This study developed methods to assess the cost of travel time variability for air and car travellers. The cost of travel time variability can be determined in three steps. First, researchers need to develop a behavioural micro economic model that is able to capture travellers’ responses to travel time variability. Second, this model needs to be validated and calibrated using empirical data. Third, the model needs to be applied using observed or simulated travel time data in order to see how large the costs of travel time variability are. Each of the chapters in this thesis contributes to at least one of these steps. The behavioural models are extensions of the scheduling model of Noland and Small (1995). In these models, travellers dislike arriving at a different time than their preferred arrival time. They anticipate travel time variability by leaving earlier from home or choosing an earlier flight, and therefore the dynamic choice of the traveller is incorporated in an intuitive way. Chapters 2 and 5 make use of discrete choice econometric methods to estimate the cost of arriving early and late for air travellers going to the airport, and for car commuters participating in a real world reward experiment. Chapters 2, 3 and 4 apply the models using observed travel time data. (Author/publisher)
Abstract