我们在表3.4中的第4列和5列分别给出了以280个城市到香港的距离及金

In the fourth and fifth columns of Table 3.4, we respectively give the results of 2SLS regression using the distance from 280 cities to Hong Kong and the lagging period of the financial efficiency variable as the instrumental variables. In the regression results, the estimated coefficient of the distance from 280 cities to Hong Kong is significantly negative, which means that the instrumental variable is highly negatively correlated with the explained variable. The regression results show that the instrumental variable has little effect on the explained variable. Therefore, after using the two-stage regression to control the endogenous problem, the significant negative relationship between the instrumental variable and the green total factor productivity is robust.

In columns 4 and 5 of table 3.4, we give 2SLS regression results with the distance from 280 cities to Hong Kong and the lag period of financial efficiency variables as instrumental variables respectively. In the regression results, the estimated coefficient of the distance from 280 cities to Hong Kong is significantly negative, which means that the instrumental variable is highly negatively correlated with the explained variable. The regression results show that the instrumental variable has little effect on the explained variable. Therefore, after using two-stage regression to control the endogenous problem, the significant negative relationship between the instrumental variable and green total factor productivity is robust.

In the 4th and 5th columns of Table 3.4, we give the 2SLS regression results with the distance from 280 cities to Hong Kong and the lag period of financial efficiency variables as tool variables. In the regression results, the estimated coefficient of the distance from 280 cities to Hong Kong is significantly negative, which means that the instrumental variable is highly negatively correlated with the explained variable. The regression results show that the instrumental variable has little influence on the explained variable. Therefore, after using two-stage regression to control endogenous problems, the significant negative relationship between the instrumental variable and green total factor productivity is stable.