Structural adjustment of industrial land use and economic resilience: Evidence from China's poverty-stricken counties

The cultivation of economic resilience in poverty-stricken areas is often constrained by capital shortage and development path locking. While existing studies primarily examine resilience through industrial composition and upgrading, the role of industrial land use structure as a spatially embedded and institutionally mediated dimension of economic organization remains underexplored. Based on data from 756 Chinese poverty-stricken counties (2010–2020), this study employs the shift-share model and XGBoost-SHAP method to examine the nonlinear and heterogeneous relationships between structural adjustment of industrial land use and economic resilience. The adopted methodological framework further uncovers threshold effects and nonlinear patterns that are difficult to detect using linear approaches. The results exhibit pronounced heterogeneity across land use types: a high concentration of resource-intensive industrial land is associated with lower economic resilience, reflecting potential technology lock-in; adjustments in capital-intensive industrial land exhibit an inverted U-shaped relationship with resilience; and higher shares of labor- and technology-intensive industrial land are consistently associated with stronger resilience, likely through employment expansion and knowledge spillovers. These findings demonstrate that industrial land use structure provides a complementary analytical lens by linking land allocation, factor mobility, and development path dependence. By highlighting the differentiated and policy-conditioned associations between industrial land use structure and economic resilience, this study offers insights into resilience-building strategies for poverty-stricken regions.