Synthetic control methods for estimating the effect of purchase incentives on plug-in electric vehicles sales in the United States

This study applied a synthetic control method (SCM) to investigate the impacts of state electric vehicle purchase incentives implemented in the United States between 2011 and 2018. The goal was to develop a model to replicate the impacts of PEV (plug-in electric vehicle) incentives on vehicle sales given the socio-economic characteristics of the treatment states. The goodness-of-fit between trajectories reproduced by SCM and the corresponding treatment region provides a clear means of visualizing the effectiveness of the model and results. For example, the results show that Georgia income tax credits appear to have induced 17,000 BEV (battery electric vehicle) sales between January 2011 and July 2015. The SCM analysis was also applied to other states, showing that the SCM methodology can represent PEV sales in many states in the United States. Accordingly, the modeling results suggest that incentives implemented in some U.S. states were effective in promoting PEV sales. The study also concluded that the SCM with a weighting factor between 0 and 1 (i.e., extrapolation avoided) may not be the approach of a choice for the states whose demographic/economic attributes significantly deviate from the control regions, or the states that implement "unique" programs, because there are insufficient system data for counterfactual analysis.