation, patents, and innovation. american economic journal: microeconomics, 3(1), 37-77 acs, z. j., & amorós, j. e. (2008). entrepreneurship and competitiveness dynamics in latin america. small business economics, 31(3), 305-322. acs, z. j., & audretsch, d. b. (1988). innovation in large and small firms: an empirical analysis. the american economic review, 678-690. acs, z. j., & sanders, m. (2012). patents, knowledge spillovers, and entrepreneurship. small business economics, 39(4), 801-817. acs, z. j., audretsch, d. b., & lehmann, e. e. (2013). the knowledge spillover theory of entrepreneurship. small business economics, 41(4), 757-774. acs, z. j., braunerhjelm, p., audretsch, d. b., & carlsson, b. (2009). the knowledge spillover theory of entrepreneurship. small business economics, 32(1), 15-30. acs, z. j., desai, s., & hessels, j. (2008). entrepreneurship, economic development and institutions. small business economics, 31(3), 219-234. aidis, r. (2017). do institutions matter for entrepreneurial development?. iza world of labor. aldrich, h. (1999). organizations evolving. sage publications amesse, f., desranleau, c., etemad, h., fortier, y., & seguin-dulude, l. (1991). the individual inventor and the role of entrepreneurship: a survey of the canadian evidence. research policy, 20(1), 13-27. amit, r., muller, e., & cockburn, i. (1995). opportunity costs and entrepreneurial activity. journal of business venturing, 10(2), 95-106. journal of applied economics and business studies, volume. 4, issue 1 (2020) 97-122 https://doi.org/10.34260/jaebs.415 115 anderson, t. w., & hsiao, c. (1982). formulation and estimation of dynamic models using panel data. journal of econometrics, 18(1), 47-82. anokhin, s., & schulze, w. s. (2009). entrepreneurship, innovation, and corruption. journal of business venturing, 24(5), 465-476. arellano, m., & bond, s. (1991). some tests of specification for panel data: monte carlo evidence and an application to employment equations. review of economic studies, 58, 277–297. audretsch, d. b., & thurik, a. r. (2001). what's new about the new economy? sources of growth in the managed and entrepreneurial economies. industrial and corporate change, 10(1), 267-315. audretsch, d. b., keilbach, m. c., & lehmann, e. e. (2006). entrepreneurship and economic growth. usa: oxford university press. audretsch, d. b., seitz, n., & rouch, k. m. (2018). tolerance and innovation: the role of institutional and social trust. eurasian business review, 8(1), 71-92. baltagi, b. (2008). econometric analysis of panel data. john wiley & sons. barasa, l., kimuyu, p., vermeulen, p., knoben, j., & kinyanjui, b. (2014). institutions, resources and innovation in developing countries: a firm level approach. dfid working paper. baum, c. f., schaffer, m. e., & stillman, s. (2003). instrumental variables and gmm: estimation and testing. the stata journal, 3(1), 1-31. baumol, w. j. (1996). entrepreneurship: productive, unproductive, and destructive. journal of business venturing, 11(1), 3-22. block, j. h., fisch, c. o., & van praag, m. (2017). the schumpeterian entrepreneur: a review of the empirical evidence on the antecedents, behaviour and consequences of innovative entrepreneurship. industry and innovation, 24(1), 61-95. blundell, r., & bond, s. (1998). initial conditions and moment restrictions in dynamic panel data models. journal of econometrics, 87(1), 115-143. blyde, j. s. (2003). the role of foreign direct investment and imports of capital goods in the north-south diffusion of technology. journal of economic integration, 545-562. caraça, j., lundvall, b. å., & mendonça, s. (2009). the changing role of science in the innovation process: from queen to cinderella? technological forecasting and social change, 76(6), 861-867. casson, m., & giusta, m. d. (2007). entrepreneurship and social capital: analysing the impact of social networks on entrepreneurial activity from a rational action perspective. international small business journal, 25(3), 220-244. casson, m., & wadeson, n. (2007). entrepreneurship and macroeconomic performance. strategic entrepreneurship journal, 1(3‐4), 239-262. alvina sabah idrees & saima sarwar 116 chang, c. l., chen, s. p., & mcaleer, m. (2013). globalization and knowledge spillover: international direct investment, exports and patents. economics of innovation and new technology, 22(4), 329-352. choi, y. r., & shepherd, d. a. (2004). entrepreneurs’ decisions to exploit opportunities. journal of management, 30(3), 377-395. christensen, t., & lægreid, p. (2005). trust in government: the relative importance of service satisfaction, political factors, and demography. public performance & management review, 28(4), 487-511 d'agostino, g., & scarlato, m. (2016). institutions, innovation and economic growth in european countries. mpra paper no. 72427. de mel, s., mckenzie, d., & woodruff, c. (2009). innovative firms or innovative owners? determinants of innovation in micro, small, and medium enterprises. the world bank. dincer, o. (2019). does corruption slow down innovation? evidence from a cointegrated panel of us states. european journal of political economy, 56, 1-10. doh, s., & zolnik, e. j. (2011). social capital and entrepreneurship: an exploratory analysis. african journal of business management, 5(12), 4961-4975. donatiello, d., & ramella, f. (2017). the innovation paradox in southern europe. unexpected performance during the economic crisis. south european society and politics, 22(2), 157-177. donges, a., meier, j. m. a., & silva, r. (2019). the impact of institutions on innovation. available at ssrn 2815541. draghici, a., & albulescu, c. t. (2014). does the entrepreneurial activity enhance the national innovative capacity? procedia-social and behavioral sciences, 124, 388396. drucker, p. f. (1985). the discipline of innovation. harvard business review, 63(3), 6772. edquist, c., & mckelvey, m. (1998). high r&d intensity without high tech products: a swedish paradox? in k. nielsen & b. johnson (eds.) institutions and economic change: new perspectives on markets, firms and technology (pp. 131-149). northampton, ma: edward elgar publishing ehrlich, i., li, d., & liu, z. (2017). the role of entrepreneurial human capital as a driver of endogenous economic growth. journal of human capital, 11(3), 310-351. elert, n., & henrekson, m. (2017). entrepreneurship and institutions: a bidirectional relationship. foundations and trends® in entrepreneurship, 13(3), 191-263. faggio, g., & silva, o. (2014). self-employment and entrepreneurship in urban and rural labour markets. journal of urban economics, 84, 67-85. fischer, b., & tello-gamarra, j. (2017). institutional quality as a driver of efficiency in laggard innovation systems. journal of globalization, competitiveness & journal of applied economics and business studies, volume. 4, issue 1 (2020) 97-122 https://doi.org/10.34260/jaebs.415 117 governability/revista de globalización, competitividad y gobernabilidad/revista de globalização, competitividade e governabilidade, 11(1), 129-144. furman, j. l., porter, m. e., & stern, s. (2002). the determinants of national innovative capacity. research policy, 31(6), 899-933. garcía, f., jin, b., & salomon, r. (2013). does inward foreign direct investment improve the innovative performance of local firms? research policy, 42(1), 231-244. gerth, h. h., & mills, c. w. (1946). from max weber: essays in sociology, (pp. 77-128). new york: oxford university press gort, m., & klepper, s. (1982). time paths in the diffusion of product innovations. the economic journal, 92(367), 630-653. graham, s. j., merges, r. p., samuelson, p., & sichelman, t. (2009). high technology entrepreneurs and the patent system: results of the 2008 berkeley patent survey. berkeley technology law journal, 1255-1327. grebel, t., pyka, a., & hanusch, h. (2003). an evolutionary approach to the theory of entrepreneurship. industry and innovation, 10(4), 493-514. hall, r. e., & jones, c. i. (1999). why do some countries produce so much more output per worker than others?. the quarterly journal of economics, 114(1), 83-116. hanusch, h., & pyka, a. (2006). principles of neo-schumpeterian economics. cambridge journal of economics, 31(2), 275-289. henrekson, m., & sanandaji, t. (2011). the interaction of entrepreneurship and institutions. journal of institutional economics, 7(1), 47-75. hirschman, a. o. (2005). the strategy of economic development (new haven, 1958). innowacje i transfer technologii. słownik pojęć, red. kb matusiak, parp, warszawa. hoff, k. (2000). beyond rosenstein-rodan: the modern theory of coordination problems in development. in: proceedings of the annual world bank conference on development economics (pp. 145-188). washington, dc: world bank publications. iyigun, m. f., & owen, a. l. (1999). entrepreneurs, professionals, and growth. journal of economic growth, 4(2), 213-232. kaasa, a., kaldaru, h., & parts, e. (2007). social capital and institutional quality as factors of innovation: evidence from europe. university of tartu faculty of economics and business administration working paper, (55-2007). kashi, k. f., & afsari, a. (2014). the impact of social capital on innovation in selected countries. iranian journal of economic studies, 3(2), 81-98. kirchhoff, b. a. (1994). entrepreneurship and dynamic capitalism: the economics of business firm formation and growth. westport, connecticut london: praeger publishers. knight, f. h. (2012). risk, uncertainty and profit. courier corporation. alvina sabah idrees & saima sarwar 118 koçak, e. (2017). does institutional quality drive innovation? evidence from system-gmm estimates. mpra paper no. 90299 available at: https://mpra.ub.unimuenchen.de/id/eprint/90299 kodila-tedika, o., & agbor, j. (2016). does trust matter for entrepreneurship: evidence from a cross-section of countries. economies, 4(1), 4. koellinger, p. (2008). why are some entrepreneurs more innovative than others?. small business economics, 31(1), 21. lee, k. (2013). schumpeterian analysis of economic catch-up: knowledge, path-creation, and the middle-income trap. cambridge university press. levie, j., & autio, e. (2008). a theoretical grounding and test of the gem model. small business economics, 31(3), 235-263. llobet, g., & suarez, j. (2013). entrepreneurial innovation, patent protection, and industry dynamics. working paper from cemfi. available at https://www.cemfi.es/ftp/ wp/1001.pdf loukil, k. (2016). foreign direct investment and technological innovation in developing countries. oradea journal of business and economics, 1(2), 31-40. lucas jr, r. e. (1988). on the mechanics of economic development. journal of monetary economics, 22(1), 3-42. lumpkin, g. t., & dess, g. g. (1996). clarifying the entrepreneurial orientation construct and linking it to performance. academy of management review, 21(1), 135-172. maltby, t. (2013). european union energy policy integration: a case of european commission policy entrepreneurship and increasing supranationalism. energy policy, 55, 435-444. mansfield, e. (1986). patents and innovation: an empirical study. management science, 32(2), 173-181. mcmullen, j. s., & shepherd, d. a. (2006). entrepreneurial action and the role of uncertainty in the theory of the entrepreneur. academy of management review, 31(1), 132-152. miller, d. (1983). the correlates of entrepreneurship in three types of firms. management science, 29(7), 770-791. mishler, w., & rose, r. (2001). what are the origins of political trust? testing institutional and cultural theories in post-communist societies. comparative political studies, 34(1), 30-62. moreno, r., paci, r., & usai, s. (2005). spatial spillovers and innovation activity in european regions. environment and planning a, 37(10), 1793-1812. moser, p. (2013). patents and innovation: evidence from economic history. journal of economic perspectives, 27(1), 23-44. https://mpra.ub.uni-muenchen.de/id/eprint/90299 https://mpra.ub.uni-muenchen.de/id/eprint/90299 https://www.cemfi.es/ftp/%20wp/1001.pdf https://www.cemfi.es/ftp/%20wp/1001.pdf journal of applied economics and business studies, volume. 