Corporate Social Responsibility and Financial Reporting Quality: Evidence from Korean Retail Industry

Abstract

Purpose - We investigate whether a firm's engagement in socially responsible activity affects the quality of financial reporting within the retail industry of Korean market. Recent studies argue that more socially responsible firms tend to show a better quality of financial reporting. Research design, data, and methodology - We use a variety of proxy variables related to the use of discretionary accruals and real activity manipulation to measure the quality of financial reporting. The total of environmental, social and governance score is used to represent the degree of socially responsible activity in the retail industry. We use regression models to examine whether more socially responsible firms show a higher quality of financial reporting. The sample of publicly traded Korea retail firms is analyzed from 2011 to 2016. Results - Our analysis finds supporting evidence for limited earning management via the use of discretionary accruals. We find, however, no significant relationship between the degree of social responsibility and the quality of financial reporting within chaebol affiliates unlike non-chaebol affiliates. Conclusions - Our results weakly support a better quality of financial reporting for more socially responsible firms. The results highlight the importance of firm characteristics in deciding the effect of socially responsible activity on corporate policies.

1. Introduction

Socially responsible activity of a corporation (CSR) has become one of major managerial objects recently. Thus, it has received a wide range of attentions how a firm’s CSR practice is related with various corporate policies. Retail firms care more significantly about their CSR performances because these firms face the needs of individual customers directly who are very susceptible to the image of brands and other social agenda of companies.

This work investigates how socially responsible activity is related to financial reporting quality. For example, Kim, Park, and Wier (2012) argue that firms conducting business based on trust have incentives to commit ethical behaviors including more transparent financial reporting. Socially responsible firms are more prone to satisfy the ethical demands of stakeholders by offering more transparent financial reporting.

This paper tests the validity of this hypothesis by using the sample of firms within the Korean retail industry. To do so, we select the sample of retail corporations listed in the Korean financial market, mostly in the Korean Exchange. The quality of financial reporting is measured by the use of discretionary accruals and the use of real activity manipulations (Roychowdhury, 2006). To capture the degree of engagement in CSR practices, this paper adopts the total of environmental, social and governance score (ESG score), which is graded by the Korean Corporate Governance Service. A set of cross-sectional regression models are adopted to test the above hypothesis.

Recent studies also emphasize that the effect of CSR activity on corporate policies is not equally applicable for firms with different characteristics. Specifically, it is widely known that the CSR performances within the group of family owned large conglomerates, chaebol affiliates have different implications on corporate policies such as Yoon, Lee, and Byun (2018). Thus, we also conduct sub-sample analysis based on the categorization of chaebol and non-chaebol affiliates in the Korean retail industry.

The main findings of our examination could be described as follows. Most of all, we find evidence for supporting a higher quality of financial reporting for more socially responsible firms in the Korean retail industry. Specifically, retail firms with more active CSR engagements are less likely to manage earnings via the use of discretionary accruals. Yet, our estimation results show no significant relationship between the ESG score and the proxy variables related to real activity management.

This result seems more closely related to the characteristics of retail industries. In contrast to manufacturing firms, retail firms may have rather simple production procedures, which makes the use of real activity manipulation including production cost or cash flow generation more difficult. Hence, retail firms have tendency to rely on discretionary accruals in their earning management tool. More socially responsible firms within the retail industry turn out to be less likely to use discretionary accruals to manage earnings but these firms may not have incentives to reduce real activity manipulation, which is already at a low level.

Our sub-sample analysis provides a set of novel results as well. More socially responsible firms within chaebol affiliates do not reduce abnormal discretionary accruals as well as the level of real activity manipulations. This finding contrasts the estimation results of firms not belonging to chaebol affiliates, which indicates a higher quality of financial reporting at least in terms of using discretionary accruals.

Economic environments of chaebol affiliates appear to be closely associated with the above findings. Chaebol affiliates are subject to continuous monitoring from the Korean Supervisory Service. Chaebol affiliates have limitations in their debt issuance, cross-guarantee, and intra trading. The Korean Supervisory Service also requires more strict auditing for chaebol affiliates as well. In fact, Park, Shin, Kang, and Kwon (2004) indicated that chaebol affiliates have better financial transparency compared to non-chaebol affiliates. Thus socially responsible firms within the chaebol affiliates may not have strong incentives to additionally enhance the quality of financial reporting because the overall standard of financial reporting is already high within chaebol affiliates.

