#3: How VTA went off track, how to get it back in gear
VTA's budget woes and dubious expenditures were all the news in 2020, but these were just the latest faceplants for one of the worst-performing transit agencies in the U.S. Randal O'Toole of the Thoreau Institute explored the issue in this deep-dive, exclusive analysis for Opportunity Now, which social media commentators described as "the only must-read analysis of local transit woes."
Part 1: A Loser Among Losers
The coronavirus has sent shock waves throughout our economy, and one of the most affected parts is public transit. While driving has recovered to more than 90 percent of pre-pandemic levels, as of September transit ridership was still less than 37 percent of what it had been the previous September.
With more people than ever working at home, many believe that the transit industry will never completely recover its lost customers. This creates an opportunity to rethink that industry and all of the subsidies it has received from taxpayers over the years.
Even before the pandemic, the transit industry as a whole was one of the worst-performing industries in the nation.
• On average, its customer revenues covered less than a third of its operating expenses and none of its capital costs.
• Tax subsidies required to keep transit running grew from $15 billion (about $26 billion in today’s dollars) in 1990 to nearly $59 billion in 2019.
• Despite having received a total of roughly $1.5 trillion in subsidies since 1970, per capita ridership declined from 49 trips per urban resident in 1970 to 37 trips in 2019.
• Worker productivity has fallen from more than 50,000 trips per operating employee in 1970 to under 25,000 in 2019.
• Despite this poor performance, the industry routinely spends billions of dollars a year on capital improvements that do little to increase ridership or revenues.
Such continuing and growing losses are enabled by taxpayer subsidies that transit agencies justify by claiming that transit provides mobility for low-income people, helps protect the environment, and relieves traffic congestion. However, close scrutiny reveals that none of these claims are true in the vast majority of urban areas in the United States.
Among the more than one thousand transit agencies that serve the nation’s urban areas, the Santa Clara Valley Transportation Authority (VTA) stands out as one of the worst performers among a crowd of losers. The percentage of its operating costs that are covered by farebox revenues is well below average. It’s light-rail cars are some of the emptiest in the country. Yet it has spent and continues to spend billions of dollars on pointless capital improvements that do little for the county’s transportation. The Federal Transit Administration’s National Transit Database, which includes information on ridership, service, finances, and energy consumption, makes it possible to compare VTA against its peer agencies.
Farebox Recovery
In 2019, the database reveals, the transit industry as a whole collected enough fares to cover 32 percent of its operating costs. VTA fares, however, covered just 9 percent of operating costs. VTA’s light-rail fares covered less than 7 percent of light-rail operating costs. Buses did a little better, covering 10 percent of operating costs. VTA’s paratransit program was average, covering 9 percent of its costs.
VTA’s farebox recovery has always been well below the industry average, but it has gotten worse in recent years. In 2011, fares covered 14 percent of operating costs. Since then, operating costs have grown by 45 percent while fare revenues have shrunk by 7 percent.
The 45 percent increase in operating costs since 2011 was translated into a modest 7 percent increase in transit service (measured in the number of miles buses and railcars operate in revenue service). Despite the increase in service, bus ridership dropped by 13 percent and light-rail ridership dropped by 16 percent.
Filling Seats
To accomplish any of its social objectives, a transit agency must be able to fill the seats it offers to the public. This can be measured by dividing the passenger miles carried by the vehicle-revenue miles of service offered.
Nationwide transit occupancies have declined as most transit agencies have maintained or increased bus and rail services despite declining ridership. The average transit bus carried 13.0 passengers in 1980, and in some cities the average was as high as 20.0. By 2019, the nationwide average was just 8.0. Light rail didn’t really exist in 1980, but since 1995 light-rail car occupancies declined from 25.3 to 20.5 in 2019.
VTA does slightly better than average when it comes to filling bus seats, filling an average of 8.6 seats on its buses in 2019. But VTA filled only 13.9 seats on its average light-rail car, well below the national average. Since the national average is already below what it once was, VTA’s light-rail performance is an embarrassment for the industry.
Environmental Quality
VTA makes so little contribution to Silicon Valley’s environment that, from a green standpoint, the region would be better off without it. Nationally, the average car consumes less than 2,900 British thermal units (BTUs) of energy and emits about 200 grams of carbon dioxide per passenger mile. Greenhouse gas emissions from cars in Santa Clara County are well below the national average due to the high percentage of electric vehicles used in the region.
