Urban innovation effects and transmission mechanisms of east-west interregional co-built industrial parks in China

Under China’s strategy for coordinated regional development, east-west interregional co-built parks have become an institutional vehicle for optimizing the cross-regional allocation of innovation factors, yet their innovation effects and underlying mechanisms remain underexamined. Using a city-year panel of all Chinese prefecture-level cities from 2010—2023, this study employs Double Debiased Machine Learning (DDML) within a treatment-control framework to identify the causal policy effects of park establishment on urban innovation level and quality and to examine transmission mechanisms. DDML improves robustness by debiasing treatment effect estimation with high-dimensional controls. co-built parks significantly increase both the quantity and the quality of local innovation outputs. The effects display a nonlinear gradient with respect to collaboration distance: long distance > short distance > medium distance, and are more pronounced in resource-based cities and old industrial bases. Policy guidance intensity, information transmission strength, and physical transport efficiency jointly constitute the key transmission pathway through which innovation factors are reorganized across jurisdictions. These findings suggest that effective implementation requires coordinated strengthening of policy steering, information connectivity, and transport links to reduce cross-regional frictions. Overall, this study advances the theoretical understanding of innovation enablement in cross-regional parks and provides empirical evidence for innovation-oriented spatial governance and policy design under east-west collaboration.