San Francisco Bay Area politicians have a well-earned reputation for hating cars and loving mass transit, especially when it runs on rails. It was thus a pleasant surprise to see the Bay Area Rapid Transit district board reject a costly plan to extend the BART commuter rail system another 5.5 miles further out into suburban Alameda County. While we recognize that this decision does not mean that BART board members have suddenly seen the wisdom of Reason’s transportation policy research, the vote does serve as a teachable moment about the drawbacks of new suburban rail projects generally and for BART. It also provides an opportunity for policymakers in the eastern end of the Bay Area to consider the benefits of managed express lanes and opportunities to integrate express buses into the local mass transit landscape.
A Legacy of Costly Car Replacement Policies
The City/County of San Francisco has had a transit-first policy since 1973. It implements this approach by increasing transit capacity and by making driving more difficult. In the aftermath of the 1989 Loma Prieta earthquake, two sections of an elevated freeway near the urban core were demolished. More recently, restrictions have been imposed on passenger cars using Market Street — the city’s main downtown thoroughfare — with a complete ban on the horizon. Throughout the city, traffic calming, dedicated bike lanes and bus rapid transit projects have resulted in the loss of automobile lanes and lower speed limits.
Meanwhile, San Francisco is making costly investments in subway construction. A 1.7-mile extension of the Third Street light rail line through Union Square to Chinatown (known as the Central Subway) has a $1.6 billion budget and was expected to open in December 2018. The project has faced repeated delays and service is now expected to begin in January 2020, with a further postponement possible due to a dispute between the San Francisco Municipal Transportation Authority (SFMTA) and the prime contractor for the project, Tutor Perini. SFMTA is accusing Tutor and its subcontractor of installing the wrong type of rail tracks in the tunnel, using standard strength steel as opposed to longer lasting high-strength steel. Tutor has countered that it was merely following SFMTA’s original instructions.
Another subway project planned for the city is an underground connection between a CalTrain station and the new Transbay Transit Center. That project was recently estimated to cost $2.6 billion and could not be completed until at least 2026. Meanwhile, the new transit center, built at a cost of $2.2 billion has experienced repeated delays and will only service a few dozen bus routes unless and until the proposed subway is completed (although SFMTA, CalTrain and BART all operate train service in San Francisco, they are separate agencies).
In suburban Bay Area counties, the bias against cars and toward transit is only somewhat less intense. Plan Bay Area 2040 — developed by the Association of Bay Area Governments and the region’s Metropolitan Transportation Commission — focuses on Transit-Oriented Development. Rather than allow more single-family houses to be built in outlying areas, the plan calls for more residential density near transit stations. This policy has produced several “transit villages” — mixed-use complexes with apartments — that have been replacing surface parking lots near BART stations. Almost two-thirds of Plan Bay Area 2040 investments are focused on transit, despite minimal growth in ridership for most services.
It is also consistent with the idea of extending BART to more communities. In 2017, BART added service to South Fremont in the East Bay and construction is well underway on a further extension to Milpitas and Berryessa, which is within San Jose’s city limit. By 2026, the Valley Transportation Authority plans to extend BART through downtown San Jose to Santa Clara at an incremental cost of $4.9 billion. Further north, in Contra Costa County, BART recently added service from Pittsburg to Antioch (albeit with a twist as discussed later).
Livermore: An Extension Too Far for BART?
Given the Bay Area’s strong bias for rail transit, it seems odd that BART’s board would reject an expansion of service to Livermore, a growing community 5.5 miles away from the system’s present terminus in Dublin. The proposed extension would have run in the median of I-580 and would not have required tunneling. Still, the $1.6 billion price tag and lack of available funding scared off BART directors, who voted down the proposal 5-4.
But this vote does not mean the end of plans to build a line between Dublin and Livermore. The board’s decision is best seen as a passing of the buck to the newly created Tri-Valley San Joaquin Regional Rail Authority. This new agency was created by the state to connect BART with the Altamont Commuter Express (ACE), a commuter rail line that runs between California’s Central Valley and San Jose.
Right now, ACE service is relatively slow and infrequent. Trains take 132 minutes to cover the full 86- mile route from Stockton to San Jose, with just four departures in each direction on weekdays and no service on weekends. Because ACE does not own its right of way, most of which is leased from Union Pacific Railroad, its ability to increase and accelerate service is limited.
But a new ACE line built in the median of I-580 would be free of these limitations. It could also provide commuters from Livermore and the Central Valley two-seat transit access to Oakland and San Francisco. Connecting the low-cost Central Valley with strong labor markets at the Bay Area’s core would create new career options for those currently unable to afford Bay Area housing or the costs of a long commute by car. A 2017 AECOM study estimated that an ACE connection to BART would increase annual ridership from 1.3 million in 2015 to 3.6 million in 2025.
New ACE service would also be less expensive than a BART extension because ACE trains use conventional rail technologies. When BART’s designers created the system back in the 1960s they implemented a number of concepts thought to be forward-looking but have not been adopted elsewhere.
For example, while most US rail services have a track gauge of 4 feet, 8.5 inches, BART’s tracks are 5 feet, 6 inches apart. Although this allowed wider and more comfortable cars, it also means that BART requires unique rolling stock. This is a major reason why new BART cars cost about $3 million each, about double the price tag incurred by other US rapid transit systems.
