최근 사회적으로 관심이 급증하고 있는 기업의 사회적 책임활동(Corporate Social Responsibility, 이하 CSR)을 회계적 관점에서 살펴보면, 이는 지속가능성으로 정의되며 회계 투명성에서 비롯된다. 본 연구는 CSR을 하는 기업과 그렇지 않은 기업에 대한 회계투명성의 차이를 알아보기 위하여 두 집단간의 이익조정 및 이익지속성의 차이를 검증하였다. 추가적으로 CSR의 실시 여부뿐만 아니라 CSR의 연속성 여부에 따라서도 회계투명성에 어떠한 차이가 있는지를 검증하였다. 실증분석결과, CSR을 하는 기업이 그렇지 않은 기업보다 이익조정은 적게 하며 이익지속성은 높았다. 이러한 결과는 CSR의 내생성을 통제한 후에도 동일하게 나타났다. 또한, CSR의 연속성에 따른 회계투명성을 분석한 결과 연속적으로 CSR을 실시하는 기업은 비연속적으로 CSR을 실시하는 기업보다 이익조정은 적게 하며 이익지속성도 높았다. 이러한 결과는 CSR을 하는 기업이 그렇지 않은 기업보다 회계투명성이 높으며 CSR은 연속적으로 수행했을 때 회계투명성이 더 높음을 의미한다. 본 연구는 CSR을 회계적 관점에서 접근한 것으로서 CSR과 회계투명성간의 관계를 검증한 것이다. 지금까지 선행연구가 재무정보의 관점이나 지배구조 관점에서 회계투명성과의 관계를 검증하였다면, 본 연구는 비재무적 정보로서 CSR에 초점을 두고 회계투명성과의 관계를 실증분석 하였다는데 의의가 있다.

Corporate social responsibility (CSR) is an issue of growing interest in the academia and business world after issuance of the ISO 26000 as the international guidelines and principals for CSR by International Organization for Standard (2010). International Federation of Accountants (2006) defines CSR as “Sustainability” from the accounting point of view and stresses that it originates from accounting transparency. The existing literature offers the conflicting views on the relationship between CSR practices and accounting transparency. Some propose that CSR reflects the firm`s value and culture which in turns affect managers` reporting behaviors (Aguilera et al. 2007; Trevino 1986). They argue that managers in firms with strong CSR commitments do not have incentives to manipulate accounting numbers through earnings management since these managers are more likely to pursue ethical expectations by stakeholders rather than achieve their self-interest. In contrast, others posit that managers implement CSR activities for their self-interest to advance their careers and reputation or to cover up unethical practices such as earnings management. The underlying assumption is that these managers tend to perform CSR practices in order to avoid the scrutiny from third parties (McWilliams et al. 2006; Prior et al. 2008). Primarily motivated by the conflicting views on the relationship between CSR activities and accounting transparency, we first investigate whether accounting transparency differs between firms doing CSR (CSR firms) and other firms lacking CSR (non-CSR firms). Using absolute performance-adjusted discretionary accruals and earnings persistency as alternative constructs for accounting transparency, we examine the level of accounting transparency between the two groups of firms. Considering that managerial efforts in CSR further enhance the quality of reported earnings when CSR activities have been conducted continuously for several years, we also examine whether sustained CSR activities affect the relationship between CSR and accounting transparency. Our initial sample consists of all Korean firms listed in the KOSPI stock market. We obtain the information on CSR from the Korean Economic Justice Institute (KEJI) and the CSR ratings are labeled as the KEJI index. We define CSR firms as the firms ranked in the KEJI index and non-CSR firms as the rest of the firms not ranked in the KEJI index. We further partition CSR firms into two other groups by sustainability of CSR activities. CSR firms with four consecutive, concurrent years of conducting CSR activities are included in the continuous CSR group. The remaining firms in the CSR firms belong to the temporary CSR group. After merging the CSR data with the accounting data, our primary sample includes 4,573 firmyear observations between 2002 and 2009. Consistent with our expectations, we find that CSR firms have lower discretionary accruals and higher earnings persistence than non-CSR firms, which indicates that firms doing CSR engage in less earnings management than other firms lacking CSR practices. More importantly, when we partition CSR firms into two groups based on sustained CSR activitities, we find that the continuous CSR group has lower discretionary accruals and higher earnings persistence than the temporary CSR group. Our results suggest that CSR practices are related with greater accounting transparency and such a relation is even stronger when firms sustain CSR activities over longer periods. We also conduct a number of additional sensitivity analyses. First, considering that CSR activities could be endogenously determined, we use a two-stage least squares estimation procedure to avoid possible endogeneity problems. Second, instead of using absolute discretionary accruals and earnings persistence as accounting transparency measures, we use deferred tax and special items as alternate proxies for accounting transparency. Third, we use a matching sample of non-CSR firms instead of the universe of all non-CSR firms. Fourth, we use a continuous measure of CSR scores rather than dichotomous CSR measures. Fifth, to test the causality between CSR activities and earnings management, we examine changes in absolute discretionary accruals in the subsequent year after firms are ranked first in the KEJI index. We find that our results are robust to these additional sensitivity analyses. This study contributes to the literature in several ways. First, our study provides new evidence for the view that managerial efforts in CSR actitivies lead to more transparent accounting information when such activities have been conducted consecutively. Second, we incorporate the endogoneity issue of CSR practices by using the two-stage least squares estimation procedure. Prior literature appears to treat CSR as exogeneous even though CSR practices could be determined endogenously. Finally, we show the causal relationship between CSR activities and earning management. Using firms first ranked by KEJI during our sample period, we offer evidence that CSR activities cause managers to provide more transparent accounting information.