Measuring the Spatial Heterogeneity in Environmental Externalities from Driving: A Comparison of Gasoline and Electric Vehicles∗

Electric vehicles offer the promise of reduced environmental externalities relative to their gasoline counterparts. We determine the spatial heterogeneity in these externalities and evaluate several spatially-differentiated policies to correct them. To do this, we combine a discrete-choice model of new vehicle purchases, an econometric analysis of the electric power industry, and the AP2 air pollution model. We find three main insights. First, there is considerable spatial variation in the environmental benefit of electric cars, ranging from a positive $2144 in California to a negative $2607 in North Dakota. Second, the vast majority of environmental externalities from driving an electric car in one place are exported to other places, implying that electric cars may be subsidized locally, even though they may lead to negative environmental benefits overall. Third, spatially differentiated policies can raise welfare, but the effect is much stronger for taxes on miles driven than for subsidies on vehicle purchases. JEL Codes: Q48, Q53 Keywords— electric vehicles, spatial heterogeneity, air pollution, subsidy policy