We extend existing models of the commuting time-of-day choice in order to analyze the effect of uncertain travel times. Travel time includes a time-varying congestion component and a random element specified by a probability distribution. We compare results from the uniform and exponential probability distributions and derive the optimal "head start" time that the commuter chooses to account for travel time variability; i.e., a safety margin that determines the probability of arriving late for work. Our model includes a one-time lateness penalty for arriving late as well as the per minute penalties for early and late arrival that other investigators have included. It also generalizes earlier work by accounting for the time variation in the predictable component of congestion, which interacts with uncertainty in interesting ways. A brief numerical analysis of the model reveals that uncertainty can account for a large proportion of the costs of the morning commute.

This research examines the behavioral reactions to the impact of changes in the probability of non-recurrent incident and how this effects the expected costs of a commute trip. This basic approach combines the estimation of a travel demand model (estimated with data collected from a stated preference survey) with a supply side model of a congested highway. We also examine the impact of various socio-economic variables, including a detailed classification of occupational groupings. Our demand model is based on a theoretical model developed to explain how unreliability in travel times affects expected travel costs. We find that expected schedule delay (early and late), lateness probability, and expected travel time influence the expected costs of travel. Our parameter estimates confirm the anticipated values of these parameters: lateness probability has a high disutility, while expected schedule delay early is preferable to expected schedule delay late, and the disutility of expected travel time is between these two. We do not find a high level of significance for planning costs, as expressed by the variance in travel times. Our simulation model shows that schedule costs and lateness probability represent a large fraction of the total cost to the commuter; these are generally not affected by capacity increases but can be reduced by decreasing the probability of a non-recurrent incident.

We present a simulation model designed to determine the impact on congestion of policies for dealing with travel time uncertainty. The model combines a supply side model of congestion delay with a discrete choice econometric demand model that predicts scheduling choices for morning commute trips. The supply model describes congestion technology and exogenously specifies the probability, severity, and duration of non-recurrent events. From these, given traffic volumes, a distribution of travel times is generated, from which a mean, a standard deviation, and a probability of arriving late are calculated. The demand model uses these outputs from the supply model as independent variables and choices are forecast using sample enumeration and a synthetic sample of work start times and free flow travel times. The process is iterated until a stable congestion pattern is achieved. We report on the components of expected cost and the average travel delay for selected simulations.