Pii: S0965-8564(99)00047-6

The theory of induced travel demand asserts that increases in highway capacity will induce additional growth in trac. This can occur through a variety of behavioral mechanisms including mode shifts, route shifts, redistribution of trips, generation of new trips, and long run land use changes that create new trips and longer trips. The objective of this paper is to statistically test whether this e€ect exists and to empirically derive elasticity relationships between lane miles of road capacity and vehicle miles of travel (VMT). An analysis of US data on lane mileage and VMT by state is conducted. The data are disaggregated by road type (interstates, arterials, and collectors) as well as by urban and rural classi®cations. Various econometric speci®cations are tested using a ®xed e€ect cross-sectional time series model and a set of equations by road type (using ZellnerÕs seemingly unrelated regression). Lane miles are found to generally have a statistically signi®cant relationship with VMT of about 0.3±0.6 in the short run and between 0.7 and 1.0 in the long run. Elasticities are larger for models with more speci®c road types. A distributed lag model suggests a reasonable long-term lag structure. About 25% of VMT growth is estimated to be due to lane mile additions assuming historical rates of growth in road capacity. The results strongly support the hypothesis that added lane mileage can induce signi®cant additional travel. Ó 2000 Elsevier Science Ltd. All rights reserved.