BART also has a unique and antiquated train control system. Integrating new track into this system adds complexity and delay to extension projects. Train control software integration caused a multi-month delay in opening the South Fremont extension and is pushing back the addition of service to Milpitas and Berryessa.
Knowledge of these complexities informed BART’s decision to use alternative technology when extending service to Pittsburg and Antioch. Rather than add track to the existing line terminating at Bay Point, BART management opted for a separate system using standard gauge track and diesel-powered trains – mirroring the technologies used on ACE rather than on the rest of BART. The decision compels Antioch commuters to change trains at Bay Point but produced significant cost savings.
Time to Get on Board the Bus
Running standard ACE trains down the I-580 corridor is preferable to extending BART, but it is still an unnecessarily expensive approach. A new ACE rail station in the I-580 median at Isabel Avenue in Livermore would not connect with existing ACE service: ACE’s Livermore station is downtown, about two miles away from the proposed BART terminus. ACE’s right of way intersects with I-580 near Greenville Road, four miles east of Isabel Avenue. Thus, to achieve a rail connection between ACE and BART, the track must be extended into Livermore’s downtown in the face of intense local opposition or must continue much further along I-580. Either choice would undoubtedly add hundreds of millions to the cost of adding new diesel train service between Dublin and Isabel Avenue, which estimated by BART staff to exceed $1 billion.
A more cost-effective option would be to add one or more lanes to I-580 that could accommodate high-speed bus service. The highway already has High Occupancy Toll (HOT) Lanes, a technology that could be used to ensure that new buses travel at full highway speeds despite rush hour congestion. The existing I-580 HOT lanes allow carpools, vanpools, clean energy vehicles and buses to all travel for free.
To accommodate new express service to Dublin BART, authorities might consider adding one reversible managed lane in the middle of I-580. It could be restricted to inbound buses in the morning and outbound buses in the evening, or a limited number of single occupant vehicles could be given access to the lane at premium prices. At other times of the day, the lane would be available for emergency vehicles or to offset the loss of lanes due to accidents. A more expensive, but easier-to-manage alternative would be the addition of one express lane in each direction.
This approach is not without precedent in California. The 11-mile I-110 Harbor Transitway in Los Angeles County accommodates buses, carpools and toll-paying vehicles. The Transitway opened for buses only in 1996 with disappointing ridership. But, more recently, the lanes were opened to other vehicles and bus ridership has grown.
Managed lanes that incorporate buses have been implemented elsewhere around the country. Reason Foundation’s Bob Poole and Ted Balaker coined the term “Virtual Exclusive Busways” to describe new managed lanes in which transit vehicles could operate at full speed while also accommodating carpools and toll-paying single-occupant vehicles. Such projects have been implemented in the median of Houston’s Katy Freeway (I-10) in 2009, as well as in the Atlanta and Washington, DC, metropolitan areas.
In the Livermore region, rapid bus transportation could be phased in quickly and inexpensively. The initial investment might be limited to the addition of a special bus entrance to the Dublin station on I-580, allowing a quick bus to BART transfer. Buses could use existing HOT lanes on I-580 while offering direct connections to Downtown Livermore and the Central Valley. As demand increases and funds become available, one or more additional managed lanes could be added to I-580 and connecting portions of I-205.
The BART board rejected a bus rapid transit option at the same time it passed on expanding rail service to Livermore, despite the publication of a well-argued op-ed supporting express buses by BART Director Nick Josefowitz.
Critics in Livermore and points east dismiss bus alternatives because they do not provide a seamless (one-seat) commute and are perceived to be slow, unreliable and uncomfortable. But, as explained earlier, a seamless rail option is also not going to be available for most users of the proposed BART Livermore extension, since it would terminate in the I-580 median.
With respect to other alleged bus drawbacks, these concerns can be alleviated by proper design of both the routes and the vehicles themselves. Out-of-hand rejection of bus options may be the result of the stigma often attached to bus transportation in the US. Unfortunately, we too often dismiss local and intercity buses as last resort transit for low-income, minority and/or elderly transit riders.
But the Bay Area has counterexamples that show buses can be a high-end option. Apple, Facebook, Google and numerous other tech companies operate shuttle buses for their employees. In its most recent survey, the Metropolitan Transportation Commission reported that 35 shuttle operators carried over 9.6 million passengers in 2014. The fact that these buses have become a focal point of protests against gentrification show that they are perceived to be an elite form of transit.
Meanwhile, AC Transit and other regional bus lines provide comfortable commuter service to downtown San Francisco. Unlike BART, these commuter buses generally provide reliable wi-fi service, which is a valuable amenity for commuters in the tech-oriented Bay Area. In the near future, these buses will begin arriving at the costly new Transbay Transit Center described earlier.
Unlike a conventional BART extension or new diesel trains, express buses operating on express toll lanes could quickly and cost-effectively help reduce traffic congestion and improve commute times along the I-580 corridor. The Bay Area’s political leaders should consider the positive possibilities buses can offer and put aside any biases toward rail transportation so that the region’s commuters, employers and urban planners can examine and evaluate the effectiveness of this available solution.