4.2.4. Robustness Test

Replacing the regression model. This study uses the Tobit model and Poisson model. The number of dependent variables, green innovation, is generally scattered in the positive range, but there are a considerable number of zero values and a non-negative integer

skewed distribution, which is often estimated using the Tobit and Poisson models. Therefore, the Tobit and Poisson models were used to test the robustness of the relationship between corporate hiring of executives with environmental backgrounds in green innovation. Columns (1) and (3) of Table 7 show the regression results of the impact of whether firms hire executives with environmental background (hbbjdum) and the number of executives with environmental protection background (lnhbbj) on green innovation when tested using the mixed Tobit model. Columns (2) and (4) show the regression results of the impact of whether firms hire executives with environmental protection background (hbbjdum) and the number of executives with environmental protection background (lnhbbj) on green innovation tested using the Poisson model. The results show that after changing the estimation method, the independent variables of corporate hiring of executives with environmental protection backgrounds are all positive at the 1% level, confirming the robustness of the previous benchmark regression results.


**Table 7.** Robustness test of the replacement regression model.

Notes: \*\* *p* < 0.05, \*\*\* *p* < 0.01.

Replacing the measurement method of green innovation. This study classifies the application of green invention patents (GRInva) and utility patents (GRUma). As a replacement variable for the original dependent variable green innovation, the promotion effect of executives with environmental protection backgrounds on green innovation is examined. The results are shown in Table 8. The results show that the regression results of whether listed firms hire executives with environmental protection background (hbbjdum) and the number of executives hired with environmental protection background (lnhbbj) on green innovation are both significantly positive at the 5% level, indicating the robustness of the study findings.


**Table 8.** Robustness test of the replacement the measurement method of green innovation.

Notes: \*\* *p* < 0.05, \*\*\* *p* < 0.01.
