Robustness tests of the OLS analysis.

This table presents the results of the robustness tests of the OLS analysis. Panel A reports the results when alternative measures of institutional cross-blockholding are used as independent variables. The alternative measures of institutional cross-blockholding include the average number of peers in the same industry that are held by the same blockholders (AVGNUM), the total percentage of shares held by cross-blockholders (CROSS_OWN), the number of unique cross-blockholders (NUMCROSS), the number of peer firms in the same industry that share any common blockholder with the firm (NUMCONNECT). Panel B shows the results when applying CSP with different horizons. CSP at t+2 (CSRt+2), 2-year average of CSR (AvgCSR[t+1,t+2]), and 3-year average of CSR (AvgCSR[t+1,t+3]) are used in this analysis. Panel C reports results when the dependent variable is the adjusted CSP (AdjCSRt+1) calculated according to Deng, Kang, and Low (2013). Panel D shows the results when excluding observations during the 2008 financial crisis. Panel E shows the results when firms’ CSP is rated by Sustainalytics database. SOSICALS and AvgSOCIALS and denote the social score and 2-year average social score in Sustainalytics, respectively. The control variables are the same as those in baseline regressions. Industries are classified by Fama-French 48 industries. Standard errors are adjusted for heteroskedasticity and clustered by firm. *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively. Standard errors are shown in the parentheses.