The determinants of renewable energy production in China: A machine learning approach

Abstract

China is a global leader in renewable energy (RE) production, yet understanding the key drivers behind its RE growth remains critical for shaping effective policies. This study employs machine learning techniques to analyze the socioeconomic, environmental, and technological factors influencing RE production in China. Using a dataset spanning 2000-2022, we incorporate 17 key variables alongside newly engineered features, such as lagged values, inter-period differences, and moving averages. The findings reveal that greenhouse gas (GHG) emissions, urbanization, financial development, RE investments, and energy consumption per capita are significant drivers of China's RE expansion. Foreign direct investment (FDI) is negatively correlated with RE production, suggesting a potential pollution haven effect. This study also demonstrates that advanced machine learning models, particularly gradient boosting and random forest models, outperform traditional econometric approaches in predicting RE trends. Policy recommendations include strengthening China's carbon trading system, expanding green finance initiatives, integrating RE into urban planning, and directing FDI toward sustainable projects. These insights provide a data-driven foundation for future energy policies aimed at accelerating China's transition to a low-carbon economy.