Does data exchange platform boost green innovation?—Empirical evidence from heavily polluting enterprises with the integration of machine learning and econometric methods

This paper investigates the impact of Data Exchange Platform (DEP) establishment on the green innovation of heavily polluting enterprises, with a focus on spatial spillover effects. Using panel data on Chinese listed firms from 2014 to 2023, we employ a hybrid framework that combines machine learning techniques, causal econometric models with instrumental variables, and spatial econometric and regression methods. The results indicate that DEP significantly promotes green innovation in host cities. The main empirically verified mechanism lies in restraining greenwashing through enhanced data transparency. However, at the spatial level, DEP shows a significant negative spatial spillover effect under the baseline spatial weight matrix, indicating a possible resource siphon effect that concentrates data and innovation resources in platform-hosting locations. Marked regional heterogeneity is observed: local effects dominate in the Beijing–Tianjin–Hebei and Pearl River Delta regions, while the Yangtze River Delta exhibits positive spillovers due to stronger regional integration. Firm-level analyses further reveal that DEP effects are strongest for firms in highly concentrated markets, small and medium-sized enterprises, and mature firms. Overall, the findings highlight data as a core production factor in green innovation and emphasize the need for spatially coordinated data governance to mitigate siphon effects and promote regional collaborative innovation.