. International Review of Management and Marketing ISSN: 2146-4405 available at http: www.econjournals.com International Review of Management and Marketing, 2020, 10(3), 1-6. International Review of Management and Marketing | Vol 10 • Issue 3 • 2020 1 Post Tax Reform and Corporate Effective Tax Rate: Evidence from Tunisia Kamel Naoui1, Abdelkader Kasraoui2* 1University of Manouba, Higher School of Business, 2010, La Manouba, Tunisia, 2PhD Student, University of Manouba, Higher School of Business, 2010, La Manouba, Tunisia. *Email: kasraouiabdelkader@gmail.com Received: 12 January 2020 Accepted: 25 March 2020 DOI: https://doi.org/10.32479/irmm.9414 ABSTRACT This study examines the impact of the tax reform on corporate effective tax rate (ETR) and firm-specifics in Tunisia for the post tax reform period (after the fiscal year 2014). The corporate effective tax rate is a component by major firm-specific characteristics, especially firm size, capital structure (leverage), inventory intensity, capital intensity. The ETR provides information about the tax burdens and can be used as a political instrument to boost the economic reliance. The post tax reform period reflects the impact of lower corporate tax rate on the firm characteristics. The sample consists of 112 firm-year observations from 16 listed companies in Tunis Stock Exchange (known Bourse de Tunis- BVMT) covering 7 years from 2010 to 2016. Our result indicates that the tax reform had a significant impact only on the inventory variable but no significant results on the others firm characteristics for the post-tax reform period. These findings urge the Tunisian’s tax authority to reformulate the corporate tax system. Keywords: Tax Reform, Corporate Effective Tax Rate, Tunisia JEL Classifications: G30, G32, M4 1. INTRODUCTION The researchers are recognizing that the Tunisian corporate tax system is a major barrier to economic growth by the tax burden on companies (Alm, 2015). According to Alm (2015), the object of the corporate tax reform should be to create a system that has an efficient and competitive tax rate. Two times reductions of the corporate tax rate by 10 percentage points between 2007 and 2014. The corporate tax rate reduction was carried out by two categories (from 30% to 25% for the companies listed before 2014, from 25% to 20% for the new listings companies), the increasing of the Minimum Tax (MT)1 from 0.1% to 0.2% of the total gross turnover and 0.1% of the turnover from exports. 1 A corporation has to pay a minimum tax liability referred to in paragraph II of article 44 of the Code of the Income Tax and of the Corporation Tax. For example to make a comparison with other countries like South Africa, Carreras et al. (2017) observe an average ETR of 5.3% for the manufacturing sector, 4.8% for the trade sector and 4.9% for the construction sector from the year 2010 to 2013. The Tunisian’s tax reform, related to the Finance Law for the year 2014, is an occasion to test the efficiency of the tax policy on the corporate economy and finance, by the assessing of corporate ETR on the firm characteristics. The aim this study in this area is that corporate tax reform is infrequent in Tunisia but the studies of impacts on determinants’ ETR after the reform are limited. It’s an opportunity to examine how Tunisian’s companies respond to a tax reform. In many cases we are as much interested in how specific-firm characteristics respond to the tax reform, i.e., a specific-firm characteristics multiplied by tax reform. To make generalizable statements, it is preferable to model with random effects model. This Journal is licensed under a Creative Commons Attribution 4.0 International License Naoui and Kasraoui: Post Tax Reform and Corporate Effective Tax Rate: Evidence from Tunisia International Review of Management and Marketing | Vol 10 • Issue 3 • 20202 2. LITERATURE REVIEW AND HYPOTHESIS DEVELOPMENT In this section, we examine the relevant literature on the effect of the tax reform on the specific-firm characteristics such as financing, investment operations and capital structure via the corporate effective tax rate. 2.1. Corporate Effective Tax Rate Many papers provide a theoretical and empirical framework about the impact of the corporate effective tax rate on the firm