VTA’s transit services consume far more energy and its buses are huge generators of greenhouse gases. In 2019, VTA’s light-rail system consumed well over 5,100 BTUs per passenger mile, compared with the national average for light rail of less than 4,000. Because little of California’s electricity is generated from burning fossil fuels, only about 100 grams of carbon dioxide were emitted per light-rail passenger mile. As noted, any electric or plug-in hybrid car is likely to do better than that.
In 2019, VTA’s buses consumed nearly 3,600 BTUs per passenger mile. This is below the national average for buses, which was 4,800, but well above the national average for cars and even SUVs (which was about 3,300 BTUs per passenger mile). VTA buses also emitted 260 grams of carbon dioxide per passenger mile, again below the national average for buses (350) but above the average for either cars or SUVs (240).
Ridership Trends
Even before the pandemic, nationwide transit ridership had been declining, having fallen 8 percent in the five years between 2014 and 2019. Once again, VTA is worse than average, having lost 19 percent of its riders in that same period. Clearly, VTA is not serving the needs of the region’s transportation users.
Another measure of how well a transit agency services its region is the number of trips carried per resident of that region. Transit in the New York urban area, for example, carried 230 trips per resident in 2019.
Nationally, transit carried 40 trips per urban resident in 2010, declining by 9 percent to 37 trips in 2019. Counting residents of the San Jose urban area, VTA carried 26 trips per resident in 2010, dropping by 21 percent to 20 trips in 2019.
Serving Low-Income People
The Census Bureau’s annual American Community Survey reveals information about commuter incomes. The survey breaks out commuters by income ranging from under $10,000 a year to more than $75,000 a year.
The 2019 survey shows that less than 5 percent of Santa Clara County workers take transit to work, but they tend to have higher incomes than average. About 46 percent of Santa Clara County workers earn more than $75,000 a year, but more than half of all Santa Clara County transit commuters earn more than this amount.
The median income of Santa Clara County workers was $65,424 in 2019, but the median income of Santa Clara County transit commuters was $75,441. That’s more than the median income of Santa Clara County residents who commute by any other method or work at home.
Some low-income workers in Santa Clara County take transit to work, but not many. Fewer than 12,000 commuters with incomes below $25,000 a year relied on transit in 2019. That’s just 6 percent of that income category.
Most of the local subsidies to VTA come from sales taxes, which are regressive. This means that 94 percent of low-income workers are disproportionately paying for a transit system they don’t use that disproportionately serves high-income commuters. VTA transit is one of the most socially unjust institutions in Santa Clara County.
Conclusions
VTA is not the worst-performing transit agency in the country, but it is well below average. At 9 percent, its farebox recovery rate is the second-lowest of the nation’s fifty largest transit agencies. The only one lower is the New York Department of Transportation whose main transit service is the Staten Island Ferry, for which there is no fare. Its per capita ridership is well below average and is declining faster than the national average. The system does more harm than good to both the region’s environment and to its low-income workers. Truly, the hundreds of millions of dollars of tax revenues that VTA collects in the region cannot be justified in any way.
An Analysis of VTA, Part 2: Why Is VTA So Bad?
The Santa Clara Valley Transportation Authority is one of the worst-performing transit agencies in the country, and that’s saying a lot in an industry in which terrible performance is the norm. The reasons why VTA stands out as such a poor performer have to do with Silicon Valley’s growth pattern and the agency’s fascination with expensive yet inappropriate rail transit systems.
Transit and Job Densities
Many people believe that population density is the key to high transit ridership. Through the urban-growth boundary and the often-subsidized construction of hundreds of high-density housing projects in Santa Clara County, urban planners have tried to increase the region’s density.
In fact, they succeeded, as in 2019 the San Jose urbanized area had, at 6,300 people per square mile, the third-highest density of all urban areas in the United States. Only Los Angeles (7,300) and the San Francisco-Oakland area (6,800) were higher; the New York urban area (which includes northern New Jersey, most of Long Island, and urban parts of Westchester County) had the fourth-highest density at 5,400 people per square mile. Yet the region’s transit ridership is low despite its high population density.
The real key to transit ridership, it turns out, is not population density but job density. Transit works best when it is a hub-and-spoke system with lots of jobs at the hub allowing downtown workers who live just about anywhere in the region to get on a bus or train and go to work without having to transfer.
In the United States, transit works best in the New York urban area, carrying 30 percent of the region’s commuters to work in 2019, because there are nearly 2 million jobs located in a small part of south Manhattan. Transit works moderately well in Boston, Chicago, Philadelphia, San Francisco-Oakland, Seattle, and Washington DC, taking more than 10 percent of commuters to work, because these areas all have more than 200,000 downtown jobs. No other urban area in the United States has anywhere close to 200,000 downtown jobs, and in no other urban area did transit carry 10 percent or more of commuters to work in 2019.