This work contributes to the previous literature in a large number of aspects. Most of all, we find generally supporting evidence for the hypothesis predicting a higher quality of financial reporting quality within the firms actively engaging in CSR practices. This hypothesis is partly supported by the case of using discretionary accruals in the Korean retail industry. The evidence from real activity, at least, does not argue against the hypothesis.

Furthermore, our work emphasizes the importance of industry characteristics or firm characteristics in shaping the effect of CSR performances on corporate policies. The characteristic of retail industry in the use of real activity manipulation appears to be closely associated with the weak evidence for a higher quality of financial reporting within socially responsible firms. Furthermore, severe supervision from the Korea Supervisory Service appear to drive statistically insignificant relationship between CSR practice and the quality of financial reporting. Such a heterogenous effect of CSR practice is consistent with recent studies such as Yoon et al. (2018) and Yoon and Lee (2019).

The organization of this paper is as follows. Section 2 reviews extant literature. Section 3 summarizes the sample and our empirical strategy. Section 4 provides the main empirical estimation results. Section I5 concludes.

2. Literature Review

2.1. Related Literature

Extant studies concerned about CSR performance have mainly focused on relation between CSR and corporate performance. Many studies have recently analyzed the effect of firm’s CSR practices on the quality of financial earnings management. Specifically, Aguilera, Rupp, Williams, and Ganapathi (2007) argue that managers with strong CSR commitments firm do not have incentives to manipulate earnings. Kim et al. (2012) provide evidence that CSR practices are negatively associated with firm’s earnings management. Their study finds an enhanced transparency of financial reporting within more socially responsible firms in the U.S. financial market. Similarly, Kim and Venkatachalam (2011) find that the financial reporting quality of firms within ‘sin’-industry that engaged in tobacco, gambling, and alcohol industries is superior compared to various control groups.

However, some studies also argue a poor quality of financial reporting for more socially responsible firms. This prediction relies on the argument that CEOs implement CSR practices for their self-serving purpose to advance their careers and reputation (McWilliams & Siegel, 2000). Specifically, a CEO engages in CSR practice more actively to cover up their unethical practices including earnings management.(Hemingway & Maclagan, 2004).

In Korean market, the relationship between CSR performances and earnings management has been recently discussed. Choi and Moon (2013) argue that firms under the assesment of Korea Economic Justice Institute (KEJI) are less likely to engage in earning management and hence have a better quality of financial reporting compared to the other group of firms. Chun and Cho (2017) find that a firm’s implementation differentiation strategy is negatively associated with real activities manipulation and great CSR performances strengthen the negative relation between differentiation strategy and real activity manipulation.

This study is closely associated with a branch of literature emphasizing the firm/industry level heterogeneity in deciding the relationship between CSR and corporate policies. Moreover, Lin, Chang, and Dang (2015) argue that each individual firm confronts heterogenous CSR requirements in accordance with the characteristics of firms. Miralles-Quirós, Miralles-Quirós, and Valente Gonçalves (2018) empirically show that the valuation effect of CSR performance is significantly more substantial for environmentally sensitive industries.

In Korea, such an emphasis on firm characteristics is tightly associated with the literature related to the corporate policies of chaebol affiliates. For instance, Yoon et al. (2018) argue that the valuation effect of CSR performance differ across the categorization of chaebol and non-chaebol affiliates. This categorization is in line with the result of Yoon and Lee (2019) as well. Park et al. (2004) show that the transparency of financial reporting is higher for chaebol affiliates probably because of their continuous monitoring from the Korea Supervisory Service.

2.2 Hypothesis Development

As highlighted in Kim et al. (2012), if managers are ethical and engaged in CSR practices based on moral imperative, these managers are more likely to constrain earnings management, which results a more transparent financial reporting. Accordingly, more socially responsible firms are less apt to participate in earning management.

H1: More socially responsible retail firms show a higher quality of financial reporting.

However, as emphasized in Park et al. (2004), the sample of chaebol affiliates already has a higher quality of financial reporting due to the strict supervision from the Korean Supervisory Service. In particular, firms belonging to chaebol affiliates have restrictions in their debt issuance, cross-guarantee, and intra trading; they are also required a higher standard of financial reporting. Thus socially responsible firms within the chaebol affiliates may not have strong incentives to improve the quality of financial reporting because the overall standard of financial reporting is already higher within chaebol affiliates.

H2: More socially responsible retail firms within chaebol affiliates may not show a higher quality of financial reporting in the retail industry.

