Intra-household interactions over time : an activity-based approac

Abstract

To jointly predict the dynamic decisions of the household and interactions among members, the household activity-based model is revisited from a time series perspective. A panel nested logit model is proposed with the group decision mechanism, and is calibrated with the data from 2011 Hong Kong Travel Characteristics Survey. In this study, the time dependence of joint activities between household members is empirically discovered. It shows that members who had more joint activities in the past, are likely to have more in the future. In addition, the individual and family taste heterogeneities are also investigated. It was found that the employed household members and students have the largest weights in family decision making. In the households, members of unemployed, retired, and homemakers have the largest extra gain from the joint activities as compared to solo activity participation. Incorporating the time dependence and taste heterogeneity significantly improves the proposed model's goodness-of-fit.