Statistically, the correlation between per capita transit ridership and the number of downtown jobs in the nation’s 66 largest urban areas is a near-perfect 0.87 (a perfect correlation would be 1.0 while 0.0 is entirely random) The correlation between ridership and population density is only 0.55, and probably much of that is because areas with high population densities tend to have lots of downtown jobs.
San Jose’s low transit ridership is directly related to the region’s wide dispersal of jobs. Despite its high residential density, an analysis of census data by demographer Wendell Cox found that fewer than 29,000 jobs, or 2.7 percent of those in the region, are located in downtown San Jose. This means transit cannot work as a hub-and-spoke system.
To its credit, VTA’s predecessor agency, the Santa Clara County Transit District, made some innovative attempts to compensate for the region’s lack of a concentrated job center. Early in its history, it started a dial-a-ride system that allowed anyone to telephone the agency to have a shared vehicle come to their door and drop them off at their destination. Essentially, this was Uber Pool or Lyft Shared before smart phones.
One of the problems with this system was that it was too successful. Demand swamped the transit center’s phone lines and most people who wanted rides were unable to get them. The region’s taxi systems sued the transit agency saying that the door-to-door service the agency offered was unfair competition. When an analysis revealed that the dial-a-ride cost per rider was higher than that of ordinary bus service, the transit district placated the taxi companies by dropping the service. Today, with smart-phone apps, VTA could probably offer a service at a cost competitive with its buses, but it hasn’t tried.
The other thing the transit district did was give up on the hub-and-spoke model. It had taken over transit service from three private bus companies—Peerless Stages, Peninsula Transit, and San Jose City Lines—all of which focused their services on downtown. But the transit district designed its bus routes to be more of a grid system, with many east-west and north-south routes that did not go downtown. Theoretically, this system should allow transit riders to get anywhere in the region while transferring from one vehicle to another no more than once. But low ridership numbers today show that this doesn’t work any better than the hub-and-spoke model for a region whose jobs are spread throughout the area.
The Light-Rail Fallacy
The transit district compounded the problems of Silicon Valley’s transit-unfriendly urban design when it decided to build light-rail lines. In 1978, Ron Diridon, who is sometimes called the father of Silicon Valley’s light-rail system, argued that, “a light rail vehicle carries four times as many people as a packed bus, three times as quickly. That gives us 12 times the commute productivity for only about two-thirds of the cost.”
As it turns out, Diridon was wrong about all of those numbers. A packed bus carried about 60 people while a light-rail car carries about 150, which is less than four times as much. But neither of these numbers matter much if there isn’t enough demand for transit to fill either kind of vehicle. While VTA’s light-rail cars may carry more people during rush hour, the overall average of 14 is less than 10 percent of the rail cars’ capacities. Since light-rail systems in New Jersey (Hudson-Bergen), Phoenix, and Seattle carry, on average, more than twice as many people as VTA’s, San Jose’s light-rail cars are probably not exceeding the capacity limits of a bus at any time.
Second, light-rail cars do not go three times as fast as buses. VTA’s light-rail cars average about 15.6 miles per hour. While that’s a little faster than its buses, which average 11.6, it is not three times as fast.
Third, light rail does not cost two-thirds as much as buses to operate. In 2019, VTA spent $16.82 per vehicle-revenue mile operating its buses while it spent $36.13 per mile operating light rail. That’s just counting operating costs. When capital costs are amortized and added to operating costs, the cost of each bus rider is under $10 while the cost of each light-rail rider is more than $26.
The real problem with light rail is that, given Silicon Valley’s widely dispersed jobs, big-box transit makes even less sense than 40-foot buses. The transit district was on the right track when it experimented with dial-a-ride, which used smaller buses. If the district could have overcome the call system problems, it might have been able to reduce costs to be less than its fixed-route buses. But the objections from the taxi companies prevented that from happening. Going from dial-a-ride to light rail was going from a system that had bugs in it to a system that absolutely did not apply to Silicon Valley, like using a supercomputer as a word processor.
Diridon’s support for light rail was based on his optimism bias, a common problem in government planning in which planners make overly optimistic assumptions about future costs of and demand for the projects they want to build. The opening of San Jose’s first light-rail line in 1988 should have revealed that Diridon’s cost and ridership assumptions were not valid and ended the experiment with big-box transit. That didn’t happen partly due to continued optimism bias but also due to the creation of VTA in 1995.