4, issue 1 (2020) 97-122 https://doi.org/10.34260/jaebs.415 119 murray, f., & budden, p. (2017). a systematic mit approach for assessing ‘innovationdriven entrepreneurship’in ecosystems (iecosystems).working paper, september 2017. mit lab for innovation science and policy, mit innovation initiative. nagaoka, s., motohashi, k., & goto, a. (2010). patent statistics as an innovation indicator. in handbook of the economics of innovation (vol. 2, pp. 1083-1127). northholland. north, d. c. (1991). institutions. journal of economic perspectives, 5(1), 97-112. oecd (2013), “trust in government, policy effectiveness and the governance agenda”, in government at a glance 2013, oecd publishing, paris. doi: https://doi.org/10.1787/gov_glance-2013-6-en osano, h. m., & koine, p. w. (2016). role of foreign direct investment on technology transfer and economic growth in kenya: a case of the energy sector. journal of innovation and entrepreneurship, 5(1), 31. picot, a., schneider, d., & laub, u. (1989). transaktionskosten und innovative unternehmensgründung. schmalenbachs zeitschrift für betriebswirtschaftliche forschung, 358-387. portela, m., vázquez-rozas, e., neira, i., & viera, e. (2012). entrepreneurship and economic growth: macroeconomic analysis and effects of social capital in the eu. in h. burger-helmchen (ed.), entrepreneurship–born, made and educated (pp. 249264). rijeka, coratia: intech. prodan, i. (2005). influence of research and development expenditures on number of patent applications: selected case studies in oecd countries and central europe, 1981-2001. applied econometrics and international development, 5(4), 5-22. robinson, l. (2009). a summary of diffusion of innovations. enabling change, 5(10), 110. rodrik, d. (2004). institutions and economic performance-getting institutions right. cesifo dice report, 2(2), 10-15. rodrik, d., subramanian, a., & trebbi, f. (2004). institutions rule: the primacy of institutions over geography and integration in economic development. journal of economic growth, 9(2), 131-165. romer, p. m. (1986). increasing returns and long-run growth. journal of political economy, 94(5), 1002-1037. roodman, d. (2009). how to do xtabond2: an introduction to difference and system gmm in stata. the stata journal, 9(1), 86-136. samadi, a. h. (2019). institutions and entrepreneurship: unidirectional or bidirectional causality? journal of global entrepreneurship research, 9(1), 1-16. https://doi.org/10.1787/gov_glance-2013-6-en https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1238323## alvina sabah idrees & saima sarwar 120 sarafidis, v., yamagata, t., & robertson, d. (2009). a test of cross section dependence for a linear dynamic panel model with regressors. journal of econometrics, 148(2), 149-161. sarbah, a., & xiao, w. (2013). the importance of trust in the development of entrepreneurship. international journal of advancements in research & technology, 2(12), 230-244. sautet, f. e. (2005). the role of institutions in entrepreneurship: implications for development policy. mercatus policy primer no. 1. available at ssrn: https://ssrn.com/abstract=1264033 savrul, m., & incekara, a. (2015). the effect of r&d intensity on innovation performance: a country level evaluation. procedia-social and behavioral sciences, 210(december), 388-396. schumpeter, j. a. (1934). the theory of economic development: an inquiry into profits, capital, credits, interest, and the business cycle. transaction publishers, piscataway. shaffer, m. j. (2011). entrepreneurial innovation: patent rank and marketing science. unpublished dissertation, washington state university, pullman, washington. shane, s., & venkataraman, s. (2000). the promise of entrepreneurship as a field of research. academy of management review, 25(1), 217-226. solow, r. m. (1956). a contribution to the theory of economic growth. the quarterly journal of economics, 70(1), 65-94. speakman, j., & rysova, a. (2015). the small entrepreneur in fragile and conflictaffected situations. the world bank. stam, e., & nooteboom, b. (2011). entrepreneurship, innovation and institutions. in d. b. audretsch, o. falck, s. heblich & a. lederer (eds.), handbook of research on innovation and entrepreneurship (pp. 421-436). massachusetts, usa: edward elgar publishing. tebaldi, e., & elmslie, b. (2013). does institutional quality impact innovation? evidence from cross-country patent grant data. applied economics, 45(7), 887-900. thirlwall, a. p. (2005). growth and development (vol. 2). palgrave macmillan. van praag, c. m., & versloot, p. h. (2007). what is the value of entrepreneurship? a review of recent research. small business economics, 29(4), 351-382. van stel, a. j., storey, d. j., & thurik, a. r. (2006). the effect of business regulations on nascent and actual entrepreneurship (no. 0406). papers on entrepreneurship, growth and public policy. von mises, l. (1996). human action. a treatise on economics, 4. walz, u. (1997). innovation, foreign direct investment and growth. economica, 64(253), 63-79. https://ssrn.com/abstract=1264033 journal of applied economics and business studies, volume. 4, issue 1 (2020) 97-122 https://doi.org/10.34260/jaebs.415 121 wennekers, s., van stel, a., thurik, r., reynolds, p. (2005). nascent entrepreneurship and the level of economic development. small business economics, 24, 293-309. williamson, c. r., & kerekes, c. b. (2011). securing private property: formal versus informal institutions. the journal of law and economics, 54(3), 537-572. wong, p. k., ho, y. p., & autio, e. (2005). entrepreneurship, innovation and economic growth: evidence from gem data. small business economics, 24(3), 335-350. wooldridge, j. (2001). applications of generalized method of moments estimation. journal of economic perspectives, 15(4), 87-100. wooldridge, j. (2002). introductory econometrics analysis of cross section and panel data. the mit press alvina sabah idrees & saima sarwar 122 appendix: table 4: list of countries australia, barbados, belgium, brazil, chile, china, colombia, croatia, denmark, ecuador, egypt, finland, france, germany, greece, guatemala, hong kong, hungary, india, iran, ireland, italy, jamaica, japan, kazakhstan, korea, latvia, lebanon, macedonia, malaysia, mexico, morocco, netherlands, norway, panama, peru, poland, portugal, romania, russian federation, saudi arabia, slovakia, slovenia, south africa, spain, sweden, switzerland, thailand, trinidad and tobago, tunisia, turkey, uganda, united arab emirates, united kingdom, united states of america, uruguay table 5: estimated results of pooled ols, fixed effect and instrumental variable regression (dependent variable: inn) variable pooled ols fixed effect 2sls /iv regression coefficient std. error coefficient std. error coefficient std. error innt-1 0.414*** 0.037 -0.061** 0.029 0.191*** 0.068 po -0.003 0.023 0.051 0.032 0.348*** 0.079 ffr -0.069* 0.036 -0.099*** 0.037 0.143 0.222 tea -0.455*** 0.072 -0.279*** 0.085 -1.882*** 0.320 ebo 0.292*** 0.072 0..151 0.108 1.492*** 0.036 fdi 0.043 0.042 0.009 0.035 0.133* 0.069 pr 0.155*** 0.018 0.169** 0.066 0.057 0.036 ses -0.499*** 0.187 -0.127 0.462 -1.647*** 0.432 constant 22.96*** 2.248 40.59*** 4.169 24.205** 11.21 diagnostics test: breusch-pagan test χ2(1)=0.74 prob.= 0.389 hausman test χ2(8) = 194.10 prob.= 0.000 endogeneity test: durbin (score) χ2(4) = 114.463 prob. = 0.000 wu-hausman test f-statistic= 39.67 prob. = 0.000 ramsey reset test f-statistic = 4.49 prob. = 0.004 (presence of omitted variable bias) pesaran test of cross sectional independence prob. = 0.000 f-test prob. = 0.000 modified wald test for heteroskedasticity χ2(55)=5906.79 prob. = 0.000 test for over identifying restrictions: sargan (score) χ2(5) = 9.068 prob. = 0.106 basman test χ2(5) = 8.947 prob. = 0.111 wooldridge test for autocorrelation f-statistic = 99.05 prob. = 0.000 pagan-hall test statistic for heteroskedasticity: χ2(13) = 8.947 prob. = 0.015 total observations 384 384 377 ***, ** and * indicates the significance level at 1%, 5% and 10% level, respectively. pooled ols estimates are biased due to omitted variable bias. the f-test suggests that fixed effect model be applied. the hausman test also validates the fixed effect model. though, the diagnostics of fixed effect model show serious flaw in the validity of estimates. the post estimates of iv/2sls regression show the validity of instruments as well as endogeneity of variables but there is strong evidence of heteroskedasticity. journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 41 journal of applied economics and business studies (jaebs) journal homepage: https://pepri.edu.pk/jaebs issn (print): 2523-2614 issn (online) 2663-693x institutions, digitization, innovation and venture capital: evidence from europe and the asia-pacific muhammad zubair khan1*, zafir ullah khan2 & affan hameed 3 1 institute of management sciences, university of science and technology bannu, pakistan 2 department of economics, university of science and technology bannu, pakistan 3 london college of fashion, university of arts london, uk abstract the article examines the impact of information and communication technologies (ict), innovation, and formal and informal institutions on venture capital (vc) investment. the analysis is based on 28-year data spanning 1990-2017 from 19 european and 13 asia-pacific countries using generalized two-stage least square instrumental variable technique. after controlling for endogeneity, the results show that ict, innovation, and informal institutions hold a strong impact on vc investment. ict and innovation exert a positive and significant influence on vc investment whereas formal institutions exert a positive yet insignificant effect on vc investment. among the informal institutions, power distance and individualism exert significant and positive influence whereas uncertainty avoidance has significant and negative influence on vc investment. the interaction analysis demonstrates that the association between ict and vc is strong when institutional quality is high. moreover, the impact of innovation on vc is pronounced in highly digitized and highly uncertainty-tolerant environments. explanation of vc capital investment also vary with geography as the effects of trend, ict and uncertainty avoidance on vc investment are noticeable in the asia-pacific region whereas power distance is prominent in the european region. the article makes important contributions to the literature of vc by revealing novel interactions between formal and informal institutions, ict and innovation depicted in a conceptual model. the study also brings in important highlights to the policy debate on vc development by showing how exactly vc investments are tangled with the different dimensions of institutional and technological environment. keywords venture capital, private equity, formal institutions, informal institutions, national cultures, innovation, ict jel classification l32, l33, g24 * email: mzkhan@ustb.edu.pk muhammad zubair khan, zafir ullah khan & affan hameed 42 1 introduction venture capital (vc) originated in the us more than 70 years ago, but it was only in the late 1970s and early 1980s that it became institutionalized. later, it spread to western europe and other parts of the world including asian economies. the world vc market survived the dot-com bubble and then the 2008 financial crisis despite the risky nature of such investments. today, the world vc industry stands at around us$2 trillion in terms of dry power1. the us vc investment reached us$130 billion in 2018 surpassing the dot-com era for the first time (pitchbook-nvca, 2018). us and some of the western european economies have more established vc markets but in rest of the world, there is still cross-country disparity in vc activity. researchers have explored different factors influencing the development of vc markets around the globe, however, most of those studies have examined the established markets. in their analysis of 314 vc research articles since 2011, 52% of articles still rely on us vc data based on meta-analysis of 314 vc prior studies (tykvova, 2017). asian vc market has grown tremendously after the 2000 high-tech market crash. for example, asia-pacific region experienced vc investment twice than its european counterpart in 2015 (see figure 1a and 1-b). there might be different institutional patterns in the economies that have nascent vc markets particularly eastern europe and some of the asian markets. but this is not the sole reason to re-visit vc investment. the primary motivation of the paper is as follows. vc is a knowledge and information intensive market which is affected by factors related to information asymmetry and knowledge creation. the factors tied by common features of information and knowledge include ict, patents, and institutions. ict has become essential part of the modern life. the landscape of the world business has changed, and digitally powered technology firms dominate the world business. the top three billionaires crossing us$100 billion mark for the first time in world history in personal wealth are the owners of digital firms namely microsoft, amazon, and facebook2. most of the vc determinants research have used data of web1.0 era before the transformation of internet in the context of web2.0 (aldrich, 2014; see the appendix 1 for previous studies and the periods covered). ict makes vc processes more efficient through cost reduction and information symmetry, on the one hand, and boost entrepreneurial activities on the other hand. patents catalyze entrepreneurial activity and promotes demand side of vc. institutions also impact vc markets because “they reduce transaction costs, provide information under uncertainty, and stabilize expectations about the behavior of others” (risse, 2000, p.4). north 1 dry powder includes venture capital, buyout, real estate, and infrastructure (bain & company, 2019) 2 while bill gates and jeff bezos had already passed the us$100 billion-mark, mark zuckerberg became the third billionaire passing that milestone in august 2020. url: https://www.bloombergquint.com/markets/zuckerberg-s-fortune-surpasses-100-billion-with-facebooksurge#:~:text=(bloomberg)%20%2d%2d%20mark%20zuckerberg's%20net,of%20its%20tiktok%20competitor%20reels. accessed on september 5, 2020. https://www.bloombergquint.com/markets/zuckerberg-s-fortune-surpasses-100-billion-with-facebook-surge#:~:text=(bloomberg)%20%2d%2d%20mark%20zuckerberg's%20net,of%20its%20tiktok%20competitor%20reels https://www.bloombergquint.com/markets/zuckerberg-s-fortune-surpasses-100-billion-with-facebook-surge#:~:text=(bloomberg)%20%2d%2d%20mark%20zuckerberg's%20net,of%20its%20tiktok%20competitor%20reels journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 43 (1991) argues that institutions are created to reduce uncertainty and maintain order and therefore determine transaction and productions costs. formal institutions are not only associated with vc but also with ict. for example, studies have found that ict, particularly, e-governments reduce corruption (sturges, 2004; srivastava, teo and devaraj, 2016) and that formal institutions are important to take advantage of ict (andrés, amavilah, & asongu, 2017). while innovation and formal institutions are well studied in the extant research, ict and informal institutions have been less recognized previously. only a few studies have explored the impact of informal institutions (li & zahra, 2012) and the ict (khan et al., 2020). figure 1(a): vc investments in usd in asia-pacific and europe 1990-2015 figure 1(b): vc investments % gdp in asia-pacific and europe 1990-2017 (source: asia-pacific data from avcj, european data from eurostat and evca yearbooks). 