Efficiency over expansion: spatial spillovers and mechanism-based drivers of agricultural output in China

Abstract

Agricultural productivity remains central to food security and sustainable development, particularly in China, where persistent regional disparities, resource constraints, and input inefficiencies continue to shape production outcomes. This study therefore examines how fertilizer input (AGFU), agricultural mechanization (TPAM), and irrigation efficiency (EFIA) jointly influence agricultural output (AGOV) within a spatially interconnected production system. Grounded in endogenous growth theory and the production function framework, the study addresses key gaps related to spatial dependence, input efficiency, and heterogeneous agricultural structures. Using provincial panel data from 31 Chinese provinces over the period 2005–2023. The Spatial Durbin Model (SDM) is used to estimate both direct and spillover effects. Robustness is ensured through IV, 2SLS, and System GMM estimations to address potential endogeneity. A mediation analysis is also conducted using agricultural patent counts (AGPC) to capture the role of technological diffusion. The results indicate that AGFU, TPAM, and EFIA significantly enhance AGOV, with mechanization (TPAM) showing the most stable and consistent positive effect across specifications. Irrigation efficiency (EFIA) contributes significantly to agricultural output and generates positive spatial spillovers. At the same time, fertilizer input (AGFU) shows strong local effects but relatively weaker or negative spillover effects, suggesting diminishing returns and inefficiencies linked to overuse. Mediation results confirm that AGPC significantly transmits the effects of agricultural inputs to AGOV, highlighting the importance of technology diffusion as a key channel in agricultural productivity. Heterogeneity analysis further shows that the effects of AGFU, TPAM, and EFIA vary significantly across regions, farm operation scales, policy environments, and planting structures, confirming strong structural differences in agricultural production dynamics.