decisions (for example financing, investment, capital structure…). To measure the corporate ETR many studies used a quotient produced by the division of the company’s tax liability and the accounting income, Gupta and Newberry (1997), Plesko (2003) and Richardson et al. (2012). Further, Gupta and Newberry (1997) analyzed that tax expenses varying across firms is an indicator to suggest that the tax system is inequitable, a justification for tax reform. The ETR used by Shevlin and Porter (1992) to identify the level of neutrality of the tax system and to identify the feed-back of firms with higher and lower tax expenses. Indeed, Alm (2015) claims that “…relative to international competitors, Tunisia has a high marginal effective tax rate for businesses.” 2.2. Tax Reform For Swenson (1994) and Desai and Dharmapala (2005) the aim of the corporate tax reform is to reduce the cost of doing business and increasing the firm’s value. According to Duanjie and Jack (2011), tax reforms would help to create economic growth and ultimately more corporate income. In the same context, Alm (2015) concludes that the Tunisian corporate tax system requires a tax reform by reducing the statutory tax rates. 2.3. ETR, Tax Reform and Firm Size The existence of empirical studies unable to conclude the relationship sign between the ETR and the firm size but the other studies prove a positive variability (Zimmerman, 1983). For Siegfried (1972) states that under the political power theory, firms with large size have a low ETR because they have more resources than the smaller firms to influence the political process, for example the investment on long term assets generate more depreciations and decrease the income tax. Another direction of political theory states that giant firms have lower ETR because they have an arsenal of resources available for the managers to manage the political environment and organize their activities to achieve an optimal tax position (Richardson and Lanis, 2007). For the reason of missing corporate tax data on the Tunisian context, the current study uses a micro backward-looking approach to examine the variability of corporate ETR based on the financial statements. In other words, the numerator is the tax expense, a proxy for income tax charge paid or payable by the company and the denominator is pre-tax income, a proxy for company’s accounting income. The measure of ETR captures the average rate of tax per unit money of accounting income. Under the political cost theory, the large companies are targeted by governments in order to make the most of the tax revenue, that is the case in Tunisia (20% contribute more than 80% of the country’s tax revenue). As the tax burden is a tool of political costs, this theory claims shows that firm size is positively associated with the ETR (Watts and Zimmerman, 1986). Historically, (Zimmerman, 1983) finds a positive relationship between ETR and firm size. On the other hand Porcano (1986) observes a negative relation between the dependant and the independent variables (size, ETR). So too, Gupta and Newberry (1997) claim significant negative effect between the size of the company and the effective tax rate. Moreover, Tran (1998) observes a negative association between ETR and firm size. In another study the causes of this negative association due to that larger firms benefit more from tax-planning (tax management) than smaller firms. Richardson et al. (2012) conclude that large companies are likely to be aggressive for tax purposes which leads to a lower effective tax rate. We expect a negative versus a positive association between ETR and firm size for the two periods of the study. H1: ETR is negatively associated with firm size at the pre-period tax reform H2: ETR is positively associated with firm size at the post-period tax reform 2.4. ETR Impact on Financial and Investment Functions The first corporate function is the financing decisions with a direct impact on ETR because the capital structure between debt and equity financing generates its tax cost (Gupta and Newberry, 1997). Consider, for example, the situation in which the choice between the sources or the techniques of financing the assets of the firm gives two situations: the interests of the debt are deductible on the other hand the dividends