Investigating Nonlinear Causal Effects and Determinants of Urban Rail Transit on Commercial Vitality

As gateways to urban rail transit networks, stations not only facilitate population mobility but also stimulate adjacent commercial development and enhance commercial vitality. Nevertheless, existing research has not yet elucidated this mechanism comprehensively, with most studies remaining at the level of correlation analysis rather than rigorously examining the causal link between station openings and improvements in commercial vitality. Therefore, this study leverages diverse socioeconomic and transportation data to explore the causal impact of opening stations on commercial vitality in the surrounding areas. First, the propensity score matching method was employed to address potential confounding bias. Additionally, we develop a Causal XGBoost algorithm, integrating geographic weights and time-fixed effects to capture spatiotemporal dependencies, thus analyzing the heterogeneity of causal effects. Finally, we employ the Shapley additive explanations (SHAP) tool to enhance result interpretability. An empirical case study in Shanghai, China, demonstrates that station opening has positive causal effects on commercial vitality, with effects varying spatially across a 1,500 to 2,000 m range. This causal relationship is evident in the significant enhancement of positive effects for population density, Gross Domestic Product (GDP), and road network density on regional commerce driven by station openings. Notably, station opening enhances commercial vitality only when located in a sufficient population base area; otherwise it exerts a negative impact. Furthermore, GDP and road network density enhance commercial vitality more strongly beyond 2,000 m from stations. These results offer insights into transportation planning for urban economic development, emphasizing strategies to improve the efficiency and sustainability of urban rail transit.