Bechetti, Ciciretti, and Hasan (2007)

Bechetti, Leonardo, Rocco Ciciretti, and Iftekhar Hasan. "Corporate Social Responsibility and Shareholder's Value: An Event Study Analysis." Federal Reserve Bank of Atlanta Working Paper, April 2007.

This strong event study reviews the returns of stocks added to and deleted from the Domini Social Index for the 1990-2004 time period, and finds positive abnormal returns for companies added to the index, and negative ones for deleted firms.

The sample includes 327 events covering 278 firms (27 firms were both added and deleted during the sample period). After eliminating changes due to mergers, buyouts, and other non-social factors, the number of events analyzed drops to 263.

The study tests several hypotheses:
1) That the market impact of additions and deletions will have risen over time. The authors argue that, as social funds have grown in size and influence, changes in the composition of the index are more likely to reflected in the market. They find evidence that this is the case, showing a statistically significant positive trend in "three out of four combinations of event windows and estimation periods considered."

2) That deletion from the index will cause negative abnormal returns. A cross-sectional regression analysis demonstrates that "deletion from the Domini index has a significant and negative effect around the announcement date for all the different event windows considered." The deletion effect was significant in the 1999-2004 time period, but is not observed in the 1990-1998 time period. The effect persisted, however, "even after controlling for concurring financial distress shocks and stock market seasonality."

3) That addition to the index will cause positive abnormal returns. The event study methodology failed to show a positive impact here. The authors attribute this to the fact that companies deleted from the index are usually deleted as the result of a notably negative CSR event. With additions, however, "the change in [C]SR of the observed stock anticipates and does not coincide with the event of entry into the index. Furthermore, the concurring shock rationale does not apply to entry events."

The authors argue that their results support the idea that social investors' influence in the marketplace is growing. They comment that "these findings, when considered together, suggest that the penalty for exit from social responsibility might depend more on the reaction of ethically screened funds than from an expected show on shareholders' value."