Back to top

A Closer Look at Congestion Pricing on the Bay Bridge

WINTER 2012 -- As anyone who drives across the Bay Bridge knows, tolls went up in July 2010. In an effort to raise revenues for seismic retrofitting and to speed travel during commute hours, the Bay Area Toll Authority (BATA) hiked fares to $6 during peak weekday commute hours but kept them at $4 during off-peak hours.

Casual carpoolers who previously traveled free with three people in a vehicle, also felt the pain: they were asked to pony up $2.50. On weekends all tolls went to a flat $5.

So were these changes effective? To find out, the Metropolitan Transportation Commission (MTC) turned to UC Berkeley’s transportation experts. A team led by City and Regional Planning Professor Elizabeth Deakin  and Karen Frick, co-director of the University of California Transportation Center, went to work, and in November 2011 presented the findings of a complex, year-long study.
Among the results, the researchers determined that carpool lane usage declined by 26 percent; that travel times improved substantially during certain hours and from some approaches to the bridge, but not all; that some drivers and carpoolers moved to off-peak periods to avoid the higher toll; and, that BART picked up 10 to 15 percent more riders across the Bay.
The task required was formidable, requiring independent analysis, mixed methods for capturing the complexity of traveler decisions, the use of surveys, focus groups, and interviews. There were complicating factors: traffic varies depending on the season, carpooling over the Bay Bridge was declining before the toll changes, and the recession affected traffic levels as business locations and activities changed.
For a behind-the-scenes look at how the team tackled the project, the Berkeley Transportation Letter queried Elizabeth Deakin.
Why did MTC and BATA come to UC Berkeley’s transportation experts for this analysis? Why not go to a consulting company?
We have a lot of relevant background. In the mid-1990s, my late husband Greig Harvey and I did an initial analysis of a much higher Bay Bridge congestion toll. That toll proposal received support from MTC and local business groups but was blocked in the state Senate. At that time, Karen Frick, now Assistant Director of the UCTC Transportation Center, was the project manager for the study and Steve Heminger, now executive director of the MTC, was MTC’s link to the legislature. The new project reunited this team, with Karen and myself now at Berkeley, and Steve at the helm of MTC. Alex Skabardonis, who in the 1990s had worked with my husband and me on a project analyzing a variety of pricing strategies for the four largest California urban areas, joined the team to provide expertise in traffic operations. Robert Cervero, Director of the UC Transportation Center, was added to look at longer-term economic impacts.
What were you asked to do, and how long were you given to do it?
UC Berkeley was asked to do an independent analysis of the effects of the peak/off-peak toll differential. The project started in late June 2010, shortly before the toll change went into effect, and was finished in October 2011.
Why did MTC need this analysis?
MTC and BATA had detailed data on Bay Bridge traffic and toll revenues before and after the toll changes, but they wanted a dispassionate investigation of the changes in travel behavior, traffic flow, and impacts resulting from the traffic and revenue changes.
You led a group of researchers that included Cervero, Skabardonis, Frick and a large and varied group of graduate students. If this were a Mission Impossible episode it would seem each of these individuals played an important, discrete role in analyzing and pulling together the findings of this report. Can you explain why you picked the team you did?
Robert Cervero’s expertise in how transportation affects urban economic performance and his ability to develop models that could distinguish the effects of the recession as well as other exogenous events--like the Giants' games--from the toll's effects were critical to the study. Alex Skabardonis’ deep knowledge of traffic operations and his experience with sensor data allowed us to look at how pricing changes on a single link affected traffic volumes, and helped us understand how bottlenecks at other locations can lead to high day-to-day variability in the traveler’s overall trip times. Karen Frick’s strong relations with MTC helped us secure MTC data quickly, and her ability to manage the project’s day to day workloads on a very tight schedule were essential. She also worked closely with me in designing the surveys and focus groups that allowed us to understand why travelers changed their behavior – or didn’t. The grad students were drawn from both planning and engineering, and we gave them major responsibility for investigating specific topics, carrying out analyses and writing up their findings. I carried out interviews with the business community, environmentalists, and social equity advocates, and led focus groups with Transbay travelers, whom we recruited using FasTrak lists, BART rider contacts, handouts at carpool sites, and broad media blitzes.  We all participated in meetings with MTC and in putting together the project reports.
How many students were involved in this project, and what sorts of things were they learning as they participated in a study of this kind?