do not have any more (Richardson et al., 2012). Research by Gupta and Newberry (1997) finds a negative association between ETR and leverage. Naoui and Kasraoui: Post Tax Reform and Corporate Effective Tax Rate: Evidence from Tunisia International Review of Management and Marketing | Vol 10 • Issue 3 • 2020 3 Firms’ investment decisions might also impact on ETR. For example, the effect of investments on the ETR lies in the depreciation charge for legally deductible assets. Another effect of investment is the inventory intensity who is a substitute for capital intensity, with the inventory intensive, firms should have higher ETR (Zimmerman, 1983). Gupta and Newberry (1997) provide that firms with a value of fixed assets have lower ETR due to tax deductions, while firms with a greater inventory intensity ratio have higher ETR. Finally, R&D expenses have a direct impact on ETR, as soon as the amounts are disbursed by the firm. Presented as a deductible expense from the tax income, R&D has a negative sign with ETR (Gupta and Newberry, 1997). The hypotheses to be detected are as follows: H3: ETR is negatively associated with firm leverage. H4: ETR is negatively associated with firm capital ratio. H5: ETR is positively associated with firm inventory ratio. H6: ETR is negatively associated with firm R&D ratio. 2.5. ETR, Tax Reform and Corporate Functions To investigate whether the tax reform had an effect on ETR, and if it impacted on the associations between ETR and variables reflecting the outcomes of firms’ financing and investment decisions, Richardson and Lanis (2007) estimate this impact on the Australian corporate tax system. For the financing mechanism firms rely more heavily on debt in the structure of their capital, this is expected to increase ETR after the tax reform, as the reduction in the corporate tax rate decreases the tax savings on interest. For investment, where firms are capital-intensive, this is also expected to increase ETR after the tax reform, due to the removal of accelerated depreciation and a reduction in the corporate tax rate. Finally, the tax reform proposed no change in tax policy for the R&D tax concession, so the tax reform have an impact on the association between ETR and R&D expenditure (Richardson and Lanis, 2007). Following from the above discussion, we hypothesize that: H7: ETR is negatively associated with the tax reform. H8: The association between ETR and firm leverage is positively impacted upon by the tax reform. H9: The association between ETR and firm capital intensity is positively impacted upon by the tax reform. H10: The association between ETR and firm inventory intensity is negatively impacted upon by the tax reform. H11: The association between ETR and firm R&D expenditure is not impacted upon by the tax reform. 3. METHOD 3.1. Sample and Data Using the panel data on 16 publicly-listed companies in Tunisian Stock Exchange of Securities over the 7 years period 2010- 2016 (112 firm years), after elimination the following categories of companies: a. Banks, insurance and leasing companies. b. Firms realized negative accounting income (Richardson and Lanis, 2007). c. Firms that have tax losses carryover2 because their ETR tend towards a negative direction. d. Firms with ETR higher than one causes an econometric problem (Gupta and Newberry, 1997. p6). e. Firms with missing data. The data for this study were collected through the financial statements available on the Tunis Stock Exchange of Securities (known Bourse de Tunis- BVMT) web site3. 3.2. Dependent Variable The dependent variable is represented by ETR. In conventional research, ETR is measured on the basis of the financial data shown in the financial statements of the companies. ETR is the quotient of two headings of the income statement as tax due divided by accounting result before taxes. However, the choice of the numerator and the denominator opens a scientific debate on the components of each heading (Plesko, 2003). For example, the use of the cumulative taxes or only the corporate tax is used since any increase in a monetary unit of taxation may affect the tax burden of this economic entity. At the level of the second part, the question to ask is what type of income or result we can admit? The choice is between three propositions: the accounting income before taxes, the taxable income and the operational cash flow. The taxable result should not be used with a presence of tax deductions that have an impact on ETR. We select the numerator (income tax due) after tax deductions and the denominator (accounting income) are before tax deductions (Gupta and Newberry, 1997). We use accounting income (accounting result) as the income measure in the denominator. 