Household Automobile and Motorcycle Ownership Analyzed with Random Parameters Bivariate Ordered Probit Model

Abstract

This paper investigates the factors that affect household automobile and motorcycle ownership in large metropolitan areas. Extensive geocoded trip data from Athens, Greece, were modeled with the random parameters bivariate ordered probit model. This model accounts for unobserved heterogeneity in the data population and commonly shared characteristics with automobile and motorcycle ownership. The random parameters bivariate probit model provided a statistically superior fit compared with its fixed parameters counterpart. The study's results indicate that vehicle (automobile and motorcycle) ownership is determined by a number of factors, such as traveler characteristics, the population density at the origin and destination, the distance and time to the destination for several trip purposes, and access to public transit.