We test these two hypotheses in the sample of Korean retail industry.

3. Sample Descriptions and Empirical Methods

3.1. Data and Empirical Models

We use two different sources of datasets to test a higher quality of financial information for socially responsible firms. The first source of data is associated with a firm’s financial information to measure the quality of financial reporting from the perspective of earning managements. By following Kim et al. (2012), the annual cross-sectional industry regression model is to estimate a firm i's discretionary accruals, which calculates the abnormal discretionary accruals as the residual of the following equation:

DA = α₀ +α₁(∆REV -∆REC) +α₂(PPE) +α₃(IBIX)+ε

where DA represents discretionary accruals; REV represents revenue; REC represents receivables; PPE represents fixed capital; and IBIX represents income before extraordinary items. All variables are normalized by the value of lagged book assets.

Abnormal level of real activity manipulation variables are also constructed from the method of Kim et al. (2012) by obtaining the residuals from the following regression:

Y_t /A_t-1 = α₀ +α₁(1/A_t-1) +α₂(S_t /A_t-1) +α₃(∆S_t /A_t-1)+ε

where the set of Y variables includes operating cash flow, cost of goods sold, changes in inventory and discretionary expenses. The S variable indicates a firm’s sales. All variables are normalized by the lagged book value of asset.

The second source of data is used to proxy the degree of CSR practice in the retail industry. A firm’s CSR performances in terms of environmental, social and governance practices are evaluated by the Korean Corporate Governance Service, mostly for the firms listed in the Korean Exchange. We select the total score of environmental, social and governance practices to represent a firm’s overall level of engagements in CSR practices.

To be precise, the environmental performance assess the managerial activities related to green marketing, pollutants from production, and sales of environmentally friendly goods. The social score measures a corporation’s attitude and investment toward business ethics and sustainable growth. A firm’s governance structure is related to well-functioning board structures, the quality of auditing, shareholder protections and so on. We combine these three scores to measure the overall effectiveness of CSR performance within a corporation.

We adopt the following empirical models to test our hypothesis. The following equation illustrates our empirical models:

Quality = α +βESG+γ^′(control) +ε

where the Quality variable represents the quality of financial reporting. ESG here indicates the score of ESG, which is defined as the total of environmental, social and governance scores. The quality variable includes the absolute value of abnormal discretionary accruals (AB_DA) and the positive (P_DA) and negative values of abnormal discretionary accruals (N_DA). Kim et al. (2012) predict a significantly negative β on the ESG score for a higher quality of financial reporting with absolute and positive abnormal discretionary accruals and a significantly positive β with negative abnormal discretionary accruals.

The quality variable also contains the proxy variables for real activity manipulations. Real activity manipulation variables are the level of abnormal operating cash flows, A_CFO, the level of abnormal production costs, A_PROD, and the level of abnormal discretionary expenses, A_EXP. The production cost refers to direct production cost (cost of goods sold) and inventory stock changes. The discretionary expenses indicate the total of advertising expenses, R&D expenses, and other indirect costs. C_RAM indicates its combination which is defined as A_CFO-A_PROD+A_EXP. Kim et al. (2012) expect positive coefficients on the ESG score for the case of abnormal operating cash flow, and the case of abnormal discretionary expenses and a negative coefficient for the case of abnormal production costs.

We use the following control variables for our empirical examinations. A firm’s SIZE is the natural logarithm of a corporation’s book value of assets. The market equity to book equity ratio is the ratio between the market value of a corporation’s equity and its book value of equity. ROA is measured by operating cashflow, scaled by the lagged book value of assets after industry adjustments. BIG indicates that a firm-year observation is audited by major auditing companies. Leverage, LEV is the debt to total asset ratios. RD represents a corporation’s R&D intensity, where the measure is represented by R&D expenditure divided by current sales. ADV represents the advertising intensity for the two digit industry codes. AGE is firm age variable, which is natural logarithm of firm age.

The sample of Korean retail firms mostly listed in the Korea Exchange is analyzed because the evaluation of CSR performance is restricted to these firms. The sample period covers from 2011 to 2016. All firm-level variables are winsorized at 1% level to mitigate outlier problems in our empirical examinations. For the estimation of these model, we use the method of cross-section regression models by adopting the ordinary least square method.

3.2. Descriptive Statistics

Table 1 presents the summary statistics for the variables used in our empirical models. The ESG score and the proxy variables related to the quality of financial reporting are our main variables of interests. The table further contains summary statistics for the set of firm control variables used in our empirical models. Table 1 documents the mean of each variable with its standard deviation. It also contains the first quartile, the second quartile and the third quartile for each variable.