3.3. Independent variables The independent variables represent the different characteristics of Tunisian firms, namely: SIZE: Natural logarithm of total assets (accounting value); INDLT: Capital structure, this variable represents the financing operation, is measured by the non-current liabilities divided by total assets (accounting values). INVEST: Capital intensity ratio, this variable measured by non- current assets divided by total assets. INVENT: inventory intensity, the value of inventory divided by total assets. For the independent 2 Losses carried forward for five years. However, deferred amortization carried forward for an unlimited period (paragraphs IV of article 48 of the Code of the Income Tax and of the Corporation Tax). 3 www.bvmt.com.tn Naoui and Kasraoui: Post Tax Reform and Corporate Effective Tax Rate: Evidence from Tunisia International Review of Management and Marketing | Vol 10 • Issue 3 • 20204 variable R&D represents the spending charges divided by net sales (Richardson and Lanis, 2007). 3.4. Control Variables Firms’ operations could also impact on ETR. Wilkie (1988) finds that ETR is a function of the ratio of tax deductions to accounting income, where tax deductions (e.g. amortizations) are items that cause accounting income to differ from taxable income. To the extent that tax deductions are not proportionately related to accounting income, ETR can change simply due to changes in accounting income. Hence, we use return on assets (ROA) to control for changes in firms’ operating results. Gupta and Newberry, 1997) find that ROA is measured as pre-tax income divided by total assets. We expect ROA to have a positive sign because an increase in return on assets leads to an increase in ETR. The company’s activity sector is a firm-specific factor that influences the ETR. The industry sector (INSEC) is measured with dummy variable. The dummy equals 1 if the company’ sector is “industries” and 0 otherwise to control for the impact of the activity on ETR. 3.5. Tax Reform Variable To test the Tax Reform’s impact on ETR, a period dummy variable (TAXREFORM) is included in our study (coded 1 for the post- tax-reform period, 0 otherwise), along with interaction terms comprising the TAXREFORM dummy variable multiplied by each of the firm-specific variables and control variables. The TAXREFORM coefficient provides a test of the mean shift in ETR after tax reform. The interaction term coefficients permit testing for slope shifts in each of the firm-specific and control variables after tax reform and therefore determine whether these associations changed after the tax reform (Richardson and Lanis, 2007). However, the firm-specific and control variable coefficients provide evidence on their associations with ETR for the period before and post tax reform (Richardson and Lanis, 2007). 3.6. Regression Model Our empirical analysis involves estimating the following regression model is based on the method developed by Swamy and Arora (1972). Having random effects means having a group– (or time–, or both) specific component in the error term. ETR SIZE INVEST INVENT INDLT RD it it it it it it � � � � � � � � � � � � � � 0 1 2 3 4 5 66 7 8 9 1 ROA INSEC TAXREFORM TAXREFORM �*�SIZE it it it it it � � � � � � � � � 00 11 12 TAXREFORM *�INVEST TAXREFORM �*INVENT TAXR it it it it� � � � � � EEFORM *INDLT TAXREFORM *RD TAXREFORM *ROA � it it it it it it� �� �13 14 �� �� �15TAXREFORM �*INSECit it it The independent variables: Size (SIZE), balance-sheet characteristics assets versus capital (INVEST, INVENT and RD), capital structure (INDLT), firm operations activity (ROA), the company’s activity sector (INSEC), the tax reform (TAXREFORM) and interaction terms (TAXREFORM * SIZE, TAXREFORM * INVEST, TAXREFORM *INVENT, TAXREFORM * INDLT, TAXREFORM *RD, TAXREFORM *ROA and TAXREFORM *INSEC). 