Metropolitan Area Heterogeneity and the Impact of Road Infrastructure Improvements on Vmt

Abstract

Past studies have found that the average elasticity of vehicle miles traveled (VMT) to the stock of highways is close to one. This result is often interpreted to mean that an increase in the stock of highway miles is likely to be accompanied by a commensurate increase in VMT, leaving congestion unaffected. In this study, we explore the heterogeneity in this elasticity due to both observed and unobserved city characteristics. We begin by using a simple model to demonstrate how cities with different initial congestion levels may respond differently to added road capacity. These differences give rise to heterogeneity in the elasticity of VMT to highway capacity. We then conduct an empirical analysis using the instrumental variable quantile regression (IVQR) model to incorporate variation in the elasticity due to the presence of unobserved differences across metropolitan statistical areas (MSAs). The IVQR results imply that expanding road capacities has a greater impact on MSAs with low levels of VMT than on MSAs with high VMT. Estimates of the elasticity using generalized quantile regression (GQR) mirror the IVQR estimates. We further simulate the mechanisms using a spatial equilibrium model with an extensive road network calibrated to the Greater Los Angeles region by treating commuting and shopping trips, mode, and route choices. We find that the elasticity of VMT to capacity—which decreases with initial VMT—in LA is 0.32. One important policy implication is that, when building more roads, the mean or median congestion level across cities remains unchanged, although it can reduce the number of cities experiencing high congestion levels.

Keywords: congestion, heterogeneity, vehicle miles traveled, travel demand, quantile regression