Estimating Drivers’ Willingness to Pay by Using Empirical Data from a Variably Priced Freeway Facility

Abstract

This paper presents a simplified method for estimating the percentage of drivers who would choose to use a priced facility in place of a free alternative by means of a year of revealed preference data from a variably priced freeway facility in Southern California. The resultant model showed a strong fit to the set of 5,071 bins of data (with an R2 value of .869) while using only three easily obtained predictor variables: the expected travel time savings by use of the managed lanes, the toll price being imposed, and the relative speeds between the two alternatives at the time the decision is made. This model contrasts sharply with other existing methods of analysis, which often use logit models and require much more detail about the population of drivers. The author then illustrates how these results can be used for planning analyses of other priced facilities with different operational conditions and roadway configurations (specifically, converting a current high-occupancy-vehicle facility to a high-occupancy-toll facility) by using only the level of data that would be available from common inductive loops, subject to assumptions about the transferability and generalizability of these results from Southern California to the facility being analyzed.