4. RESULTS 4.1. Descriptive Results To make a comparison with other studies for the Tunisian context Menchaoui et al. (2017) find that an average ETR of 18.03% and a median ETR of 17.56% for the period 2007-2011 for the Tunisian corporate groups. The dependent variable, ETR has a mean of 17.26% and a median of 19.16% different from the statutory company tax rates (25% or 20%) due to the existence of the accounting-tax divergence in the Tunisian context (Table 1). For the independent variables, SIZE has a mean of 7.883 and a median of 7.888, about the size We conclude that all of the companies of our sample have a similar size with a restraint market, INVEST has a mean of 0.598 and a median of 0.583, INVENT has a mean of 27.7% of assets and the cumulate of the mean of the two variables is 87.5%, so We can conclude that the total assets in our sample composed by net property, plant, equipment and inventory presented by 87.57% of companies’ assets. INDLT has a mean of 0.070 and a median 0.033. RD has a mean of 7%, especially for the companies of technology and pharmaceutical industry having the higher percentage and a median of 0.010, and ROA has a mean of 0.090 (9%) and a median of 0.063 (6.3%). 4.2. Measurement Model Our estimation about the impact of the tax reform on the firm- specifics is presented in Table 2. The model is significant with a level of α=10% due to a large number of independent variables. The results show that the Tunisian company’s has a specific reaction for the size variable and for the value of inventory in the post-tax reform period with a significant coefficient positive and negative signs. The positive sign of the reaction between the company size and the tax reform can be explained by the impact of reducing corporate tax rate on large companies. These findings converge with our hypothesis (H1 and H2). The coefficient on the inventory is negative and significant, we confirm the hypothesis H10, suggesting that the inventory value increases after the tax rate reducing. Table 1: Descriptive statistics ETR SIZE INVEST INVENT INDLT RD ROA Mean 0.1726 7.883 0.5984 0.277 0.070 0.017 0.090 Median 0.1916 7.888 0.5831 0.235 0.033 0.010 0.063 S.D. 0.332 0.045 0.021 0.065 0.045 0.008 0.046 Naoui and Kasraoui: Post Tax Reform and Corporate Effective Tax Rate: Evidence from Tunisia International Review of Management and Marketing | Vol 10 • Issue 3 • 2020 5 4.3. Testing Panel Data We run the Hausman test to select between fixed or random effects where the null hypothesis is that the best model is “random effects” versus the alternative hypothesis is “the fixed effects”, the P = 0.401, we adopt the null hypothesis with a presence of random effects. We use the Breusch-Godfrey test (Breusch, 1978; Breusch and Pagan, 1980) for detecting a serial correlation in panel models. The null hypothesis is that there is no serial correlation of any order up to the P = 0.894. We can conclude that no serial correlation in our model, so the term error’s observations are uncorrelated with each other (Wooldridge, 2010). A second test, is to detect the cross-sectional dependence in panels, if individuals respond to common shocks, (Pesaran, 2004; 2015), the P = 0.892, We conclude that no cross-sectional dependence for the Tunisian’s listing companies the model is specified correctly resulting pure disturbance, idiosyncratic and uncorrelation across companies. A third test to detect the heteroscedasticty is the Breush-Pagan test. The test is also no significant, the P = 0.217, the model is homoscedastic and the variance of estimated residuals is fixed. 5. DISCUSSION The study provides an empirical evidence to evaluate the association between the tax reform and the corporate effective tax rate in Tunisia. We have to focus on the impact of the tax reform on ETR, and its impact on the impact between ETR and variables reflecting the financing and investment functions by reducing of the corporate tax rate and increasing minimum tax on the income. For the two periods – pre-tax reform and post tax reform- We find a significant effect between the ETR and the three independent variables such as the SIZE, RD and the INSEC, but no significant effect on firms characteristics in the post-tax reform for the fiscal year 2014 except the inventory variable INVENT with a negative significant coefficient. To relate our results to the empirical taxation literature, the ETR is an indicator of tax planning Dyreng et al. (2008), We find that the association between ETR and Firm-specific variables changes in the two periods. For Alm (2015) using the “Allocative Effects” term to find that taxes affect the ability to create resources. Indeed, this inefficiency results a weak employment of resources in the Tunisian context. 