Valuation of travel time reliability in an extended expected utility theory framework

Valuation of travel time savings (VTTS) is a critical measure in transport infrastructure appraisal, traffic modelling and network performance. It has been recognised for some time that the traditional measure of VTTS should be complemented by a valuation of travel time reliability (VOR). An alternative approach, proposed in this paper, promotes the view that a single revised estimate of VTTS that accounts for the observed variability in travel time for a specific trip and the associated likelihood of such variations in travel times occurring, might be a more sensible way of accounting for the amount that an individual is willing to pay to save time; that is the notion of separate estimates of VTTS and VOR may be unnecessary. The most widely used approach for valuing travel time reliability is the mix of Random Utility Maximisation (RUM) and Expected Utility Theory (EUT) (i.e., linear utility specification with linear probability weighting). We extend the EUT approach (i.e., a non-linear utility specification with a linear probability weighting function), by applying a non-linear probability weighting function to accommodate choice made under risk, referred to as Extended EUT. The empirical findings recognise the extent of attitudes towards risk in estimates of ‘reliability embedded VTTS’ or REVTTS.