This research incorporates time-of-day effects into the analysis of inter-city business travel behavior. Towards that end, a model of traveler behavior incorporating time-of-day effects is developed, tested, and empirically calibrated. The model is then compared with a conventional mode choice model to predict the market share of high speed rail in a hypothetical market.; We extend De Serpa's 1971 time allocation model to a multiple time regime environment. Time-of-day effects can then be modeled as differences in the resource value of time for each regime. A model is specified in terms of a three regime daily cycle: work, leisure and sleep. For each time regime, the resource value is specified as the sum of two components: a spatial component and an activity component.; The empirical research has two elements: an exploratory set of open-ended interviews and a stated preference survey used to estimate model parameters. The interview results offered qualitative support to the multiple regimes theory. Survey data were analyzed using a random parameters logit model. The estimates indicated that in the sample population, the activity component of the value of time changed significantly across time regimes. In particular, sleep time was valued six times as much as leisure time, and thrice as much as work time.; The spatial component of the resource values of sleep and leisure (for one evening) was jointly estimated. It was found to be equivalent to the disruption of 40 minutes of sleep time for single individuals, and on average equivalent to 80 minutes of sleep disruption for individuals with families.; A mode choice simulation experiment was conducted to analyze the aggregate effects of neglecting time-of-day effects in conventional analyses. This experiment indicates that under a broad range of assumptions, conventional models underpredict market shares for high speed rail. This bias can be traced to the trade-off between spending a night away from home and beginning travel early in the morning.
展开▼
