Life Course-oriented Approach of Modeling Vehicle Ownership State and Vehicle Type Choice Using Panel Data

This study presents a biographical research approach of modeling households’ vehicle ownership states and vehicle type choice behavior, using data from a retrospective survey. A random parameter accelerated hazard-based model is developed to investigate vehicle ownership states. One of the unique features of the study is that it evaluates the determinants of termination probability of ‘no-car ownership states’ and subsequent ‘transient states’, explicitly examining the effects of life-cycle events. Model results suggest that a birth, member move-in, and addition of a job in the household accelerate the termination of ‘no-car state’, leading to the first car ownership. In contrast, those life-cycle events delay transaction events for the transient owners. The paper also examines vehicle type choice for both types of households using a random parameter logit model that accounts for the panel effects due to repeated choices. The results reveal that vehicle attributes, including space and performance factors identified by a principal component analysis, are strong predictors for vehicle type choice. However, life stages and life cycle events also significantly influence the choice of vehicle. For example, the formation of a new household substantially increases the probability of choosing compact and mid-size cars by the first-time owners. A birth of a child increases the preference for SUVs and vans among transient vehicle owners. The study also confirms considerable preference differences among first-time owners and transient owners with respect to other variables. Therefore, a longitudinal approach of modeling vehicle ownership is necessary to comprehensively assess behavioral dynamics, and adopt policies to influence behavioral changes on a longer-term basis.