Digital Economy and Agricultural Resilience: Evidence from China Using Double Machine Learning

As the foundation for stable national economic development, examining how the digital economy (DE) transforms agricultural production relations holds significant importance for improving modern agricultural industrial systems and building resilient agricultural industrial chains (AIC). Based on data from 31 Chinese provinces (2012–2022), this study evaluates China’s AIC resilience through three aspects: resistance, recovery capacity and sustainability, and investigates the impact of the DE from the perspective of production relations transformation using a double machine learning (DML) model. The results show that: (1) Regarding direct effects, the DE significantly enhances the resilience of AIC in terms of resistance, recovery capacity and sustainability, a conclusion that persists across a range of robustness and endogeneity assessments. (2) Mechanism analysis identifies a dual pathway: Horizontally, the DE promotes integration between agriculture and secondary/tertiary industries, expanding the breadth of production relations. Vertically, it optimizes the four links of value creation and transmission in the AIC, which are production, circulation, consumption, and distribution, and transforms the stages of social reproduction, reflecting the deepening of production relations. (3) Heterogeneous effects are significantly stronger in regions that are major grain producers, implement the industrial integration pilot policy, have better-educated farmers, or feature more mature agricultural insurance systems. These results underscore that digital transformation reshapes agricultural production relations through structural reorganization, not merely productivity gains. The study provides targeted policy insights for leveraging digital technologies to strengthen agricultural resilience in a differentiated and institutionally informed manner.