Effects of Political Connections and Corporate Governance on Tax Aggressiveness in Indonesian Service and Banking Sectors

We investigate whether political connections and corporate governance have any effects on tax aggressiveness in service and banking sector in Indonesia. Corporate governance act as the independent variable on the first model, and as moderating variable on the second model. Tax aggressiveness is measured using effective tax rates. Political connection is measured with the amount of any connection between the company and its board member's political background. While the measurement of corporate governance are among others: board independence, board size, CEO duality, institutional ownership, and external auditor's reputation. We took the samples service companies and banks listed on the Indonesia Stock Exchange for the period 2013-2017. The analysis used in this study is the multiple linear regression analysis. The finding is that political connection does not influence tax aggressiveness in both sectors. In the service sector, corporate governance measured with CEO duality and institutional ownership has a negative effect on tax aggressiveness, while the other measurements have no effect. While in banking sectors, board size has a negative effect, institutional ownership and external auditor's reputation have a positive influence. Corporate governance did not moderate the influence of political connections on tax aggressiveness in both sectors.


Introduction
The governmental purpose of maximizing revenue from taxes is not the same as the company purpose as a taxpayer, in which the company attempts to minimize the tax burdens, so it gets higher net income. The actions of tax-saving conducted by a number of companies in Indonesia are aimed not to do tax evasions, but considerably to the purpose of saving many tax expenses paid by the company by way of utilizing gaps available within tax regulations in Indonesia (Suandy, 2014, Pranoto, andWidagdo, 2016). Wahab, Ariff, Marzuki, Zuraidah, & Sanusi (2017); Wicaksono (2017); Pranoto & Widagdo (2016); Boussaidi & Hamed (2015); Putri (2014); Mulyadi, Anwar, & Krisma (2014); Kim dan Zhang(2013), investigated the influences of political connection and corporate governance towards tax aggressiveness. According to Wahab, Ariff, Marzuki, Zuraidah, & Sanusi (2017), they found that political related positively towards tax aggressiveness. The more who have political connections, so it can enable to do tax aggressiveness. Good governance can prevent the company from practicing company tax aggressiveness policies. Corporate governance not only have an effect on tax aggressiveness but also moderating the effect of political connection toward tax aggressiveness (Wahab et al. 2017) negatively. Whereas according to Pranoto & Widagdo (2016), political connection probably may be used by the taxpayer for helping to reduce the possibilities of tax inspections or reduce any tax sanctions by utilizing the relationship with government officials.
From the researches of Boussaidi & Hamed (2015); Putri (2014); Mulyadi, Anwar, & Krisma (2014), it is known that corporate governance also influences toward tax aggressiveness. The research of Boussaidi & Hamed (2015) found that corporate governance has positive influences toward tax aggressiveness, whereas the research of Putri (2014) found that corporate governance has negative influences toward company tax aggressiveness. Kim and Zhang (2013) stated that a company with a political connection could easily access for getting information about tax regulations or law enforcement in the future, so the company could arrange more complex tax strategies.
The samples taken are from service and banking sectors listed in Indonesian Stock Exchange in the period of 2013 to 2017. Compared with manufacturing sectors, the service sector has a higher percentage of growth, and then its tax aggressiveness tends to be higher. The banking sector is one of the biggest country tax contributors after the manufacturing company.

Literature Review
Agency theory explains about the relationship between management and the company owner. In agency theory, there are two parties, i.e., the management who acts as an agent and the stakeholders as principal. Jensen and Meckling (1976) defined agency relationship as contract or agreement conducted by one or more people (principals) for binding other person (agent) in order to manage services in company activities in which the agent parties have the authorities upon the management.
