THE IMPACT OF CORPORATE SOCIAL RESPONSIBILITY ON CEO COMPENSATION

This study aims to determine the factors that affect the CEO compensation. Variables used in this study include corporate social responsibility, return on assets, CEO ownership, independent board, tobin's q, firm size, and leverage. The approach used in this research is quantitative approach. The samples in this study are companies in the mining sector listed on the Indonesia Stock Exchange (BEI) in the period of 2012-2016. Based on the results of the test, the results obtained that the variables of corporate social responsibility, return on assets, CEO ownership, board independent, firm size, and leverage have a significant influence on CEO compensation Return on assets variables have a significant positive impact on CEO compensation. CEO ownership variables have a significant negative impact on CEO compensation. Variable board independent, firm size, and leverage have a significant positive impact on CEO compensation.

p-1412-3789 www.journalmabis.org e-2477-1783 According to Cai et al. (2011) in his study discussed The Impact of Corporate Social Responsibility on Executive Compensation. Independent variables used by firm size (FIRMSIZE), Tobin Q is measured as the ratio of asset market value to asset book value (TOBQ), leverage (LEV), return on assets (ROA), share ownership percentage (CEOOWN), Total number of directors on board (BSIZE), and board independence (BINDEP). The dependent variable used in this study is CEO compensation. The results of this study indicate that FIRMSIZE, TOBQ, LEV, ROA, BINDEP have a significant positive influence on CEO compensation. CEOOWN has a significant negative impact on CEO compensation whereas BSIZE has no significant negative impact on CEO compensation.
Study by Jian et al. (2015) stated that CSR has a significant negative impact on CEO compensation. These results showed that when CSR investments deviate from optimal levels, CEOs receive lower compensation for excessive investment in CSR. A strong corporate governance structure punishes other CEOs reducing CEO compensation if the CEO invests more in CSR. Meanwhile, Karen & Robert (2014) stated that CSR has a significant negative impact on CEO compensation. The results showed that crisis conditions have an impact on the relationship between the amount of compensation and cash compensation with CSR. The crisis has a negative and significant moderate impact for cash compensation and the most important is that during the non-crisis period, the relationship between CEO compensation and CSR is not significant (for each compensation component).
The research by Jian et al. (2015) stated that size has a significant positive impact on CEO compensation. Size indicates the size of the company, the bigger or more complex a company, the compensation received by the CEO will also increase. Cai et al. (2011) stated that size has a significant positive impact on CEO compensation. The bigger a company, the market reputation will also increase and the compensation practices received will also be greater than other companies with smaller sizes.
Study conducted by Jian et al. (2015) and Cai et al. (2011) stated that return on assets has a significant positive correlation to CEO compensation. ROA reflects how much return is generated on every rupiah invested in assets (Murhadi 2013). The higher the ROA, the better it will be for the company so that if the company's performance increases indirectly it can affect the CEO's compensation. Whereas in the research of Rekker et al. (2014) stated that return on assets has no significant positive correlation for the long term to CEO compensation. The amount of return generated on every rupiah invested in the form of assets does not always affect the company's performance in the long run.
In the research by Rekker et al. (2014) showed that leverage has no significant positive impact on CEO compensation. Leverage here shows the company's long-term debt. The compensation received by the CEO has a strong influence on the strength of the CEO itself so that leverage is not a variable that significantly correlates with the compensation received by the CEO. Meanwhile, according to Cai et al. (2011) showed that leverage has a significant positive impact on CEO compensation. Leverage shows that the debt ratio, where if the debt ratio of a large company shows that the company cannot manage the company's debt properly so that it will affect the compensation received by the CEO. In the research of Rekker et al. (2014) showed that leverage has no significant positive impact on CEO compensation. Leverage here shows the company's long-term debt. The compensation received by the CEO has a strong influence on the strength of the CEO itself so that leverage is not a variable that significantly correlates with the compensation received by the CEO. Meanwhile, according to Cai et al. (2011) showed that leverage has a significant positive impact on CEO compensation. Leverage shows that the debt ratio, where if the debt ratio of a large company shows that the company cannot manage the company's debt properly so that it will affect the compensation received by the CEO.
