Smarter Spending on Electric Vehicle Incentives: A Machine Learning Approach to Boost Rebate Efficiency Across Diverse Local Contexts

Research Poster Engineering 2025 Graduate Exhibition

Presentation by Helia Mohammadi Mavi

Exhibition Number 177

Abstract

Electric vehicles (EVs) aim to transform the automotive industry by reducing fossil fuel reliance and emissions. However, the higher initial costs deter buyers, leading to the need for financial incentives such as rebates. This research explores the effectiveness of such rebates across diverse communities, using California’s statewide Clean Vehicle Rebate Project (CVRP) as a case study. We collected the CVRP participation data (2010–2023) and supplemented it with local county-level information, including sociodemographic and economic indicators, infrastructure characteristics, and geographic attributes. Using the K-Means method, an unsupervised clustering algorithm, we grouped counties based on these local characteristics and applied the Analysis of Variance (ANOVA) to examine whether clusters differed significantly from one another in average rebate participation. Finally, a Random Forest machine learning model was employed to model and predict county-level participation. Our findings indicate that higher participation rates are prevalent in counties with higher education levels, income, GDP, and employment rates, suggesting that awareness and financial capability are critical drivers. Furthermore, counties with robust infrastructure, particularly extensive charging station networks and well-developed public transit systems, showed significantly higher participation. These results suggest that uniform statewide incentive spending may be inefficient, as certain locations may first require infrastructure development for incentives to effectively stimulate EV adoption. In conclusion, the study highlights the importance of considering local factors when designing EV rebate programs. Programs tailored to the specific characteristics and needs of diverse populations could enhance the impact of financial incentives on EV adoption, contributing to broader environmental and economic goals.

Importance

This study reveals the complexities behind electric vehicle (EV) adoption, showing that it depends on more than just financial incentives. By investigating the variability of EV rebate program effectiveness across California counties, this study guides policymakers in customizing interventions, such as targeted infrastructure investments or tailored incentive programs, to address diverse regional needs. Identifying these localized needs can accelerate EV adoption, a critical step toward reducing carbon emissions and achieving sustainability goals. Thus, the research contributes to crafting more effective policies and advancing the broader goal of sustainable transportation. It further highlights potentially more effective ways to allocate incentive funds, ensuring that resources are used efficiently to increase EV adoption.

DEI Statement

This research delves into the impacts of California’s Clean Vehicle Rebate Project (CVRP) on electric vehicle (EV) adoption across counties with varied sociodemographic and infrastructural characteristics. It reveals significant disparities in EV rebate uptake, highlighting the different needs of diverse populations. By analyzing how distinct groups respond to the rebate program, this study underscores the need to tailor transportation policies more effectively. Such customization helps all communities, regardless of economic standing or infrastructure quality, have equitable access to sustainable and clean technologies. This approach embodies a deep commitment to diversity, equity, and inclusion (DEI) principles within the realm of transportation planning and sustainable solutions.