BART to San Jose (Volume 3): Wicked, Tricksy, False

In a Merc editorial last week, Caltrans director Will Kempton urged us to support BART to San Jose because “transportation professionals with 70 years of combined experience” support the project. Kempton’s piece also proclaims that the inflated ridership projection, in the vicinity of 100,000 daily riders, is “solid,” a statement that appears highly dubious in light of all available evidence. In 2000, the Metropolitan Transportation Commission studied many potential Bay Area capital improvements for transit, ranging from rail projects to rapid bus corridors — including a commuter rail option from Union City BART to San Jose, but the report also contemplated a BART extension south of Warm Springs to San Jose. BART and VTA carried out another study at about the same time, but the two studies yielded vastly different results. MTC’s report produced an approximate projection of 11,500 daily riders. Unimpressive, perhaps, but at least it was honest. Even in 2000, the extension was estimated to cost over $4 billion; at $4 billion for just 11,500 riders, the MTC projection exposed the San Jose extension for the cost-ineffective project it was (and still is).

Vancouver? Or Downtown San Jose in 2030? Photo Credit: Flickr user Tallyn.

Meanwhile, the VTA study projected 78,000 daily riders. The pronounced difference between the two projections is explained by the fact that the latter was the product of a fantastical land use and growth model that was based on a multifold increase in both the residential population and the daytime work population, housed in new skyscrapers that would both blanket downtown and push its boundaries. And while we might be momentarily enchanted with the vision of a truly dense urban center to crown Silicon Valley’s self-proclaimed “capital city,” we cannot endorse predicating the construction of an overpriced BART line on a fabricated growth model that was based on numbers not one inch short of outlandish: 180,000 new residents, 176,000 new workers, and some 144 million square feet of additional office space — all in the space of two decades. In 2005, the projection time scale was cleverly expanded from 2025 to the year 2030, making room for an even higher projection of 111,500 riders that was allegedly based on a more realistic model that used ABAG’s projections for transit-oriented development. And in 2006, the figure retreated to about 104,000 daily riders after the two downtown San Jose stations were consolidated into one longer station box spanning a few blocks of Santa Clara Street.

Low-density residential development just to the north of Berryessa Station.

Okay. 104,000 riders. Six new stations. Shall we try one on for size? Berryessa Station, proposed for a site near the intersection of Berryessa Road and Lundy Avenue / North King Road in San Jose, is projected to have 7,932 boardings by the year 2030. This ridership projection at Berryessa is comparable to the weekday ridership last year at two existing BART stations in Oakland: 19th Street (8,891 riders) and Fruitvale (7,479 riders) — perplexing, to be sure, since the Berryessa site has basically nothing in common with either Downtown Oakland or the Fruitvale District. 19th Street Station serves a large concentration of jobs downtown and an emerging neighborhood and entertainment destination. Fruitvale Station, meanwhile, is located in the midst of a recently constructed transit village (albeit one that still has yet to experience great residential influx), provides convenient access to a substantial commercial district once dubbed Oakland’s second downtown, and it sits adjacent to the most densely populated neighborhoods in the East Bay. Meanwhile, the site of Berryessa Station is adjacent to the Coyote and Upper Penitencia Creeks, near a golf course, the world-famous San Jose Flea Market, and suburban cul-de-sacs. But how about a few actual numbers?

¹ Daily ridership in FY01 (to correspond with Census data).
² Projected daily ridership in 2030.
³ Reported densities (both in this table and throughout this post) are an average of the densities for census tracts adjacent to the station or proposed station site. Data is extracted from the 2000 Census.
⁴ The numbers represent transit lines that connect to each station, in order: (# of local bus lines) / (# of limited/rapid service bus lines) / (# of all-night routes).

The table lists the population density surrounding each station and the number of connecting transit lines, both of which are factors that contribute to a station’s high rider volume. Berryessa fares unfavorably with respect to both factors: significantly lower density, and less connecting transit, both in terms of number of lines and diversity of service. Another factor is proximity of landmarks and other traffic-generating uses near stations. In this respect, Berryessa clearly fares unfavorably once again, but a more important question is whether that will change by the year 2030; more on that later in the post. Yet another factor that is especially important for downtown stations like 19th Street is the concentration of jobs near the station: the Downtown Oakland CBD contained about 63,100 jobs in 2000. Aside from the Flea Market, the current land uses around Berryessa consist primarily of a golf course, low-density and industrial and office buildings, and residential subdivisions. We should also note that a few suburban BART stations last year daily ridership levels comparable to that of the Berryessa projection, notably: Fremont (7,164 riders), Dublin/Pleasanton (7,858 riders), Daly City (8,590 riders) and El Cerrito del Norte (8,213 riders). So: if these stations can achieve this level of ridership, then why not Berryessa? El Cerrito del Norte, which has traditionally been the highest ridership station north of Downtown Berkeley, is a regional bus hub that is served by AC Transit, Golden Gate Transit, Fairfield/Suisin Transit, Vallejo Transit, and WestCAT. Fremont and Dublin/Pleasanton ridership are partly explained by the fact that both are terminal stations that attract riders from a larger surrounding area who seek a convenient ride into Oakland or San Francisco. Ridership at Daly City is explained by higher population density local to the station (comparable to Fruitvale, at about 24,785 persons/square mile in 2000). Notably, Berryessa would not be a regional hub or a terminal station, nor would it share the urban characteristics of 19th Street or Fruitvale.