In the research by Rekker et al. (2014) stated that TOBQ has a significant positive influence on CEO compensation. This research shows that the alternative used in assessing the p-1412-3789 www.journalmabis.org e-2477-1783 13 value of a company is to use Tobin's Q. Tobin's Q shows that investment in assets generates profits that give a higher value than investment in expenditure, this will stimulate new investment. Cai et al. (2011) stated that TOBQ has a significant positive impact on CEO compensation. In his research, Tobin's Q is a more rigorous measure of how impactively management utilizes economic resources in its power.
In the research by Rekker et al. (2014) showed the percentage of share ownership of CEO (CEOOWN) has no significant negative influence on CEO compensation. The level of CEO ownership of the company's shares has no significant negative relationship to CEO compensation. Meanwhile, according to Cai et al. (2011) stated that CEOOWN has a significant negative impact on CEO compensation. The level of CEO ownership of the company's shares has a significant negative relationship to CEO compensation, meaning that in a company whose CEO ownership is still dominant, a CEO is paid less than the CEO in a company where public ownership is more dominant.
Study by Jian et al. (2015) stated that an independent board of directors (board independence) has a significant negative influence on CEO compensation. The board independence is a main part of the company in supervising the company's management carried out by the directors and advising the directors in running the company's operations, where it has a negative influence on CEO compensation. Meanwhile, according to Rekker et al. (2014) and Cai et al. (2011) stated that the board independence has a significant positive influence on CEO compensation. The board independence has an important role as a bridge between the interests of management and the interests of shareholders. The board independence carries out a monitoring function that is independent to the company's performance and aims to maximize returns for shareholders; if the shareholders are satisfied then the company's performance is better which eventually affecting the compensation received by the CEO.
The research by Rekker et al. (2014) stated that Tobin's Q has a significant positive impact on CEO compensation. Tobin's Q is used as a performance measurement tool that carries an important role related to financial decisions that can affect CEO compensation. Cai et al. (2011) stated that Tobin's Q has a significant positive impact on CEO compensation. If the Tobin's Q value of the company is greater than it reflects the better performance of the company where the better the company's performance will affect the compensation received by the CEO.
In this study, the objects used are all mining companies listed on the Indonesia Stock Exchange (BEI) to be able to test the independent variables used as a reference with the dependent variables that have been determined.
This study certainly has advantages and differences when compared with other or previous studies considering research on the influence of corporate social responsibility on CEO compensation is rarely examined in Indonesia. This study uses and combines journals by Jian et al. (2015) and Cai et al. (2011) by using the independent variables return on assets (ROA), debt ratio (LEV), CEO (CEOOWN) share ownership, independent board of directors (BINDEP), company size (FIRMSIZE), Tobin's Q (TOBQ), and corporate social responsibility (CSR) and the dependent variable is CEO compensation.

Research Method
The population in this study are all mining sector companies listed on the Indonesia Stock Exchange (IDX) for the period of 2012-2016 with the following criteria: (1) business entities listed on the IDX for five consecutive years, (2) business entities publish audited financial statements annually during the period of 2012 to 2016, (3) data was available for all variables needed during the period of 2012 to 2016. While the characteristics of the population are: (1) Elements: Consisting of all elements in the company's financial statements published on the IDX period of 2012-2016, (2) Sampling Unit: The sampling unit drawn from the population element and the basis for this research is total assets, return on assets, CEO p-1412-3789 www.journalmabis.org e-2477-1783 14 ownership, independent board, market value, corporate social responsibility, firm size and leverage, (3) Scope: The scope of this research is all companies whose financial statements are on the IDX period of 2012-2016, (4) Time: The period used to assess the company's financial performance is five years, namely 2012-2016. This study uses a time sample by entering data for all mining sector companies listed on the IDX for the period of 2012-2016.
