The spatial network structure and drivers of provincial carbon emission reduction capacity in China: An indicator-based analysis

• A novel multi-dimensional indicator system is proposed to evaluate CERC. • China’s provincial CERC network exhibits a core-periphery structure with four functional plates. • Network structure is driven by both homophily (e.g., economy, tech) and complementarity (e.g., carbon market). • Analysis identifies eastern provinces as core hubs and central regions as critical brokers. • Findings inform differentiated regional policies for synergistic carbon mitigation in China. Measuring carbon emission reduction capacity (CERC) is a prerequisite for formulating synergistic decarbonization policies; however, the existing literature lacks a holistic indicator framework to capture the multidimensional nature of CERC. It thus cannot fully describe the complex spatial network structure of inter-regional interactions. To bridge this knowledge gap, this study develops a novel indicator system encompassing seven dimensions, namely, policy, society, market, technology, industry, energy, and emission control, to evaluate the CERC of 30 Chinese provinces from 2013 to 2022. We constructed a spatial correlation network of provincial CERC utilizing a modified gravity model that incorporates both geographical and socioeconomic distances. Social network analysis (SNA) was then employed to investigate the overall and individual network characteristics, as well as the spatial clustering patterns through block model analysis. The quadratic assignment procedure (QAP) method was applied to decipher the driving factors behind the network structure. The results showed that (1) the network connectivity strengthens over time, exhibiting a clear “core-periphery” structure that evolves towards a more polycentric system, with developed eastern provinces (e.g., Shanghai and Jiangsu) at the core. (2) All of the provinces can be classified into four functionally distinct plates, namely, bidirectional spillover, primary inflow, net spillover, and agent plates, that align with the Hu Line, revealing a profound functional division of labor. (3) QAP regression indicates that the network formation is dually driven by both homophily (e.g., similarities in economic development and technology) and complementarity (e.g., differences in carbon market activity and industrial structure). These findings provide a more nuanced understanding of regional carbon synergy, moving beyond the limitations of traditional spatial econometrics. The results offer a scientific basis for policymakers to design differentiated and synergistic regional carbon reduction strategies, leveraging both peer-like collaboration and complementary partnerships to accelerate China’s progress towards its dual carbon goals.