0 20000 40000 60000 80000 100000 120000 140000 160000 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 1 9 9 7 1 9 9 8 1 9 9 9 2 0 0 0 2 0 0 1 2 0 0 2 2 0 0 3 2 0 0 4 2 0 0 5 2 0 0 6 2 0 0 7 2 0 0 8 2 0 0 9 2 0 1 0 2 0 1 1 2 0 1 2 2 0 1 3 2 0 1 4 2 0 1 5 asia europe muhammad zubair khan, zafir ullah khan & affan hameed 44 note: the chart is based on 13 asia-pacific countries and 19 european countries included in the study within the literature on vc, previous research lacks the moderating role of institutions. there is a scope for research on how informal institutions affect decisions in the vc market as vc processes take place in an environment of uncertainty and vcs invest in ventures that are opaque. in that context, culture of uncertainty avoidance might have an impact on the chance that a patent is translated into an innovative product that subsequently attracts vc. furthermore, the effect of ict embeddedness may vary with formal institutions as institutions protect rights of minority shareholdings and property rights. moreover, it is also likely that ict may have impact the association between innovation and vc as ict acts as efficiencyenhancing technology that can create a competitive advantage for smes through product innovation (higón, 2012). the current study examines the impact of formal and informal institutions, innovation and ict on vc investments in europe and asia. it also examines how these factors interact with each other in their influence on vc activity. the present study also examines how different factors vary in their impact on vc investment across the two regions i.e. europe and asia-pacific. before proceeding to the next section, it is important to define vc. vc and private equity are two closely related concepts. strictly, in the us, vc involves only seed, start-up, and expansion investments whereas private equity includes vc, buyouts, consolidations, and turnarounds. outside the us, the concept of vc and private equity converge, and vc is usually referred to what is called private equity in us context (jeng and wells, 2000). in europe, many vc firms offer equity investments would be called private equity financing in the us (black and gilson, 1998). the concept gets further complicated in japan where venture capitalists (vcs) extend loans based on interest rather than equity alone. in the current study, consistent with wright, pruthi and lockett (2005), we take the broader definition of vc, which covers seed, start-up, later stage, expansion, growth, replacement and buyouts which is synonymous to private equity. the current article cites both vc and private equity literature – or in other words literature of venture capital in both narrower sense and broader sense – which is a standard practice in academic work on vc (kumar and orleck, 2002). 2 hypothesis development 2.1 formal institutions formal institutions consist of constitutions, laws, and property rights (north, 1991). since institutions differ across regions, they exert varying influences. for example, investment patterns of various investor types are enormously different in the us from those practiced in europe (bertoni and colombo, 2015). groh and von liechtenstein (2009) find that corporate governance and the protection of investors’ rights have positive impact on the attractiveness of journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 45 a vc investment in a country. rule of law and government effectiveness have a negative effect on cleantech vc activity (cumming, henriques and sadorsky, 2016). freedom from corruption has a significant negative effect on early stage vc (cherif and gazdar, 2011). m&a investment volume, the shareholder suits index, and the legal rights index have a significant positive effect on vc investment (groh and wallmeroth, 2016). armour and cumming (2006) find that temperate bankruptcy laws for entrepreneurs trigger demand for risk capital. institutions also affect international allocations. ragosa and warren (2019) show that regulatory support measures and feed-in tariffs, coupled with political stability, are strong drivers of cross-border investment in renewable energy in developing countries. guler and guillén (2010) find that host countries’ characteristics such as technological, legal, financial, and political institutions attract international venture capital firms to invest in those destinations. they posit that these institutions offer technological opportunities, protect property rights, catalyze exits, and bring legal stability, respectively. in a survey of limited partners world-wide, groh and liechtenstein (2011) find the protection of property rights is the primary concern while allocating capital internationally. h-1: quality of formal institutions are likely to have a direct significant positive effect on vc investments. 2.2 informal institutions informal institutions consist of “sanctions, taboos, customs, traditions, and codes of conduct” (north, 1991, p.97). in the current study, we take hofstede’s national cultures as informal institutions. surprisingly, informal institutions have received little attention from vc scholars. black and gilson (1998) refute the role of culture in the development of vc markets. however, kenney, han and tanaka (2003) and kumar and orleck (2002) show, that the cultural factors influence the development of vc. kenney, han and tanaka (2002) contend that korean entrepreneurs insist on deals that eventually lead to transfer of control back to their family, indicating a strong collectivist culture. li & zahra (2012) find that formal institutions interact with hofstede cultural dimensions, particularly, uncertainty avoidance and collectivism. in the current research, we will use hofstede cultural dimensions to examine the impact of culture on vc. hofstede dimensions of national cultures are frequently used to investigate the influence of culture in business and economics research. initially found in ibm study, the four cultural dimensions include uncertainty avoidance, power distance, individualism, and masculinity. uncertainty avoidance is defined as ”the extent to which the members of a culture feel threatened by ambiguous or unknown situations” (hofstede, hofstede and minkov, 2010, p.191). countries high on uncertainty avoidance resist change and innovation because people have rigid beliefs and do not welcome new ideas. moreover, security is an important muhammad zubair khan, zafir ullah khan & affan hameed 46 component in individual motivations. societies that are high on uncertainty avoidance rely more on written rules and regulations, rely on informal structures to deal with uncertainty, and show less tolerance for ambiguity and change (kreiser et al., 2010). this might have implications for vc development as societies that are more uncertainty tolerant are prepared to take more entrepreneurial risks (spencer and gómez, 2004). h-2: uncertainty avoidance is likely to have a significant negative impact on vc investments. power distance represents the degree to which less powerful people in a hierarchy expect and accept the influence of more powerful people (hofstede, hofstede and minkov, 2010). this dimension may be an important factor influencing the relationship between vcs and entrepreneurs, given the notion that control has an element of struggle over power. vcs and entrepreneurs could have issues over the control of the firm in societies where power sharing is resisted. at the time of deal structuring, entrepreneurs are likely to hand over control more easily in high power distance societies compared to entrepreneurs in low power distance cultures. h-3: power distance is likely to have a significant positive impact on vc investments. hofstede’s individualism represents the cultural norms regulating actions in society pertaining to interdependence of people. to what extent do they support or depend on each other based on their blood relationship, community feelings, or friendship. in collectivist societies, personal relationships, friendships, family connections, and networks play a significant role. moreover, violations of promises and rules/norms are corrected by shame mechanisms (hofstede, 2012). in collectivist societies, firms rely more on informal institutions than formal markets (world development report, 2002). this might have implications for vc development. for example, in collectivist societies, entrepreneurs may want to pass their business on to their children and therefore resist handing over control to the vc firm (spencer and gómez, 2004). thus, it is hypothesized: h-4: individualism is likely to have a significant positive impact on vc investments. hofstede (2020) notes that “in masculine countries people “live in order to work”, managers are expected to be decisive and assertive, the emphasis is on equity, competition and performance and conflicts are resolved by fighting them out”3. in a masculine society, people are assertive and want to be the best, and perhaps these attitudes could be problematic considering the issue of control between the vcs and entrepreneurs. additionally, the competition between individual firms could be fierce and therefore collective action and cooperation which is essential for vc development might be a problem at industry level. cooperation is essential to develop vc as the key of vc institutionalization in the us in the 3 available at https://www.hofstede-insights.com/country/south-africa/ accessed on september 5, 2020. the date shows the date of access not the date of publication on the website as it does not mention any date of publication. https://www.hofstede-insights.com/country/south-africa/ journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 47 1980s was the high level of cooperative behavior (bruton et al., 2005). thus, a masculine culture is likely to be problematic for vc development. the formal hypothesis is: h-5: masculinity is likely to exert a significant negative impact on vc investment. 2.3 ict another important element of an entrepreneurial environment is the information and communication technology (ict) that has changed the landscape of business and technology in the last two decades, the period in which vc has also flourished enormously. ict has enhanced financial participation (pradhan, arvin, nair, bennett & bahmani, 2017) financial access, inclusion (asongu & acha-anyi, 2017; gabor & brooks, 2017) and improved financial development (pradhan, arvin and norman, 2015; asongu and moulin, 2016; pradhan et al., 2018; lechman and marszk, 2019)4. financial systems are deemed as information systems (ocampo, 2018). vc investment is knowledge-intensive instrument where such investors highly depend on information about the portfolio companies before and after the investment is finalized (carey, prowse, rea & udell, 1993). thus, internet is not only the tool that has made vc and vc processes more efficient, but it has also created the most dynamic entrepreneurial environment. so far, vc studies have ignored the power of internet and related technologies while in this research, we expect it to have a strong influence on the vc investment. formally it is proposed that: h-6: ict is likely to have a significant positive impact on vc investments. 2.4 innovation and technological opportunity schertler (2007) finds a strong positive impact of number of patents, number of r&d researchers and r&d expenditure on vc investments. besides, the impact of total knowledge capital in terms of volume is not highly significant. also, government-financed knowledge capital has weak explanatory power for the size of vc investments, probably due to commercial applications (keuschnigg and nielsen, 2003) or due to its long-term focus which is translated into commercial applications in more than two years. schertler (2007), however, finds that businessfinanced gerd does not affect the volume of investments. schertler (2003) also finds that human capital endowment plays a positive role in vc development. however, according to da rin, nicodano and sembenelli (2006), r&d has no effect on early stage and high-tech vc investment. we expect a positive impact of innovative potential (represented by patents) to have positive impact on vc investment because innovative activities boost entrepreneurial activities that ultimately attracts vc investment. more formally it is hypothesized: h-7: innovation is likely to have a significant positive impact on vc investments. 