Green Finance, Stage of Economic Development, and Low-Carbon Transition: Evidence from Dynamic Panel and Threshold Regression Models Using Chinese Prefecture-Level-City Data

Green finance is widely regarded as a key policy instrument for China’s carbon neutrality goals, yet identifying its effect on low-carbon transition requires addressing green finance’s high correlation with regional development, and its nonlinear features remain insufficiently tested at a fine geographic scale. Using a balanced panel of 325 Chinese prefecture-level cities (2010–2022), this paper applies a fixed-effects model, Difference GMM, and System GMM, validated via an expanded Monte Carlo simulation, and Hansen threshold regression using the level of economic development as the threshold variable. Green finance is significantly and positively associated with low-carbon transition (2.175 under fixed effects, 3.294 under Difference GMM, both p < 0.01), with significant dynamic persistence; its direct effect is concentrated in the deployment period but propagates forward through this persistence. Two threshold values are identified (CNY 78.40 billion and CNY 160.44 billion), with effect strength rising from 0.558 to 2.180 as development advances, though the lower threshold is sensitive to the trimming proportion used, and System GMM’s insignificant coefficient reflects genuine, unresolved estimation uncertainty. These findings indicate that the green finance–low-carbon transition association is a structural process shaped by development stage, offering an evidence-based, though not precisely fixed, quantitative basis for staged, differentiated policy design.