Congestion Management
In 1984, Santa Clara County voters agreed to a ten-year half-cent sales tax for building new highways in the region. This money was used to, among other things, build a freeway on state route 85, expand state routes 87 and 237, and add lanes to U.S. 101 and Interstate 880. While there are reasons to believe that new roads should be funded out of user fees rather than general taxes, it is clear that this new construction led to an actual decline in arterial congestion in the region, showing that it is possible to build for a region to build its way out of congestion.
In 1990, state voters agreed to proposition 111, which increased gas taxes in order to relieve congestion. The measure also required each county in the state to form a congestion management agency that was supposed to find the most cost-effective ways to relieve congestion.
In 1995, Santa Clara County decided to merge its transit district with its congestion management agency, thus forming VTA. This created a conflict of interest. If VTA decided to relieve congestion by building roads, most of the money for those roads would go to some other agency. Only if VTA decided to relieve congestion through transit would VTA get to keep the money itself.
As a result, when the ten-year sales tax that voters agreed to in 1984 expired, VTA put a measure on the 1996 ballot to keep the sales tax for congestion relief. But it planned to have only half the money go to new roads; the other half would go to transit including new light-rail construction. In fact, light rail is more likely to increase congestion, but voters approved the extension of the tax. When that tax expired, VTA convinced voters to extend it for 30 years and to dedicate all of it to transit and none to roads.
This left VTA flush with cash to build new light-rail lines, whose mileage doubled between 1996 and 2006. The 30-year time period for the sales tax also allowed VTA to fund the BART extension to San Jose.
San Jose’s modern urban form meant that transit, especially big-box transit, doesn’t work in the region. Despite this, optimism bias led the transit district to build a light-rail line. Perverse incentives created by merging the transit district with the congestion management agency led to VTA empire building as it built more light-rail lines despite their proven unsuitability to the region. This has led VTA to where it is today: a high-cost, low-productivity transit system dependent on heavy taxpayer support despite the fact that it provides few benefits for anyone who is not a contractor or union construction worker.
An Analysis of VTA, Part 3: Fixing the Problems
The Santa Clara Valley Transportation Authority (VTA) is one of the worst-performing transit agencies in the country, partly because Silicon Valley is unsuited for a traditional transit model and partly because VTA is focused more on expanding its rail empire than it is on providing cost-effective transportation.
The Dot-Com Crash
The dot-com crash of 2000 revealed one of the problems with empire building a rail transit system. VTA had borrowed heavily to build new light-rail lines in anticipation of future sales tax revenues. But between 2001 and 2003, sales tax revenues declined by 28 percent and most of VTA’s other revenues declined as well. Regardless of the lost revenues, VTA still had to meet interest payments on light-rail debt, and those payment more than doubled because of the new borrowings.
The situation got so bad that in 2002 VTA was forced to sell land that it had intended to use for a park-and-ride station. “VTA is completely illiquid,” said a VTA board member. “They sold that piece of property to make payroll.”
Between 2001 and 2003, VTA had to cut light-rail service by a third and make cuts to bus service as well. By 2005, VTA had lost a third of its riders and most of them never came back.
In 2007, the VTA board commissioned an outside auditor called the Hay Group to look at the agency’s finances. “VTA has lost its regional focus and strayed from its core business,” the report concluded. It built capital projects without ensuring that it had funding to operate and maintain those projects. Those projects were “political solutions to address the needs of certain local neighborhoods at the expense of regional congestion management.” The result is that “VTA has built transportation systems that have low ridership and are also expensive to operate and maintain.” The Hay Group cited VTA’s low farebox recovery as evidence of these mistakes.
This 38-page report itself could be cited as an example of VTA mismanagement, as it cost VTA $500,000 or more than $13,000 per page to tell the board things it should already have known.
In response to the report, VTA’s chief financial officer resigned. VTA replaced him with a new temporary CFO who had never worked in the transit industry or in any transportation organization; instead, he had worked in the mining industry. Despite his lack of qualifications, VTA agreed to pay him $530,400 for 39 weeks of work—that’s about $13,600 a week. Such shenanigans led some people to describe the VTA as “the nation’s worst-managed transit agency” and a 2019 Santa Clara County Grand Jury report seemed to agree.
The BART Fiasco
Instead of changing its policies in response to the Hay Group report, VTA completely ignored its implications as it planned to build a BART line to downtown San Jose. An early study of this line found that it would cost taxpayers $100 for every trip taken on the line that would otherwise have been taken in an automobile.
