Forecasting Demand and Values of Travel Time Savings for Freeway HOV, Toll and HOT Facilities: Incorporating Attitudes and Perceptions into Commuter Choice Models
Project Direct
David Brownstone Thomas F. Golob
Department of Economics Institute of Transportation Studies
University of California University of California
Irvine, CA 92697-5100 Irvine, CA 92697-3600
949-824-6231 949-824-6287
dbrownst@uci.edu tgolob@uci.edu
Abstract:
Accurate forecasts of demand for restricted roadway facilities – high occupancy vehicle (HOV) lanes, toll lanes (including congestion pricing), or combined HOV and toll (HOT) lanes on freeways and bridges – are key to the success of such projects. Yet the track record for predictions for such projects throughout the U.S. is dismal; transportation professionals have not been successful in understanding traveler behavior regarding such choice alternatives. The objective of the proposed research is to explore reasons for these failings and to make recommendations regarding priorities for better models. Alternative model specifications documented in the literature will be compared on a common dataset. The most effective dataset for this purpose is the panel survey collected in 1997-1999 for evaluation of the San Diego I-15 Congestion Pricing Project, combined with recorded toll data and traffic speed data from freeway loop detectors and floating car measurements. The key new feature of this work is the joint modeling of commuters’ choices, perceptions of key trip attributes, and attitudes about road pricing. These variables will be related to commuter’s sociodemographic information as well as objective traffic network data typically used in demand analysis. This new model is designed to predict both the economic and political feasibility of a project.
Key words: Travel demand forecasting, HOT lanes, HOV lanes, Value Pricing
Work Completed to Date:
We have completed merging the freeway loop detector, floating car, and panel survey data. These merged data have been used to estimate a number of route choice models, and these results are described in Steimetz and Brownstone (2003). The median value of time saved on the HOT lane is about $30/hour from actual behavior and $10/hour from responses to hypothetical questions. These estimates are very close to those obtained using data from the SR91 private toll facility linking Riverside and Orange counties. The comparisons between the I15 and SR91 results are described in Small and Brownstone (2003), and they represent one of the few cases where transportation demand models from different studies have produced quantitatively similar estimates. Since hypothetical (SP) studies are used frequently in actual work, the finding that they yield very different conclusions than those from actual traveler behavior needs further careful study. The remainder of this project will look at whether attitudes and perceptions about toll roads and traffic conditions can explain these differences.
Papers to Date:
Percent Complete: 100%