According to Saeed (2013), corporate governance is a system applied in the company for directly controlling operational activities. Sandy and Lukviarman (2015) stated that corporate governance for some company would support operational activities. Implementational of corporate governance mechanism should be company main attention, then it will make agent follow the policy and regulations (Lestari and Putri, 2017). Gomez (2009) in Wicaksono (2017) explained that the special relationship with the government enables the company getting different treatment from the government compared with the one which did not have. The indicator of this political connection is whether the independent commissioner currently has past experience or currently still being a politician who affiliates to a political party and/or governmental official, and/or military official. Tiaras and Wijaya, (2015) explained that company tax aggressiveness action is some action of engineering incomes charged by taxes conducted by the company. The enforcement from tax aggressiveness is for the purpose of minimizing tax debts, and the ways conducted do not violate the valid determinations. Wahab, Ariff, Marzuki, Zuraidah, Sanusi (2017) showed that there are positive influences of political connection towards tax aggressiveness. The company related by political connection has benefits of risk decreases, accesses to information about tax changes, so the company related politically has lower detection risks because of protected by political connection (Kim and Zhang, 2016). The company related to the political connection in paying taxes has effectiveness level far lower than the company unrelated (Adhikari et al., 2006). Based on the explanations, this research assumes that by the company which has a political connection, so it will increase the company tax aggressiveness. H 1 . Political connection influences towards tax aggressiveness Sandy and Hamed (2015) found that corporate governance has positive influences towards tax aggressiveness. The positive influences can be explained that the higher the corporate governance will lead to the lower the tax aggressiveness. Corporate governance or management has important roles in the company's tax attitudes. It Companies that meet the criteria for research are 61 companies because there are two main requirements that must be fulfilled, namely: completeness of financial statements published on the Indonesia Stock Exchange from 2013-2017. During that period the company did not experience losses. There were also several companies not publishing on the website. The number of data that meet the criteria for 5 years is 305 company data. But outliers were made so that only 270 data were used. Companies that meet the criteria for research are 37 companies because there are two main requirements that must be fulfilled, namely: completeness of financial statements published on the Indonesia Stock Exchange from 2013 to 2017 and during that period the company did not experience losses, but there were also several companies not publishing on the website. The number of companies that meet the criteria for 5 years is 185 company data. But outliers were made so that only 115 data were used.

Variables Operationalization
The other independent variable in this study is political connections. Political connections that can be used in this study refer to the research of Pranoto and Widagdo (2016) because they are adjusted to companies in Indonesia, that political connection variables will be measured using numbers, by counting the number of directors of companies, external auditors and board independence who have political connections Corporate governance measurement refers to research by Wahab et al. (2017), which is categorized into internal and external governance mechanisms. For internal governance mechanisms, including: 1. Board Independence (BINDit), measured by the comparison of the number of independent directors and the overall company director. 2. Board Size (LBSIZEit) is measured from the Ln transformation of the total number of commissioners in the company. 3. CEO Duality (DUALITYit) means that there are multiple positions between the board of commissioners and the board of directors. DUALITYit in its measurement uses a dummy, with a value of "1" when the role of the CEO and chairman is separated and "0" when CEO duality in the company exists.
For external governance mechanisms, including: 1. Institutional Investors (INSTOWNit), Institutional Investors can be seen from the percentage of share ownership by the institution to the total number of shares. The indicator used for the presentation is the percentage of shares held by other companies, the data is represented by the percentage of 5 shareholders of the top institutional investors. 2. External auditor (BIGNit), is an independent company audit of financial statements in Indonesia. Represented by the value "1" when the company uses an audit service by one of the BIG 4; if not the value "0".
The dependent variable in this study is tax aggressiveness. Measurement of tax aggressiveness by using effective tax rate (ETR) with the following formula (Wahab, Ariff, Marzuki, Zuraidah, Sanusi, 2016): The control variable in this research adopts the research of Wahab et al. (2017), and it is adjusted by the research objects. The following ones are control variables used in this research: 1. Firm Size is a measurement scale used for classifying big or small company. Size measurement is measured by using natural logarithm from the total assets. ASSETS = Ln (Total Assets) 2. Leverage, is a ratio that describes company capital structure and describes company financing decisions. Leverage can be calculated by using proportions of the total debts divided by the total assets owned by the company.
3. Market-to-book ratio, is comparison or ratio between market value and book value. (4)

The Model of Analysis
The analysis models used in this research use the research models available in the research of Wahab et al. (2017). The first model uses two independent variables, i.e., Corporate Governance and political connections and one dependent variable, i.e., tax aggressiveness. Whereas, the second model uses one independent variable, i.e., political connections, one moderation variable, i.e., Corporate Governance and one dependent variable, i.e., tax aggressiveness.
