Search for a command to run...
To accurately reveal the real level and dynamic evolution of agricultural production efficiency across China’s eastern, central, and western regions, and address the limitation of traditional DEA models in separating environmental variables and random errors, this study employs a three-stage DEA model combined with Bootstrap and Malmquist index. Based on static measurement and dynamic decomposition of agricultural production efficiency in 15 representative provinces from 2014 to 2023, it isolates environmental and random noise interference to explore real efficiency, regional disparities, and growth drivers. The findings are as follows: (1) Methodologically, the three-stage DEA-Bootstrap model outperforms traditional DEA by effectively distinguishing management and environmental inefficiency, with significantly reduced corrected efficiency values—indicating traditional DEA overestimates comprehensive efficiency and risks misjudging regional performance, thus verifying the former’s superiority in accuracy and robustness. (2) Static efficiency exhibits the spatio-temporal pattern of “eastern leadership, western catch-up, and central steady progress” but with intra-regional differentiation: eastern provinces like Guangdong and Zhejiang owe part of their efficiency advantage to the external environment, revealing scale efficiency shortcomings; central and western provinces such as Anhui, Xinjiang, and Gansu have overestimated efficiency, with prominent gaps in real management and technical capacity after correction. (3) Dynamically, total factor productivity (TFP) grew at an average annual rate of 9.0%, driven primarily by technological progress, while technical efficiency improvement contributed weakly, with TFP fluctuations highly synchronized with technological progress. (4) Regional growth drivers differ spatially: western provinces (e.g., Xinjiang, Gansu) lead in productivity via outstanding technological progress; eastern provinces (e.g., Shandong, Jiangsu) face sluggish technological progress and weak growth; central and some western provinces (e.g., Sichuan, Hubei) demonstrate both efficiency catch-up and technological progress. The results show that China’s agricultural production efficiency is statically overestimated by environmental factors and dynamically dominated by technological progress, with regional disparities in technological progress being the key to efficiency differentiation. Policy recommendations include strengthening frontier technological innovation in the eastern region, and prioritizing management optimization and scale efficiency improvement alongside sustained technological progress in the central and western regions.