Using a sample of 22,389 US firm-year observations over the period from 1991 to 2012, we find that high CSR firms pay more dividends than low CSR firms. This is consistent with our expectation that socially responsible firms may use the dividend policy to manage the agency problems related to overinvestment in CSR. The analysis of individual components of CSR provides strong support for this main finding: five of the six dimensions used in the analysis are also associated with high dividend payout, namely, corporate governance, community, diversity, employee relations, environment. Furthermore, by analyzing the stability of dividend payout, we find that socially irresponsible firms adjust dividends quicker than do socially responsible firms: dividend payout is more stable in high CSR firms than in low CSR firms. Additional results show that firms involved in two controversial activities: military and alcohol are associated with low dividend payout which is likely due to the high cost of external funding for these firms as highlighted by Goss and Roberts (2011). Our findings are robust to alternative assumptions and model specifications, alternative measures of dividend payout, additional control variables, and several approaches to address endogeneity and selection bias issues.Keywords: Corporate social responsibility, dividend policy, agency theory JEL classification: G32, M14 *PhD Candidate in Finance -Grenoble University -CERAG UMR CNRS 5820 -BP 47, 38040 Grenoble Cedex 9, France -Mohammed.benlemlih@upmf-grenoble.fr.
2
Why do socially responsible firms pay more dividends?Abstract Using a sample of 22,389 US firm-year observations over the period from 1991 to 2012, we find that high CSR firms pay more dividends than low CSR firms. This is consistent with our expectation that socially responsible firms may use the dividend policy to manage the agency problems related to overinvestment in CSR. The analysis of individual components of CSR provides strong support for this main finding: five of the six dimensions used in the analysis are also associated with high dividend payout, namely, corporate governance, community, diversity, employee relations, environment. Furthermore, by analyzing the stability of dividend payout, we find that socially irresponsible firms adjust dividends quicker than do socially responsible firms: dividend payout is more stable in high CSR firms than in low CSR firms. Additional results show that firms involved in two controversial activities: military and alcohol are associated with low dividend payout which is likely due to the high cost of external funding for these firms as highlighted by Goss and Roberts (2011). Our findings are robust to alternative assumptions and model specifications, alternative measures of dividend payout, additional control variables, and several approaches to address endogeneity and selection bias issues.