Spatio-temporal Evolution of the Agricultural Eco-efficiency Network and Its Multidimensional Proximity Analysis in China

As a traditional agricultural country, China has always prioritized agricultural development, and has increasingly focused on green and sustainable agricultural development. Based on the inter-provincial panel data for China from 1997 to 2019, this study divided these data into five periods according to the Five-Year Plan (FYP) of China, measured the agricultural eco-efficiency (AEE) values using the Super-SBM model, and then determined the spatial association network of the inter-provincial AEE of China using the improved gravity model. Finally, social network analysis (SNA) was used to further analyze the evolution process of AEE, and we developed a framework of how multidimensional proximity, which includes geographical, economic, technological, cognitive, and institutional proximity, made an influence on the formation of AEE spatial relation network. The findings indicated that: 1) in 1997–2019, the AEE in China was present in some spatial and temporal differences characteristics at the provincial scale, and we specifically found that national macro-regulation and policy incentives played a positive role in the long-term development of AEE. 2) The spatial correlation of AEE development among provincial regions were becoming closer and exhibits obvious spatial correlation and spillover effects. The evolution of the AEE network has clearly observable trends of hierarchization and aggregation, and the complexity of the correlation network continues to increase and exhibits spatial clustering characteristics that are dense in the east and sparse in the west. The network structure has changed from monocentric radiation to a multicentric network, and network nodes select the more advantageous nodes with which to connect. 3) Finally, the geographical proximity had a significant negative effect; the economic, technological, and institutional proximities were all observed to contribute to the AEE network formation, and cognitive proximity did not significantly influence this network formation.