This study uses multiple linear regression data processing methods to determine the impact of independent variables on the dependent variable. The variables used in this study are dependent and independent variables. The dependent variable in this study is CEO Compensation, while the independent variables are return on assets, CEO ownership, independent board, Tobin's Q, firm size, leverage and corporate social responsibility.  Table 1 shows the statistics of each research variables namely corporate social responsibility, return on assets, CEO ownership, independent board, Tobin's q, firmsize, and leverage as independent variables and CEO compensation as the dependent variable. Compensation CEO (CEO_COMP) based on Table 1 shows that on 160 observation data used has an average value of 9,294 and has a standard deviation of 0,551. CSR has an average value of 0.0023 with a standard deviation of 0.006. Return on assets based on Table 1 has an average value of 0.004 and has a standard deviation of 0.119. CEO ownership in Table 1 shows that it has an average value of 0.015 and has a standard deviation of 0.052. The Independent Board in Table 1 has an average value of 0.390 with a standard deviation of 0.102. Tobin's q in Table 1 has an average value of 1.473 with a standard deviation value of 1.372. Firmsize in Table 1 has an average value of 29,094 with a standard deviation value of 1,809. Leverage in Table 1 has an average value of 0.498 with a standard deviation value of 0.288.

Multiple Linear Regression Analysis
Multiple linear regression analysis was conducted to examine the impact of corporate social responsibility (CSR), return on assets (ROA), CEO ownership (CEOOWN), board independent (BINDEP), Tobin's q (TOBQ), firm size (SIZE), and leverage (LEV) to CEO compensation (CEO_COMP) in mining sector companies. The following is a table that shows the results of multiple linear regression analysis: This equation has a constant value of 7.884567. This means that when the independent variable is 0, the CEO compensation value will increase by 7.884567.
Corporate social responsibility variables have a regression coefficient of 1.043118. This value indicates that there is a positive influence between changes in corporate social responsibility variables and CEO compensation changes. This means that if there is an increase or decrease of 1 unit of corporate social responsibility variables, the CEO compensation variable will move in the same direction of 1.043118 assuming other variables remain the same.
Return on assets variable has a regression coefficient of 0.120577. This value indicates that there is a positive influence between changes in the variable return on assets and changes in CEO compensation. This means that if there is an increase or decrease of 1 variable unit of return on assets, then the CEO compensation variable will move in the direction of 0.120577 assuming other variables remain the same.
The CEO ownership variable has a regression coefficient value of -1.465662. This value indicates that there is a negative influence between changes in CEO ownership variables and CEO compensation changes. This means that if there is an increase or decrease of 1 unit of CEO ownership variable, then the CEO compensation variable will move opposite at 1,465662 assuming other variables remain the same.
Independent board variables have a coefficient value of 0.194173. This value indicates that there is a positive influence between independent board variable changes and CEO compensation changes. This means that if there is an increase or decrease of 1 independent variable board unit, then the CEO compensation variable will move in the direction of 0.194173 assuming other variables remain the same.
The Tobin's q variable has a regression coefficient of 0.003107. This value indicates that there is a positive influence between changes in the variable Tobin's q and changes in CEO compensation. This means that if there is an increase or decrease of 1 tobin variable's q unit, the CEO compensation variable will move in the direction of 0.003107 assuming other variables remain the same.
Variable Size has a coefficient value of 0.044877. This value indicates that there is a positive influence between changes in variable size and changes in CEO compensation. This means that if there is an increase or decrease of 1 unit of variable Size, then the CEO compensation variable will move in the same direction as 0.044877 assuming other variables remain the same.
The leverage variable has a coefficient value of 0.087287. This value indicates that there is a positive influence between changes in leverage variables and CEO compensation changes. This means that if there is an increase or decrease of 1 unit of the leverage variable, then the CEO compensation variable will move in the same direction as 0.087287 assuming other variables remain the same.
