Abstract
This paper investigates the relationship between commuters’ trip-chaining behaviors on their round-trip commutes: from home to work and from work back home. Two types of joint econometric models—the bivariate ordered probit model and the bivariate Poisson regression model—are developed on the basis of household travel survey data collected from Xiamen, China, in 2003. Nonwork trip frequencies and round-trip commutes are considered as two interrelated dependent variables, whereas independent variables in models include commuters’ age, gender, occupation, and residency status; household composition; commute mode choice; and transit accessibility at the workplace. The error correlations in both types of models are estimated to be positive, which strongly supports the existence of a promotive relationship between trip-chaining behaviors on round-trip commutes. However, the conjecture of a substitutive relationship between round-trip commutes is not supported by the joint models. In addition, it is found that automobile use for commuting plays a highly positive role in commuters’ trip-chaining behaviors for both directions of a round-trip commute.
Get full access to this article
View all access options for this article.