Model 1: TAX_AGRR it : tax aggressiveness is measured by effective values of tax rates (ETR). POLCON it : political connections, is measured by dummy, the value variable is "1" if the company is related politically and "0" if not CGOV it : corporate governance, this variable can be categorized into internal and external mechanisms ASSETS it : firm size, measured by Ln_Total Asset. DEBT it : leverage, measured by the total debts divided by the total equities. MTBV it : market-to-book ratio, measured by the capitalization market ratio divided by the total book values.
The following ones are the data results of Descriptive Statistics for service and banking sectors:

Results of Analysis and Discussion
The feasibility test of the model in this study includes: Determination Coefficient Test, Simultaneous Significance Test (F Test) and Individual Parameter Significance Test (t test). The test was conducted with a model without moderation and with moderation of the interaction test (MRA). The following is a summary of the table of test results for each of these analyzes:  The Determination Coefficient Test (R 2 ) is a test to see whether the regression model is appropriate or not using the adjusted R 2 value as the Determination Coefficient Value. Based on the results of the analysis of the Service sector in table 2, it shows that the adjusted R 2 value or unmoderated coefficient is 0,105. From these results indicate that the independent variables together affect the dependent variable by 10,5% and the remaining 89,5% is explained by other variables that have not been studied.
The moderating results of the BIND variable (board independence), LBSIZE (board size), and DUALITY (CEO duality) are seen in the table that the adjusted R 2 value between moderation and R 2 uses the first test moderation, the second test, and the third test decreases to 0,103 % (10,3%). This explains that independent board, board size, and CEO duality are not moderating variables.
The results of INSTOWN moderation (institutional investors) in the table can be seen that the adjusted R 2 value has increased from R 2 before moderation to 0,017 (17%). This explains that institutional investors as moderating variables strengthen the regression model by 2%.
The moderation results of BIGN (external auditor) in the table can be seen the adjusted R 2 value between without moderation and R 2 using moderation in the fifth test decreased to 0,104 (14%). This explains that the external auditor is not a moderating variable.  Sig.F 0,000 0,000 0,000 0,000 0,000 0,000 The correlation coefficient is a coefficient to explain the two-way relationship between the independent variables studied and the dependent variable. Based on the results of the regression analysis in the Banking sector in   In the first regression model shows the existence of positive and negative coefficients. Positive coefficient shows a unidirectional change between the independent variables on the dependent variable and vice versa if the negative coefficient shows the opposite change between the independent variables on the dependent variable.
1. The t test value of the POLCONit variable in the service sector is 0.357 (p> 0.05), and the regression coefficient value is 0.017, and it can be concluded that political connections have a positive and insignificant influence on Tax Aggressiveness. For the t test value of the POLCONit variable in the banking sector at 0.762 (p> 0.05) and the regression coefficient value of -0.001 (negative), the political connection has a negative but not significant influence on tax aggressiveness. Political connections are not influential because companies whose shares are mostly owned by the government are defined as low-risk taxpayers, in accordance with Minister of Finance Regulation No. 71 / PMK.03 / 2010. In this regulation, it illustrates the belief that companies whose majority shareholders are the government do not carry out tax avoidance. The closeness of the company makes the company more careful in making any decisions in order to continue to get an award from the government as an obedient taxpayer. Compliant companies often get awards from the government so that they can be said to improve the company's reputation. This encourages companies to always follow various government regulations issued. Utilizing closeness to political parties can indeed be said to provide several benefits for the company, but the company must consider of the long-term impacts. The poor reputation of the company will have an impact in the long term so that it can reduce people's trust and could cause any business's loss. 2. The value of the BINDit variable in the service sector is 0.357 (p> 0.05) and the regression coefficient value of -0.006, the board independence has a negative influence but is not significant on tax aggressiveness. The value of the BINDit variable t test in the banking sector is 0.973 (p> 0.05), and