## I. INTRODUCTION Taxes are an important element for a country, especially for developing countries. In most countries like Tunisia, a major part of government's revenue sources is financed through taxes. A country will always make efforts to optimize state revenues through the tax sector. However, efforts to optimize tax revenue may have problems, this happens because of the efforts made by companies to avoid paying taxes. In fact, companies undertake tax avoidance to reduce their tax expenses to the desired level. The phenomenon of conflicting interests between taxpayers and countries and the underperformance of average tax rates indicate the existence of significant tax avoidance activity. Tax avoidance policy is inseparable from the policies of company owners and management where company owners can consist of various ownerships and one of the most noticeable and influential factors in tax avoidance is corporate ownership structure. According to agency theory, ownership structures are control mechanisms to reduce profits between principals and agents (Jensen and Meckling, 1976). The ownership structure mechanism is one factor that narrows the gap between information used by management in the name of opportunistic behavior and the willingness of shareholders to run the company according to their wishes. An ownership structure is the component that a company's network can incorporate into company performance and decision-making. A centralized ownership structure can be a source of tax avoidance. Tax avoidance practices make it a strategic choice for executives to reduce a company's tax burden and improve profitability. Corporate tax avoidance seems to be the biggest problem facing today's generation. This represents a serious loss of revenue for governments in many developed and developing countries. Tax avoidance is therefore defined as any strategy, activity, or decision taken with the aim of lowering a company's effective tax rate. Tax avoidance is the ability of a company to pay less tax than it owes. Shareholders are expected to weigh the benefits of tax avoidance against the company's costs of possible prosecution, penalties and reputational damage if its strategy is compromised by tax (Hanlon and Heitzman, 2010). Corporate managers have a significant individual effect on tax avoidance (Dyreng et al., 2008). (Shackelford and Shevlin, 2001) stated that examining ownership structure is a determinant of potential corporate tax avoidance as it is important for corporate governance. Managers have a significant impact on a company's level of tax avoidance. A company can be owned by several shareholders where the percentage of ownership can vary. The ownership structure is believed to have the ability to influence the company which in turn can affect the company's performance. The ownership structure, namely Institutional ownership is ownership of company shares by financial institutions such as insurance companies, banks, pension funds, and investment banking (Veronica and Utama, 2005). The existence of ownership by financial institutions will encourage an increase in more optimal supervision of management performance, because share ownership represents a source of power that is used to support the existence of management. so that This study aims to empirically examine the effect of institutional ownership on tax avoidance. There is still not much research on the ownership structure of tax avoidance in the Tunisian context. The contribution of this research is to help and provide a new understanding of how the existing ownership structure, which in this study is institutional ownership on tax avoidance. ## II. LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT Institutional ownership plays a vital role in monitoring the performance of management because institutional ownership can improve better control. After all, it is considered capable of overseeing management decisions (Yuliawati and Sutrisno 2021). The larger the shares owned by an institution or institution, the higher the level of supervision of the company will be, so that it can reduce the occurrence of tax avoidance activities by company management (Andini et al. 2021). Although there have been many studies on tax avoidance, there are many differences from the results of previous studies due to differences like both dependent and independent variables, besides that there are differences in the subjects and objects studied, the research methods chosen, and the use of research methods is of course different. In this study, we will expose studies that found positive and negative effect of institutional ownership on tax avoidance and those who do not found any effect. ### a) Studies that Found an Effect of Institutional Ownership on Tax Avoidance Institutional ownership plays an important role in overseeing the performance of company management to make it more optimal (Sintyawati & Dewi, 2018). A large percentage of institutional ownership will have a big impact on monitoring company activities in achieving large profits so that companies tend to avoid tax. However, Tehranianet al. (2006) find that supervisory actions carried out by a firm and institutional investors