Methodology for modeling household time allocation behavior

Abstract

It is well known that members of a household often interact in making decisions about different activities that they perform and the time that is involved. However, little research on such group decision-making issues has been conducted to date in transportation research. In the context of time allocation, the authors previously proposed two types of household time allocation models within a random utility maximization framework. The first one assumes a multi-linear type of household utility function, while the second one adopts an iso-elastic class of welfare function as the household utility function. fn representing household decision-making mechanisms, the two types of functions functionally overlap and theoretical comparisons suggest that the iso-elastic utility function seems more general and flexible. In this paper, these two models are estimated and empirically compared by using the same activity diary data collected in a depopulated region in Japan. The study confirms that both models have a relatively high and nearly equal goodness-of-fit. It also provides further evidence that 1) households interact in making decisions, 2) male household members do have most influence on the decision-making process and 3) the household time allocation model seems good at describing female's time allocation behavior. Moreover, similar household decision-making rules are observed in the sense that the household members first form their own separate utility function as a weighted average of part-worth utilities and then maximize the resulting function. However, in spite of nearly equal model accuracy and the similarity of household decision-making rules, the two models estimate a different relative influence of household members. This implies that household members might follow different paths during the negotiation process to arrive at their most preferred outcomes.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 7 citations exist upstream.