Travel Demand Models for Road Pricing in the U.S.

Travel Demand Models for Road Pricing in the U.S.


P Vovsha, PB America US; M Bradley, Mark Bradley Consulting, US; K Lawton, Keith Lawton Consulting, US


The paper provides a constructive synthesis of the State of the Art & Practice in modeling toll roads in US based on examples of advanced models developed in New York, Columbus, San-Francisco, Montreal, Sacramento, Denver, and other cities.


The paper provides a constructive synthesis of the State of the Art & Practice in modeling toll roads in US. Road pricing is one of the major transport policy areas in the US. Road pricing, and especially a variety of new pricing forms, represent a growing challenge to travel modelers. The paper includes the following three sections:
1. An overview of traveler behavior responses to pricing and the corresponding applied models.
2. A review of recent revealed preference (RP) and stated preference (SP) surveys and evidence in the US regarding drivers? responses to pricing.
3. A discussion of the most promising directions for the improvement of modeling practices for road pricing.

Traveler behavior responses to pricing and applied model structures.
The nature of the road pricing project under study and its potential impacts on various dimensions of travel in the region dictates the requirements to travel model. In this regard, the paper provides a classification of road pricing forms with the linkages to possible modeling techniques. In general, the following travel dimensions can be affected by pricing:
? Trip-level decisions:
o Route itinerary in the highway network (modeled by traffic simulation),
o Principal decision to take a toll route versus non-toll route (pre-route choice)
? Tour-level decisions:
o Auto occupancy (individual vs. joint travel),
o Mode choice
o Time-of-day choice (including peak spreading effects)
o Destination choice
? Day-level decisions:
o Trip/tour/activity frequency
o Activity re-sequencing as result of time-of-day shifts
? Mid-term mobility decisions and household/person attributes:
o Acquisition of transponder, transit pass, parking ticket (payment type choice)
o Free parking eligibility at workplace / school
o Car ownership
? Long-term location choices:
o Residential location and dwelling type
o Usual workplace location

The paper analyzes existing US forecasting models with respect to the coverage of these dimensions. It is shown that in most cases only route itinerary and binary pre-route choice models are employed. There is also a growing number of applications where mode and/or occupancy choices are included. In some regional model systems that were specifically developed for congestion pricing projects, peak-spreading models were applied. However, very few models to date have incorporated all trip and tour-level dimensions in a consistent way, and there have not yet been any practical examples of incorporation of pricing impacts on the day-level, mid-term, and long-term choices, even in the activity-based models now that have recently come into use.

A review of RP and SP survey methods and evidence
Development of model for road pricing requires supporting data collection and travel surveys. A comprehensive Revealed Preference (RP) household travel survey is needed to develop a regional transportation model. The RP data is often supported by complementary or project-specific Stated Preference (SP) surveys. SP surveys are typically designed to address willingness-to-pay factors relevant for road pricing (value of time savings, value of reliability). Survey data collection can also support other model development data needs, including HOV/HOT lane usage and payment media. The authors have been involved in several such recent surveys, and have also written syntheses of recent survey evidence. The paper will provide a review of this work.

Most important and promising directions for improvement.
The most promising directions for the improvement of road pricing models are shown to be associated with advanced activity-based tour-based demand models and advanced network simulation tools (dynamic traffic assignment and micro-simulation). More specifically, these major breakthroughs provide for the incorporation of the following model features and components essential for road pricing:
? Heterogeneity of road users with respect to their VOT and willingness to pay. It requires a consistent segmentation through all demand modeling and network simulation procedures ensuring compatibility of implied VOTs. In addition to explicit segmentation, random coefficient choice models represent a promising tool for capturing heterogeneity.
? Accounting for reliability of travel time associated with toll roads. Incorporation of travel time reliability in applied models requires quantitative measures that could be modeled on both demand and supply sides.
? More comprehensive modeling of time-of-day choice based on the analysis of all constraints associated with changing individual daily schedules.
? Proper incorporation of toll road choice in the general hierarchy of travel choices in the model system. Additional travel dimensions (such as the whether to pay a toll, car occupancy, and payment type/technology) and associated choice models should be properly integrated with the other sub-models in the in the model system.

The paper provides a discussion on recent advances in these directions and first examples of incorporation of advanced features in regional models developed in New York, Columbus, San-Francisco, Montreal, Sacramento, Denver, and other cities.


Association for European Transport