Causal Effects and Spatial Spillovers of Agricultural Talent Policies on Urban–Rural Integration in China: A Double Machine Learning Approach

The dual urban–rural structure severely restricts sustainable and equitable development in China. While agricultural talent policies aim to break down these barriers, their actual net effects and spatial spillovers remain unclear due to the limitations of traditional linear models in handling complex, high-dimensional confounding factors (the research gap). The objective of this study is to accurately identify the causal impacts of these policies on urban–rural integration across 31 major Chinese cities from 2011 to 2024. We utilized text mining to construct a continuous policy intensity index and established a multidimensional integration evaluation system. Crucially, a double machine learning (DML) approach was employed to isolate the net policy effects. The results demonstrate the following: (1) Overcoming the negative bias of classical methods, the DML reveals a significant positive causal effect of talent policies on local urban–rural integration. (2) However, strong policies in central cities create significant negative spatial spillovers (siphon effects) on neighboring peripheral regions, exacerbating spatial injustice. (3) Feature analysis indicates that agricultural productivity is the primary driver of integration, while policies serve as essential catalysts. These findings emphasize that sustainable urban–rural integration requires a shift from zero-sum local competition to coordinated regional talent governance.