can limit manager behaviour. Shleifer (2009) also proves that institutional ownership plays an important role in monitoring, disciplining, and influencing managers so that it can force managers to minimize tax avoidance. Consistently, Feranika et al. (2016), Wijayani (2016), and Krisna (2019) prove that the greater the institutional ownership, the more likely it is to minimize tax avoidance. Abdillah et al. (2016) prove that the existence of institutional ownership will professionally monitor the development of its investment resulting in a high level of control over management actions so that the potential for fraud can be suppressed. Institutional controlling shareholders often sacrifice the interests of other shareholders. On the other hand, high executive profits affect the amount of tax the company has to pay, hampering company performance (Darsani and Sukartha, 2021). The high tax rates imposed on businesses, managers are looking for ways to reduce the taxes they pay. The information presented by the manager in the financial statements does not reflect the state of the company due to information asymmetry. Tax avoidance is practiced by management to serve the interests of shareholders and management. Small or great institutional ownership in the company has an effect on the company's tax policy (Khurana and Moser, 2013). (Bird and Karolyi, 2017) found that institutional ownership was positively correlated with tax avoidance. They argue that the involvement of corporates with institutional investors with tax planning skills will lead to better tax planning and greater use of tax relief. Research conducted by (Ngadiman and Puspitasari, 2014) indicates that the institutional ownership variable has a significant impact on tax avoidance. The higher the institutional ownership, the higher the tax burden the company must pay. This is because companies are less likely to engage in tax avoidance. Institutional owners can force managers to focus on economic performance and avoid opportunities for selfish behaviour through actions of voice and power. The results of the analysis done by Damayanti and Wulandari (2021) show that institutional ownership has a significance level below $5\%$. The results of this study are consistent with Feranika et al. (2016), Wijayani (2016), and Krisna (2019). This study proves that the actions taken by institutional shareholders tend to minimize the risk of being detected as tax avoidance which can reduce the company's reputation so that institutional shareholders. Hoda Eskandar and Poya Ebrahimi (2017) have divided total institutional ownership into active and passive. The findings illustrated that institutional ownership had a positive effect on tax avoidance. It means that firms which institutional owners are their owners are highly likely to avoid form-paying tax. They may be motivated to use their cash in investment projects instead of paying tax. This finding is in line with the results of Hassana et al. (2016) and Khurana and Moser (2013) and contradicts the results of Khan, Srinivasan and Tan (2016), Having also dividing total institutional ownership into active and passive, it became clear that active institutional owners also had a positive effect on tax avoidance and inspired corporates to avoid paying taxes but the effect of passive owners on tax avoidance was negative. According to these findings firms whose institutional owners are active, are highly likely to avoid paying tax. In spite of such firms, those ones whose institutional owners are more passive, are highly likely to pay tax because their owners may not place particular emphasis on profitable projects and better future performance. These results are also in line with the results of Khurana and Moser (2013), Hassana et al. (2016). With regard to obtained results, investors and tax authorities in the country can observe institutional ownership and active institutional ownership as a signal of the company's willingness to further tax avoidance. Darsani and Sukartha, (2021) find that the greater the institutional ownership, the lower the tax avoidance action taken by the company. The results of this study are in line with the research of Merslythalia & Lasmana (2017), Romadona & Setiyorini (2018) and Maraya & Yendrawati (2016) which state that institutional ownership has a negative effect on tax avoidance. The results of this study also support agency theory which explains that institutional ownership can solve agency problems. This condition can occur because institutional ownership is needed in monitoring the activities of company management. Institutional investors are investors who come from outside the company and are not affiliated with the company concerned, so that institutional investors tend to obey the rules made by the government. In addition, institutional investors as supervisors who come from external parties will supervise company management in carrying out tax management because basically institutional investors tend to avoid the risk of tax avoidance activities that can damage the company's reputation. Therefore, a high percentage of share ownership by institutional institutions will increase control over company