Type Choice Behavior of Alternative Fuel Vehicles: A Latent Class Model Approach

Abstract

Abstract This study presents the findings of modeling alternative fuel vehicle type choice behavior in the case of a hypothetical scenario of 100% increase in gas prices in Halifax, Canada. A latent class model (LCM) is developed utilizing a stated response component from the Household Mobility and Travel Survey, conducted in Halifax, Canada, in 2012-13. The study considers a comprehensive set of alternative vehicle type choices, including: Diesel Powered Vehicles, Hybrid Electric Vehicles, Plug-In Hybrid electric Vehicles, Plug-In electric Vehicles, and regular gasoline vehicles. The LCM model developed in this paper captures latent heterogeneity among the sample households by developing a flexible latent class allocation model within the LCM framework. In this paper, the LCM model assumes two latent classes, where the classes are defined using socio-demographics, accessibility, and neighborhood characteristics. The model results suggest that considerable heterogeneity exists across the two classes. For instance, presence of children in the household shows a higher probability to choose hybrid electric vehicles in class two. On the other hand, households in class one show a negative relationship. High income households show a lower likelihood of choosing alternative vehicles and exhibit a higher propensity to continue with regular gasoline vehicles. The elasticity effects suggest that significant variation in the magnitude of effects of different variables exist across the two classes, which needs to be addressed within the policies for promoting alternative fuel vehicles as alternate choice for consumers during a sudden increase in gas price.