Joint Disaggregate Modeling of Car and Motorcycle Ownership

Abstract

This paper analyzes the determinants of household car and motorcycle ownership and the relationship between them in Bogotá, Colombia. The basis of the analysis was disaggregate household data from a cross-sectional household travel survey undertaken in 2011. Ordered probit models were developed independently for car and motorcycle ownership by considering explanatory variables related to income, household size, distance to work, age, gender, and access to public transportation. The relationship between car and motorcycle ownership was evaluated by considering correlation and substitution effects. The results show that, in accordance with previous empirical findings, income has an unequivocal positive effect on car ownership. Meanwhile, households with medium- to low-income levels present the higher levels of motorcycle ownership. Furthermore, even though negative correlation and substitution effects exist, they are not highly significant, and therefore independent modeling of car and motorcycle ownership would not generate important biases. Suggestions for further research indicate the need to replicate the analysis in cities with higher levels of motorcycle ownership.