the regression coefficient value is -0.001 (negative), the board independence has a negative but not significant influence on tax aggressiveness. The results of this study support the results of research conducted by Ridha and Martani (2014), Pradipta and Supriyadi (2015), Utami and Setyawan (2015). According to Indonesia's Stock Exchange advisory board (BAPEPAM) Regulation Number IA concerning General Provisions for Listing of Equity Securities in the C-1 letter exchange, where in order to implement good corporate governance the listed companies are required to have independent commissioners with a composition of at least 30% of the board of commissioners. The existence of independent commissioners in companies in Indonesia has not been able to meet the composition regulated by the guidelines of good corporate governance, namely 30% (Laily, 2017) so that it could be the reason why the board independence does not have a significant impact on tax effectiveness. Viewed from the average value of BINDit variables in the service sector (12%) and for the banking sector (58%, it) can be said that although small or large the independence of the boards in the company does not affect decision making in tax aggressiveness. 3. The LBSIZEit variable t test value in the service sector is 0.927 (p> 0.05), and the regression coefficient value is 0.019, it can be concluded that board size has a positive and insignificant influence on tax aggressiveness. Positive influence can be explained that the more the number of company directors causes higher tax aggressiveness. However, the influence of board size on tax aggressiveness was stated to be insignificant in concluding that it was not significant. The results of this study support the results of research conducted by Budiman, 2012 (in Ni Nyoman andI Ketut, 2014) that the board size of directors does not influence on tax aggressiveness. The reason for the board size of directors does not influence on tax aggressiveness because of the action of tax aggressiveness by the company with its determination through policies taken by the company's leaders themselves. For the LBSIZEit variable t value in the banking sector is 0.046 (p <0.05) and the regression coefficient value is -0.008 (negative), the board size has a negative and significant influence on tax aggressiveness. The results of this study are similar to previous studies conducted by Wahab et al. (2017). This negative relationship indicates that the larger size of the board acts as a monitoring mechanism and prevents tax aggressiveness. Hubaid (2016) argues that small councils may look more effective in improving performance and limiting director incentive avoidance because the performance of each member is easier to be monitored and decisions can be made faster. 4. The t test value of the DUALITYit variable in the service sector is 0,000 (p <0.05), and the regression coefficient value is -0,076, CEO duality has a significant negative influence on tax aggressiveness. The results from the data above show that CEO duality has been shown to have a significant influence on tax aggressiveness. Negative influences indicate that when a company owner joins the ranks of management or the board of commissioners will cause tax aggressiveness to decline. Decreasing tax aggressiveness, which means that the higher the tax burden paid compared to pre-tax profits. The explanation can be revealed that when a company owner is actively involved in the composition of the board of directors or the board of commissioners shows a tendency to pay taxes higher than pre-tax profit. For the t test value, DUALITYit variable in the banking sector is 0.744 (p> 0.05), and the regression coefficient value is 0.004 (positive), the CEO Duality has a positive but not significant influence on Tax Aggressiveness. From the descriptive results show that the average value of CEO Duality is 0.8 or close to 1. That is, the company has a CEO and board of directors separately. In Indonesia rarely there are CEO Duality, because of Indonesia more use of two-tiers board system. In the data the banking sector having CEO Duality no more than 5 company, so is it can be said that CEO Duality does not affect in plan tax aggressiveness company. 