HOW CORPORATE SOCIAL RESPONSIBILITY EFFECTS COMPANY’S FINANCIAL PERFORMANCES

A successful business must pay attention not only to how to improve the welfare of shareholders, but also the welfare of all stakeholders. This study aims to see the effect of disclosure of Corporate Social Responsibility (CSR) on corporate financial performance. Financial performance is measured using three indicators, which are return on Equity (ROE), total stock return and Tobin‟s Q. ROE represents company‟s profitability, total stock return reflects company‟s performance in the market and tobin‟s Q shows firm value. CSR is measured using the Corporate Social Disclosure Index (CSRDI) which is based on the ISO 26000 standard. By using two controlling variables, company size and leverage, the results show that there is a negative and insignificant effect between CSR and ROE ratio. Although CSR practice together with control variables also has no effect on total stock returns, but there is a significant positive effect between CSR disclosures and firm value. These results can motivate company‟s managers to include CSR as a management strategy to increase firm value by paying serious attention to stakeholders and sustainability issues.

integrate social and environmental concerns into their operations and their interactions with their stakeholders (Latofi, 2008). The 2002 World Summit on Sustainable Development encouraged all companies in the world to carry out CSR in order to create a sustainable development. The role of CSR is seen as an effort to realize good corporate governance, good corporate citizenship and good business ethics of a business entity (Amba-Rao, 1993;Anderson, Jr., 1989;Kim, 2000;and Raynard & Forstater, 2002).
On the other hand, Henderson (2001) argues that CSR cannot be generalized to every situation because there may be different interpretations of what is to be achieved.
Furthermore, when CSR activities are enforced and competitors follow suit, this can lower the overall performance of the industry. Research conducted by Rumengan (2017) also shows that there is a significant negative effect between CSR and return on equity.
In The food and beverage industry faces its own social responsibility challenges, especially for three reasons (Hartmann, 2011). First, they have a high impact and dependence on natural, human and physical resources. Second, because food is a basic human need, humans have strong views about what they consume. Third, the food industry has a unique and diverse organizational structure that a small and large companies differ in their approach to social responsibility, this creates a potential conflict regarding social responsibility involvement in the food industry.
The purpose of this study is to link CSR disclosure with the financial performances in the consumer goods industry using two control variables, namely firm size and leverage. Firm size is used as a controlling variable because previous research stated that firm size has a significant positive effect on financial performance (Tristiarini, 2013). While, leverage is used due to growing companies will need more funds that can be obtained from liabilities and equity. SCIENTIFIC JOURNAL OF REFLECTION: p-ISSN 2615-3009 Economic, Accounting, Management and Business e-ISSN 2621-3389 Vol. 4, No. 3, July 2021

Return on Equity
Return on equity (ROE) is one measure of management performance. It is one of the key aspect of company"s profitability in accounting based-measure and denotes what proportion of the earnings of corporation is generated using its own equity, that is money raised from shareholders. A higher ROE indicates a higher level of management performance (Australian Shareholders' Association, 2010). In this study, ROE is measured by dividing company"s net income after tax with company"s total equity for a year period.

Total Stock Return
The total stock return is the entire return on investment over a period. It reflects company"s market performance. The total stock return consists of capital gains (losses) and dividend yields. Capital gain (loss) is the percentage change in share price.
Dividend yield is the percentage of cash receipts (dividends) to the stock price during a certain period (Hartono, 2011).
Stock return that will be used as dependent variable in this research is realized stock return in given a year period. This ratio is obtained by summarizing capital gains (losses) with dividend yields, then, dividing it with share price at initial year period.

Tobin's Q
Tobin's Q ratio is the ratio of the market value of the company's assets (measured by the market value of outstanding shares and debt) to the replacement cost of the company's assets (Carlton and Perloff, 2005). This ratio reflects firm"s value. It requires cross-sectional differences in investment decisions and diversification decisions, the relationship between managerial equity ownership and firm value, the relationship between managerial performance and tender offer returns, investment opportunities and tender offer response, financing, dividends, and compensation policies (Wolfe and Sauaia, 2003).

Corporate Social Responsibility
The International Organization for Standardization (ISO) defines CSR as "… organizational responsibility for the impact of its decisions and activities on society and the environment, … that contributes to sustainable development, including public health and well-being, taking into account the expectations of stakeholders…." SCIENTIFIC JOURNAL OF REFLECTION: p-ISSN 2615-3009 Economic, Accounting, Management and Business e- ISSN 2621-3389 Vol. 4, No. 3, July 2021 In measuring CSR, the authors use Corporate Social Dislosure Index (CSRDI) that is based on standard of ISO 26000 about Guidance on Social Responsibility. Based on the standard, there are seven core subjects (Kritkausky & Schmidt, 2011). Each core subject includes a range of issues of social responsibility. There are 37 issues described related to CSR. Each issue will be given 1 point if disclosed and 0 point if not disclosed in the annual financial report. Then, all points are collected and divided by 37 to get the final score for each company. The following is the formula used: CSRDj,t represents disclosure of social responsibility ratio in company j at period t and Xij,t is total issues disclosed in the annual report in company j at period t,

Control Variables
Leverage is measured by the Debt to Equity Ratio (DER). While, firm size is measured by the logarithm of total assets.

