Alternative Approaches for Estimating Annual Mileage Budgets for a Multiple Discrete- Continuous Choice Model of Household Vehicle Ownership and Utilization: An Empirical Assessment

1 This paper presents an empirical comparison of the following approaches to estimate annual 2 mileage budgets for multiple discrete-continuous extreme value (MDCEV) models of household 3 vehicle ownership and utilization: (1) The log-linear regression approach to model observed total 4 annual household vehicle miles traveled (AH-VMT), (2) The stochastic frontier regression 5 approach to model latent annual vehicle mileage frontier (AH-VMF), and (3) Other approaches 6 used in the literature to assume annual household vehicle mileage budgets. For the stochastic 7 regression approach, both MDCEV and multiple discrete-continuous heteroscedastic extreme 8 value (MDCHEV) models were estimated and examined. When model predictions were 9 compared with observed distributions of vehicle ownership and utilization in a validation data 10 sample, the log-linear regression approach performed better than other approaches. However, 11 policy simulations demonstrate that the log-linear regression approach does not allow for AH12 VMT to increase or decrease due to changes in vehicle-specific attributes such as changes in fuel 13 economy. The stochastic frontier approach overcomes this limitation. Policy simulation results 14 with the stochastic frontier approach suggest that increasing fuel economy of a category of 15 vehicles increases the ownership and usage of those vehicles. But this doesn’t necessarily 16 translate into an equal decrease in usage of other household vehicles confirming previous 17 findings in literature that improvements in fuel economy tend to induce additional travel. In view 18 of policy responsiveness and prediction accuracy, we recommend using the stochastic frontier 19 regression (for estimating mileage budgets) in conjunction with the MDCHEV model for 20 discrete-continuous choice analysis of household vehicle ownership and utilization. 21 Augustin, Pinjari, Eluru, and Pendyala 2