Standard practice in econometric estimation of the value of travel-time changes divides travellers into market segments and assumes linear utility functions in each category. Here instead we derive a RUM-model from a second-order approximation to the utility function. The model is estimated by probit on data from a recent SP value-of-travel-time study. The influence of socio-economic variables, trip duration, and so on, on the value of travel-time changes is estimated and tested. Our conclusions differ in several ways from those of a previous study evaluating the same dataset with the conventional method. We find a considerably lower value of travel-time savings and signs of a cognitive threshold; respondents decline travel time savings smaller than 10-15 minutes irrespectively of the bid price. The non-linear approach also provides a convenient means for benefits transfer, that is, to calculate the mean value of time savings for a specific road project, depending on how it will be used, and by whom. (Author/publisher).
Abstract