Table 1: Descriptive Statistics.

Table 1 documents significant variations in the proxy variables for the transparency of financial reporting. For instance, the mean of absolute value of abnormal discretionary accruals is 3.77 but its standard deviation is quite large at 4.18. Such significant variations are also observed even when we split the sample with positive and negative abnormal discretionary accruals. The proxy variables related to real activity manipulations also provide large variations as well. Such a large variation helps us examining the relationship between the ESG score and the quality of financial reporting.

Table 1 also implies that almost all of the control variables related to firm characteristics have right skewed distributions. The average value of these control variables are generally larger than the corresponding median values. If we exclude the binary variables of indicating the auditing company and equity offerings, only one exception is the logarithm of firm age variables. This implies a large proportion of relatively young firms in the retail industry, which consistent to the survivorship patterns of a company.

Table 2 provides correlation coefficients among the variables used in our examinations. The set of variables encompasses the ESG score, proxy variables related to the transparency of financial reporting, and other control variables in the empirical model. The table excludes correlations related to positive and negative abnormal discretionary accruals and combined real activity manipulation because these values are calculated from the construction of other variables of interests.

Table 2: Pairwise Correlations.

Table 2 presents a wide range of interesting findings. Most importantly, the ESG score shows negative relationship with absolute value of abnormal discretionary accruals, which implies a higher quality of financial reporting. In contrast, the correlation coefficients of ESG scores with real activity manipulation variables are inconsistent with the prediction of Kim et al. (2012). For example, the ESG score has a negative correlation with the abnormal operating cash flow and positive correlations with the abnormal production costs, both of which are not in line with the hypothesis of Kim et al. (2012). Finally, we also observe relatively weak correlations between the proxy variables related to discretionary accruals and real activity manipulations. Such weak correlations are also in line with the contrasting correlation patterns between the quality of financial reporting and the ESG score, as documented above.

4. Empirical Results

4.1 Entire Sample

Table 3 documents our estimation results for the entire firm-year observations in the Korean distribution industry. The table tests H1 from the perspective of the use of discretionary accruals as earning management tool. The use of discretionary accruals are divided into three subcategories: the absolute value of abnormal discretionary accruals, the positive abnormal discretionary accrual case, and the negative abnormal discretionary accrual case. The total of environmental, social and governance score, the ESG score is employed as the benchmark CSR measure. A higher quality of financial reporting implies negative coefficients on the absolute and positive values of abnormal discretionary accruals and positive coefficients on the negative value of abnormal discretionary accruals. The set of control variables that are depicted in the empirical model is included as well. The table reports the estimated coefficients with their corresponding standard errors from the model. For all following tables. the symbols of *, **, and *** captures the level of statistical significance at 10%, 5% and 1%, respectively.

Table 3: CSR and Discretionary Accruals

Table 3 reports a negatively significant relationship between CSR performance and the absolute value of abnormal discretionary accruals, which supports a higher quality of financial reporting for socially responsible firms. Specifically, the estimated coefficient on the ESG score is significantly negative at –1.793. Such a negative coefficient implies a more transparent financial reporting in socially responsible firms.

The next two columns indicate that CEOs in the retail industry constrain to manage earnings significantly if they need to use discretionary accruals negatively. The ESG score shows significantly positive correlation with the dependent variable, when we consider the negative abnormal discretionary accrual case. Such a significantly positive coefficient implies more transparent financial information for socially responsible firms. The coefficient on the main proxy variable (ESG score) is not significant relationship in case of the positive discretionary accrual case

The results of Table 3 generally support our hypothesis H1 predicting a higher quality of financial reporting for socially responsible firms via the traditional way of earning management. The use of discretionary accruals for earning management is generally considered as a traditional and easier way in earning management. This finding is also consistent with the U.S evidence of Kim et al. (2012) as well.

Note that the set of control variables have limited power to explain the use of discretionary accruals in our examination. For instance, except the market to book ratio, the coefficients on these firm characteristic variables are statistically insignificant. Such insignificant results rather highlight the importance of the ESG score in shaping the use of discretionary accruals.