6. CONCLUSION The Tunisian companies competition was distorted due to a high ETR comparison to other countries4, they are less competitive especially after taxes. For example the dependant variable RD has a significant link to the ETR only in the pre-tax reform. Another factor can explain the inefficiency of the tax reform is the increasing of the inflation rate on the post tax reform period 2014-2016 Rouissi and Frioui (2014)5 at this approach Boadway et al. (1984) find that inflation reduced the effective tax rate on net property, plant and equipment but raise the tax of inventories. An additional factor proves this negative impact on firms- characteristics is the Tunisian tax system is ad hoc (Alm, 2015) due to the excess of taxation in Tunisia. For Alm (2015) suggests reducing the statutory rate by reforming the tax system. 4 (Alm, 2015): Level of Taxes: When the level of taxation (including Social Security contributions) in Tunisia (as a percent of GDP in 2010) is compared to a subset of relevant Middle East and North Africa (MENA) countries (e.g., Algeria, Bahrain, Djibouti, Egypt, Iran, Iraq, Jordan, Kuwait, Lebanon, Libya, Malta, Morocco, Oman, Qatar, Saudi Arabia, Syria, United Arab Emirates, West Bank and Gaza, Yemen), taxes relative to GDP are generally higher for Tunisia than for the median of other MENA countries. 5 (Rouissi & Frioui, 2014) Increase of the inflation: The inflation rate, which continues to climb, reaching to end of April 2012 to 5.7% with a peak of 8.2%. Table 2: Random –effects Model results of interaction between tax reform, ETR and firms-characteristics for the period 2010-2016 (n=112) Variable Predicted sign Estimate Std. Error t-value Pr(>|t|) (Intercept) 1.805 0.716 2.518 0.013* SIZE − −0.209 0.089 −2.331 0.021* INVEST − 0.039 0.088 0.451 0.652 INVENT + 0.132 0.203 0.650 0.516 INDLT - −0.408 0.482 −0.846 0.399 RD − −6.276 2.136 −2.937 0.004** ROA + −0.420 0.663 −0.633 0.527 INSEC ? 0.006 0.002 2.761 0.006** TAXREFORM − −0.451 1.513 0.297 0.766 TAXREFORMSIZE + 0.022 0.026 0.857 0.006** TAXREFORMINVEST + −0.086 0.136 −0.635 0.526 TAXREFORMINVENT − −0.656 0.390 −1.683 0.095 TAXREFORMINDLT + 0.168 0.713 0.236 0.813 TAXREFORMRD + 0.638 2.694 0.236 0.813 TAXREFORMROA − 0.142 0.778 0.182 0.855 TAXREFORMINEC ? 0.000 0.003 0.156 0.875 ‘***’, ‘**’, ‘*’ Statistical significance at levels of 0%, 1%, 5% and 10%, respectively Naoui and Kasraoui: Post Tax Reform and Corporate Effective Tax Rate: Evidence from Tunisia International Review of Management and Marketing | Vol 10 • Issue 3 • 20206 The third reason to explain the null impact of the tax reform for the fiscal year 2014 and after is the increasing of the MTA from 0.1% to 0.2% of the total gross turnover and 0.1% of the turnover from exports. 6.1. Limitation The limitations of this study to more explain the impact of the tax reform on the firm’s characteristics. First, we have to use the ratio of the income tax expenses to the operating cash flow to appear the firms’ added value. Second, Plesko (2003) argues that the use of financial statements for the ETR measure has shortcomings and limitations and should be interpreted with a precautionary measure. We couldn’t include unlisted firms to our sample because of data unavailability (accounting and tax data). Indeed, the ETR model may be incomplete many firms have negative accounting income and tax losses carry forward are excluded from the sample despite they pay income tax (Zimmerman, 1983; Richardson and Lanis, 2007) with MTA equals to 0.2% of the gross sales and revenues plus the authority tax about up to 0.2% for the same taxable base, the registration taxes on the corporate assets, the payroll taxes and others contributions imposed on Tunisian’s firms for example the total rate for the social security contributions paid by the employer is 16.57% of the gross salaries. Third, we6 eliminated the sum of the reinvested earnings, the deduction capped at 50 % of net tax income subject to corporate tax6. Finally, a Tunisian company subject to withholding tax on corporate tax, the rate is 1.5% on the gross revenue receipts. 