Seattle traffic is bad, but ranking may be bad, too
Seattle Times staff reporter
Everyone may be wrong.
When the Texas Transportation Institute (TTI) awarded Seattle the No. 2 ranking last year in its widely publicized annual Urban Mobility Report, regional leaders embraced the dubious distinction as a sign of just how serious the Puget Sound region's congestion problem had become, and just how badly action was needed.
They still cite the ranking regularly, often without attribution. It's simply accepted as fact.
But some of this region's leading transportation researchers say the high ranking is misleading. The way TTI calculates congestion makes Seattle's traffic appear worse in relation to other cities than it really is, they contend.
The state Department of Transportation agrees with them. It has helped finance TTI's report for years. Last month, it withdrew its support.
In a May 14 letter, Secretary Doug MacDonald informed the institute it would get no more money from Washington. He also asked it to take his agency's name off the list of sponsors of the 2002 report, scheduled for release next week. "I want to move far, far away from the TTI measures," MacDonald says. "They don't tell us a thing."
This isn't retribution for the high ranking, MacDonald insists, or a backdoor attempt to convince Seattle commuters traffic really isn't a problem. He says he just wants an accurate measure of the effectiveness of his department's efforts to combat congestion, and that TTI's report doesn't provide it.
Some of Seattle's innovations in improving freeway flow may actually penalize the city in the institute's rankings, MacDonald and others maintain.
"The better we do, the worse we look," says Mark Hallenbeck, director of the Washington State Transportation Center at the University of Washington.
Tim Lomax, a research engineer at TTI and co-author of the Urban Mobility Report, says he's disappointed Washington backed out. The state's $40,000 annual contribution is about one-eighth of the study's budget.
But he also says MacDonald and Hallenbeck are right about the limitations of TTI's methodology.
"We have to use the databases that are available, and they don't have the detail," Lomax says. "If we had waited until we had perfect data before we began presenting information, we'd still be waiting."
TTI, part of Texas A&M University, began publishing the Urban Mobility Report in the early 1980s. Since then, the report has become the best-known study of traffic congestion in the nation.
When the city rankings are unveiled each year, the media splash is like a small tsunami.
Last year's report was Page 1 news in Seattle; St. Louis; Denver; Las Vegas; Kansas City, Mo.; Indianapolis; Charlotte, N.C.; Minneapolis; Atlanta; Dallas; Washington, D.C.; Milwaukee; Tampa, Fla.; and San Diego.
It was featured on "ABC World News Tonight" and "NBC Nightly News." Lomax was interviewed live by Bryant Gumbel on "CBS News' Early Show."
TTI's rankings get plenty of attention. The way it calculates them rarely attracts much notice.
The institute measures congestion several ways. Last year, the Seattle-Everett area placed second, behind only Los Angeles, in two of the most closely watched categories: the "travel time index," which compares travel times when traffic is flowing freely with travel times at rush hour, and "average delay per capita," which estimates how many hours a typical resident spends stuck in traffic each year.
But TTI doesn't base its conclusions on what's really happening on freeways at rush hour. Lomax and his colleagues don't consider how fast traffic is actually moving on the Evergreen Point Floating Bridge or the San Diego Freeway at 5 p.m.
Instead, they use a database from the Federal Highway Administration that contains relatively limited information: miles and lanes of freeway and arterial in each city, and the average daily traffic volumes on each road segment. They produce their measures and rankings by applying assumptions and estimates to that data.
That's where the problem lies, critics say.
"It's a relatively gross, overall, blanket kind of measure," says John Niles, senior fellow with the Seattle-based Discovery Institute, whose analysis of TTI's methodology helped persuade MacDonald to pull Washington's sponsorship of the annual report.
"These are crude assumptions," Hallenbeck agrees.
One estimate, in particular, distorts TTI's statistical portrait of Seattle, they say. It underlies all the institute's congestion calculations.
TTI assumes, not unreasonably, that as volumes increase on highways, traffic inevitably will slow. But in calculating that delay, the institute applies the same speed estimates nationwide, from Seattle to Miami.
For instance, if a mile of freeway carries between 15,000 and 17,500 vehicles per lane on an average day, TTI's model assumes traffic at rush hour will travel at 45 mph. If daily volume exceeds 25,000 per lane, it plugs a rush-hour speed of 32 mph for that road segment into all its equations.
Lomax says speed estimates are based on computer simulations and some real-world studies. Niles says they're probably pretty accurate as national averages.
But Seattle isn't average, he and other critics argue: The city has been a leader in finding ways such as ramp metering to keep freeways moving while cramming in more cars.
As a result of those innovations, they say, Seattle freeways probably move more volume at faster speeds than do freeways in many other cities. TTI's methodology doesn't give Seattle credit for that.
"Maybe what we're measuring is efficiency rather than congestion," says Larry Blain, principal planner for the Puget Sound Regional Council and another TTI critic.
The state Department of Transportation plans to do more to squeeze more cars onto freeways, effectively increasing capacity by increasing efficiency. Last month, for instance, the agency announced plans to double the number of tow trucks along Seattle-area highways to pull disabled vehicles out of traffic.
But, perversely, if the trucks succeed in clearing accidents more quickly and volumes increase as a result, Seattle's TTI ranking could suffer.
"Growth in daily traffic volumes (is) automatically calculated by the TTI computer model as greater delays, whether or not this tells the story of what is actually happening on the highways," a recent state Department of Transportation memo charges.
The institute's methodology rewards only one approach to coping with greater volumes, the memo says: building more roads.
The only apples-to-apples data
Lomax doesn't disagree with any of this. The volume/speed estimates TTI uses are "relatively simplistic," he admits: "The kinds of efficiencies that Seattle is doing don't get accounted for."
The problem is lack of information, he says. The database used, while "less than perfect," is the only one that allows apples-to-apples comparisons across the nation. Information on what's actually happening on the road at rush hour in every city — data that could make Seattle's case — isn't available yet in any comprehensive form, Lomax says.
Hallenbeck, a longtime acquaintance of Lomax's, agrees. TTI's database "may be rotten apples to rotten apples," he says, "but it's all there is. Everything else is kumquats and pomegranates."
TTI shares its critics' concerns and is trying to address them, Lomax says. It is gathering and analyzing information on real-world freeway and arterial performance from cities that collect such information. A report is due this fall.
Lomax says he spoke with the Washington Department of Transportation about adding a section to this year's Urban Mobility Report discussing methodological limitations, such as its failure to account for ramp metering and other operational improvements.
That offer got sidetracked as the state weighed whether to continue its sponsorship, Lomax says. "I'm really disappointed that they pulled out."
MacDonald is unrepentant. Congestion can be measured more accurately and more usefully than how TTI does it, he says. One example: The department's new Web page that provides "real-time" travel times for 11 common commutes, updated every five minutes with information from detectors embedded in the freeway pavement. (See www.wsdot.wa.gov/pugetsoundtraffic/traveltimes/)
Traffic is a big problem in the Puget Sound region, MacDonald says: It doesn't really matter if Seattle ranks second, third or 70th.
But it makes little sense for the state to invest its limited research dollars in a report that doesn't measure some of its key initiatives, he says.
"The public needs to know what we are doing to fight congestion, and whether it's working or not. Everybody gets so mesmerized by TTI, but it doesn't tell us what we need to know."
Eric Pryne can be reached at 206-464-2231 or firstname.lastname@example.org.