The Oregon Department of Transportation conducted a test of an innovative technology to replace fuel taxes with mileage fees. In the test, some vehicles were charged a flat fee per mile and others were charged differential fees that were higher for travel in the Portland metropolitan area during weekday peak hours and lower for other travel. I was charged with developing the database for the project and analyzing the behavioral responses of the participants. The database is quite extensive, including responses to three surveys as well as information on mileage in various categories. The experiment ended on April 1, 2007 and the final report on the project is due June 30, 2007. This is also the time when funding for the project will end. The final report will include information on key behavioral changes, such as the reduction in miles traveled per day during the peak for the congestion charge group. However, the database has the potential for much more extensive analysis. In addition, a different type of congestion charging experiment was conducted by the Puget Sound Regional Council, and a public use dataset from that experiment is expected to become available within the next six months. The objective of this proposal is to extend the analysis of changes in behavior by subjects in the ODOT Road User Fee Pilot Project, and to draw on other sources, such as the Puget Sound project, to compare the behavioral changes observed in this experiment with those found in other contexts.
A substantial amount of data was collected on participants in the Oregon Road User Fee Pilot Project. Basic analysis of the data is being conducted as part of the project. However, there is potential to gain further information on characteristics that caused or prevented changes in participants’ driving patterns. For example, participants were asked the distance from their home to the nearest transit stop, but that was the only information on transit service that was available for analysis. It would be possible to use the participant’s address to generate much more comprehensive measures of transit service availability to determine whether these other dimensions of transit service affected the response to road pricing. In addition, information on day of week and morning versus evening peak travel is being aggregated for the existing study, but this data holds the potential for further information on the nature of the responses.
A variety of statistical analyses will be conducted to evaluate both the extent of response to a vehicle mileage fee and the interaction with both demographic and attitudinal characteristics of the participants. A GIS analysis will be used to link household location with better measures of transit service.
Results would include a better understanding of how pricing interacts with other factors in affecting driving patterns and in particular in affecting driving during peak periods. It would also provide a better understanding of the revenue potential from such charges.