Can green finance reduce agricultural carbon emissions? evidence from China’s green finance reform and innovation pilot zones

Background – Agriculture contributes approximately 17% of China’s greenhouse gas emissions. In 2017, China established Green Finance Reform and Innovation Pilot Zones (GFRIPZ) to promote green development, yet its effectiveness in reducing agricultural carbon emissions remains unclear. Objectives – This study examines whether GFRIPZ reduces agricultural carbon emissions, identifies transmission mechanisms, and explores regional heterogeneity in policy effectiveness. Methods – Using GFRIPZ establishment as a quasi-natural experiment, we employ a Double/Debiased Machine Learning (DML) framework with provincial panel data from 30 Chinese provinces (2011–2024, N = 420). Three algorithms (Lasso-CV, Elastic Net-CV, Random Forest) ensure robustness. Parallel trend tests, placebo tests (500 iterations), and mediation analysis validate identification and mechanisms. Results – GFRIPZ reduces agricultural carbon emissions by approximately 41.2 × 104 tons annually (10.9% reduction, p < 0.01). Three transmission mechanisms are identified: input structure optimization (19.8%), green technology innovation (15.6%), and industrial structure adjustment (11.8%), collectively explaining 47.2% of total effect. Heterogeneity analysis reveals stronger effects in eastern regions (β = −54.0, p < 0.01) than central/western regions (β = −31.4, p < 0.10), and in economically developed provinces. Conclusion – GFRIPZ effectively reduces agricultural carbon emissions, with input structure optimization as the primary channel. Policy recommendations include expanding pilot coverage, prioritizing fertilizer reduction investments, and strengthening financial infrastructure in less developed regions.

Text 126994.pdf - Published Version Available under License Creative Commons Attribution. Download (2MB) Official URL or Download Paper: https://www.frontiersin.org/articles/10.3389/fenvs...

Actions (login required)

View Item