4 for detailed literature read (lechman and marszk, 2019). muhammad zubair khan, zafir ullah khan & affan hameed 48 2.5 ict and formal institutions there are studies showing that formal institutions have an impact on ict adoption. martinez & williams (2010) support the institutional view to explain ict adoption and shows that formal institutions exert influence on ict adoption and therefore electronic commerce. the institutional view is based on the premise that quality institutions stimulate the confidence of private actors in the ability of government to set and enforce ‘rules of the game’ for protection of commercial transactions (shareef, kumar and kumar, 2008; martinez and williams, 2010). research also shows that ict and institutional quality play a mediating role between electronic government and corruption (adam, 2020). more importantly, studies demonstrate that institutions have moderated the impact of ict on development and anticorruption practices. andrés, amavilah, & asongu (2017) also show that formal institutions are essential to promote ict adoption for development. sassi & goaied (2013) show that the effect of ict diffusion on corruption in african economies depends on the rule of law. therefore, it is expected that institutions may also moderate the effect of ict on vc. hence, we formulate the following hypothesis: h-8: the effect of ict on vc investments is likely to vary significantly with quality of formal institutions 2.6 ict, informal institutions and innovation the key point in this section is that the impact of innovation on vc investment is moderated by culture of uncertainty avoidance and ict. the rationale is that all these are important from information, knowledge and risk-taking point of view. as pointed out in khan et al. (2020) that vc is an information-problematic and knowledge-intensive industry so much that vcs had to rely on informal informants (fiet, 1995; lockett et al., 2002). vcs highly depend on pre-investment information to avoid adverse selection and post-investment information to evade moral hazard (wright and robbie, 1998). they invest in opaque, high risk, high growth-potential smes with little or no transactional history. since uncertainty in private placements is relatively high, public intermediation is more reluctant to commit resources to such ambiguous transactions. resultantly, such private investments involve more due diligence and monitoring compared to other financing alternatives (carey et al., 1993). ict has made information sharing very easy that has facilitated processes of information collection for industry selection, firm selection, deal origination, monitoring and exits. there is also evidence that ict has a positive impact on firm innovativeness. for example hall et al. (2013) demonstrate that investment in ict and r&d have a strong impact on innovation and productivity in manufacturing firms in italy. ollo-lópez & aramendía-muneta (2012) also show that ict enhances innovation in the glass, ceramics, and cement concrete journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 49 industry. therefore, it is expected that ict may moderate the impact of innovation on vc investment. the proposition is as follows: h-9: impact of innovation on vc investments is likely to vary significantly with level of ict. culture also has a significant impact on innovation (efrat, 2014). three of the hofstede’s cultural dimensions have relevance to innovation, namely, uncertainty avoidance, individualism, and power distance. however, there is strong theoretical relevance of uncertainty avoidance to vc research because of importance of information and knowledge in the vc processes particularly where innovative start-ups and innovative products are involved. naturally, investors in high uncertainty avoidance cultures would commit more time and money to secure deals than their counterparts in high uncertainty tolerance cultures. there is evidence in previous research that innovation varies with culture of uncertainty avoidance. high uncertainty avoidance exerts a negative influence on innovation (shane, 1995), innovation diffusion takeoff (tellis, stremersch and yin, 2003) and new product adoption (yeniyurt and townsend, 2003). uncertainty avoidance has a negative relationship with innovation diffusion and that the impact of uncertainty avoidance changes from negative to positive in later stage diffusion (he and lee, 2020). the key argument is that cultures where people can deal better with uncertainty are better at converting basic research into innovative start-ups and risk-taking. subsequently, the more there is tolerance for ambiguity and the more there is research and innovation, the more there will be demand for vc. the underlying theory is that such uncertainty tolerance facilitates the process of converting patents into vibrant start-ups and start-up opportunities provide fuel to the vc market. thus, it is hypothesized as follows: h-10: the impact of innovation on vc investment is likely to vary significantly with uncertainty avoidance. 2.7 the conceptual model formal and informal institutions, ict and innovation are likely to have a significant direct impact on vc investment. in addition to direct effect, formal institutions also moderate the association between ict and vc investment. the key reason is the institutional theory that strong protection of property rights of private actors in the cyberspace get more ict compared to weak legal systems. as discussed earlier in section 2.5, if ict vary with institutions, then there is an expectation that institutions impact the relationship between ict and vc investment. on the other hand, informal institutions and ict moderate the impact of innovation on vc investment. muhammad zubair khan, zafir ullah khan & affan hameed 50 the rationale is that vc is an information intensive market. vc needs information technology when it invests in risky businesses converting patents into innovative products and certainly such processes would be more facilitated in a culture that is tolerant of uncertainty. the conceptual model depicted in figure 2 shows the relations between formal institutions, information institutions, ict, innovative potential and vc. the arrows indicate the direction of the influence of independent variable on dependent variable (direct impact represented by black lines) or on relationship between another independent variable and dependent variable (indirect impact represented by blue lines). figure 2: conceptual model (source: author) 2.8 regional disparities we are also interested to know if there are any regional differences. we expect differences as the asian context is different in several respects. most prominent of them are high income inequalities and poverty and lower per capita income in asian context (aoyagi and ganelli, 2015) compared to europe. while gdp growth has been controlled for, other factors may have led the investors of the two regions to respond to different market and institutional stimuli. the overall cultural and institutional context not captured in this study might also be in play that makes asian vc activity distinct from the european one. it will, thus, be useful to conduct regional analysis to take in to account those contextual differences for the sake of interest. 2.9 control variables vc researchers have investigated the favorability of an environment for investment across countries using different factors. they find liquidity of stock markets and ipos (black and gilson, 1999; jeng and wells, 2000) exert a strong effect on vc markets as they facilitate exits from such investments though the impact varies across different stages of investments formal institutions ict informal institutions innovation venture capital investments h6 h8 h1 h9 h7 h2-5 h10 journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 51 (schertler, 2003a; schertler, 2003b; schröder, 2009). tax rates also matter. some show that taxation negatively determines vc (cherif and gazdar, 2011) while others aver that taxation is of little or no importance as an explanatory variable in determining vc (romain and pottelsberghe, 2004). research using us and the asian data show a lesser impact of taxation while studies taking european countries observe taxation as the significant determinant. this indicates a regional disparity. labor market rigidities also restrict vc investment, particularly the european markets. employment protection legislation (cherif and gazdar, 2011; bonini and alkan, 2012). yet some show positive impact of labor market rigidities (schertler, 2003a; bozkaya and kerr, 2014). fundamental changes in economic situations such as high-tech bubble and 2008 financial crunch have dramatic effects on vc financing particularly risk preferences and investment strategies (ning, wang and yu, 2015). previous research has also shown that gdp growth rate contributes positively to the promotion of vc (cherif and gazdar, 2011). others show that gdp is not linked to vc development (jeng and wells, 2000; kumar and orleck, 2002). the article also takes in to account the currency exchange rate. vc investments include both national and cross border investments. international investors calculate their investments and returns in us dollars. they convert us dollars into local currencies while investing in portfolio companies and back into us dollars to receive proceeds upon exits (minardi et al., 2017). although, minardi et al. (2017) find no evidence that currency exchange rates have any significant influence on vc returns in the long term in brazil, it is still not clear how much is the effect of the exchange rate over the years on country’s vc investment. we expect a negative effect of the exchange rate on vc investment. considering this, the present article uses the recently introduced novel financial development index5, gdp growth, taxation, employment, and exchange rate as control variables. financial development index, gdp growth and employment are expected to exert a positive impact on vc activity whereas taxation and exchange rate are expected to have negative impact on vc activity. moreover, following li & zahra (2012), this study uses bubbles and trend variables as control variables. bubbles variable captures the cyclical fluctuations caused by dot cot bubble during 1999-2000 and market crash triggered by property prices in 2007-2008. bubbles is a dummy variable equal to one for the years 1999, 2000, 2007 and 2008. trend has been introduced to capture the impact of all other variables that vary over time but have not been considered in the regressions. 5 which consists of financial markets index and financial institutions index. further, the financial markets index consists of access, depth, and efficiency of financial markets whereas financial institutions index consists of access, depth and efficiency of financial institutions. muhammad zubair khan, zafir ullah khan & affan hameed 52 3 methodology 3.1 model specification and robustness tests the wooldridge test for autocorrelation in panel data shows the first order autocorrelation is insignificant. the breusch-pagan/cook-weisberg test for heteroskedasticity rejects the null that there is constant variance in the data or in other words indicate the presence of heteroskedasticity. to tackle endogeneity, both durbin-wu-hausman and wu-hausman f tests were conducted using ivendog routine in stata stata 15.0 which shows the presence of endogeneity in gdp growth. this is consistent with the findings of ning, wang and yu (2015) showing that gdp growth is endogenous regressor of vc investment. fixed effects estimator is not an option because the cultural regressors are time-invariant and that it does not tackle endogeneity. gmm is the most suitable estimator in the presence of endogeneity and heteroskedasticity (roodman, 2009). however, this method has limitations particularly in panels with larger t and smaller n. our sample consists of 32 countries and 28 years of data enough for instruments to outnumber groups even after using collapse option – that reduces number of instruments — in stata 15.0. in longer time periods and small group size, the number of instruments explode in system gmm estimations making it inconsistent (roodman, 2009) which is the key reason to avoid system gmm. breusch and pagan lagrangian multiplier test was conducted for random effects and the result suggests the use of random effects against pooled ols. following li & zahra (2012), the article uses generalized two stage least square (g2sls) iv estimator to tackle endogeneity. 3.2 econometric model consider the following random effects model: 1 2 3 4 6 5 6 p it it it it it i it i it vci pr ict patents gdpgr culture z e        = + + + + + + + + the subscripts " "i signifies countries and " "t represents time in years. itvci denotes venture capital investment as percent of gdp which is the dependent variable that varies across countries and time. 1 is the intercept whereas 2 to 6 measure the slopes of explanatory variables. the exogenous predictors that change across countries and over time include protection of property rights denoted by p itpr , ict symbolized by itict and patents as percent of population represented by itpatents . gdp growth, symbolized as itgdpgr , is the endogenous variable instrumented with gross capital formation (gcf) as percent of gdp. gcf, also called gross domestic investment or investment ratio consists of “outlays on additions to the fixed assets of the economy plus net journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 53 changes in the level of inventories”6. in the ‘neoclassical growth models in a closed economy’, it equals savings ratio (barro, 1996). in line with the basic solow model, the ratio of investment to output bears positive impact on economic growth by virtue of equilibrium level of output per effective worker (long and summers, 1991; mankiw, romer and weil, 1992; barro, 1996). adams (2009) show that gross domestic investment exerts a positive impact on gdp growth. this offers theoretical base to the gcf % gdp as an instrument for gdp growth. the cultural variables are represented by iculture that vary only across countries but not over time. control variables, denoted by it z , include time trend, a dummy variable to capture market crashes of 1999-2000 and 2007-2008 labeled as “bubbles”, tax burden, employment, exchange rate and financial development index. cross-sectional errors are denoted by i and observation-level error is represented by ite . the data used in this study have been gathered from various sources and covers period from 1990 to 2017 (appendix 2). data for vc investment % gdp has been extracted from avcj, eurostat and evca yearbooks. the data of ict variables, gdp growth, patents, and gross capital formation (gcf) % gdp have been sourced from word bank database. data of financial development index has been obtained from imf while data on exchange rate and employment have been taken from penn world tables version 9.1 (feenstra, robert and timmer, 2015). tax burden, property rights, and business freedom data have been borrowed from heritage foundation whereas data for the variable legal system and property rights has been obtained from frazer institute which was available only for 1990, 1995 and 2000-2017. data of cultural variables come from hofstede, hofstede and minkov (2010). the data pertains to 19 european countries and 13 asia-pacific countries. the european countries include austria, belgium, czech republic, denmark, finland, france, germany, hungary, ireland, italy, netherlands, norway, poland, portugal, romania, spain, sweden, switzerland, united kingdom. the asia-pacific countries are australia, china, hong kong, india, indonesia, japan, malaysia, new zealand, philippines, singapore, south korea, thailand, and vietnam. 