A copula-based joint model of commute mode choice and number of non-work stops during the commute

In this paper, in the spirit of a tour-based frame of analysis, we examine the commute mode choice and the number of non-work stops during the commute. Understanding the mode and activity stop dimensions of weekday commute travel is important since the highest level of weekday traffic congestion in urban areas occurs during the commute periods. The paper employs a copula-based joint multinomial logit - ordered modeling framework in which commute mode choice is modeled using a multinomial logit formulation and the number of commute stops is modeled using an ordered response formulation. The data used in this study are drawn from the "Time use" multipurpose survey conducted between 2002 and 2003 by the Turin Town Council and the Italian National Institute of Statistics (ISTAT) in the Greater Turin metropolitan area of Italy. The results highlight the importance of accommodating the inter-relationship between commute mode choice and commute stops behavior. The results also point to the stronger effect of household responsibilities and demographic characteristics in the Italian context compared to the US context.