Household Vehicle Ownership by Vehicle Type : Application of a Multivariate Negative Binomial Model

Household vehicle ownership is a critical demographic characteristic influencing many aspects of travel demand. Using the 1995 NPTS data set, this research proposes a multivariate negative binomial model to investigate households’ distinctions in vehicle purchases among passenger cars, SUVs, pickups, and minivans. This model structure is capable of capturing unobserved heterogeneity across the vehicle ownership levels and is equivalent to a multinomial distribution of the combinations of vehicles owned, conditioned on a negative binomial of the total vehicles owned. The results suggest that vehicle ownership decisions are firmly related to household size, income, population density (of zone of residence), and vehicle prices. For example, larger households exhibit a preference for minivans over SUVs, while SUVs are preferred to passenger cars, which are preferred to pickups. Higher-incomes households prefer SUVs most, and pickups least. In contrast, households living areas of lower population density favor pickups the most. The results of this work inform various vehicleclass-specific policy applications while providing a new statistical approach for the simultaneous vehicle-choice decision that households make.