The F test is useful to find out whether the independent variables together have a significant impact on the dependent variable. To find out about this, F test can be done on multiple linear regression models with Fixed Impact Model. The results of this F test can be seen from the F-statistical probability. If the F-statistical probability value gets smaller, the stronger the independent variable will be on the dependent variable Table 2 showed that the probability of F-statistics is below 1% that is equal to 0%. Thus it can be stated that the variables of corporate social responsibility, return on assets, CEO ownership, independent board, Tobin's q, size, and leverage together have a significant impact on CEO compensation at level 1%. Table 2 shows that the coefficient of corporate social responsibility variables is 1.043118 and the significance level is 0.0861. This means that corporate social responsibility variables have a significant positive relationship to CEO compensation. This is supported by research by Rekker et al. (2014) stating that corporate social responsibility has a significant p-1412-3789 www.journalmabis.org e-2477-1783 17 positive influence on CEO compensation. This is because companies that carry out CSR activities regularly are able to build a good reputation and consumers increasingly know the company that is always doing activities that are beneficial to the community, so as to create customer loyalty and increase the demand for company products. With the increase in sales, the company's profit also increases and causes more and more investors who are interested in investing their funds in the company. This resulted in the company's stock price also increasing. Some companies with the highest CSR scores show an increase in the companies' stock price. The implementation of CSR will increase the company's value seen from the stock price and company profit (earnings) as a result of investors who invest in the company. Rekker et al. (2014) stated that with good CSR practices, it is expected that the value of the company will be properly assessed by investors. This indicates an increase in company performance and compensation given to the CEO, as a form of appreciation of the company for the CEO's performance in improving the company's performance (Rekker et al. 2014). Table 2 showed that the coefficient of the return on assets variable is 0.120577 and the significance level is 0.0001. This means that the variable return on assets has a significant positive relationship with CEO compensation. This is supported by the study of Jian et al. (2015), Cai et al. (2011), Jaiswall & Bhattacharyya (2016, Lazarides et al. (2008), and Iqbal and Shehzad (2010) stated that return on assets has a significant positive relationship on CEO compensation. Return on Asset reflects how much return is generated on every rupiah invested in assets. The higher the ROA, the better it will be for the company because it improves the company's performance (Murhadi 2013). Corporate executives have an obligation to increase ROA because this will provide an assessment of the performance of a company. Therefore, executive compensation is very much based on the rate of return on assets (Iqbal and Shehzad, 2010). When the ROA value gets higher, it reflects the company's better performance and causes the increasing in the amount of compensation given. Table 2 showed that the coefficient of CEO ownership variable is -1.465662 and the significance level is 0.0748. This means that CEO ownership variables have a significant negative relationship on CEO compensation. This is supported by research by Cai et al. (2011), Raithatha & Komera (2014), and Chen et al. (2012) who found a significant negative relationship between CEO ownership and CEO compensation. This is because the conflict of interest between managers and owners becomes greater when managerial ownership of the company becomes smaller. In this case the manager will try to maximize his interests compared to the interests of the company. Conversely, the greater the manager's ownership in the company, the more productive the manager's actions are in maximizing the value of the company, so that the contract and supervision costs will be low. Table 2 showed that the independent board variable coefficient is 0.194173 and the significance level is 0.0338. This means that the independent board variable has a significant positive relationship on CEO compensation. This is supported by the research of Rekker et al. (2014), Cai et al. (2011), andAyadi andBoujèlbène (2013) who found significant and significant relationships between independent boards and CEO compensation. This is because the existence of independent commissioners in the company can help to monitor managers because they are more neutral and do not have personal interests so that they can reduce agency costs. In addition, independent commissioners have the responsibility to proactively encourage the commissioners to carry out their duties as supervisors and advisors to the board of directors to ensure that the company has a business strategy so that the performance produced by the company will increase. With the increasing performance of the company, the company will appreciate the performance of the CEO by providing rewards in the form of greater