The Impact of the green Finance Reform and Innovation Pilot Zone on the Low-Carbon Transformation of High-Energy-Consuming Enterprises
Whether green finance policies can drive the low-carbon transformation of high-energy-consuming industries is a crucial issue under the "dual carbon" goals. This paper uses the green finance reform and innovation pilot zones established in batches since 2017 as a quasi-natural experiment. Based on panel data from 814 listed companies in six major high-energy-consuming industries in the A-share market from 2011 to 2024, we employ a multi-period difference-in-differences (DID) approach to examine the impact of green finance policies on the environmental protection investment intensity of high-energy-consuming enterprises. Both the baseline regression and the robust estimation using staggered DID (Callaway-Sant'Anna, Sun-Abraham) fail to find a significant average treatment effect. However, heterogeneity analysis reveals a clear differentiated picture: the policy significantly enhances the environmental protection investment intensity in the electricity and heat industry (+1.19 percentage points, p=0.037), but has a significant negative effect on the non-metallic mineral products industry (-0.45 percentage points, p=0.027). Large-scale enterprises exhibit a positive marginal effect at the 10% statistical level (p=0.063), while small-scale enterprises show a negative marginal effect at the 10% statistical level (p=0.057), suggesting a potential "Matthew effect" in the policy. The policy synergy test indicates that the environmental protection investment promotion effect is significantly stronger in enterprises classified as heavy pollution (interaction term p=0.013). Placebo tests and time counterfactual tests support the robustness of the baseline conclusions. The findings of this paper suggest that the effects of green finance policies are highly heterogeneous. Policy design needs to pay attention to industry adaptation and differences in enterprise size to avoid unintended consequences that may arise from a "one-size-fits-all" approach.