Value of schedule delays by time of day evidence from usage data from high-occupancy toll lanes on state road 167

This study estimated the value of schedule delays (VSD) along the highoccupancy toll (HOT) lanes on State Road 167 in the state of Washington on the basis of revealed preference data from toll-paying users. More than 5 months of tolling records that included more than 1 million lane choices were used in the analysis. A mixed logit model was estimated to account for heterogeneity at the individual level. Drivers were assumed to evaluate reliability relative to their desired or expected arrival times (i.e., reference points). Three empirically determined reference point specifications were employed to calculate schedule delays, and the sensitivity of the results with respect to reference points was reported accordingly. The model estimations showed significant variations in the estimated VSDs for different times of day and also in different travel directions. The differences in VSD estimations could be as high as 17/h in a single time period depending on the reference point assumption. Reference point assumption was also shown to have a significant effect on the VSD when scheduling delays were used for measuring travel time reliability. Empirical findings of this study provide useful insights into the variability of the value of travel time of users traveling in different time periods. In particular, these results can help in the development of policies for more effective allocation of traffic capacity to HOT lanes, especially during peak congestion periods.