Electric vehicle adoption and planning: The increasing importance of the built environment

Abstract

Understanding how the built environment influences electric vehicle (EV) adoption is critical for EV-related investment decisions and policy-making, but the influence of built infrastructure and regional context on encouraging EV adoption and preparing infrastructure for EVs is less understood. To address this fundamental gap, we rigorously analyzed US EV sales data from 2012 to 2019, using state-of-the art supervised machine learning techniques. Our results indicate that accounting for non-linearities is especially important for explaining EV adoption. Our results outperformed prior studies by up to 25 %, with an out-of-sample adjusted $R^2$ of 0.92. Our analysis reveals that EV adoption is most influenced by owner-occupied residences, travel costs, charging infrastructure, and home solar installations. We also find that as adoption matures, the built infrastructure and regional context become increasingly more important for understanding EV adoption, suggesting that investment in EV-supporting infrastructure should be prioritized. Our results have implications for decision-makers who aim to understand the key drivers of EV adoption nationwide and how those drivers are changing.