The U.S. is working harder to ensure Sound Transit and other regional agencies include high budget reserves and avoid overruns, says Bent Flyvbjerg, a groundbreaking Oxford scholar.
Transit agencies are getting better at making realistic cost estimates, offering hope that programs such as Sound Transit 3 (ST3) can avoid disastrous overruns, a leading world expert says.
The U.S. Federal Transit Administration (FTA) in particular is taking seriously its mission to supervise local projects that receive federal grants, following a string of embarrassments years ago, said Bent Flyvbjerg, professor of management at Oxford University.
“It’s an improvement from a disastrous situation” in the past, he said in a phone interview Monday. Yet the state of worldwide megaprojects has merely progressed from “extremely high overruns” to “big overruns,” he said.
Seattle-based Sound Transit is asking voters this fall to raise taxes for a third round of projects, at the same time its initial light-rail corridor from the University District to SeaTac is nearing completion — at a cost 86 percent higher than the $1.7 billion pitched to voters on the 1996 ballot, The Seattle Times reported Sunday.
Most Read Local Stories
- Coronavirus daily news updates, April 10: What to know today about COVID-19 in the Seattle area, Washington state and the world
- A reckoning is due for Seattle's dark side, as hate crimes and bias incidents soar 63%
- Three North Seattle light-rail stations to open Oct. 2
- Cops for $1,000 a day: How Seattle spends millions hiring off-duty police officers but does little to monitor their moonlighting
- Gov. Inslee alters criteria for COVID-19 reopening phases
That overrun is nearly double the worldwide average a Flyvbjerg-led team found in a landmark 2002 study, including 58 rail projects that blew their budgets by an average of 45 percent.
But Seattle-area projects have tracked close to Sound Transit’s revised estimates made during an agency reorganization in 2001, and for second-round projects in its 2008 ballot measure.
U.S. transit data show average overruns shrank from 54 percent, on an agency’s first project, to 17 percent on later extensions, according to Carole Turley Voulgaris, a Ph.D. candidate at UCLA not affiliated with Flyvbjerg’s research.
Flyvbjerg said he agrees with her findings that agencies learn as they build new lines, based on his newer data. Nonetheless, his expanding database of 264 world rail projects shows an average 40 percent overrun worldwide.
At $54 billion including seven light-rail extensions, commuter trains, buses and park-and-ride garages, the 25-year ST3 program dwarfs the 1996 plan. Even opponents say the numbers are so high that overruns are unlikely.
Some factors to consider:
• Most of ST3’s track miles in Lynnwood, Everett, Federal Way, Bellevue and Tacoma are proposed as elevated spans, and the agency knows how to build those. Contractors previously built 7 miles overhead in Tukwila and SeaTac on time and on budget.
• FTA has awarded $2.3 billion already in grants for construction to segments to Tukwila, the University of Washington and Lynnwood, and former FTA administrator Peter Rogoff now is CEO at Sound Transit, so the feds will keep an eye on projects here.
The FTA has compiled cost data to check the accuracy of proposals and requires an after-opening studyof each line’s performance, a federal spokesman said. The feds showed some bite in Honolulu this year by putting $1.55 billion on hold for that city’s elevated-rail project, until leaders there find cost savings.
• The ST3 numbers include a budget cushion, for both individual segments and the entire regional program, totaling up to 32 percent.
“You need a very high risk contingency,” Flyvbjerg said. Government sponsors have gotten into trouble in the past by limiting contingencies to 10 or 15 percent at the outset to make projects look affordable, he said.
• Perhaps the trickiest piece of ST3 is a new tunnel from Stadium Station in Sodo to east downtown and Uptown, which would pass under the existing transit tunnel north of Westlake Station, and enlarge the hub station there — something Sound Transit hasn’t tried yet. There’s also a deep station proposed in South Lake Union, next to where tracks must dive under Highway 99.
For uncommon jobs like the Westlake passage, Flyvbjerg urges “reference class forecasting,” where planners must scour the globe to find similar work and budgets. Otherwise, leaders are tempted to show a “uniqueness bias” and settle for budgets based on conjecture.
“The general public are more concerned about cost overruns than costs,” he said. “If you get a more realistic cost at the outset, people tend to respect that.”