Table 4 provides our estimation results for the entire sample of Korean distribution corporations with a different measure of financial reporting quality. The table considers the proxy variables for real activity manipulation as the dependent variables. The abnormal cashflow, abnormal production costs, abnormal discretionary expense, and their combinations are examined separately for each column. The total of environmental, social and governance score, the ESG score is employed as the performance measure of CSR practices. A better quality of financial reporting implies significantly positive coefficients on the ESG score all the empirical models, except the case of abnormal production costs; a higher quality of financial information implies a negative coefficient on the ESG score. and positive coefficients on the negative value of abnormal discretionary accruals. The set of control variables of Table 3 is included as well. Table 4 contains the estimated coefficients from the cross-sectional model and their standard errors.

Table 4: CSR and Real Activity Manipulations.

Table 4 shows statistically insignificant relationship between our measure of CSR performances and the proxy variables for real activity manipulations. For all cases of abnormal cashflow, abnormal production costs, and abnormal discretionary expenses, the estimation does not find any significant relationships. Accordingly, the combined measure of real activity manipulation does not have a significant relationship with the ESG score, either.

The results of Table 4 do not support our hypothesis H1 predicting a higher quality of financial reporting for socially responsible firms in terms of real activity manipulation. These results are also in line with the pairwise correlation coefficient reported in Table 2; the signs of correlation coefficients are all inconsistent with the prediction of Kim et al. (2012).

The contrasting results in between Table 3 and Table 4 are tightly related to the characteristics of retail firms. In contrast to ordinary manufacturing firms, retail firms tend to focus on the distribution of products, which makes the use of real activity manipulation by using production cost or cash flow generation difficult. Retail firms may have tendency to rely on the traditional way of earning management, the use of discretionary accruals. Thus, more socially responsible firms within the retail industry may focus on the use of discretionary accruals rather than real activity manipulation, which increases the transparency of financial reporting only in terms of discretionary accruals.

4.2. Subsample Analysis: Chaebol and Non-Chaebol Affiliates

Now, we conduct a subsample analysis based on the categorization of chaebol and non-chaebol affiliates. As highlighted in recent studies such as Yoon et al. (2018), firm level heterogeneity affects the effectiveness of CSR performance on various corporate policies.

Table 5 presents the examination results from our empirical model when we use the subsamples of chaebol affiliates. Table 5 examines the use of discretionary accruals as a measure for financial reporting quality. Similar to Table 3, Table 5 separately examines the absolute abnormal discretionary accruals, positive abnormal discretionary accruals, and negative abnormal discretionary accruals cases. The ESG score is also used to capture the degree of engagements in socially responsible activities. The set of control variables are identical to those of Table 3. The table shows the estimated coefficients, the standard errors, the total number of observations and the estimated R-square value. The estimation results for EO and BIG variables are missed because chaebol affiliates did not conduct equity offerings and were audited by major auditing companies during the sample periods.

Table 5: CSR and Discretionary Accruals: Chaebol Affiliates.

Table 5 clearly indicates that the retail firms within chaebol affiliates do not increase the transparency of financial reporting even in the use of discretionary accruals. For all three cases of abnormal discretionary accrual values, the coefficients on the ESG scores are not statistically significantly. The estimated values for the absolute and negative abnormal discretionary accrual cases are quite smaller than the corresponding values from the entire sample regressions documented in Table 3.

Table 6 further shows thee examination results from our empirical model when we use the subsamples of chaebol affiliates with regard to the real activity manipulations. The table considers the 4 proxy variables for real activity manipulation as the dependent variables including the abnormal cashflow, abnormal production costs, abnormal discretionary expense and their combinations. The ESG score is still used as the proxy variable for a firm’s CSR performances. The set of control variables remain the same as those of previous tables. The table includes the coefficients, corresponding standard errors, the number of observations and the adjusted R-square as well. The estimation results for EO and BIG variables are omitted because chaebol affiliates did not conduct equity offerings and were audited by major auditing companies during the sample periods.

Table 6: CSR and Real Activity Manipulations: Chaebol Affiliates.

Table 6 shows no significant relationship between a firm’s CSR performances and real activity managements within chaebol affiliates. All of the coefficients on the ESG score are not statistically significant in line with the entire sample examination reported in Table 4. Such insignificant relationships are consistent with our hypothesis, H2, which predicts limited quality improvement of financial reporting for more socially responsible firms within chaebol affiliates.

