Assessing the impact of new energy demonstration city policy on industrial carbon intensity using machine learning

Abstract

Industrial carbon intensity (ICI) is a key indicator for evaluating carbon dioxide emission reduction efficiency. Given that the industrial sector has historically dominated urban carbon emissions in China, its energy transition constitutes a pivotal pathway for achieving national "dual carbon" objectives. To address this imperative, the Chinese government launched a nationwide New Energy Demonstration City (NEDC) pilot program in 2014. However, its impact on ICI remains unexplored. To fill this gap, utilizing panel data from 274 Chinese cities between 2006 and 2022, we employ dual machine learning models to evaluate NEDC's influence on ICI. Our findings indicate that NEDC significantly reduces ICI, bolstered by various robustness tests. Mechanism analysis reveals that NEDC fosters inventive and improved green technology innovations, and alleviates factor market distortions of capital and labor, leading to reduced ICI. Heterogeneity analysis shows that NEDC exerts a stronger inhibitory effect on ICI in resource-based cities compared to non-resource-based ones. It effectively reduces ICI in non-old industrial base cities and large urban areas, while its impacts are insignificant in old industrial base cities as well as medium & small cities. These findings provide valuable theoretical insights and empirical evidence to guide the strategic advancement of NEDC initiatives and sustainable urban industrial development.