Investigating intra-household interactions between individuals' time and space constraints

The activity space of an individual is defined as the activity-travel environment which a traveller is using for his or her activities (Axhausen et al., 2002). It is limited by this individual's ability and resources, such as available time for travel as well as his or her anchor points. However, most existing studies have focused on single individual activity space, ignoring the fact that individuals' activities often interact with that of his or her family members'. In this paper a multivariate model is proposed where the correlation between travel time of fathers and mothers, and the correlation between the activity space and travel time are modelled explicitly. The estimated correlations from these joint distributions provide insights into both the intra-household interactions in daily travel and the intrinsic relationships of the hidden limits in the dimensions of space and time. The travel time limits are modelled using a stochastic frontier model component, which can estimate an unobserved upper or lower limit for travel time expenditure. This limit usually refers to the maximum travel time budget or minimum travel time need, which denotes the maximum or minimum amount of travel time that an individual is willing or able to allocate per day. The concept of the confidence ellipse is used as a measure of activity space constructed from the multi-day travel diary data. It is hypothesised that the unobserved travel time limits and activity space sizes of fathers and mothers are correlated with each other, due to a similar spatial knowledge and accessibility to various facilities. The daily variations in the travel time expenditure of parents are also assumed to be correlated because of daily household task allocation and joint household travel. Data collected from a three-week household travel diary in the Bandung Metropolitan Area in Indonesia are used for estimation in this study. The estimated frontier model component shows that neither parent has reached their maximum travel time budget and/or minimum travel time need that they inherently must spend. Compared with other attributes, the perceived accessibility attributes play the most important role in influencing the activity space limits. For households with fully employed fathers, a trade-off mechanism is found in travel time expenditure between parents, which is likely due to the redistribution of household tasks. On the other hand, for households with fathers who are not fully employed, a complementary effect is found, arising from the joint travel among household members. The travel time budget and activity space limits of fathers are positively correlated with those of mothers. These findings call for the formulation of transport policies that consider the household as a unit, especially in developing countries such as Indonesia, to fulfil the mobility needs of different market segments, e.g., households with fully employed fathers and those with fathers who are not fully employed.