Spatial variabilities in factors affecting electric vehicle adoption across Virginia: A county-level analysis

Adoption of electric vehicles (EV) has been increasing in recent years in the U.S. Studies have investigated the determinants of EV adoption, such as income and housing structure. However, few studies have examined the spatial variation in the effects of such factors on EV adoption rates. Using Virginia as a case, this study evaluates how the effects of factors commonly associated with EV adoption vary geographically and investigates the influence of two understudied factors — highway density and political preferences — at the county level. Using standard regression, spatial lag regression, and geographically weighted regression (GWR) models, this study assesses how highway density, percent of urban population, percent of 1-unit housing structures, commute time, percent of population 65 and older, median household income, and percent votes for the Republican candidate in 2020 affect EV adoption rates at the county level. Results show that highway density and urban environment are insignificant, and all other factors are significant based on standard and spatial lag regression models. However, GWR adds housing structure to the list of insignificant factors at the local scale, while the impacts of other significant factors vary across Virginia counties differently. Thus, local policies facilitating EV adoption may have different effectiveness levels across Virginia counties, a conclusion likely applicable to other states. The current study also ascertains the importance of commute time, income and age in affecting EV adoption, and highlights the significance of political preference, a factor that has not been assessed previously.