Data governance is increasingly recognized as a critical institutional instrument for facilitating energy system transformation, yet its role in reshaping the structural characteristics of urban renewable energy transitions remains insufficiently understood. This study exploits China’s National Big Data Pilot (NBDP) policy as a quasi-natural experiment and employs panel data from 291 Chinese cities spanning 2008–2022 to investigate the impact of data governance on urban renewable energy transitions. Employing a double machine learning (DML) framework, we identify the causal effects of the NBDP on multidimensional transition outcomes. The results show that the policy significantly accelerates urban renewable energy transition, with the strongest effects observed in transition breadth, followed by depth and level. Mechanism analysis reveals that renewable energy technological innovation serves as a key mediating channel, promoting substitution deepening while partially constraining short-term diversification due to path dependence. In contrast, heightened government attention to green development exerts a negative indirect effect on transition level and depth and does not mediate diversification outcomes. Further analysis indicates that the policy effects are significantly amplified in cities with more advanced digital infrastructure and higher energy demand intensity. By introducing a dual-dimensional “depth–breadth” framework, this study extends existing measures of energy transition and provides robust causal evidence on how digital governance reshapes the structural evolution of urban energy systems. • NBDP promotes urban renewable energy transition, enhancing depth & breadth. • Data governance drives transition via niche, regime, and landscape pathways. • “Digital empowerment of marginal regions” effect observed in transition. • A dual-dimension framework assesses substitution and diversification. • Big data governance breaks carbon lock-in and fosters energy complementarity.
Song et al. (Wed,) studied this question.