How close to rail transit projects come to meeting their promises of being completed on budget and attracting the projected number of passengers? If you listen to transit agencies, almost every project is completed on time and beats its ridership goals. But those numbers aren’t very reliable as the transit agencies base their claims on projections made shortly before the projects were completed, not when the decision was made to build them.
In the early years of the Bush administration, the Federal Transit Administration commissioned a study to find out how well they were doing. The study was completed in 2003 — and the FTA then sat on it for four years. Now, they have finally released it, and you can download it here (4.4 MB Word file).
Is this light-rail project really necessary?
Flickr photo by brewbooks.
The report was “primarily authored” by Frank Spielberg of SG Associates and Steven Lewis-Workman of the FTA. Based on a review of more than 20 different projects, the report finds that things have improved since US DOT research Don Pickrell released his 1989 report showing that most rail projects went way over budget. But there are still significant problems.
Projects that began in the 1980s, the report estimates, went an average of 40 percent or more over budget. By the 1990s, this was down to around 20 percent. However, this should not be much cause for cheer among rail proponents. For one thing (as documented here), some newer projects that were not evaluated in the Spielberg report have gone far more than 20 percent over budget. The Hudson-Bergen light rail, whose planning began in 1993, went 78 percent over budget, while the San Juan Tren Urbano, whose planning began in 1995, went 113 percent over budget. Seattle’s light-rail project, now under construction, is breaking records for light-rail cost overruns.
The patient this website on sale now online viagra was at the hands of Pfizer, a multinational company the origin or which is in USA. Urinary infection is caused due to bacteria that can have an effect on any part of cheapest cheap viagra the body. You might have not given a thought at least once about shop viagra the process by which food is converted into energy by the body. The Nightforce ATACR is viagra ordination laced with multi-coated ED glass elements Nightforce ATACR rifle scopes come with a variety of features and can be used for this purpose. 6. Spielberg and Lewis-Workman find that, on average, projections of operating costs were on target. But this average betrays a wide spread. The cost of operating two projects — the Jacksonville Skyway and the San Jose Guadalupe light rail — were more than 70 percent more than projected, and a San Francisco BART extension cost nearly 40 percent more than projected.
Several others cost 40 percent or more less than projected. However, some of the numbers sound suspicious: can Baltimore really run a light-rail line for less than 2 percent of the projected cost? According to the report, Baltimore planned to run the line every five minutes but is only running it every eight minutes, but that hardly seems enough to reduce operating costs by 98 percent.
If projections of operating costs turned out to be okay on average, projections of ridership were drastically overestimated. On average, actual ridership is falling short of projections by 35 percent. To be fair, most of the projections were for 2005 while the actual numbers were for 2002. But most transit agencies did not experience a huge increase in riders between 2002 and 2005, so updating to 2005 numbers would not significantly improve the results.
The report does not compare projections of new riders with actual new riders, giving the excuse that new riders are difficult to calculate based on published numbers. But we know that the Dallas light-rail line that opened in 1996 seemed to generate some new riders, while the line Dallas opened in 2001 did not: they lost as many bus riders as they gained rail riders.
The St. Louis light-rail extension that opened in 2001 did even worse. While the Spielberg report indicates that it carries an average of 16,000 passengers a day, the sad truth is that total light-rail ridership hardly changed after the line opened while bus ridership fell by more than 15 percent. (See the Dallas and St. Louis pages of this worksheet for actual numbers.)
On top of the cost overruns and ridership shortfalls, the report notes that the average transit project took eight years to complete — that is, eight years after publication of the environmental impact statement. The median was six years. That’s a long time to wait for supposedly improved transportation.
The Spielberg report is nearly 200 pages long, most of it devoted to an in-depth evaluation of each of the 21 projects it examines. This should prove a valuable source of information when looking at new rail transit proposals.
Would you have any explanation as to why St. Louis has a slight downtick in 2004-2005 MDVMT while an uptick in transit PM? Honolulu’s planners point to St. Louis as an example of where rail has reduced traffic congestion and I am only guessing they point to this statistic (of course the overall picture is dismal).
A quick look at the FTA report begs the question of the relationship between actual capacity offered and capacity used. Unlike the Pickrell document from 1990, this latest analysis at least mentions the issue of capacity, but unfortunately doesn’t go beyond description into real analysis, such as identifying the “best practices” leading to accuracy in capital and operating expenses, and matching or exceeding ridership forecasts. On the other hand, the report spares us the sort of inane editorializing that Pickrell engaged in.
As Demery and I have contended for quite a long time now, a major shortcoming of the AA/EIS-FEIS process is a consistent failure to consider comparing the levels of peak period crowding people will consider, as opposed to the unrealistic capacities claimed by transit vehicle manufacturers. In turn, vehicle capacity in actual practice and the level of patronage that can be expected directly determines the fleet size needed. In a number of cases, the fleet sized allowed by FTA turned out to be significantly smaller than demand required. Failure to consider these factors probably explains much of the under-achievement in meeting ridership forecasts.
To cite Randal’s least favorite example, the MAX East Line didn’t actually achieve its predicted patronage of 42,500+ per day until after 1998, when enough LRVs became available at the time the Hillsboro line opened, after adjusting for Airport Line ridership beyond the Gateway station. A similar situation occurred in Salt Lake City, when the initial small LRV fleet was pushed to its limits upon the original line opening in 2000, and then again when the University Line opened, leading Salt Lake several years ago to purchase 25 used LRVs from Santa Clara (Sacramento bought the other 25).
Peterkay,
Transit’s market share in St. Louis was nearly 1.5 percent in 1982, declining to less than 0.6 percent in 1993. The opening of the light rail in 1993 led transit’s share to increase unsteadily all the way to 0.77 percent in 1999.
Opening a new light-rail line in 2001 did not prevent transit’s share from declining to 0.68 percent in 2004. It may have ticked upward slightly to 0.70 percent in 2005, but still remains below the 1999 level, much less the 1982 level.
So what does this mean? Light rail’s frequencies and speeds can get a few people out of their cars. But it is very hard to sustain this, even by building new rail lines. The 2005 uptick could be the start of a great new trend for rail, but may be just due to high gas prices. If so, we’ll see another uptick in 2006, but not much more after that.
You certainly can’t consider St. Louis as a great example of a region where rail transit has relieved congestion. Those who say so obviously spend too much time reading transit agency press releases and not enough time checking out the real numbers.