6 Investment Incentives Code. Duanjie, C., Jack, M. (2011), New Estimates of Effective Corporate Tax Rates on Business Investment Public Policy. Calgary: University of Calgary. Dyreng, S., Hanlon, M., Maydew, E. (2008), Long run corporate tax avoidance. The Accounting Review, 83, 61-82. Gupta, S., Newberry, K. (1997), Determinants of the variability on corporate effective tax rates: evidence from longitudinal data. Journal of Accounting and Public Policy, 16(1), 1-34. Menchaoui, I., Rossignol, J., Omri, M. (2017), Fiscal management practices and their impact on corporate groups’ fiscal performance. Accounting and Taxation, 9(1), 73-86. Pesaran, M. (2004), General Diagnostic Tests for Cross Section Dependence in Panels. CESifo Working Paper No. 1229. Pesaran, M. (2015), Testing weak cross-sectional dependence in large panels. Econometric Reviews, 34(6-10), 1089-1117. Plesko, G. (2003), An evaluation of alternative measures of corporate tax rates. Journal of Accounting and Economics, 35(2), 201-226. Plesko, G. (2003), An evaluation of alternative measures of corporate tax rates. Journal of Accounting and Economics, 35, 201-226. Porcano, T.M. (1986), Corporte tax rates: Progessive, proportional or regressive. The Journal of the American Taxation Association, 7, 17-31. Richardson, G., Lanis, R. (2007), Determinants of the variability in corporate effective tax rates and tax reform: Evidence from Australia. Journal of Accounting and Public Policy, 26, 689-704. Richardson, G., Taylor, G., Lanis, R. (2012), The Impact of Risk Management and Audit Characteristics on Corporate Tax Aggressiveness: An Empirical Analysis. Conference: Paper Presented at the Accounting and Finance Association of Australia and New Zealand. Rouissi, C., Frioui, M. (2014), The impact of inflation after the revolution in Tunisia. Social and Behavioral Sciences, 109, 246-249. Shevlin, T., Porter, S. (1992), The corporate tax comeback in 1987: Some further evidence. The Journal of the American Tax Association, 14(1), 58-79. Siegfried, J. (1972), The Relationship Between Economic Structure and the Effect of Political Influence; Empirical Evidence from the Corporation Income Tax Program, Thesis. United States: University of Wisconsin. Swamy, P., Arora, S. (1972), The exact finite sample properties of the estimators of coefficients in the error components regression models. Econometrica, 40(2), 261-275. Swenson, D. (1994), The impact of the U.S. Tax reform on foreign direct investment in the United States. Journal of Public Economics, 54(2), 243-266. Tran, A. (1998), Causes of the book-tax income gap. Australian Tax Forum, 14(3), 253-286. Watts, R., Zimmerman, J. (1986), Towards a Positive Theory of Accounting. New Jersey: Prentice-Hall. Wilkie, P. (1988), Corporate average effective tax rates and inferences about relative tax preferences. The Journal of the American Taxation Association, 10, 175-188. Wooldridge, J. (2010), Introductory Econometrics: A Modern Approach. Michigan: Michigan State University. Zimmerman, J. (1983), Taxes and firm size. Journal of Accounting and Economics, 5(2), 119-149. REFERENCES Alm, J. (2015), Analyzing and Reforming Tunisia’s Tax System. United States, CEQ Institute: Tulane University. Boadway, R., Bruce, N., Mintz, J. (1984), Taxation, inflation, and the effective marginal tax rate on capital in Canada. The Canadian Journal of Economics, 17(1), 62-79. Breusch, T. (1978), Testing for autocorrelation in dynamic linear models. Australian Economic Papers, 17, 334-355. Breusch, T., Pagan, A. (1980), The lagrange multiplier test and its applications to model specification in econometrics. Review of Economic Studies, 47(1), 239-253. Carreras, M., Dachapalli, C., Mascagni, G. (2017), Effective Corporate Tax Burden and Firm Size in South Africa: A Firm-level Analysis. WIDER Working Paper. Available from: http://www.wider.unu.edu. Desai, M., Dharmapala, D. (2005), Earnings Management, Corporate Tax Shelters. Available from: https://www.researchgate.net//228427319. [Last accessed on 2018 Jan 04].