Near Montague Station.

Let’s consider one more example. The Montague/Capitol station, whose site next to the Great Mall would be a transfer point to VTA light rail, is projected to have 31,010 riders daily in 2030. Just as a point of comparison, the only stations in the entire BART system that currently enjoy this high level of ridership are both located in downtown San Francisco: Embarcadero (36,094 riders) and Montgomery (34,472 riders). Even Powell and Civic Center, which sit within easy reach of large government buildings, Union Square retail, thousands of jobs, universities whose combined enrollment numbers in the thousands of students, and dozens of cultural and entertainment venues — not to mention the densest, most transit-dependent neighborhoods in the western United States (some with densities in the vicinity of 100,000 persons/square mile) — even these stations, at last count, clock in at less than the projection for Montague/Capitol, with 27,897 and 20,313 daily riders, respectively. In contrast, the average population density in 2000 in the vicinity of the Montague/Capitol station site was 5,892 persons/square mile.

We could carry out similar analysis to cast doubt on the ridership projections for the other stations. One station whose projection is more realistic is the consolidated Downtown San Jose Station, with 23,474 riders in 2030. Downtown is a more traditional CBD like that of San Francisco and Oakland, although smaller than both of those in terms of both quantity and density of jobs. (Notably, the proposed BART alignment does not serve the Golden Triangle, which is a considerably larger employment center than downtown San Jose.) Still, when you consider the fact that Oakland’s three downtown stations enjoyed a combined ridership of 27,300 last year, that projection for just one downtown station does not carry the same ring of fabrication that some of the other projections do.

But as I alluded to earlier, one cannot only consider the current land use at the these proposed station sites, which is often severely lacking. After all, these are projections that are based on the assumption that land use near stations will be stepped up, so we would be remiss not to at least touch on station area plans that are in the works. But even here, we have to question whether the level of development under consideration will be sufficient to generate the proposed ridership. For example:

• As pictured above, the developable land near Berryessa Station is immediately constrained to the north and east by low-density residental neighborhoods; similar constraints exist to the west and south, though slightly more distant from the station site. The plan for Berryessa concerns about 350,000 square feet of retail and 2,855 homes on 120 acres. Let’s assume that each unit is constructed and houses, on average, 2.5 persons. Adding these units to the 2000 population counts effectuates a density increase from 7,056 to 9,304 persons/square mile in the station area. Note that this is still much less dense than the BART stations at 19th Street, Fruitvale, and Daly City. Also, 350,000 square feet of retail would not achieve nearly the same concentration of jobs as the several office towers located near Lake Merritt. Whether or not the retail space itself would attract riders is an open question. Recall that the Shops at Tanforan in San Bruno, which sit directly adjacent to the San Bruno BART station, contain over one million square feet; but the BART station only had 2,451 daily riders last year.
• For the Montague/Capitol station, the Transit Area Plan in Milpitas contemplates about 7,109 homes, 287,075 square feet of retail, 175,500 square feet of hotel, and 993,843 square feet of office space — to be sure, a significantly larger scale than Berryessa, but it is still questionable whether the new development (even when combined with the Great Mall patronage and light rail transfers) will truly generate the projected ridership. The area under study include 437 acres surrounding the proposed Montague station site; I-880 and low-density residential uses adjacent to the freeway constrain the area to the west. Adding these residential units to the 2000 population counts for the area effectuates an increase in station area density from 5,892 persons/square mile to a modest 8,502 persons/square mile. Close to one million square feet of office space is considerable; it’s more than the total square footage contained in 600-foot 50 Fremont Center in San Francisco. But the ridership projected for Montague Station is comparable to the two highest ridership stations in downtown San Francisco, an area which contains far more than just a single skyscraper per station.

And finally, in the next (and last) installment: the real issue with BART, and the better way to improve transit in the South Bay.