Analyses of polities to reduce gasoline consumption have focused on two effects, a compositional effect on the fuel economy of the automotive fleet and a utilization effect on how much people drive. However, the literature has missed a third effect: a matching effect, in which high-utilization households are matched to fuel-efficient vehicles in equilibrium. We show that higher gas prices should lead to stronger assortative matching. Empirical estimates using US micro-level data are consistent with this hypothesis. We find the effect of a gas tax through the matching effect is larger than the compositional effect, with a $1 tax saving 1.7% of gas consumption.
↧