firms with high equity-based incent

firms with high equity-based incentives and high levels of operating risk by year across subsamples with and without religious CEOs. We find that in all but two of our sample years (2007 and 2008), the percentage of observations with high CEO equity incentive is greater among firms with non-religious CEOs than the percentage among firms with religious CEOs. The percentage of observations with high operating risk, on the other hand, is greater among firms with non-religious CEOs compared to firms with religious CEOs only in the first half of the sample period (20 0 0–20 04). In columns (1)–(3) of Panel B, we find that while firms managed by religious CEOs engage in less earnings management, the effect is more pronounced for CEOs with high equity-based incentives. In column (1), the interaction term, CEO_Relsch ∗ High_CEO_Incentive , has a coefficient of −0.010, statis- tically significant at the 5% level. The effect is economically large as well. When CEO incentives are high, the effect of CEO religiosity on earnings management is more than twice as large ( −0.017 vs. −0.007). In columns (4)–(6), we find that the negative relation between CEO_Relsch and earnings management is more pronounced for firms with greater operating risks. In column (4), the interaction term, CEO_Relsch ∗High_Risk , has a coefficient of −0.014, statistically significant at the 5% level. When a firm’s operating risk is high, the effect of CEO religiosity on earnings management is almost four times as large ( −0.019 vs. −0.005).