5. The t test value of the INSTOWNit variable in the service sector is 0.015 (p <0.05) and the regression coefficient value is -0.093, Institutional Investors have a significant negative influence on Tax Aggressiveness. These results indicate that institutional ownership is proven to have a significant influence on tax aggressiveness. Negative influences indicate that when the higher the percentage of institutional ownership in the company, it can cause tax aggressiveness to decrease. Decreasing tax aggressiveness means the higher the tax burden paid compared to pre-tax profits. These results support research from (Dewi and Jati, 2014) institutional investor ownership of shares whose shares are owned by institutions such as companies, foreign investor insurance companies, except for individual ownership. Institutional owners participate in conducting supervision and management, but institutional owners entrust the board of commissioners because it is the duty of the board of commissioners who represent institutional owners. Institutional ownership has a high percentage of ownership of the company's shares so that institutional owners have the ability to pressure management to follow the interests of institutional owners. Whereas, the t test value of INSTOWNit variable in the banking sector is 0.001 (p <0.05) and the regression coefficient is 0.054 (positive), Institutional Investment has a positive and significant influence on Tax Aggressiveness. With the presence of institutional ownership or INSTOWNit in a company plays an important role to monitor, discipline and influence managers in tax management (Shafer and Simmons,2006). The greater the institutional ownership held by the institution, the greater the pressure the company management will take to carry out tax avoidance. 6. The t test value of the BIGNit variable in the service sector is 0.713 (p> 0.05), and the regression coefficient value is -0.007, the external auditor is not a significantly negative influence on Tax Aggressiveness. The Big 4 proved not to have a significant influence on tax aggressiveness. The quality of external auditors can be illustrated by the difference between the use of Big Four and non-Big Four auditors. The reason is that the Public Accountants Office in the Big Four has the competence to express errors in reporting in company management, professionalism to maintain that the public accounting. The firm remains an option for the company, as a way to avoid conflicts of interest, and ensure the integrity of the audit process and evidence for demands for transparency on company performance (Boussaidi & Hamed, 2015). Whereas, for the banking sector, the t-test value of the BIGNit variable is 0.01 (p <0.05) and the regression coefficient of 0.027 (positive), the Institutional External Auditor has a positive and significant influence on Tax Aggressiveness. External Auditor or BIGNit is a company that uses public accounting firm BIG 4 to conduct an audit. The selection of Big 4 public accounting firm services can be caused by the reputation and international credibility of the auditor. Therefore, the appointment of Big 4 auditors is a sign for the public that the reported financial statements have high credibility. Previous studies showed that auditors in the Big 4 group had higher quality than non Big 4. As in the study by Becker et al. (1998) in Alves 2013 found that Big 4 auditors provided higher audit quality than non Big 4 auditors because Big 4 auditors had greater incentives to provide higher audit quality than non Big 4. According to Maharani (2015) and Fadhilah (2014) which stated that the size of the public accounting firm has a positive influence on tax avoidance. Based on the results of the analysis and testing that has been done shows that the size of a large public accounting firm allows the company to carry out tax aggressiveness. Large public accounting firm (BIG 4) can facilitate companies in conducting tax aggressiveness, considering public accounting firm also offers non-assurance services in the form of tax services where these opportunities can be used by companies to minimize their tax burden in order to achieve their interests. 7. The t test value of the ASSETSit variable in the service sector is 0.592 (p> 0.05), and the regression coefficient value is -0.004, it can be concluded that firm size has no significant negative influence on Tax Aggressiveness. For the multiplication sector banking, the test value t ASSETSit variable is 0.619 (p> 0.05), and the regression coefficient value is 0.003 (positive), it can be concluded that ASSETSit has a positive but not significant influence on Tax Aggressiveness. 8. The t test value of the DEBTit variable on sector services is 0.006 (p <0.05) and the regression coefficient value is 0.023, it can be concluded that leverage has a significant positive influence on Tax Aggressiveness. For the banking sector Test value t DEBTit variable is 0.389 (p> 0.05), and the regression coefficient value is 0.000119 (positive), it can be concluded that DEBTit has a positive but not significant influence on Tax Aggressiveness. 9. The t test value of the MTBVit variable in the service sector is 0.355 (p> 0.05), and the regression coefficient value is -0.006, it can be concluded that the market-to-book ratio has a negative influence but not significant on tax aggressiveness. In the banking sector, the t-test value of the MTBV variable is 0.002 (p <0.05), and the regression coefficient value is -0.014 (negative), it can be concluded that MTBV has a negative and significant influence on tax aggressiveness.