compensation. Ayadi and Boujèlbène (2013) stated that the existence of independent commissioners in the company's board of directors can influence the company in determining the amount of compensation for the company's executives. Table 2 showed that the coefficient of Tobin's q variable q is 0.003107 and the p-1412-3789 www.journalmabis.org e-2477-1783 significance level is 0.4614. This means that Tobin's q variable has a positive and insignificant relationship on CEO compensation. This is supported by the research of Darmadi (2011), andLazarides et al. (2008) who found a positive and insignificant relationship between Tobin's q and CEO's compensation. According to Darmadi (2011) andLazariders et al. (2008), this shows that there is no relationship between stock premiums and the level of compensation received by the CEO. This is because Tobin's q describes market conditions as a measure of company value, which means that good market conditions will potentially increase stock prices rather than increase compensation received by the CEO. Table 2 showed that the coefficient of the size variable is 0.075902 and the significance level is 0.0098. This means that the size variable has a significant positive relationship on CEO compensation. This is supported by the research of Jian et al. (2015), Cai et al. (2011), and Sigler (2011), Gill (2014, Haid and Yurtoglu (2006) who found a significant positive relationship between size and CEO compensation. According to Sigler (2011), Gill et al. (2014, Haid and Yurtoglu (2006), it was revealed that relatively large companies will also have greater ability to make more payments to the company's CEOs. This is because there is a relationship that is directly proportional between the responsibilities of the CEO as the company leader and the size of the company itself. The greater the size of a company, the greater the responsibility that must be borne by the CEO as the leader of that company. In addition, a larger company size certainly has a high complexity as well so that it will affect the reward in the form of compensation received by the CEO. The greater the size of a company, the reputation of the company will also increase and the compensation received by the CEO must also be greater compared to other companies whose size is small. Table 2 showed that the coefficient of the leverage variable is 0.087287 and the significance level is 0.0205. This means that leverage variables have a significant positive relationship on CEO compensation. This is supported by research by Cai et al. (2011) who found a positive and significant relationship between leverage and CEO compensation. According to Myers (1977) in Dawar (2014) stated "Consequently, use of leverage in capital structure can reduce agency costs by regulating the choice of investment. Thus increasing leverage can have agency costs and have a positive impact on profitability and consequently firm performance". Companies that have high leverage ratios lead to supervision of high corporate activities carried out by debt holder. Therefore, the supervision provided will force managers to perform better. Companies that can manage their corporate debt properly will also increase the company's performance so this will also affect the compensation received by the CEO.
The function of R2 and adjusted-R2 values is to explain how much the dependent variable can be explained by the independent variable. R2 has a weakness, that is the value will be higher when the number of independent variables increases without paying attention whether the new independent variable is significant or not. To prevent this, the adjusted-R2 value is used, whose value will increase when the independent variable added has a significant impact on the dependent variable. Table 4.6, R2 and adjusted-R2 values showed high results which are 0.985732 and 0.981251. Thus it can be concluded that the CEO compensation dependent variable can be explained well by the independent variables of corporate social responsibility, return on assets, CEO ownership, independent board, Tobin's q, size, and leverage.

Conclusion
Based on results of the analysis of the sample companies from 2012 to 2016 on this study, it can be concluded as follows: corporate social responsibility (CSR), Return on assets (ROA), CEO ownership (CEO_OWN), independent board (BINDEP), size companies (SIZE), p-1412-3789 www.journalmabis.org e-2477-1783 19 and leverage (LEV) have significant impacts on CEO compensation. Based on the research used, there are several suggestions that researchers can provide to develop further research, including: (1) For further research, it is expected to increase the number of observations by examining other sectors listed on the Indonesia Stock Exchange (IDX) and the Stock Exchange in ASEAN countries. Thus, the samples used can represent all characteristics in the population and can reflect the overall capital market reaction. (2) This research can be used as a reference and consideration for investors to consider factors related to CEO compensation such as CSR, ROA, CEO ownership, independent board, Tobin's q, size and leverage. In addition, this research can also be used as a consideration for investors who want to invest in shares of mining sector companies in the Indonesia Stock Exchange (IDX).