Table 13. Moderating effect of political connection.
| VARIABLES | (1) | (2) | (3) |
|---|---|---|---|
| ESG | ESG | ESG | |
| PT | -0.0670*** | ||
| (-3.312) | |||
| Mayor | -0.0712*** | ||
| (-3.750) | |||
| Psecretary | -0.0013 | ||
| (-1.629) | |||
| PT*Policon | 0.0663* | ||
| (1.835) | |||
| Mayor*Policon | 0.0695* | ||
| (1.939) | |||
| Psecretary*Policon | 0.0875** | ||
| (2.326) | |||
| Policon | 0.0104 | 0.0166 | 0.0126 |
| (0.342) | (0.572) | (0.443) | |
| Size | 0.5496** | 0.5498** | 0.5529** |
| (2.043) | (2.038) | (2.047) | |
| Lev | 0.0909 | 0.0903 | 0.0911 |
| (1.356) | (1.346) | (1.358) | |
| ROA | 3.3664*** | 3.3657*** | 3.3691*** |
| (21.393) | (21.380) | (21.425) | |
| Growth | -0.0718*** | -0.0714*** | -0.0716*** |
| (-5.564) | (-5.533) | (-5.550) | |
| Indep | 0.0879 | 0.0869 | 0.0936 |
| (0.424) | (0.419) | (0.452) | |
| TobinQ | -0.0988*** | -0.0988*** | -0.0990*** |
| (-14.722) | (-14.725) | (-14.730) | |
| Duality | 0.0930*** | 0.0928*** | 0.0929*** |
| (4.646) | (4.638) | (4.641) | |
| Top10 | 0.8624*** | 0.8633*** | 0.8637*** |
| (16.317) | (16.322) | (16.319) | |
| Age | 0.0834 | 0.0837 | 0.0832 |
| (1.505) | (1.510) | (1.501) | |
| Per capita GDP | 0.0085** | 0.0085** | 0.0086** |
| (2.378) | (2.364) | (2.382) | |
| Constant | 4.6839*** | 4.6789*** | 4.6720*** |
| (8.356) | (8.349) | (8.334) | |
| Year & Firm | Yes | Yes | Yes |
| Observations | 30627 | 30627 | 30627 |
| R2 | 0.192 | 0.192 | 0.192 |
| Adj-R2 | 0.189 | 0.189 | 0.189 |
Note. This table reports the OLS regression results of the moderating effect of ownership on the relationship between political turnover and corporate ESG performance. We use Policon to measure whether the company has a strong political connection with the government, if any of the chairman and general manager of the enterprise is or was a government official, Policon equals 1, and 0 otherwise. PT*Policon, Mayor*Policon, and Psecretary*Policon are multiplication of PT and Policon, Mayor and Policon, and Psecretaery and Policon respectively. Firm-fixed effects and year-fixed effects are also included. The t values reported in parentheses are adjusted based on robust standard errors clustered by firm, where *, **, and *** denote significance levels of 0.01, 0.05, and 0.10 respectively. All variables are defined in Appendix A in S1 Table.