Shortly after the Hay Group report was presented to the board, VTA published an environmental impact report for the BART extension. Among other things, the report looked at the line’s impact on traffic congestion by dividing up the region’s freeways into dozens of segments and estimating the amount of rush-hour traffic those segments would carry with and without BART.
Without BART, the report found, the average freeway segment was expected to carry about 10,000 cars per hour. BART, however, was projected to take an average of 59 cars an hour off of those freeways, or about 0.6 percent.
More specifically, the report found that BART would reduce traffic by 2 percent on 60 percent of the freeway segments, but it would increase traffic by an average of 1.3 percent on the other 40 percent. Even 2 percent was not enough to measurably relieve congestion. Traffic engineers classify traffic with a letter grade of A through F, with A meaning little traffic and F meaning stop-and-go traffic. BART did not change the letter grades of any of the freeway segments.
Despite this, and completely contrary to the recommendations of the Hay Group report, VTA went ahead with its plans to build BART. This put it further into debt and made it more vulnerable to future recessions. Meanwhile, transit ridership continues to drop.
Fixing VTA
VTA’s problem is the same as described in the Hay Group report: instead of providing transportation, it is fixated on benefitting special interest groups such as certain neighborhoods (including downtown San Jose) and contractors and consultants by building urban monuments. This is a problem that afflicts many other large transit agencies, but VTA is one of the most extreme cases.
VTA’s problems will not be solved by redefining how its board of directors is selected or by hiring a new CEO or CFO. Those actions would be merely rearranging the deck chairs. Instead, VTA requires a wholesale restructuring.
The congestion management agency should be split away from the transit agency, thus ending the conflict of interest. The transit agency should be completely disbanded, ending all bus and rail service, and replaced by a new entity, with new people, new goals, and new activities. This agency should not be in the transportation business, but in the problem-solving business, specifically focused on problems related to transportation. For example:
• Problem: Some people are transit-dependent because they don’t have enough income to buy even a used car, partly because banks charge nearly 20 percent interest to people with poor credit ratings who want to buy a car. Solution: Santa Clara County should start a program of providing low-interest loans to low-income people who want to buy a car. In fact, interest should be completely waived to low-income borrowers who make their payments on time. Studies show that car ownership, not transit subsidies, will do far more to help people out of poverty because, in the San Jose area, a 20-minute auto drive can reach twice as many potential jobs as a 60-minute transit ride.
• Problem: Some people are transit-dependent because they can’t drive, either because of age, disabilities, or they lost their license. Solution: Provide such people with means-tested transportation vouchers that they can use on transit, systems like Uber or Lyft, or other common-carrier transportation.
• Problem: The California legislature has set a target of reducing greenhouse gas emissions. Solution: Santa Clara County could join the state in providing tax credits to people who buy electric or plug-in hybrid electric cars.
• Problem: Congestion is still a serious issue in Santa Clara County. Solution: Congestion is a problem because roads are poorly priced. Moreover, highways have the unique attribute that their productivity declines when demand increases. Freeway lanes that can move 2,000 or more vehicles an hour in free-flowing traffic can only move 1,000 or fewer vehicles an hour in congested traffic. That means that road pricing can effectively double the capacity of highways during rush hour. Instead of tolling people off the road, as toll critics complain, road pricing effectively tolls people onto the road. One caveat: the revenues from road pricing should be dedicated to highway maintenance and, if a surplus is available, new road construction and not, as it is now, diverted to subsidize VTA or other transit agencies.
If Santa Clara County were to get out of the business of providing subsidized transit, private transit would take its place in some corridors. This might be in the form of “Google buses” provided by specific employers or by private transit companies such as Bridj and Chariot, both of which attempted to provide bus services in major cities but failed because they couldn’t compete against heavily subsidized transit agencies.
For the most part, however, transit in general and transit subsidies in particular are not the solution to the region’s transportation issues. Transit subsidies do not help poor people who disproportionately end up paying the taxes needed to provide those subsidies. Transit subsidies do not produce greener transportation, especially when the transit vehicles are nearly empty. Transit subsidies to not relieve congestion; in fact, they may actually make it worse if they put nearly empty buses and light-rail cars in the streets.
This may sound drastic, but realistically VTA is a failed agency in a failed industry that is based on an obsolete business model that only worked when there were a lot of centrally located jobs, such as in America of more than 100 years ago. It’s time to recognize this failure and develop something that works.
Follow Opportunity Now on Twitter @svopportunity.
Photo taken by Logan Sakai.