management to comply with tax regulations, so it can be concluded that the existence of institutional ownership can reduce tax avoidance efforts made by company management. Ratnawati et al. 2018) stated that institutional ownership negatively affects tax avoidance. This statement differs from (Khan et al. 2017) who state that institutional ownership has a significant positive on tax avoidance. Anwar, Salihu, and Obaid (2014) achieved a positive relation between the state ownership variable and tax avoidance. Khan, Srinivasan and Tan (2016) found that institutional ownership increased corporates tax avoidance H1: Institutional ownership has a significant effect on tax avoidance. ### b) Studies that not Supported an Effect of Institutional Ownership on Tax Avoidance Yusri et al. 2022 stated that institutional ownership does not affect tax avoidance. Institutional ownership helps to oversee administrative policies because it serves to unify the various policies that administrators take to make decisions that bring effectiveness to their organizations. Research conducted by Vizandra, E, and Mustikasari, E (2021) stated also that institutional ownership does not affect tax avoidance. This result indicates that the presence or absence of institutional ownership has no significant effect on corporate tax avoidance activities. This study obtained results similar to those of Suprimarini and Suprasto (2017), Arianandini and Ramantha (2018), which show that institutional ownership does not affect corporate tax avoidance activities. This is because institutional ownership has the assumption that the fulfilment of high corporate tax obligations will reduce the level of welfare of company owners. High tax obligations will cause a decrease in corporate profits and will cause a decrease in the amount of dividends received by shareholders, therefore institutional owners do not take action to prevent or limit management activities to reduce the company's tax burden. Ajeng Ayuning Tyas and Bani Binekas (2023) found that the size of institutional ownership cannot influence or oversee the management's policies in carrying out tax avoidance activities. Institutional Ownership delegates oversight and management of the company to the board of commissioners, which is their job, so whether or not there is institutional ownership, tax avoidance activities still occur. The orientation of institutional ownership is how to maximize the welfare (profit) obtained at the end of the period. The company will carry out tax avoidance, or it will not become the authority of the company's management. If this activity can benefit the welfare of institutional ownership, they will continue to support every policy carried out by the company. Institutional owners have incentives to ensure that management makes decisions that can maximize the welfare of institutional shareholders so that they only focus on earnings management. This statement is supported by the results of research by (Andini et al. 2021), who found has no effect of institutional ownership on tax avoidance. But different from (Khan et al. 2017) said that institutional ownership positively affects tax avoidance, and (Ratnawati et al. 2018) said that institutional ownership negatively affects tax avoidance. H2: Institutional Ownership has no effect on Tax Avoidance. Established theories and empirical results previously found regarding the impact of institutional ownership on tax avoidance reflect the characteristics of the markets where the studies were carried out, namely European, American and Asian markets. We will test in the next section this impact on the Tunisian context. ## III. RESEARCH METHODS ### a) Types of Research This research is included in a quantitative approach, which measures the interaction between variables empirically, objectively, and systematically. ### b) Population and Sample The population of this research are companies listed on the Tunisian Stock Exchange. The sample used in this research are the chemical, cement, transportation, telecommunications, Bank, leasing, insurance, agribusiness and trading sector. Those sectors was chosen to variate the percentage of institutional ownership level. There are several criteria for selecting the sample used show of table 1 - Companies listed on the Tunisian Stock Exchange from 2018 to 2022 - The company should attach complete annual reports and have the necessary information for research. - All activities most of them are related to aspects of taxation. Table 1 <table><tr><td>Criteria</td><td>Number of Firms</td><td>Number of Data</td></tr><tr><td>Listed firms</td><td>38</td><td>190</td></tr></table> The data used in this research is quantitative data. Meanwhile, the data source used is secondary data, namely the annual reports of Tunisian companies from 2018 to 2022 which are obtained from the BVMT website or the official website of each company. The data collecting method used is documentation method from secondary data through annual reports for the years 2018 until 2022 period listed in Tunisian Stock Exchange. ### c) Variables Operational Definition and Measurement ## i. Tax Avoidance (Tax avoidance) The dependent variable in this study is