The Costs of Uncertain Travel Times for Car Travelers
P Koster, VU University Amsterdam, NL
We analyze the value of uncertainty for car travelers which is defined by the costs of not knowing the objective probabilities of the travel time distribution.
In this paper we propose an analytical framework for scheduling decisions of road travelers facing travel time uncertainty. The scheduling model has been used frequently to estimate the costs of schedule delay early and schedule delay late for commuters. The value of schedule late and early can be used to derive calculate the costs of variable travel times. The main drawback of this approach is that it is assumed that travelers have a perfect knowledge of the travel time distribution. In reality this is not the case because people use their subjective probabilities to determine their optimal departure time. The fundamental difference between scheduling under uncertainty and scheduling under risk is that for scheduling under uncertainty the probabilities of arrivals are not known. We show how scheduling decisions and departure time choice are affected by the subjective probabilities of the traveler. There could be two reasons for misperceiving the objective probabilities. The first reason is lack of experience. The second reason is lack of understaning of what the concept of probability is. We use cumulative prospect theory to derive the value of uncertainty, which is defined by the costs of misperceiving the objective probabilities. In the empirical part of the paper we use data to show how large the value of uncertainty is. If it is assumed that probability weighting is in line with the literature, the value of uncertainty is rather small. This means that policy makers can use the scheduling under risk model to evaluate the costs of variable travel times.
Association for European Transport