4 results 4.1 summary statistics summary statistics have been presented in table 1. the sample consists of 896 observations. however, due to missing observations, some of the variables such as tax burden and property rights display lesser observations that will lead to a reduction in the total number of observations and these observations vary across different models. all the time invariant 6 definition taken from world development indicators, world bank muhammad zubair khan, zafir ullah khan & affan hameed 54 variables have been log transformed except financial development index and patents % population which are already narrow in range. internet use, financial development and trend show the highest correlation with vc investment which is the dependent variable. some of the variable display high correlation. for example, power distance has correlation magnitude of 0.68 with property rights and individualism. there is no concern of multicollinearity among the variables as variance inflation factor is near 3 or less in all the models used in this study. patents and uncertainty avoidance are also highly correlated with vc investment. uncertainty avoidance and masculinity are negatively correlated whereas individualism is positively correlated with vc investment. the instrumental variable i.e. gross capital formation is highly correlated with endogenous variable gdp growth. table 1: summary statistics variable observations mean standard deviation minimum maximum vc investment % gdp 805 -2.14 1.67 -12.72 1.85 bubbles 896 0.14 0.35 0.00 1.00 trend 896 14.50 8.08 1.00 28.00 tax burden 729 4.12 0.27 3.40 4.54 employment 896 3.84 0.13 3.52 4.24 exchange rate 896 1.89 2.69 -6.10 10.02 financial development index 896 0.58 0.20 0.00 1.00 property rights 729 4.19 0.47 2.30 4.58 internet use 857 2.45 2.54 -9.86 4.58 patents % population 869 0.56 0.80 0.00 4.19 uncertainty avoidance 896 58.34 24.90 8.00 104.00 individualism 896 52.88 23.98 14.00 90.00 masculinity 896 50.72 21.49 5.00 95.00 power distance 896 54.41 22.95 11.00 104.00 gdp growth 790 1.12 0.83 -2.78 2.68 gross capital formation % gdp 890 3.21 0.21 2.53 3.84 journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 55 table 2: matrix of correlations variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (1) vc investment % gdp 1.000 (2) bubbles 0.121 1.000 (3) trend 0.386 0.227 1.000 (4) tax burden 0.033 0.048 0.255 1.000 (5) employment 0.198 0.006 0.173 0.062 1.000 (6) exchange rate 0.139 0.006 0.027 0.367 0.069 1.000 (7) financial development 0.398 0.038 0.155 0.219 0.436 0.406 1.000 (8) property rights 0.153 0.007 0.087 0.377 0.192 0.445 0.621 1.000 (9) internet use 0.481 0.051 0.634 0.070 0.289 0.266 0.522 0.373 1.000 (10) patents % population 0.224 0.009 0.031 0.193 0.365 0.247 0.374 0.300 0.254 1.000 (11) uncertainty avoidance 0.279 0.024 0.008 0.142 0.316 0.013 0.078 0.002 0.066 0.007 1.000 (12) individualism 0.100 0.006 0.022 0.589 0.085 0.501 0.337 0.568 0.296 0.163 0.066 1.000 (13) masculinity 0.180 0.005 0.023 0.376 0.150 0.055 0.055 0.123 0.110 0.119 0.179 0.012 1.000 (14) power distance 0.027 0.012 0.039 0.537 0.339 0.345 0.486 0.675 0.351 0.107 0.058 0.698 0.096 1.000 (15) gdp growth 0.007 0.092 0.158 0.304 0.086 0.227 0.331 0.373 0.333 0.061 0.268 0.366 0.034 0.352 1.000 (16) gross capital formation % gdp 0.077 0.066 0.049 0.276 0.254 0.356 0.182 0.367 0.247 0.141 0.100 0.443 0.073 0.289 0.400 1.000 muhammad zubair khan, zafir ullah khan & affan hameed 56 4.2 empirical results the main results have been presented in table 3 with reduced form equations in column 1 and 3, while the second stage regressions in column 2 and 4. gdp growth has been instrumented with the gross capital formation (gcf) % gdp. in addition to their theoretical relevance presented in section 3.2, the effect of the instrument is quantitatively very large and statistically very significant in the reduced form equation with large f statistic as shown in column 1 and column 3 of table 3. moreover, the weak instrument test was performed for the instrumental variable using condivreg routine in stata 15.0. the overall f statistic is 40 and the confidence intervals of the three size corrected tests (i.e. conditional lr, anderson-rubin, and lm score) are wider compared to the asymptotic 95% confidence intervals of the gdp growth. this indicates that gcf variable does not suffer from weak instrument effect. but the p values of the three tests are not significant at 10% which is a bit of a concern. to address the concern, the coefficients and standard errors of gdp growth using iv 2sls regressions were obtained. these coefficients and standard errors were not significantly different from the magnitudes in the condivreg test results confirming that the instrument is not weak. column 2 shows that the trend variable is highly significant, indicating that the vc industry has grown vigorously during the period under investigation. property rights display an insignificant positive impact on vc investment rejecting hypothesis 1. among the informal institutions, uncertainty avoidance has a strong negative while power distance has a strong positive impact on vc investment confirming hypotheses 2 and 3. a country 10 percentage points higher on uncertainty avoidance experiences a decrease of 0.24 percentage points in vc investment. an increase of 10 percentage points in power distance worth about 0.28 percentage points increase in the vc investment. individualism is significant only at 10% whereas masculinity is not significant. thus hypothesis 4 is partially supported whereas hypothesis 5 is not supported. internet use and patents show strong and positive impact on vc investment supporting the hypothesis 6 and hypothesis 7. with 10 percentage point increase in individuals using internet in a country, the vc investment increase by 2.7 percentage points while 10% increase in patents worth about 3.2 percentage points increase in vc investment. the second stage regression in column 4 has excluded the trend variable as it is highly correlated with internet use (0.63 in table 2 and 0.75 in pairwise correlation). though the internet use is significant in column 2, its coefficient improves after removing trend variable from equation in column 4. all the models with internet use will be presented without the trend variable to avoid potential collinearity. 4.3 interaction analysis there is a strong interaction between formal institutions and ict in their impact on vc investment. to test the hypothesis 8, column 1 in table 4 introduces the interaction term for property rights (heritage foundation) and internet use. the interaction worth about 2 percentage points share in the vc investment if the former changes by 10 percentage points. the positive sign indicates that internet use positively interacts with property rights supporting journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 57 hypothesis 8. the corresponding interaction plot in figure 3(a) signifies that internet use significantly improves vc activity in environments of strong property rights. table 3: impact of institutions, ict, innovation, and culture on country-level amount of venture capital investments % gdp g2sls first stage second stage first stage second stage (1) (2) (3) (4) constant 2.326 0.945 2.647* -1.971 (1.432) (5.433) (1.391) (5.454) bubbles 0.111 0.702*** 0.151** 0.588*** (0.0704) (0.164) (0.0657) (0.172) trend -0.0134** 0.0471*** (0.00610) (0.0146) gdp growth -0.369 -0.522 (0.632) (0.641) tax burden 0.489*** -0.651 0.426*** -0.187 (0.151) (0.671) (0.146) (0.658) employment -1.221*** -0.590 -1.296*** -0.0787 (0.310) (1.279) (0.302) (1.242) exchange rate -0.0320** 0.0186 -0.0371*** 0.0309 (0.0136) (0.0427) (0.0133) (0.0429) financial development index -0.571*** 1.591 -0.607*** 1.845 (0.210) (1.155) (0.206) (1.183) property rights -0.206** -0.225 -0.175* -0.382 (0.0887) (0.331) (0.0895) (0.332) internet use -0.0131 0.269*** -0.0510*** 0.384*** (0.0195) (0.0771) (0.0146) (0.0793) patents % population 0.0276 0.432*** 0.0495 0.321* (0.0415) (0.149) (0.0400) (0.180) uncertainty avoidance -0.00819*** -0.0240*** -0.00827*** -0.0241*** (0.00127) (0.00785) (0.00126) (0.00704) individualism -0.00326* 0.0132* -0.00353* 0.0149* (0.00189) (0.00775) (0.00188) (0.00783) masculinity -0.00173 -0.00852 -0.00184 -0.00879 (0.00145) (0.00598) (0.00145) (0.00574) power distance -0.00104 0.0275*** -0.00157 0.0302*** (0.00222) (0.00659) (0.00218) (0.00639) gross capital formation % gdp 1.173*** 1.189*** (0.119) (0.120) #observations 641 622 641 622 #countries 32 32 r-squared 0.388 0.383 wald 207.8 163.9 note: robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1. gdp growth has been instrumented with gcf % gdp. column 1 is the reduced form equation for g2sls iv regression in column 2 and similarly column 3 presents first stage regression for g2sls iv estimation in column 4. muhammad zubair khan, zafir ullah khan & affan hameed 58 table 4: impact of formal institutions, ict, innovation, and culture on country-level amount of venture capital investments % gdp: interaction analysis g2sls (1) (2) (3) (4) (5) (6) constant -1.004 -7.094** -2.472 6.730 -4.217 -0.0180 (5.122) (3.137) (3.241) (5.139) (3.992) (4.863) bubbles 0.560*** 0.380*** 0.435*** 0.541*** 0.514*** 0.708*** (0.155) (0.130) (0.154) (0.144) (0.153) (0.142) trend 0.0834*** (0.0139) gdp growth -0.208 0.472 0.0387 -0.0985 -0.0210 -0.146 (0.598) (0.598) (0.552) (0.523) (0.572) (0.590) tax burden -0.252 -0.196 -0.338 -0.236 -0.424 -0.341 (0.644) (0.686) (0.599) (0.646) (0.598) (0.528) employment -0.199 0.721 0.103 -0.168 0.289 -0.791 (1.183) (0.997) (0.985) (1.014) (0.947) (1.101) exchange rate 0.0313 0.0524 0.0774 0.0159 0.0317 0.0603 (0.0460) (0.0458) (0.0561) (0.0419) (0.0408) (0.0389) financial development index 1.466 1.290 0.584 1.551 1.419 2.931*** (1.056) (1.041) (0.955) (0.948) (1.033) (0.930) patents % population 0.301* 0.263** 0.310** 0.307** -1.540*** 1.398*** (0.165) (0.129) (0.131) (0.147) (0.372) (0.233) uncertainty avoidance 0.0213*** -0.0129** 0.0177*** 0.0197*** 0.0194*** -0.00885 (0.00663) (0.00604) (0.00619) (0.00674) (0.00622) (0.00581) individualism 0.0148** 0.0192** 0.0183** 0.0151** 0.0173** 0.0202*** (0.00751) (0.00805) (0.00806) (0.00667) (0.00723) (0.00576) masculinity -0.00751 -0.00664 -0.00657 -0.00785 -0.00622 -0.0119** (0.00539) (0.00662) (0.00616) (0.00543) (0.00458) (0.00468) power distance 0.0313*** 0.0315*** 0.0343*** 0.0291*** 0.0282*** 0.0206*** (0.00654) (0.00743) (0.00623) (0.00655) (0.00569) (0.00497) internet use -0.312* -0.630** -2.201*** 0.341*** (0.167) (0.292) (0.675) (0.0722) property rights -0.651* -0.0672 -0.519 (0.340) (0.287) (0.324) property rights×internet use 0.200*** (0.0569) legal system property rights -0.997 -1.896** (0.723) (0.803) legal system & property rights×internet use 0.619*** (0.177) mobile subscriptions -1.025*** (0.308) legal system & property rights×mobile subscriptions 0.907*** (0.203) business freedom -2.545*** (0.853) business freedom ×internet use 0.643*** (0.174) patents % population×internet use 0.456*** (0.0913) patents % population× uncertainty avoidance 0.0137*** (0.00333) #observations 622 521 524 622 622 625 #countries 32 32 32 32 32 32 wald 296.6 358.3 563.8 224.3 405.1 635.8 note: robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1. gdp growth has been instrumented with gcf % gdp. the table show only second stage regressions for g2sls iv estimation. journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 59 figure 3: interaction plots (source: author) -1 5 -1 0 -5 0 5 l in e a r p re d ic ti o n 2.3 3.3 4.3 property rights internet use=-10 internet use=-5 internet use=0 internet use=5 predictive margins with 95% cis -1 5 -1 0 -5 0 5 l in e a r p re d ic ti o n .8 1.5 2.2 legal system & property rights internet use=-10 internet use=-5 internet use=0 internet use=5 predictive margins with 95% cis -1 5 -1 0 -5 0 l in e a r p re d ic ti o n .8 1.5 2.2 legal system & property rights mobile subscriptions=-6.5 mobile subscriptions=-.5 mobile subscriptions=5.5 predictive margins with 95% cis -1 5 -1 0 -5 0 5 l in e a r p re d ic ti o n 3.6 4.6 business freedom internet use=-10 internet use=-5 internet use=0 internet use=5 predictive margins with 95% cis -4 0 -3 0 -2 0 -1 0 0 l in e a r p re d ic ti o n 0 2 4 patents % population internet use=-10 internet use=-5 internet use=0 internet use=5 predictive margins with 95% cis -4 -2 0 2 4 l in e a r p re d ic ti o n 0 2 4 patents % population uncertainty avoidance=10 uncertainty avoidance=40 uncertainty avoidance=70 uncertainty avoidance=100 predictive margins with 95% cis a e f c d b muhammad zubair khan, zafir ullah khan & affan hameed 60 for robustness, column 2 shows the interaction of internet use with legal system and property rights (frazer institute) followed by interaction of interaction of internet use and mobile subscriptions in column 3. an interaction term for internet use and business freedom (heritage foundation) has been shown in the column 4. all of them demonstrate similar pattern confirming that the results of interaction between formal institutions and ict are robust to different measures of ict and different measures and sources of formal institutions. the significant negative sign of ict variables in the first four columns show that they play a negative role in absence of strong institutions. the interaction terms are significantly positive which indicates that institutional quality favorably impacts the effect of ict on vc investment. to improve vc activity, it is ideal to have favorable ict environment and strong formal institutions at the same time. the graphs further help to interpret these interactions (see also figure 3a to 3d). the orange line in figure 3 (a) depict high digital environment whereas the blue line show that ict is less embedded in society. the wide gaps between the lines between the highest and lowest internet use on the right-hand side of the graph represent the ict differential in high property rights environments. this shows that ict exerts highest positive impact on vc when property rights are strong. rest of the graphs representing formal institutions (i.e. from figure 3b to 3d) follow the same pattern and have similar interpretations. overall, we conclude that strong formal institutions makes it sure that there is transparency in the electronic transactions protecting consumers from online market frauds particularly in e-commerce (shareef, kumar and kumar, 2008) that promotes demands side of vc. the impact of interaction term for patents and internet use has been reported in column 5. as expressed in hypothesis 9, the effect is positive which indicates that ict favorably impacts the relationship between innovation and vc activity. the negative coefficient of patents suggests that they play negative role in the absence of ict. the interaction plot in figure 3 (e) shows vc activity is at highest point when patents and internet use are high, thereby supporting hypothesis 9. finally, a significant positive interaction term for uncertainty avoidance and patents indicates that vc thrives in uncertainty tolerant environments number of patents % population are high. the blue line in figure 3 (f) representing a high level of uncertainty tolerance moves up as patents on horizontal line moves from 0 to 2 and 2 to 4. this indicates that patents generate more vc activity in low uncertainty avoidance (or high tolerance) countries. the result supports hypothesis 10. 