tax avoidance, which is a business run by a taxpayer in order to reduce the tax owed by not violating the law and is safe for the company. The ratio used to measure how much tax avoidance is done by a company using Cash Effective Tax Rate (CETR). The Cash Effective Tax Rate (CETR) can be calculated by: $$ \mathrm{C E T R} = \frac{\text{PaymentofTaxes}}{\text{EarningBeforeTax}} $$ This measure is used by Tyas, A and Bani, B. (2023) ## ii. Institutional Ownership (INST) The independent variable in this study is institutional ownership. It will be denoted by INST. Institutional ownership is expressed as a percentage by calculating the total shares owned by institutional investors divided by the number of outstanding shares (Utami, 2013). ## iii. Leverage (Leverage) Leverage is a control variable. It is a ratio that aims to determine the value of assets financed by debt is. One way to measure it is total debt divided by the company's total assets (Annuar et al., 2014). ## iv. Size (SIZE) Company size is a control variable. It is a used for classifying whether a company is large or small based on its nominal size value. The size of the company can be determined by calculating the total Ln of assets owned by the company (De George et al., 2013). ### d) Empirical Model The empirical model proposed to test the hypothesis in this research is as follows: $$ \begin{array}{c} \text{TaxAvoidance} = \alpha + \beta 1 I N S T + \beta 2 L E V E R A G E + \\\beta 3 S I Z E + e \end{array} $$ Where: $$ \alpha = \text{Constant} $$ $$ \beta 1 - \beta 3 = \text{Regressioncoefficient} $$ $$ e = Error Results $$ ### e) Descriptive Statistics Descriptive statistics can describe the research variable data that has been collected. Following are the results of variable descriptive statistical analysis, see table 2. Table 2 <table><tr><td>Variables</td><td>MEAN</td><td>MEDIAN</td><td>SD</td><td>MIN</td><td>Q1</td><td>Q3</td><td>max</td></tr><tr><td>Tax avoidance</td><td>0.178</td><td>0.179</td><td>0.117</td><td>-0.011</td><td>0.064</td><td>0.283</td><td>0.382</td></tr><tr><td>inst</td><td>13.677</td><td>0.000</td><td>18.593</td><td>0.000</td><td>0.000</td><td>30.800</td><td>61.410</td></tr><tr><td>leverage</td><td>29.916</td><td>26.115</td><td>20.008</td><td>0.060</td><td>11.165</td><td>47.350</td><td>66.690</td></tr><tr><td>size</td><td>13.324</td><td>13.032</td><td>1.314</td><td>11.412</td><td>12.301</td><td>13.952</td><td>16.151</td></tr></table> Based on table 2, the dependent variable in this study is tax avoidance. This variable has a minimum value of -0.011 and a maximum value of 0.382 it can be said that the level of tax avoidance carried out by the companies in this sample is low and the difference is not that far between those companies. Tax avoidance variable has a mean of 0.178 with a standard deviation of 0.117. It shows that in the sample of this study only a few companies do tax avoidance practices. Thus, there are already many companies that have complied with government regulations and meet their tax obligations. the independent variable in this study is institutional ownership. This variable has a minimum value of 0 which means that there are companies in this sample that are not owned by institutional investors. Meanwhile, the maximum value of 61.410 means that there are companies that is almost entirely owned by institutional investors. Institutional ownership has a mean of 13.677 with a standard deviation of 18.593. This implies that there is less variability in the data set, and the values are relatively consistent. Control variables in this study are leverage and company size. Leverage has a minimum value of 0.060 and a maximum value of 66.690. Leverage has a mean of 0.58304 with standard deviation 0.15153. The standard deviation value which is smaller than the mean value indicates that the research data is homogeneous. The size of the company has a minimum value of 11.412 and a maximum value of 16.151 Company size has a mean of 11.412 with standard deviation 1.314. The standard deviation value which is smaller than the mean value indicates that the research data is homogeneous, so that the data distribution shows normal results. Therefore, it shows that the company in this study has a fairly large company size. or more explanatory variables of the model as shown in the following matrix: ### f) The Correlation Matrix The correlation matrix gives us an idea of a perfect or almost perfect linear relationship between two Table 3: Correlation Matrix <table><tr><td></td><td>Tax Avoidance</td><td>Inst</td><td>Leverage</td><td>Size</td></tr><tr><td>Tax avoidance</td><td>1.000</td><td></td><td></td><td></td></tr><tr><td>inst</td><td>0.255**</td><td>1.000</td><td></td><td></td></tr><tr><td></td><td>(0.019)</td><td></td><td></td><td></td></tr><tr><td>leverage</td><td>-0.373***</td><td>-0.015</td><td>1.000</td><td></td></tr><tr><td></td><td>(0.000)</td><td>(0.893)</td><td></td><td></td></tr><tr><td>size</td><td>0.165</td><td>0.143</td><td>-0.175</td><td>1.000</td></tr><tr><td></td><td>(0.135)</td><td>(0.195)</td><td>(0.111)</td><td></td></tr></table> There is an absence of values that are greater than 0.8 for all explanatory variables. In summary, it