In this paper, we develop several models of electric vehicles registrations in the United States. To model registrations, we included a variety of environment, health, financial, and education variables. Data from multiple data sources at both the state-level and county-level were used. We obtained a linear regression model of 62.09% variance explained for our baseline model. We used several machine learning methods to improve the accuracy, highlight nonlinearities, and provide additional insights. The overall predictive accuracy of our Gradient Boosted Machine model is far superior to that of the linear regression model and significantly better than the other machine learning models. Consistent across the predictive models, the most important predictor variables are high school graduation rates, air quality, and solar generation of electricity. We interpret the models and discuss their implications for research and practice, highlighting opportunities for future directions.
Kamis, Arnold, "Models of Electric Vehicle Adoption in the United States" (2023). NEAIS 2023 Proceedings. 15.