4.4 comparing asia-pacific with europe to compare the two regions, the variables suspected to be different in the two regions have been examined through their interaction with the dummy variable representing region that equals 1 for europe and 2 for asia-pacific. separate regressions for europe and asia-pacific region would suffer from small sample bias (and thus avoided) particularly for the timeinvariant cultural regressors that vary only across space but not time. table 5 presents the interaction terms for region and selected variables. in column 1, the positive coefficient of journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 61 interaction of regional dummy with trend indicate that asia-pacific region has experienced strong vc investment activity in the given period compared to europe which is also indicated by figure 1. table 5: interaction of region with selected variables: dependent variable is vc investment % gdp. g2sls (1) (2) (4) (5) constant 2.488 0.997 -1.071 1.333 (6.057) (5.920) (4.706) (5.377) bubbles 0.753*** 0.747*** 0.687*** 0.693*** (0.157) (0.162) (0.157) (0.159) trend -0.0484 0.0613*** 0.0469*** 0.0481*** (0.0376) (0.0154) (0.0151) (0.0146) gdp growth -0.433 -0.447 -0.392 -0.368 (0.539) (0.572) (0.599) (0.626) tax burden -0.285 -0.468 -0.566 -0.801 (0.699) (0.705) (0.546) (0.668) employment -1.049 -0.466 -0.908 -1.005 (1.307) (1.252) (1.147) (1.293) exchange rate 0.0177 0.0428 0.0687 0.0453 (0.0467) (0.0465) (0.0467) (0.0405) financial development index 1.673 2.029* 2.096* 1.999* (1.164) (1.228) (1.075) (1.032) property rights -0.0639 -0.142 -0.256 -0.157 (0.374) (0.364) (0.309) (0.304) internet use 0.198** -0.328 0.248*** 0.264*** (0.0829) (0.280) (0.0705) (0.0729) patents % population 0.292* 0.307* 0.453*** 0.340** (0.154) (0.172) (0.125) (0.162) uncertainty avoidance -0.0224*** -0.0244*** 0.0262 -0.0343*** (0.00746) (0.00772) (0.0189) (0.0106) individualism 0.0129* 0.0144* 0.0161*** 0.00456 (0.00747) (0.00760) (0.00608) (0.0100) masculinity -0.0105* -0.0104* -0.0119** -0.00465 (0.00573) (0.00611) (0.00496) (0.00591) power distance 0.0232*** 0.0272*** 0.0174** 0.0762*** (0.00817) (0.00746) (0.00783) (0.0224) region -1.083** -0.976* 2.116*** 1.680** (0.421) (0.530) (0.728) (0.716) region× trend 0.0737*** (0.0253) region× internet use 0.308** (0.148) region× uncertainty avoidance -0.0342*** (0.0104) region× power distance -0.0348** (0.0154) #observations 622 622 622 622 #countries 32 32 32 32 wald 306.4 251.9 312.2 364.2 note: robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1. gdp growth has been instrumented with gcf % gdp. the table show only second stage regressions for g2sls iv estimation. muhammad zubair khan, zafir ullah khan & affan hameed 62 figure 4: interaction of region with selected variables: dependent variable is vc investment % gdp (note: the figures are based on regression results in table 5.) the interaction of the regional dummy and internet use in column 2 is also significant at 5% indicating that internet use is more pronounced in region 2 compared to region 1. column 3 introduces the interaction term for regional dummy with uncertainty avoidance which is statistically significant at 1%. the negative sign indicates that as we jump from region 1 to region 2, the uncertainty avoidance exerts more negative impact on vc investment. the interaction term for regional dummy and power distance also demonstrate negative impact. as we move from region 2 to region 1, the impact of power distance on vc increase by .05 percentage points whereas the impact of internet use decrease by .31 percentage points. to better interpret, these interactions have been visualized in figure 4a to 4d. region 1 represents european countries whereas region 2 on the right-side show asia-pacific region. in figure 4 (a), as we move from region 1 to region 2, the lines get wider upand down-wards. blue line represents the early periods i.e. in and around 1990, whereas orange line depicts the -4 -3 -2 -1 0 1 l in e a r p re d ic ti o n 1 2 region uncertainty avoidance=10 uncertainty avoidance=30 uncertainty avoidance=50 uncertainty avoidance=70 uncertainty avoidance=90 predictive margins with 95% cis -1 0 -8 -6 -4 -2 0 l in e a r p re d ic ti o n 1 2 region internet use=-9 internet use=-4.5 internet use=0 internet use=4.5 predictive margins with 95% cis -4 -3 -2 -1 0 1 l in e a r p re d ic ti o n 1 2 region power distance=10 power distance=25 power distance=40 power distance=55 power distance=70 power distance=85 power distance=100 predictive margins with 95% cis -4 -3 -2 -1 l in e a r p re d ic ti o n 1 2 region trend=1 trend=8 trend=15 trend=22 predictive margins with 95% cis a d c b journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 63 most recent period i.e. in and around 2017. compared to europe, asia-pacific experienced significantly lower vc investment in earlier years but significantly higher in most recent years. with some variation, the same pattens have been demonstrated by internet use that it matters more in asian than in europe to predict vc investment. in europe, internet use differential makes not much difference in vc activity as it does in asia-pacific. like others, uncertainty avoidance has been depicted by different colors with blue line representing the lower uncertainty avoidance (i.e. high uncertainty tolerance) whereas grey color line represents high uncertainty avoidance. the figure shows that asia-pacific countries with lowest uncertainty avoidance (right end of the blue line) have experienced the highest vc investment and asiapacific countries with lowest uncertainty avoidance have invested the lowest in vc. vc investment varies slightly with change in uncertainty avoidance across countries in the european region. on the other hand, the uncertainty avoidance differential in asia pacific region predicts significantly larger variation in vc investment. power distance, on the other hand, has strong and positive impact on vc investment in europe than in the asian-pacific region and that european countries high on power distance invest more vc than all other countries in both the regions. 5 discussion and conclusion in summary, the current paper has examined the impact of formal and informal institutions, ict and innovation on vc investment. informal institutions are proxied by uncertainty avoidance, power distance, individualism, and masculinity. the results show that patents, ict, power distance and individualism exert significant effect on vc investment whereas the uncertainty avoidance demonstrate strong and negative influence on vc. the impact of formal institutions and masculinity are insignificant unlike hypothesized. the impact of ict on vc investment varies with formal institutions, while the influence of innovation on vc investment varies with ict and informal institutions. these relationships were presented in hypotheses and depicted in the conceptual model in figure 2. the study concludes that institutional quality is more important when coupled with digitization of a country. moreover, patents can be further exploited to generate knowledge-based economy and innovative entrepreneurship attracting vc investment if a country is high on uncertainty tolerance and experiences more digital activity. while the impact of patents on vc is already established in the literature (schertler, 2007), the stronger impact of ict on vc confirms the recent findings by khan et al. (2020) who found ict to have a strong effect on early stage and later stage vc7. financial systems – and for that matter vc – are information systems (ocampo, 2018) and vc is informationproblematic industry (fiet, 1995; lockett et al., 2002), vc processes involve more due 7 they have used narrower definition of vc. muhammad zubair khan, zafir ullah khan & affan hameed 64 diligence and monitoring compared to other financing alternatives (carey et al., 1993). certainly, ict has facilitated vc processes of information collection for deal selection, deal origination and structuring, monitoring, and valuations. the other aspect of ict is that it enhances deal flow by creating entrepreneurial opportunities (melissa, hamidati and saraswati, 2013). the impact of informal cultural institutions on vc is also pronounced. uncertainty avoidance exerts a negative impact on vc because such cultures create circumstances where entrepreneurs and vcs repel each other to reach a risky deal. similarly, masculinity also exhibits a negative influence on vc because masculine environments are hostile and less cooperative whereas vc development needs cooperative behavior at least between the vcs and the entrepreneur. however, the impact is insignificant. consistent with our hypothesis, power distance has significant and positive impact on vc investment because in high power distance cultures entrepreneurs easily handover control to vcs compared entrepreneurs in low power distance cultures. also, individualism demonstrates significant and positive influence on vc. the reason is that the strong role of family as an institution may refrain entrepreneurs from handing over control to vc (spencer and gomez, 2004). moreover, we argue that collectivist cultures promote informal businesses because firms rely more on informal institutions than formal markets (world development report, 2002) and informal businesses do not attract formal vc. formal institutions have an insignificant positive influence on vc investment against the hypothesized significant positive impact. however, its moderating influence on the association between ict and vc is highly significant as hypothesized. while high quality institutions protect the property rights of business actors, they cannot create entrepreneurial opportunities and deal flow for venture capital. ict, on the other hand, boosts entrepreneurial activities and performance (asongu and nwachukwu, 2018; zhang and li, 2018) strengthening the demand side of vc and that ict grows when institutional quality is high. this gives us striking evidence that institutional quality alone is not important for vc development unless it is coupled with digitization reflecting the technology-dependent nature of vc. the explanatory power of interaction term for innovation and informal institutions is also very important. the association between patents and vc investment is adversely affected by uncertainty avoidance. this means that patents cannot be converted into innovative driven entrepreneurship in an environment where security is preferred over risk-taking (spencer and gómez, 2004). one possible explanation is that lower uncertainty avoidance enhances the pace of innovation diffusion (van den bulte and stremersch, 2004). even if the rate of patents is high in uncertainty avoidance cultures, such patents do not attract vc because uncertaintyavoidant entrepreneurs would not venture into risky projects to convert patents into innovative start-ups or products. ultimately, this reduces the deal flow for venture capital. another interpretation is that vcs as entrepreneurs would also avoid investing vc in risky projects in environments where there is tendency to avoid uncertainty by culture. journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 65 moreover, as hypothesized, the effect of innovation is conditional upon ict use. one possible interpretation is that unless there is widespread use of digital technology in a country, entrepreneurs are not able to fully exploit the innovative potential that would, otherwise, create entrepreneurial activity and deal flow attracting vc. this is in line with existing entrepreneurship theory supporting the view ict has a positive impact on entrepreneurship. higón (2012) find that ict acts as efficiency-enhancing technology that can create a competitive advantage for smes through product innovation. in other words, at an aggregate level, ict increases the chances