As highlighted in Yoon et al. (2018), this finding is closely associated with economic environments of chaebol affiliates. Chaebol affiliates are under continuous monitoring from the Korean Supervisory Service. For instance, firms belonging to chaebol affiliates have restrictions in their debt issuance, cross-guarantee, and intra trading. A better quality of financial reporting is required as well and a number of empirical studies verify a more transparent financial reporting within chaebol affiliates. Because these firms already have a great quality of financial reporting, more socially responsible firms within chaebol affiliates may not have strong incentives to raise the quality of financial reporting even in the case of using discretionary accruals.

Tables 7 and 8 provide the estimation results from our empirical model when we employ the subsamples of non-chaebol affiliates. Tables 7 and 8 examine the use of discretionary accruals and the use of real activity manipulation to measure finanical reporting quality. The uses of dependant variables in Tables 7 and 8 are in line with those of Tables 5 and 6. The ESG score is still used to capture the degree of engagements in socially responsible activities. The table reports the estimated coefficients and their standard errors.

Table 7: CSR and Discretionary Accruals: Non-Chaebol Affiliates

Table 7 documents a negatively significant relationship between CSR practices and the absolute value of abnormal discretionary accruals, implying more transparent financial information for socially responsible firms. The estimated coefficient on the ESG score is negatively significant at –7.520. In line with the result of entire sample, the ESG score shows significantly positive relationship with the negative abnormal discretionary accruals. This finding suggests that the results of entire sample are mainly driven by the sample of non-chaebol affiliates within the Korean retail industry.

Table 8 presents statistically insignificant relationship between the ESG scores and the proxy variables for real activity manipulations. For all of the cases of abnormal operating cashflow, abnormal cost of production, and abnormal discretionary expenses and their combinations, we are not able to find any statistically significant relationship between the ESG score and the proxy variables for real activity manipulation. These findings are consistent to the results of Tables 4 and 6, and probably capture the relatively insignificant role of real activity manipulation in the Korean retail industry.

Table 8: CSR and Real Activity Manipulations: Non-Chaebol Affiliates.

5. Conclusions

This study investigated how a firm’s CSR performance in the Korean retail industry is related to the quality of financial reporting. Specifically, Kim et al. (2012) argue that more socially responsible firms have strong incentives to improve the quality of financial reporting because they are committed to ethical behaviors to stakerholders. This paper tests whether this hypothesis applies well for the publicly traded firms in the Korean retail industry.

For this purpose, we measure transparency of financial reporting in terms of the use of abnormal discretionary accruals and real activity manipulations. The effectiveness of CSR practice is proxied by the total of environmental, social and governance scores graded from the Korean Corporate Governance Services. All other control variables in shaping the quality of financial information are constructed based on a firm’s financial statement information provided by FnGuide. The heterogenous effects of CSR practices on the quality of financial information are also evaluated in accordance with the categorization of chaebol and non-chaebol affiliates.

Our estimation results generally implies a more transparent financial reporting within socially responsible firms in the Korean retail industry. Retail firms with better CSR performances are less likely to manipulate earnings via the variation of discretionary accruals. These firms at least provide a similar quality of financial information as ordinary retail firm does in terms of real activity manipulation. However, our estimation results further confirm that chaebol affiliates do not tend to provide a greater financial reporting quality in both terms of discretionary accruals and real activity manipulation. Such insignificant results are quite distinctive from the estimation results in non-chaebol affiliates.

Our findings are closely associated with the characteristics of retail industries and chaebol affiliates. Retail firms may have simple production procedures, which restricts the use of real activity manipulation to manage earnings. Thus more socially responsible firms within the retail industry only have substantial incentives to make a higher quality of financial reporting by controlling the use of abnormal discretionary accruals, not by restricting real activity manipulations. Chaebol affiliates are subject to strict monitoring from the Korean Supervisory Service, which make them produce high quality of financial reporting. Thus even socially responsibles firms may not have incentives to additionally improve the quality of financial reporting within Chaebol affiliates.

Our work contributes to the extant literature from a number of perspectives. Most of all, we find generally supporting evidence for the hypothesis predicting a more transparent financial reporting in socially responsible firms. This finding is in line with the U.S. evidence of Kim et al. (2012) Furthermore, we also emphasize that the characteristics of retail industry and chaebol affiliates significantly affects the testing results consistent with recent studies such as Yoon et al. (2018).

However, our work still does not perfectly resolve the issue of estimation biases from the adoption of ordinary least square methods. Furthermore, our sample is limited to the firms within retail industry and we have not conducted similar analysis for other industries, while our estimation results emphasize the importance of industry characteristics in shaping the effectiveness of CSR performances on corporate policies. These topics are left for future researches.