The Mechanism of Green Finance on “Dual” Carbon Emission Goals Based on Lasso and DML Methods

Green finance faces challenges in determining its direct impact on reducing carbon emissions due to its strong endogeneity issues. This research introduces fresh insights on the relationship between green finance policies and carbon emissions through a natural experiment and the use of machine learning methods to address the endogeneity problem. Based on green finance data mainly from 1997 to 2021, the empirical results show the following: (1) Green finance development in any form can promote the reduction of carbon emission. (2) A higher level of tertiary industry and economic development stands for positive influences on the effect of green finance on reducing carbon emissions. (3) Meanwhile, regions with proper energy consumption structure performance better in suppressing carbon emissions through green finance. (4) Innovations have a positive mediating impact of green finance system on the reduction in carbon emission. According to the findings, it is recommended that decision-makers weigh the potential to align green financing with carbon emissions cuts in order to achieve economic and societal advantages. This should take into account the considerable prospects present in developing areas and the need for equilibrium among innovation, consumption, and economic levels.