that smes would attract vc because ict enhances innovation (ollo-lópez and aramendía-muneta, 2012; hall, lotti and mairesse, 2013). therefore, we can conclude that ict helps entrepreneurs to exploit patents and that patents exert a positive impact on vc investment when ict use is frequent. regarding the regional differences, uncertainty avoidance, internet use, and trend are more noticeable in their impact on vc investment in the asia-pacific region than in europe. on the other hand, the impact of power distance is more prominent in europe than in the asiapacific region. the high impact of the interaction term for trend and regional dummy indicates that the asia-pacific region has experienced more vc investment during the time under study compared to europe. the same is evident in figure 1 that shows asian countries have experienced more vc activity than its european counterpart. the regional disparities indicate structural and contextual differences such as income inequality and other cultural and institutional factors not captured in this study that needs further scrutiny in future research. the novel contribution of this research is that such interaction analysis has never been conducted previously. khan et al. (2020) show that the impact of domestic credit on vc investment vary adversely with ict embeddedness in enterprises. li & zahra (2012) show that formal institutions have a positive impact on vc investment and this effect weakens in countries high on uncertainty avoidance. khan, ferrier, and khan. (2020) show that formal institutions favorably moderate the impact of financial development on venture capital fundraising. the current research is the first in investigating the moderating effect of formal institutions on the relationship between ict and vc investment. additionally, it examines for the first time the moderating role of ict and uncertainty avoidance on the association between innovation and vc investment. furthermore, it presents a conceptual model to explain vc investment in context of formal and informal institutions, ict and innovation. finally, the present research uses gross capital formation as an instrument for gdp growth in vc investment research for the first time. the instrument is theoretically relevant and passes statistical tests of weak instruments. the article suggests strong policy recommendations. without taking the contextual and institutional environment into account, policy to boost entrepreneurship and venture capital would be less effective (li and zahra, 2012). institutional quality must be coupled with high adoption and diffusion of ict as it offers substance to the vc market in the form of more muhammad zubair khan, zafir ullah khan & affan hameed 66 entrepreneurial activities and high deal flow. while patents are important to develop vc market, they attract more vc when an economy has a strong digital base and experiences lower uncertainty avoidance. there is also a scope for future research on other aspects of vc such as exits from vc investments. research has demonstrated that digital technologies have resulted in massive improvements and an enormous increase in trading volumes. for instance, in the last ten years, the us has experienced an almost twelve-fold increase in annual stock turnover between 1988 and 2008 and increase in stock market capitalization to gdp from 58% in 1988 to 163% in 1999 (stockhammer, 2013). such digitization might have a strong impact on vc exits as well. moreover, there is a need to see how the conceptual model used in this research applies to the exits from vc investments. the conceptual model repositions the role institutions and innovation play in technology dependent vc markets. references adam, i. o. (2020) ‘examining e-government development effects on corruption in africa: the mediating effects of ict development and institutional quality’, technology in society. elsevier ltd, 61, p. 101245. doi: 10.1016/j.techsoc.2020.101245. adams, s. (2009) ‘foreign direct investment, domestic investment, and economic growth in sub-saharan africa’, journal of policy modeling. north-holland, 31(6), pp. 939–949. doi: 10.1016/j.jpolmod.2009.03.003. aldrich, h. e. (2014) ‘the democratization of entrepreneurship? hackers, makerspaces, and crowdfunding entrepreneurship view project’, in academy of management annual meeting. philadelphia, pa. doi: 10.13140/2.1.1371.6162. andrés, a. r., amavilah, v. and asongu, s. (2017) ‘linkages between formal institutions, ict adoption, and inclusive human development in sub-saharan africa’, in catalyzing development through ict adoption: the developing world experience. springer international publishing, pp. 175–203. doi: 10.1007/978-3-319-56523-1_10. aoyagi, c. and ganelli, g. (2015) ‘asia’s quest for inclusive growth revisited’, journal of asian economics. elsevier, 40, pp. 29–46. doi: 10.1016/j.asieco.2015.06.005. armour, j. and cumming, d. (2006) ‘the legislative road to silicon valley’, oxford economic papers. oxford university press, 58(4), pp. 596–635. doi: 10.1093/oep/gpl007. asongu, s. a. and acha-anyi, p. n. (2017) ‘ict, conflicts in financial intermediation and financial access: evidence of synergy and threshold effects’, netnomics: economic research and electronic networking, 18(2–3), pp. 131–168. doi: 10.1007/s11066-0179119-5. asongu, s. a. and moulin, b. (2016) ‘the role of ict in reducing information asymmetry for financial access’, research in international business and finance. elsevier ltd, 38, pp. 202–213. doi: 10.1016/j.ribaf.2016.04.011. asongu, s. a. and nwachukwu, j. c. (2018) ‘openness, ict and entrepreneurship in subsaharan africa’, information technology and people. emerald group publishing ltd., journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 67 31(1), pp. 278–303. doi: 10.1108/itp-02-2017-0033. bain & company (2019) global private equity report 2019. available at: bain.com. barro, r. j. (1996) ‘determinants of economic growth: a cross-country empirical study’, national bureau of economic research. (nber working paper 5698), (w5698). bertoni, f. and colombo, m. g. (2015) ‘the patterns of venture capital investment in europe’, small business economics. springer us, 45(3), pp. 543–560. doi: 10.1007/s11187015-9662-0. black, b. s. and gilson, r. j. (1998) ‘venture capital and the structure of capital markets: banks versus stock markets’, journal of financial economics. elsevier, 47(3), pp. 243– 277. doi: 10.1016/s0304-405x(97)00045-7. black, b. s. and gilson, r. j. (1999) ‘does venture capital require an active stock market: banks versus stock markets’, journal of corporate finance, 47, pp. 243–277. bonini, s. and alkan, s. (2012) ‘the political and legal determinants of venture capital investments around the world’, small business economics, 39(4), pp. 997–1016. doi: 10.1007/s11187-011-9323-x. bozkaya, a. and kerr, w. r. (2014) ‘labor regulations and european venture capital’, journal of economics & management strategy, 23(4), pp. 776–810. doi: 10.1111/jems.12077. van den bulte, c. and stremersch, s. (2004) ‘social contagion and income heterogeneity in new product diffusion: a meta-analytic test’, marketing science. informs , 23(4), pp. 530–544. doi: 10.1287/mksc.1040.0054. carey, m., prowse, s., rea, j. & udell, g. f. (1993) ‘the economics of private placements: a new look’, financial markets, institutions and instruments, 2, pp. 1–67. cherif, m. and gazdar, k. (2011) ‘what drives venture capital investments in europe? new results from a panel data analysis’, journal of applied business and economics, 12(3), pp. 122–139. available at: http://www.nabusinesspress.com/jabe/cherif_abstract.html. cumming, d., henriques, i. and sadorsky, p. (2016) ‘cleantech’ venture capital around the world’, international review of financial analysis. elsevier b.v. doi: 10.1016/j.irfa.2016.01.015. da, m., nicodano, g. and sembenelli, a. (2006) ‘public policy and the creation of active venture capital markets’, 90, pp. 1699–1723. doi: 10.1016/j.jpubeco.2005.09.013. efrat, k. (2014) ‘the direct and indirect impact of culture on innovation’, technovation. elsevier, 34(1), pp. 12–20. doi: 10.1016/j.technovation.2013.08.003. feenstra, r. c., robert, i. and and timmer, m. p. (2015) ‘the next generation of the penn world table’, american economic review, 105(10), pp. 3150–3182. fiet, j. o. (1995) ‘reliance upon informants in the venture capital industry’, journal of business venturing. elsevier, 10(3), pp. 195–223. doi: 10.1016/0883-9026(94)00039w. gabor, d., & brooks, s. (2017) ‘the digital revolution in financial inclusion: international development in the fintech era’, new political economy, 22(4), pp. 423–436. muhammad zubair khan, zafir ullah khan & affan hameed 68 gompers, p. and lerner, j. (1998) what drives venture capital fundraising?, nber working paper no. 6906. issued in january 1999. groh, a. p. and liechtenstein, h. (2011) ‘international allocation determinants of institutional investments in venture capital and private equity limited partnerships’, international journal of banking, accounting and finance, 3(2/3), pp. 176–206. doi: 10.2139/ssrn.977471. groh, a. p. and von liechtenstein, h. (2009) ‘how attractive is central eastern europe for risk capital investors?’, journal of international money and finance. elsevier ltd, 28(4), pp. 625–647. doi: 10.1016/j.jimonfin.2009.01.006. groh, a. p. and wallmeroth, j. (2016) ‘determinants of venture capital investments in emerging markets’, emerging markets review. elsevier b.v., 29, pp. 104–132. doi: 10.1016/j.ememar.2016.08.020. guler, i. and guillén, m. f. (2010) ‘home country networks and foreign expansion: evidence from the venture capital industry’, academy of management journal. academy of management briarcliff manor, ny , 53(2), pp. 390–410. doi: 10.5465/amj.2010.49389027. hall, b. h., lotti, f. and mairesse, j. (2013) ‘evidence on the impact of r&d and ict investments on innovation and productivity in italian firms’, economics of innovation and new technology. routledge , 22(3), pp. 300–328. doi: 10.1080/10438599.2012.708134. he, m. and lee, j. (2020) ‘social culture and innovation diffusion: a theoretically founded agent-based model’, journal of evolutionary economics. springer, pp. 1–41. doi: 10.1007/s00191-020-00665-9. higón, d. a. (2012) ‘the impact of ict on innovation activities: evidence for uk smes’, international small business journal: researching entrepreneurship. sage publicationssage uk: london, england, 30(6), pp. 684–699. doi: 10.1177/0266242610374484. hofstede, g. h. (2012) national culture dimensions, http://geert-hofstede.com/. hofstede, g., hofstede, g. j. & and minkov, m. (2010) cultures and organizations: software of the mind. 3rd edn. mcgraw-hill. jeng, l. a. and wells, p. c. (2000) ‘the determinants of venture capital funding : evidence across countries’, journal of corporate finance, 6, pp. 241–289. kenney, m., han, k. and and tanaka, s. (2002) ‘adapting to strange environments: venture capital in japan, korea and taiwan’, in annual meeting of the academy of international business. san juan, puerto rico. kenney, m., han, k. and and tanaka, s. (2003) ‘he globalization of venture capital: the cases of taiwan and japan’, in anthony bartzokas and sunil mani (eds.) financial systems, corporate investment in innovation and venture capital. cheltenham: uk and northampton, ma: edward elgar, pp. 52–84. keuschnigg, c. and nielsen, s. b. (2003) ‘tax policy, venture capital, and entrepreneurship’, journal of public economics. north-holland, 87(1), pp. 175–203. doi: 10.1016/s00472727(01)00170-0. journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 69 khan, m. z. et al. (2020) ‘on the upside or flipside: where is venture capital positioned in the digitally-powered financial market disruptions?’, submitted manuscript. khan, m. z., ferrier, w. and khan, z. u. (2020) ‘institutions, financial development and private equity fundraising’, submitted manuscript. kreiser, p. m. et al. (2010) ‘cultural influences on entrepreneurial orientation: the impact of national culture on risk taking and proactiveness in smes’, entrepreneurship theory and practice. sage publicationssage ca: los angeles, ca, 34(5), pp. 959–983. doi: 10.1111/j.1540-6520.2010.00396.x. kumar, v. and orleck, s. (2002) ‘why does private equity vary across countries and time ?’, working paper, pp. 1–38. lechman, e. and marszk, a. (2019) ict-driven economic and financial development : analyses of oecd countries. 