Des McCafferty

THE USE OF MULTINOMIAL LOGIT ANALYSIS TO MODEL THE CHOICE OF TIME TO TRAVEL

Ever scarcer public investment capital and the need for greater energy conservation have combined in recent years to force a reconsideration of traditional approaches to the problem of peak-period congestion in medium to large sized cities. Increasingly, the emphasis is shifting away from the construction of new transportation infrastructure and vehicles toward more efficient use of existing facilities, with particular attention being paid to policies designed to spread peakperiod travel demand more evenly throughout the day. In this context the individuals choice of time to travel to and from work is of crucial significance: it is also probably the least understood travel decision. This paper investigates the use of multinomial logit analysis to model the time of day choice, using data collected in Hamilton, Ontario during the summer of 1977. The exercise highlights the difficulties involved in using the logit model both as a descriptive and predictive tool, as well as indicating the need for further research into the determinants of the time of day decision. A preliminary finding is that journey time is not a significant factor in the individuals choice of time to travel.