A Statistical Analysis of Induced Travel Effects in the U.S. Mid-Atlantic Region

Abstract

We investigate the hypothesis of induced travel demand. County level data from Maryland, Virginia, North Carolina, and Washington, DC are used to estimate? fixed-effects? cross-sectional time-series models that relate travel levels, measured as daily vehicle miles of travel, to roadway capacity in lane-miles. This includes analysis of a difference (or growth) model estimated using a two-stage least squares procedure with an instrumental variable to account for simultaneity bias. Individual models for each state, a combined-state model, and a model with data from the Washington, DC/Baltimore metropolitan area are estimated. Results are generally significant and relationships robust across geographic areas and different specifications. Average elasticities of vehicle-miles of travel (VMT) with respect to lane-miles are estimated to be on the order of 0.2 to 0.6. A Granger Causality test indicates that growth in lane-miles precedes growth in VMT. Overall, the results build on recent research in this area by confirming both the range of elasticities found in other studies and the robustness of these estimates by accounting for simultaneity bias.