Abstract
To examine the leverage effects of carbon pricing on power sector decarbonization and its underlying mechanisms, in this study, a dynamic emission abatement demand model is developed that incorporates lagged effects, innovatively integrated with convergent cross-mapping (CCM) causality analysis. Leveraging daily transaction data from China's national carbon market (January 2022-January 2024) and firm-level operational data from thermal power enterprises, we systematically unravel the nonlinear incentive mechanisms of carbon pricing. The key findings include the following. First, carbon prices exert significant amplification effects through cost transmission pathways, with an average short-term elasticity coefficient of 1.78 during trading phases, indicating that a 1% price increase drives a 1.78% marginal emission reduction. Second, CCM causality tests demonstrate that historical cumulative emissions exert threefold stronger causal influence (0.63) on market trading volumes compared to incremental emissions (0.21), validating the consensus-driven emission control under China's free quota allocation regime and the synergistic efficacy of cap-and-trade mechanisms. Third, while doubling discount factors and carbon asset conversion efficiency yields proportional growth in abatement demand (<100%), a 20% improvement in market liquidity amplifies demand by up to 100%, highlighting liquidity's critical role in enhancing price signalling efficacy during later trading stages. Our findings suggest that carbon markets can effectively incentivize power sector decarbonization, yet sustained impacts require market designs that balance liquidity optimization, risk mitigation, and dynamic quota allocation to deepen market maturity. This research contributes micro-level empirical evidence and methodological innovations for enhancing carbon pricing efficacy, offering actionable insights for policymakers to refine market mechanisms and accelerate low-carbon transitions.