We had three students who worked on the project for most of the year, but many more came in to help with surveys, focus groups, carpooling studies, BART user surveys, and data entry and analysis tasks. They had the opportunity to apply the methods and techniques that they had learned about in classes, and they learned how to work on a multidisciplinary team and present their work to agency staff.
I think we were surprised that so many people had access to free parking in San Francisco, and more so that only about half were provided it by employers or businesses.
What were the greatest challenges the team encountered with this study?
We got started late due to contracting delays, and almost missed the chance to do the “before” studies. In fact, Karen and I designed the before survey pro bono and gave it to MTC to distribute before the study began.  Also, there was a lot of “noise” in the data due to construction of the new Bay Bridge. The recession’s impacts also added to the complexity of the study.
What were the most important findings?
We found that a small, $2 toll increase does have a measurable effect, resulting in some travelers moving to off-peak travel times and others changing modes, in this case mostly to BART. Casual carpoolers, who traveled free one way before the toll increase and were usually expected to contribute a dollar for the ride when the tolls were added, gave up on carpooling in large numbers. This was not entirely rational because carpools still had a time advantage and the cost per person was far lower than driving or transit use for most trips. This indicates that casual carpooling attracts users not just because of time and cost advantages but also because casual carpoolers like the social and political aspects: Traveling for free, giving a stranger in need a ride, and beating the government systems were reasons for casual carpooling mentioned in focus groups with casual carpoolers.
What surprised researchers most about the findings?
I think Karen and I were both surprised that so many people had access to free parking in San Francisco, and more so that only about half were provided it by employers or businesses; the other half found free spots on their own somehow. In a city where daily parking rates are often $20 a day or more, how can that much free parking be available? That was a surprise. On the other hand, it is not a surprise that many drivers aren’t paying the full cost of their trip; if they had to do so, I’d expect that many more would be on transit or in carpools.
Many casual carpoolers appear to have abandoned carpooling once the toll was imposed. Did they move to transit or to driving alone?
Some moved to transit, some to driving alone, but not large numbers.  Quite a few probably moved to off-peak periods as well, but it’s hard to estimate this because there are no special carpool lanes off-peak. In addition, there’s some evidence that a portion of the carpool lane users were travelers who did not actually qualify to use the lanes. Once FasTrak and a toll were added, beating the system seemed both riskier and less worth it.
The press made much of the 26% decline in carpools. But it’s important to remember that only a small share of the peak period bridge users were in the carpool lanes in the first place. So another way to look at the numbers is that carpool lane usage went from 14% of all vehicles during the peak  to 10%.  Carpool lanes lost  four percentage points on a share of total vehicles basis. That doesn’t mean that everyone quit carpooling – in focus groups, some told us that they were still carpooling with a partner or colleague and sharing costs, but were not picking up the third passenger to use the carpool lanes. For some, the regular lanes had gotten a bit faster, and the added time to get a third rider was no longer worth it.
The $2 surcharge during the peak period was not enough to clear congestion. We knew it wouldn’t be, but it was a place to start. 
Charging higher tolls during peak commute hours is fairly new to the Bay Area. Do you think this mild form of congestion pricing worked well in this very congested approach to the bridge? Has it accomplished most or all of its goals?
The toll worked much as modeling done for the corridor predicted. The $2 surcharge during the peak period was not enough to clear congestion, however. We knew it wouldn’t be, but it was a place to start. The Bay Bridge toll increase also had no impact on other bottlenecks along the way, so from a door-to-door travel time perspective most users saw little change.
Is the congestion pricing on the Bay Bridge a harbinger for other congested corridors in the Bay Area?
The Bay Area has been implementing High Occupancy Toll lanes in congested corridors, and I expect that we will see more of this.  We also need to start a serious discussion of tolling as an alternative to the gas tax. Recent nationwide surveys indicate that voters actually prefer tolls and road user fees to gas tax increases, so tolls may be the way to pay for roads in the future.
Finally, will you be continuing research on the Bay Bridge, or have you pretty much provided the answers to the questions MTC/BATA needed to know?
We are finished for now, but we think it would be important to keep an eye on the Bridge and its impacts, because some – like business and residential location shifts - may only appear after a few years.