This study uses interaction tests or moderation tests. Interaction Test or moderation test is the application of multiple linear regression where the equation contains an element of interaction (multiplying two / more independent variables). The reason can be explained that the task of board independent in the company is to ensure the performance of the company in accordance with good governance in accordance with applicable regulations. But the number of board independent in each company in Indonesia is relatively small, that is, on average only 1 person so that when faced with the scope of company operations with complicated financial reports, the existence of the independent board does not have an impact on the high and low tax efficiency. The supervision carried out by a board independent on financial reporting is relatively limited because the average number is only 1 person and the presence of the board of commissioners in this company still does not have the full power to participate in corporate policy making. 2. Based on the test results, it shows that the significance value for the POLCONit*LBSIZEit variable in the service sector is 0.576 (p> 0.05) and the significance value for the POLCONit*LBSIZEit variable in the banking sector is 0.396 (p> 0.05). This shows that the size of the board does not moderate the influence of political connection on tax aggressiveness. In this study, it was found that the size of the board had no influence on aggressive tax in the service sector seen from the average value in the service sector which has a board of commissioners of more than 5 people. This can explain that the more members of the board of commissioners, the more difficult it will be in making a policy decision that will be made. Because the role of the board of commissioners is to assess and direct the company's strategy, monitor the implementation and make changes if needed, monitor the process of change going forward. 3. Based on the test results that show the significance value for the POLCONit * DUALITYit variable of 0.686 (p> 0.05) and for the significance value for the POLCONit * DUALITYit variable in the banking sector at 0.933 (p> 0.05). This shows that CEO duality is not able to moderate the influence of political connection on tax aggressiveness. The results of testing the data show that CEO duality is not proven to mediate the influence of political connection on tax aggressiveness. The presence or absence of CEO duality does not affect the influence of political connections on tax aggressiveness. This is because the owner who is actively involved in company management does not try to utilize his political connections to influence tax aggressiveness. It is assumed that it relates to transparency that is increasingly strong in financial reporting and because of the flow of information so that all proximity to power holders is very risky when utilized for tax aggressiveness. This condition causes owners who are actively involved in company management (CEO duality) not to try to take advantage of the influence of political counseling on tax aggressiveness.  (2017), institutional ownership does not affect the relationship between political connections to tax aggressiveness, because institutional ownership only pays attention to the company's profits received and does not think of the company's good reputation. 5. Based on the test results, it shows the significance value for the POLCONit * BIGNit variable in the service sector of 0.439 (p> 0.05) and the banking sector shows a significance value for the POLCONit * BIGNit variable of 0.879 (p> 0.05). This shows that external auditors are not able to moderate the influence of political connection on tax aggressiveness. The results of testing the data indicate that audit quality is not proven to mediate the influence of political connection on tax aggressiveness. The auditor's status does not affect the influence of political connection on tax aggressiveness. This is because public accounting firm conducts audits in accordance with the applicable guidelines and regulations and runs them as auditors, namely actions in terms of auditing financial statements so that public accounting firm does not have an influence between political connections on tax aggressiveness.

Conclusion
Based on the results of hypothesis testing and discussion, the conclusions of this study are as follows: 1. Political Connection has a positive and not significant influence directly on tax aggressiveness in the service sector company, while in the Banking sector Political Connection has a direct negative influence on tax aggressiveness. 2. Board independence has a negative and not significant influence directly on tax aggressiveness in the Services and Banking sector companies. 3. Board size has a positive and not significant influence directly on tax aggressiveness in service sector companies, whereas in the banking sector Board size has a direct negative and significant influence on tax aggressiveness. 4. CEO duality has a direct negative influence on tax aggressiveness in the service sector company, while in the banking sector CEO duality has a direct and not significant positive influence on tax aggressiveness. 5. Institutional investment has a negative and significant direct influence on tax aggressiveness in service sector companies, while in the banking sector institutional investment has a direct positive and significant influence on tax aggressiveness. 6. External auditors have a negative and not significant influence directly on tax aggressiveness in service sector companies, while in the banking sector, external auditors have a direct positive and significant influence on tax aggressiveness. 7. Firm size has a negative and not significant influence directly on tax aggressiveness in service sector companies, while in the Banking sector Firm size has a positive and not significant direct influence on tax aggressiveness. 8. Leverage has a positive and significant influence on tax aggressiveness in service sector companies, while in the banking sector leverage has a positive and not significant influence on tax aggressiveness. 9. Market-to-book ratio has a negative and not significant influence on tax aggressiveness in service sector companies, while in the Banking sector Market-to-book ratio has a direct and negative influence on tax aggressiveness. 10. Corporate governance which includes Board Independence, board size, CEO duality, institutional investors, and external auditors does not mediate the influence of political connections on tax aggressiveness in the Services and Banking sector companies.