can be seen that the correlation among the current study's variables is generally low; hence, this is an indication that there are no problems arising from the multi-collinearity between the variables used in the regression models. ## IV. REGRESSION ANALYSIS To obtain the final results, the extreme values in the data are removed. This brings us to data where the inst variable is no longer variable over time and therefore impossible to use fixed effect panels. Instead, the MCO method is used while controlling the fixed-year effect. The test results can be seen from the following table 4. Table 4: Regression Analysis <table><tr><td>Constante</td><td>0.129</td></tr><tr><td></td><td>(0.289)</td></tr><tr><td>inst</td><td>0.003**</td></tr><tr><td></td><td>(0.023)</td></tr><tr><td>leverage</td><td>-0.002***</td></tr><tr><td></td><td>(0.001)</td></tr><tr><td>size</td><td>0.005</td></tr><tr><td></td><td>(0.528)</td></tr><tr><td>F test</td><td>2.84**</td></tr><tr><td>p-value</td><td>0.011</td></tr><tr><td>R2 (%)</td><td>20.76</td></tr><tr><td>Adj. R2 (%)</td><td>13.46</td></tr></table> According to table 4 the regression equation can be determined as follows: $$ \begin{array}{l} \text{Tax} = 0.1 2 9 + 0.0 0 3 \text{INST} - 0.0 0 2 \text{LEV} \\+ 0.0 0 5 S I Z E \\\end{array} $$ The interpretation of the regression results will be explained as follows: The constant value $(\alpha)$ is 0.129, which means that if all the independent variables used do not change, the value of tax avoidance will increase 0.971 times due to other variables outside the research. The regression coefficient value $(\beta 1)$ of institutional ownership (INST) is 0.003, which means that when institutional ownership increases 1 time, the value of tax avoidance also increases by 0.003 times, assuming the other variables are constant. The regression coefficient (β2) of leverage (Leverage) is -0.002 which means that when the leverage increases 1 time value of tax avoidance decreases by 0.002 times, assuming the other control variables are constant. The regression coefficient value (β3) of company size (SIZE) is 0.005, which means that when the size of the company increases 1 time, the value of tax avoidance also increases by 0.005 times, assuming the other control variables are constant. According to table 4, the coefficient of determination (R2) is $20.76\%$. Thus, it can be said that the dependent variable in this study can be explained by the independent variable of $20.76\%$, and $79.24\%$ is explained by other variables not examined. According to the results in Table 4, it can be seen that the p- value of the institutional ownership variable is 0.023 with a significant value of $0.023 < 0.05$, so it can be concluded that institutional ownership has an effect on tax avoidance for companies listed on the Tunisian stock exchange for the 2018- 2022 period. The likelihood ratio (F-statistics), which statistically is significant $(p < 0.001)$, indicates that the model is fit to describe the change in the explanatory variables. ## V. DISCUSSION The results of the analysis show that institutional ownership has a significance level below $5\%$ so that H1 is accepted and H2 is rejected. The results of this study are consistent with (Khurana and Moser, 2013). (Bird and Karolyi, 2017). This study proves that the actions taken by institutional shareholders affect tax avoidance in the Tunisian context. However and by observing the regression coefficient relating to the variable" inst" which is equal to 0.003 we can conclude that the effect of institutional ownership on tax avoidance is minimal. This interpretation is confirmed by the value of $\mathsf{R}^2$ which confirms that $79.24\%$ of the dependent variable in this study is explained by other variables not examined. The results show also that firm size doesn't affect tax avoidance. In fact, According to the results in Table 4, it can be seen that the p- value of size variable is $0.528 > 0.05$, so it can be concluded that firm size has no effect on tax avoidance. The findings of this study have several implications. This finding can help government to add insight into the effect of company size and institutional ownership on tax evasion. ## VI. CONCLUSION This study assesses the effect of institutional ownership on tax avoidance of listed Tunisian companies. In this paper, we found a significantly positive relation between institutional ownership and tax avoidance. The effect of institutional ownership on tax avoidance is minimal and Size Company has no effect on tax avoidance. We also found that there is negative relationship between leverage and tax avoidance. This study proves that tax avoidance tends to be not explained by institutional ownership. The findings of this study have several implications. This finding can help government to add insight into the effect of company size and institutional ownership on tax evasion. For future researchers, it is suggested to increase the research period and object of research from various sectors to know more about the development of tax avoidance from year to year. Also is suggested to add other independent variables that are different from this research.

Generating HTML Viewer...