Research on spatial differentiation, source decomposition and convergence of agricultural carbon emission efficiency in China

To advance China’s carbon neutrality goals, this study introduces a novel integrated framework to analyze agricultural carbon emission efficiency with unprecedented granularity. Moving beyond traditional provincial-level analyses, we utilize panel data from 266 prefecture-level cities between 2003 and 2023. Our primary innovation lies in the comprehensive methodology, which combines an improved global dynamic super-efficiency SBM model for robust efficiency measurement, the Dagum Gini coefficient for precise decomposition of regional disparities, and spatial convergence models to investigate dynamic trends. The empirical results reveal several key insights: (1) A counter-intuitive spatial pattern of “high efficiency in the west, low in the east” is identified, with inter-regional gaps being the primary source of overall disparity. (2) Kernel density estimates uncover a clear trend of multipolar polarization, indicating widening inequalities. (3) A crucial finding is the paradox of convergence: while no σ-convergence exists, significant absolute and conditional β-convergence is confirmed. By providing a multi-dimensional and city-level perspective, this research offers a more scientifically robust basis for designing targeted, region-specific policies to foster coordinated and low-carbon agricultural development.