Agent-based modeling of electric vehicle diffusion under the phase-out of charging infrastructure subsidies in China

Government subsidies for electric vehicle charging infrastructure (EVCI) in China have accelerated the deployment of charging stations and promoted the diffusion of electric vehicles (EVs). However, these subsidies have also imposed a substantial fiscal burden on public finances. While much of the existing literature compares different types of EVCI subsidies, few studies explore the implications of phasing out EVCI-related subsidies for government spending and EV diffusion. This paper develops an agent-based model (ABM) incorporating EVCI operator, heterogeneous EV consumers, and the government to analyze how EVCI subsidies influence EV diffusion and proposes tailored phase-out policy combinations. A key innovation of this study is the integration of private charging pile-related factors into the consumer decision-making process through a discrete choice experiment. Additionally, regional disparities in EV diffusion between urban and suburban areas under EVCI subsidies are explored, and we find that by 2030, the EV penetration rate could reach 79.78 %, with suburban EV ownership surpassing that of urban areas. While EVCI subsidies significantly influence early and mid-stage EV adoption, their effectiveness diminishes in the later stages. Implementing phase-out subsidies under current standards can reduce cumulative government spending by approximately 91 % compared to a no-phase-out scenario, with only a marginal decline of 0.05 % in EV ownership. A comparative analysis of 50 subsidy phase-out policy combinations reveals that those featuring high initial operating subsidies with low initial construction subsidies under a rapid phase-out mode are the most cost-effective. The policy recommendations proposed alleviate fiscal burdens and promote more balanced EV development between urban and suburban areas.