The Impact of Digital Government on Regional Scientific and Technological Innovation Capacity

Background and Purpose: Contemporary digital government initiatives in China face a well-documented real-world paradox: massive fiscal outlays on digital governance coexist with marked disequilibrium in regional innovation returns. Two structural mismatches define this paradox. First, local governments overwhelmingly prioritize high-visibility hardware investments such as data centers and large AI models, while neglecting deep-seated institutional reforms including cross-departmental business process reengineering and factor market liberalization. The pervasive phenomenon of “aggregated but non-interoperable data, and interoperable data left unused” reflects a severe asynchrony between rapid technological deployment and lagging institutional restructuring. Second, comparable digital investments yield vastly divergent innovation dividends across eastern, central, and western regions, with regional divergence entrenching into a rigid “higher in the east lower in the west, higher in the south lower in the north” pattern. Extant literature, largely confined to the lens of “instrumental rationality,” reduces digital government to an exogenous technological variable, leaving it unable to explain this core practical puzzle of “homogeneous inputs generating heterogeneous returns.” Moving beyond the narrow “technology-enabled governance” narrative, this study draws on the Digital-Era Governance (DEG) paradigm to investigate the actual impact of institutional restructuring on regional scientific and technological innovation capacity, aiming to provide empirical evidence to unlock the inefficiency lock-in prevalent in digital governance practices. Research Methods: This study uses 280 prefecture-level cities and above in China from 2018 to 2023 as the research sample and constructs a two-way fixed-effects model for benchmark regression analysis. To address endogeneity, the average level of digital government development in other cities within the same province is used as an instrumental variable, and the 2SLS method is employed to identify the causal effect. On this basis, a series of robustness checks are conducted, including excluding the special impact of the COVID-19 pandemic, substituting core variable measures, and decomposing the dimensions of the core explanatory variables, to ensure the reliability of the research conclusions. For mechanism identification, the Bootstrap sampling method is used to test the dual mediating effects of “digital industry agglomeration” and “resource misallocation alleviation”; furthermore, moderating effects and heterogeneity analysis models are introduced to reveal the boundary constraints of regional economic development levels and city types on the empowerment effect. Main Findings: Empirical results show that: (1) Digital government construction significantly improves regional scientific and technological innovation capacity, and this conclusion remains valid after endogeneity treatment and robustness checks. (2) Mechanism analysis demonstrates that digital government drives innovation through the dual paths of “promoting digital industry agglomeration” and “alleviating resource misallocation,” with the marginal contribution of alleviating resource misallocation being significantly higher than that of industrial agglomeration. This suggests that, in transitional economies, eliminating institutional frictions in factor mobility brings greater innovation dividends than simply building physical spatial clusters. (3) Moderating effects indicate that the higher the level of regional economic development, the stronger the innovation empowerment effect of digital government. (4) Heterogeneity analysis further reveals that the innovation dividends of digital government are significant only in non-resource-based cities, non-central cities, and large and medium-sized cities, while in resource-based cities, central cities, and small cities, the effects are systematically absorbed and not significant. Research Conclusions and Contributions: This study breaks through the ontological limitations of existing research that views digital government as a technological tool, grounding it within the DEG theoretical framework and confirming that digital government is an institutional force in the reconstruction of regional innovation ecosystems. The findings suggest to policymakers that digital government construction should promote a shift from a “technology-oriented” to an “institution-oriented” approach. The policy focus should shift from mere infrastructure expansion to the elimination of deep-seated institutional frictions, the improvement of factor allocation efficiency, and the advancement of gradients and the implementation of classified governance, all guided by regional economic foundations and heterogeneity characteristics.