The Impact of China’s National Independent Innovation Demonstration Zone Policy on Provincial-Level Innovation

• Published in: Procedia computer science Vol. 266; pp. 731 - 738
• Main Author: Zhu, Meihong
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 2025
• Subjects: Double Machine Learning; Effect; Heterogeneity; National Independent Innovation Demonstration Zone; National Independent Innovation Demonstration Zone; Heterogeneity; Effect; Double Machine Learning
• ISSN: 1877-0509; 1877-0509
• DOI: 10.1016/j.procs.2025.08.092

Based on panel data encompassing 25 provinces over a 19-year period from 2005 to 2023 and employing the Double/Debiased Machine Learning (DML) framework, this work examines the impact of China’s National Independent Innovation Demonstration Zone (NIDZ) pilot policy on provincial-level innovation capability. In applying the DML model, this study accounts for the characteristics of panel data and eliminates the influence of confounding variables, along with individual and time fixed effects. Multiple robustness tests confirm that the NIDZ policy generates a significant short-term positive impact on provincial innovation capacity, which also exhibits strong persistence. Regional heterogeneity analysis indicates that the average policy effect is largest in the eastern region, although the inter-provincial variation in effects is also relatively greater. The policy effect in the western region is significantly positive and, comparatively, inter-provincial variation is smaller. Conversely, the short-term policy effect in the central region is not significant. An economic interpretation of the plausibility of these research findings is provided.