Abstract
This paper relies on generalized structural equation modeling (SEM) to explore the relationship between land use, gasoline prices, and driving. The authors analyze data from the Southern California subsample of the 2009 National Household Travel Survey (NHTS), which has a quasi-experimental nature thanks to large exogenous variations in gasoline prices during the conduct of the NHTS. The authorsâ?? joint models of residential urban form, vehicle efficiency choice, and vehicle use account for residential self-selection and endogeneity of vehicle preferences in order to explain vehicle miles traveled (VMT) for both work and non-work trips. Residential urban form is treated as a latent construct that reflects variables such as population density, land use diversity and distance to employment centers. The authorsâ?? results suggest that in the short run, households drive 0.15% less for non-work trips when gas prices increase by 1%, while work trips are not responsive to gasoline price changes. Moreover, the direct effect of residential urban form on driving is statistically significant for total and non-work VMT, but it has no impact on work trips. The authors also find that owners of more fuel efficient vehicles tend to be more educated, Asian and younger (under 30). Moreover, households in low density neighborhoods are more likely to have a higher income, to be older than 65 and Caucasian. Finally, the results show that accounting for the nature of trips is important for understanding the short term price elasticity of travel.