Electric vehicle and supply equipment adoption dynamics in the United States

In the present era of climate change concerns, plug-in electric vehicles (PEV) and their associated infrastructure are strategic investments by many governments. Investment discussions often include the dynamic supply-demand relationship between PEV adoption and charging station deployment, but rarely is it given an empirical treatment. We make use of publicly available charging station installation and PEV vehicle registration data by U.S. County between the years 2012 and 2021 to examine the phasing and causal relationship between adoption trends in these two markets. Non-linear Granger causality tests suggest a feedback relationship between PEV vehicle adoption and charging station installations. It can be further explained that when individuals adopt PEV, the demand for the charging infrastructure increases to meet the demand for PEV. Conversely, when more charging station installations become available, it encourages adoption of this technology making it convenient for individuals to have access to these charging facilities. It is observed in aggregate that PEVs are penetrating across the U.S., but the penetration rate differs across states and counties. We examine this interaction using a generalized propensity score (GPS) identification strategy for the effect of charging station installations on PEV registrations, coupled with a generalized additive model (GAM) to construct a causal dose-response curve. We observe statistically significant effects on per capita charging station rates for income and total population. We suggest a policy package that includes charging infrastructure investments but also provides wider adoption of subsidies for low-income households across all states.