1st edn. academic press. li, y. and zahra, s. a. (2012) ‘formal institutions, culture, and venture capital activity: a cross-country analysis’, journal of business venturing. elsevier b.v., 27(1), pp. 95– 111. doi: 10.1016/j.jbusvent.2010.06.003. lockett, a. et al. (2002) ‘venture capital investors, valuation and information: a comparative study of the us, hong kong, india and singapore’, venture capital, 4(3), pp. 237–252. doi: 10.1080/13691060213715. long, j. b. de and summers, l. h. (1991) ‘equipment investment and economic growth’, the quarterly journal of economics. oxford university press (oup), 106(2), p. 445. doi: 10.2307/2937944. mankiw, n. g., romer, d. and weil, d. n. (1992) a contribution to the empirics of economic growth a contribution to the empirics of economic growth*, source: the quarterly journal of economics. martinez, c. a. and williams, c. (2010) ‘national institutions, entrepreneurship and global ict adoption: a cross-country test of competing theories’, journal of electronic commerce research, 11(1), pp. 73–91. available at: http://0search.proquest.com.innopac.up.ac.za/docview/236658764/fulltextpdf/64e96f7dde 7e47f3pq/7?accountid=14717. melissa, e., hamidati, a. and saraswati, s. (2013) ‘social media empowerment: how social media helps to boost women entrepreneurship in indonesian urban areas’, the iafor jounal of media, communications and film, 1(1). minardi, a. m. a. f. et al. (2017) ‘currency impact in brazilian pe/vc deals’ performance’, journal of private equity, 20(3), pp. 45–51. doi: 10.3905/jpe.2017.20.3.045. ning, y., wang, w. and yu, b. (2015) ‘the driving forces of venture capital investments’, small business economics, (june 2014), pp. 315–344. doi: 10.1007/s11187-014-95913. north, d. c. (1991) ‘institutions’, journal of economic perspectives, 5(november 1), pp. 97– 112. ocampo, j. a. (2018) ‘international asymmetries and the design of the international financial system 1’, in critical issues in international financial reform. routledge, pp. 45–74. muhammad zubair khan, zafir ullah khan & affan hameed 70 doi: 10.4324/9781351323765-2. ollo-lópez, a. and aramendía-muneta, m. e. (2012) ‘ict impact on competitiveness, innovation and environment’, telematics and informatics. pergamon, 29(2), pp. 204– 210. doi: 10.1016/j.tele.2011.08.002. pitchbook-nvca (2018) venture monitor, by pitchbook and national venture capital association, usa. pradhan, r. p., arvin, m., nair, m., bennett, s., & bahmani, s. (2017) ‘ict-financegrowth nexus: empirical evidence from the next-11 countries’, cuadernos de economia, 40(113), pp. 115–134. pradhan, r. p. et al. (2018) ‘mobile telephony, economic growth, financial development, foreign direct investment, and imports of ict goods: the case of the g-20 countries’, economia e politica industriale. springer international publishing, 45(2), pp. 279–310. doi: 10.1007/s40812-017-0084-7. pradhan, r. p., arvin, m. b. and norman, n. r. (2015) ‘the dynamics of information and communications technologies infrastructure, economic growth, and financial development: evidence from asian countries’, technology in society, 42, pp. 135–149. doi: 10.1016/j.techsoc.2015.04.002. ragosa, g. and warren, p. (2019) ‘unpacking the determinants of cross-border private investment in renewable energy in developing countries’, journal of cleaner production. elsevier ltd, 235, pp. 854–865. doi: 10.1016/j.jclepro.2019.06.166. report, w. development (2002) building institutions for markets. risse, t. (2000) ‘rational choice, constructivism, and the study of international institutions’, in annual meeting of the american political science association. washington dc. romain, a. and pottelsberghe, b. van (2004) ‘the determinants of venture capital: additional evidence’, discussion paper series 1: studies of the economic research centre, 1(19). roodman, d. (2009) ‘how to do xtabond2: an introduction to difference and system gmm in stata’, stata journal, 9(1), pp. 86–136. doi: 10.1177/1536867x0900900106. sassi, s. and goaied, m. (2013) ‘financial development , ict diffusion and economic growth : lessons from mena region’, telecommunications policy. elsevier, 37(4–5), pp. 252– 261. doi: 10.1016/j.telpol.2012.12.004. schertler, a. (2003) ‘a comparative overview of venture capital in europe and the united states’, international journal of entrepreneurship education, 1(4), pp. 539–584. schertler, a (2003) ‘driving forces of venture capital investments in europe: a dynamic panel data analysis’, working paper, (03). available at: http://www.econstor.eu/handle/10419/17695. schertler, a. (2007) ‘knowledge capital and venture capital investments: new evidence from european panel data’, german economic review. john wiley & sons, ltd (10.1111), 8(1), pp. 64–88. doi: 10.1111/j.1468-0475.2007.00134.x. schröder, c. (2009) ‘financial system and innovationsdeterminants of early stage venture capital in europe’, in schumpeter discussion papers, sdp 2009-04 issn 1867-5352. shane, s. (1995) ‘uncertainty avoidance and the preference for innovation championing journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 71 roles’, journal of international business studies, 26(1), pp. 47–68. doi: 10.2307/155477. shareef, m., kumar, u. a. and kumar, v. (2008) ‘role of different electronic-commerce quality factors on purchase decision: a developing country perspective’, journal of electronic commerce research, 9(2), pp. 92–113. spencer, j. w. and gómez, c. (2004) ‘the relationship among national institutional structures, economic factors, and domestic entrepreneurial activity: a multicountry study’, journal of business research. elsevier, 57(10), pp. 1098–1107. doi: 10.1016/s01482963(03)00040-7. srivastava, s. c., teo, t. s. h. & and devaraj, s. (2016) ‘you can’t bribe a computer: dealing with the societal challenge of corruption through ict’, mis quarterly, 40(2), pp. 511– 526. stockhammer, e. (2013) ‘financialization and the global economy’, in wolfson, m. h. and epstein, g. a. (eds) the handbook of the political economy of financial crises. oxfor university press, p. 770. sturges, p. (2004) ‘corruption, transparency and a role for ict?’, international journal of information ethics, 2(11). available at: http://www.lboro.ac.uk/departments/ls/staff/psturges.html (accessed: 9 august 2020). tellis, g. j., stremersch, s. and yin, e. (2003) ‘the international takeoff of new products: the role of economics, culture, and country innovativeness’, marketing science. institute for operations research and management sciences, 22(2). doi: 10.1287/mksc.22.2.188.16041. tykvova, t. (2017) ‘venture capital and private equity financing : an overview of recent literature and an agenda for future research’, journal of business economics, volume 88, pp. 325–362. doi: 10.1007/s11573-017-0874-4. wright, m., pruthi, s. and lockett, a. (2005) ‘international venture capital research : from comparisons to crossing borders’, international journal of management reviews, 7(3), pp. 135–165. wright, m. and robbie, k. (1998) ‘venture capital and private equity: a review and synthesis’, journal of business finance, 25(5–6), pp. 521–570. doi: 10.1111/14685957.00201. yeniyurt, s. and townsend, j. d. (2003) ‘does culture explain acceptance of new products in a country?: an empirical investigation’, international marketing review. mcb up ltd, 20(4), pp. 377–396. doi: 10.1108/02651330310485153. zhang, f. and li, d. (2018) ‘regional ict access and entrepreneurship: evidence from china’, information and management. elsevier b.v., 55(2), pp. 188–198. doi: 10.1016/j.im.2017.05.005. muhammad zubair khan, zafir ullah khan & affan hameed 72 appendix 1: previous studies on vc research authors region/countries period factors influencing vc jeng and wells (2000) 21 countries (us, europe, japan and australia) 1986-1995 ipo and labor market rigidity bonini and alkan (2012) 16 countries (us, europe, japan and australia) 1995-2002 favorable sociopolitical, entrepreneurial environment and legal system black and gilson (1998) us, japan and germany 1992-1995 stock-market centered capital market kumar and orleck (2002) 9 countries (usa and europe) 1986-1999 exit options, cost capital, legal traditions, patent regulation, and the transparency of markets ning, wang and yu (2015) us 1995-2011 stock market, gdp growth rate and industry production index have positive impact on vc; global financial crises also affect vc félix et al. (2013) 23 european nations 1992–2003 gdp growth, interest rate, ipo, m&a, market-to-book ratio and r&d have sig positive influence unemployment, stock market capitalization and tea have negative impact groh and wallmeroth (2016) 118 emerging and developed markets 2000-2013 m&a investment volume, disclosure index, shareholder suits index, legal rights index, bribery & corruption index, innovation index and ip protection have positive effects on vc inv groh and liechtenstein (2009) 27 european countries 2000-2005 corporate governance and the protection of investors’ rights have positive impact on the attractiveness of a country for vc investment da rin, nicodano, and sembenelli (2006) 14 european countries 1988-2001 stock market; capital gains tax and labor regulation have impact on high-tech investment while capital gains tax has also impact on early stage investment romain and pottelsberghe (2004) 16 oecd countries 1990-2000 interest rates, stock of knowledge and the number of triadic patents affect positively and significantly the relative level of vc baygan and freudenberg (2000) oecd countries 1990-2000 barriers to entrepreneurship cherif and gazdar (2011) 21 european nations 1997–2006 market capitalization, research and development expenditures, gdp growth and unemployment bozkaya and kerr (2014) us, uk, and europe 1990–2008 labor market expenditures (as the mechanism for providing worker insurance) cumming et al. (2016) 31 countries of four continents 1996-2010 increase in oil prices, media coverage, formal institutions (particularly rule of law and government effectives) have sig positive effect while uncertainty avoidance has negative effect on cleantech vc activity cumming and knill (2012) 34 countries from asia, europe, north america and south america 2000-2008 more stringent securities regulations – particularly disclosure requirements – positively affect vc supply armour and cumming (2006) us and 14 eu countries 1990-2002 temperate bankruptcy laws for entrepreneurs promote vc investment while government programs crowd out. schertler (2007) 15 western european countries 1991–2001 countries’ knowledge capital i.e. number of patents, or the number of r&d researchers, or gross domestic expenditures on r&d. schertler (2003b) 14 western european countries 1988-2000 liquidity of stock markets, human capital endowment, and labor market rigidities affect early stage vc li and zahra (2012) 61 countries 2000-2011 informal institutions of uncertainty avoidance and collectivism moderate the impact of formal institutions on vc. journal of applied economics and business studies, volume. 4, issue 2 (2020) 41-74 https://doi.org/10.34260/jaebs.423 73 appendix 2: data descriptions and sources variables description source vc investment % gdp it includes seed, start-up, later stage, expansion, growth, and buyout investments. avcj, eurostat, evca yearbooks 1990-2017 internet use individuals using internet per 100 population world bank 1990-2017 mobile subscriptions per 100 people mobile cellular telephone subscriptions are subscriptions to a public mobile telephone service that provide access to the pstn using cellular technology. world bank 1990-2017 financial development index it is an index that includes financial markets index and financial institutions index. imf 1990-2017 exchange rate exchange rate, national currency/usd penn world tables (feenstra, robert and timmer, 2015) 1990-2017 gdp growth annual percentage growth rate of gdp. aggregates are based on constant 2010 u.s. dollars. world bank 1990-2017 employment number of persons engaged divided by total population penn world tables (feenstra, robert and timmer, 2015) 1990-2017 patents % population patent applications (resident and non-resident) are worldwide patent applications filed through the patent cooperation treaty. world bank 1990-2017 tax burden tax burden is a composite measure that reflects marginal tax rates on both personal and corporate income and the overall level of taxation as a percentage of gross domestic product (gdp). on the 100-point scale, the highest score indicates the favorability of a taxation system. freedom of the world (heritage foundation) 1995-2017 property rights property rights is the sub-index of rule of law and measures the degree to which a country’s laws protect private property rights and the extent to which those laws are respected. an ideal country with 100 score means that private property is guaranteed by the government. heritage foundation 1995-2017 business freedom it measures the extent to which the regulatory and infrastructure environments constrain the efficient operation of businesses, particularly ease of starting, operating, and closing a business. the business freedom score for each country is a number between 0 and 100, with 100 indicating the freest business environment. heritage foundation 1995-2017 legal system and property rights it is the comprehensive index that covers judicial independence, impartial courts, protection of property rights, military interference in rule of law and politics, integrity of legal system, legal enforcement, regulatory costs of the sale of real property, reliability of police, and business costs of crime. it is based on scale 0 to 10, the highest representing highest quality institutions. frazer institute 1990, 1995, 2000-2017 uncertainty avoidance the index measures the degree to which people feel ambiguous situations and take measures to deal with uncertainty. score of 0 on represents high uncertainty tolerance whereas 100 score shows the society is doing best to cope with uncertainty. (hofstede, hofstede and minkov, 2010) timeinvariant individualism individualism represents the lifestyle or culture where people live without much dependence on others opposite to collectivism where people live in an integrated system. score of 100 is assigned if country is highly individualistic and 0 to highly collectivistic. (hofstede, hofstede and minkov, 2010) timeinvariant muhammad zubair khan, zafir ullah khan & affan hameed 74 power distance power distance is the degree to which targets are susceptible to the influence/power of agents. score of 0 on power distance means targets do not accept the influence of agents whereas the score of 100 reflects that people in the society as targets completely conform to the views of the agents. (hofstede, hofstede and minkov, 2010) timeinvariant masculinity a society is called masculine when men are “more assertive, tough and focused on material success” than women who are more “modest, tender and concerned with quality of life” (page.140). country-level highest gap in the gender roles is assigned score of 100 whereas gender roles greatly overlapped is assigned score of 0. (hofstede, hofstede and minkov, 2010) timeinvariant gross capital formation % gdp gross capital formation (formerly gross domestic investment) consists of outlays on additions to the fixed assets of the economy plus net changes in the level of inventories. world bank 1990-2017