Essays on gasoline demand, gasoline taxes and consumer behavior



Journal Title

Journal ISSN

Volume Title



This study includes three empirical essays that use economic theories and a variety of econometrics techniques to explore the behavioral responses of households and individuals to the innovations and technologies. The first essay uses the endogenous switching regression model with data from the 2009 National Household Travel Survey (NHTS) and finds that households with hybrid vehicles respond more strongly to increases in gasoline prices (with an estimated elasticity of -0.16) than do households with non-hybrid vehicles (with a price elasticity of -0.04). This disparity in elasticity is verified using regular OLS, the instrumental variable method and the propensity scoring match technique, and is analyzed by decomposing the elasticity of traditional demand for gasoline into the elasticity of demand for vehicle miles traveled (VMT) and the elasticity of demand for household composite fuel efficiency (MPG). This analysis reveals that the change in MPG of non-hybrid-vehicle households accounts for 4% of the change in demand for gasoline, while the change in MPG of hybrid-vehicle households accounts for 44%, perhaps because members of hybrid-vehicle households can bring up their household’s composite MPG when gasoline prices increase by either switching vehicles or sharing their hybrid vehicle. These findings imply that the increased use of hybrid vehicles could be one cause of the current decrease in gasoline tax revenue and that a VMT tax might be a feasible solution. The second essay examines in three stages the impact on the distributional effects of substituting a VMT tax for Texas’s gasoline tax. Using a discrete-continuous choice (DCC) model and the 2009 NHTS, the first stage presents an integrated analysis of households’ decisions about car ownership and use. The second stage estimates the price elasticities of the demand for VMT conditional on households’ vehicle decisions. The third stage constructs four VMT tax scenarios: a revenue-neutral tax, a conservative tax, a revenue-increasing tax, and an efficient-vehicle-encouraging tax. Results indicate that while the design of a VMT tax could certainly address concerns regarding distributional effects, gas tax revenue, and hybrid vehicle purchases, it would likely be unable to address the concerns. The third essay examines the relationship between online purchasing and shopping-related travel behavior. Using data from the 2009 NHTS, this essay explores whether online purchasing relates to higher or lower in-store shopping and examines how online purchasing relates to vehicle miles traveled and the demand for gasoline. Results indicate that online purchasing significantly relates to increased in-store shopping, shopping-related VMT, and gasoline use. Results also show that this relationship is influenced by how frequently shoppers access the internet and differs across states. These results could guide policy makers in travel demand management, congestion mitigation and even environment pollution from vehicles.



Hybrid vehicle, Gasoline demand, vehicle miles traveled (VMT), Switching regression, Propensity Scoring Match, National Household Travel Survey (NHTS), Gasoline Tax, vehicle miles traveled (VMT), Discrete-Continuous Choice, Simulation, Online purchase, Shopping-related travel