Despite the prominent role of corporate social responsibility (CSR) in today's business environment, there is substantial debate over how it affects financial performance. Advocates maintain CSR contributes to financial success. Critics argue CSR diverts resources from more profit-maximizing projects and potentially obfuscates business activity. Academic research has failed to provide clear answers. As companies continue to invest in CSR and communicate CSR information, corporate managers, investors, and creditors need to understand how CSR affects financial performance and observable measures of financial performance to make more efficient resource allocation and contracting decisions.

The current study synthesizes research across disciplines, applies an economic framework to delineate costs and benefits of CSR, and translates results into an accounting context. This provides an explicit model of how CSR affects financial performance, along with a description of how this relationship is manifested in accounting-based measures of performance. In doing so, it shows much prior research has misspecified the CSR/financial performance relationship. Many studies use concurrent measures of CSR and accounting earnings, though theoretical arguments suggest otherwise. The theoretical framework developed here suggests CSR plays a more important role in the predictability of accounting earnings, a critical area of concern to managers and users of financial statements. This framework allowed for a rich empirical analysis, integrating well-developed streams of accounting research within the CSR context.

CSR is measured along two dimensions—strengths and concerns. Results show more socially responsible firms, in terms of concerns (firms with lower levels of concerns), are associated with increased levels of accounting earnings, higher ERCs, less analyst forecast error, and less forecast dispersion. With respect to CSR strengths, more socially responsible firms are associated with lower levels of accounting earnings, higher analyst forecast error and forecast dispersion; however, evidence is at times only marginally significant.

Overall results suggest CSR may be associated with information uncertainty, consistent with the argument that more (less) socially responsible firms have less (more) uncertainty over future cash flows. This raises important empirical questions concerning CSR and firm valuation, earnings quality, corporate governance, and requirements for social and environmental disclosures.

KEYWORDS: Corporate Social Responsibility, Financial Performance, Accounting, Earnings Predictability, Information Uncertainty