METHOD
This study is a quantitative study that uses secondary data collected from annual financial reports published on the Indonesia Stock Exchange (IDX) for the period of 2016-2018. The authors use Eviews version 9.0 in testing the regression analysis. There are three types of panel data modeling which are fixed effect, common effects and random effects. The model was chosen based on the assumption that certain residual characters are random (Ariefianto, 2012). To determine which method is more appropriate to use, the several tests (the Chow test, the Lagrange Multiplier test and the Hausman test) must be conducted. In this study, hypothesis tests are performed by conducting F significance test, test of the coefficient of determination (R-Squared) and T significance test.

Sample Data
Using the purposive sampling method, the relevant samples were selected based on the specified criteria. There are 51 observations obtained that consist of 17 companies from 2016 to 2018. conclude that fixed effect method will be used in model 1 and model 3, while common effect method is more appropriate for model 2.

Normality test and Classical Assumption tests
Authors conduct normality test and classical assumption tests consisting of multicollinearity test, heteroscedasticity test and autocorrelation. Tests find that three models have passed the normality test, multicollinearity test and heteroscedasticity test. However, only model 2 that is free from autocorrelation. Using Durbin Watson value, the authors detect autocorrelation for model 1 and model 3.
According to Nachrowi and Usman (2006), in panel data, the fixed effect model is a model that does not require classical assumptions because the model is free from correlation so that autocorrelation can be ignored. This is also reinforced by Basuki (2015), not all classical assumption tests in the OLS method are used in panel data regression, only the Multicollinearity Test and Heteroscedasticity Test are needed. So in this case, the autocorrelation can be neglected for model 1 and model 3.  Table 1 found that the F-statistical probability is below 1%, so the model used has a significance level of 99% confidence. This indicates that all variables used simultaneously have an influence on the ROE variable. Next, the value of R 2 and adjusted R 2 are R 2 0.865 and 0.782, respectively. Thus, the proportion of explanatory variables in explaining the ROE variable is 78.2%.

Hypothesis Testing Model 1
The CSR variable indicates that CSR has a negative effect on ROE. It means that when the company strictly implements social responsibility, it will cause a decrease in profitability (Rumengan, 2017). According to Henderson (2001), the adoption of CSR carries a high probability of cost increases and lower performance. Managers have to take account of outside stakeholders concerns and this process is timeconsuming.
Besides, new accounting systems, monitoring and auditing is needed as the results of the adoption of CSR.

Source: Research Data, 2020
The results of the regression shows the probability value of F of 0.196375. Thus, the model used has no significance at the 99% level. This shows that all the variables used together have no effect on the total stock return variable.
Furthermore, the results show that the adjusted R 2 is 0.0361 which means that the proportion of explanatory variables in explaining the total stock return variables is only 3.61%. The authors suspect that there are other fundamental factors that have a more significant influence on total stock returns, such as earnings that determine dividends on stocks and interest rates that discount future cash payments to date (Siegel, 2008). The coefficient value indicates that CSR has a positive effect on total stock returns. However, the probability value of the CSR t-statistic generated is 0.871 so that CSR does not have a significant effect on the total stock return at a significant level of 10%. This means that there is no significant difference in total stock returns between companies that disclose more and companies that disclose less. Author assumes that there is high volatile market share so that CSR cannot be enough in giving an effect to total stock return. This is consistent with Ntim and Soobaroyen (2013) that CSR alone has weak positive effect on total stock return. The probability F value of 0.0000 means that the model used is significant at the 99% significance level. This shows that all the variables used together have an influence on the Tobin's Q variable. The adjusted R 2 value of 0.8578 indicates that the proportion of independent variables in explaining the firm value variable is 85.78%.

Hypothesis Testing Model 3
The coefficients of explanatory variables indicates that the three variables have a positive influence on firm value. However, only CSR variable has a significant effect on firm value at a significance level of 10%. This can be seen from the probability value of CSR t-statistics of 0.085, while the two control variables have a probability value of tstatistics above 10%. This evidence implies that CSR is one of the factors that increase firm value. CSR disclosure also gives a positive sign to the market that companies that implement CSR activities well indicate that the company has a good internal management system. In addition, these results can also be consistent with management theory that CSR is one of the corporate strategies used to shape the company's image and reputation (Tristiarini, 2013). For example, a company that has implemented the use of environmentally friendly materials will certainly attract attention to the public. The public will assume that the company cares about the environment so that a positive image will appear towards the company and encourage consumers to buy its products.

CONCLUSION
Model 1 examines the effect of CSR on the ROE ratio using the fixed effect method. The findings show that there is a negative effect of CSR disclosure toward company"s profitability.
Model 2 uses the common effect method to examines the effect of CSR on total stock returns. The results showed that the practice of CSR together with the control variables had no effect on the total stock return. The author assumes that there is a high trend of fluctuating market share. Thus, CSR alone is not sufficient to explain the dependent variable in this model.
The results of the Model 3 regression using the fixed effect method show that there is a significant positive influence between CSR disclosure and firm value. This can motivate company managers to include CSR as a management strategy by paying serious attention to stakeholders and sustainability issues.
To support CSR activities performed well by companies, the government needs to make clear standards and assessments of corporate social responsibility, such as clear references on how much money should be allocated for CSR activities so that it can evaluate the company's contribution to social activities and determine strict sanctions given to companies that do not carry out social responsibility.
Authors also suggest that because Indonesia data might not be relevant to other countries, further research should try to adopt the same framework in cross-country contexts. SCIENTIFIC JOURNAL OF REFLECTION: p-ISSN 2615-3009 Economic, Accounting, Management and Business e-ISSN 2621-3389 Vol. 4, No. 3, July 2021