When Green Accounting Fails to Drive Green Energy: Institutional Quality and China’s Renewable Energy Transition

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:Provides empirical evidence on the role of institutional quality in the energy transition, using both econometric and ML methods for robustness.

🏢実務担当者:Highlights that corporate sustainability teams should consider institutional and regulatory factors when assessing renewable energy investment risks.

🏛政策担当者:Suggests that effective energy transition policy requires combining environmental accounting with strong governance and deployment-oriented R&D spending.

This study examines how institutional quality, government R&D expenditure, and SEEA-aligned environmental asset indicators are associated with renewable energy adoption in China from 1999 to 2023. It addresses the puzzle that environmental accounting signals and innovation investment do not automatically translate into renewable energy use. Annual data from the World Development Indicators and Worldwide Governance Indicators were analyzed using a hybrid empirical strategy. The Autoregressive Distributed Lag approach was applied to estimate short-run and long-run relationships, while Double Machine Learning, Random Forest, Gradient Boosting, and SHAP interpretability were used as supplementary tools for robustness and predictive importance. Given the limited sample size, machine-learning results are interpreted as orthogonalized associations and as predictive evidence rather than as definitive causal effects. ARDL results show that regulatory quality has a positive and statistically significant long-run association with renewable energy adoption. In contrast, government R&D expenditure and adjusted net savings show negative associations, suggesting that innovation spending and fiscal capacity may not support renewable energy adoption unless directed toward deployment and energy-system substitution. Energy resource depletion is statistically insignificant, while natural resource rents show a weak positive long-run association. Machine-learning results identify government R&D expenditure as the strongest predictor, although its direction remains negative. The findings indicate that China’s renewable energy transition depends less on fiscal or technological inputs alone and more on the institutional capacity to convert these inputs into adoption outcomes. The study implies that SEEA-based indicators should be integrated with regulatory mechanisms, deployment-oriented innovation policy, and outcome-based evaluation of energy transition.

• openalex https://doi.org/10.36923/ie-frontiers.v29i1.426first seen 2026-06-08 04:37:35