Nonlinear and spatiotemporal dynamics of land finance and urban expansion in China: A GeoXAI-enhanced analysis

While land finance is a recognized institutional driver of urban expansion, the nonlinear and spatially heterogeneous dynamics of this fiscal mechanism remain underexplored. This study integrates geospatial explainable artificial intelligence (GeoXAI) with econometric models to decode the spatiotemporal evolution of the land finance – urban expansion nexus across 204 Chinese cities (2005−2020). By overcoming the limitations of traditional methods, we reveal that the impact of land finance on urban expansion is profoundly nonlinear, transitioning from a mitigating effect to a powerful reinforcing driver over time. Furthermore, this fiscal stimulus exhibits stark spatial disparities, with positive spatial interactions diffusing from coastal regions to inland second-tier cities. Ultimately, this research provides a novel methodological paradigm for spatial policy evaluation and extends the theoretical understanding of land urbanization for the emerging economies. These findings offer actionable insights for implementing zoned regulations and polycentric governance to curb inefficient urban sprawl driven by fiscal incentives.