Traffic congestion contributes to longer travel times and increased travel time variability. We account for the dynamic nature of travellers' choices, by deriving a closed-form solution for the costs of travel time variability. The resulting travel delay cost function is linear in the mean travel delay. Then, we use a semiparametric estimation approach to analyse observed and unobserved heterogeneity in the value of travel time and reliability. Using data from a stated choice experiment, we show that there is substantial heterogeneity in the willingness to pay for fast and reliable travel. About 5-25 percent of the heterogeneity in the value of time and reliability is attributable to observed characteristics of individuals, implying that unobserved heterogeneity is much more important than heterogeneity related to observable characteristics. It is furthermore shown that schedule delay costs are on average 24 percent of the total costs of travel delays.
local maximum likelihood heterogeneity value of reliability scheduling semiparametric Logit