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 non- agricultural 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 non- farm 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: 2469- 9062 (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 (2015- 2025). 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): 47- 53. 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 1818- 4952 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): 360- 366. 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-for- estimation-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, 4850- 4869. 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. (2- tailed) 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