With policy makers scrambling to make their projects "shovel ready" for Obama's new New Deal and with end-of-the-year decisions (supposedly) imminent on the multi-billion-dollar Alaskan Way Viaduct replacement options, it's worth it to revisit research on why large projects cost more than promised and deliver less. An article from earlier this year in Miller-McCune magazine, "Derailing the Boondoggle," provides some good guidelines for how to be a skeptical consumer of megaproject promises.
David Brewster mentioned this story when it first appeared in August because the magazine listed Sound Transit among eight international megaprojects (costing $1 billion or more) that were major boondoggles. These are projects that cost way more than planned, often take longer than expected to complete, and have fewer benefits than pledged.
But the story wasn't simply about failure. It looks at the work of a Danish professor, Bent Flyvbjerg, who has become a world authority on such projects. He has made detailed studies of how they go wrong. Are there ways we can avoid a new Big Dig in Seattle?
Whether cost-overruns can be avoided is the first real question, and Flyvbjerg's research on this has gained international attention. Basically, the answer looking at history is: Usually not. The article reports that "[i]n a worldwide study of 258 rail, bridge and road projects over 70 years, Flyvbjerg found that nine of 10 went over budget. In another analysis, he showed that the average transit system carries less than half as many riders as preconstruction forecasts predicted."
Expanding on that work, Flyvbjerg began building a database on projects around the world to see what the patterns were:
A decade later, by collecting data from 20 nations on five continents, Flyvbjerg...produced the first statistically significant analysis to show [that] the vast majority of public works projects go drastically over budget and aren'êt as well patronized as proponents claim. He also found that modelers didn'êt seem to be improving their estimates over time; the scale of overruns remained relatively constant. Rail and highway projects are often the worst boondoggles, and they form the bulk of Flyvbjerg'ês research.
Why the flat learning curve? Why the repetition of error in project after project? Granted they are complicated, sometimes unique, but is there something else? Turns out, the perp is who you'd first suspect: politicians, public officials, and the folks who work for them (and value their employment). Researchers have found that projects are victim of several forces. One is so called "optimism bias," which results when proponents get carried away with the vision and downplay the risks (insert picture of Joel Horn here). Politicians and public officials like to say "yes" rather than "no," but they have incentives to push ahead. It makes for good politics:Local officials predict low costs and big benefits to persuade skeptical citizens and to compete with other local governments seeking federal funds. Flyvbjerg calls the result survival of the unfittest: Instead of approving the best projects, officials end up funding those that look best on paper. And by the time accurate figures rear their ugly head — and megaprojects routinely last longer than a decade, from conception to completion — the officials who launched them are long gone from office.
So the pols get maximum immediate benefit with little long-term exposure once the real bills start flowing in. By that time, there's a large hole in the ground that has to be filled.
Given such a climate, the article observes, planners are inclined to tell officials what they want to hear. Pessimists or skeptics get bounced from the payroll, or at best ignored, while everyone who is on board is rewarded with work, contracts, and jobs (that's the economic stimulus we keep hearing about). The bottom line, then, is that megaprojects go over budget and under perform because either their proponents are lying, what Flyvbjerg identifies as "strategic misrepresentation," or they're deluded by dreamy optimism where every hand is a straight flush.
The Green Line Monorail, the downtown Seattle bus tunnel, and Sound Transit have all given us a local look at this phenomenon. And beware the smaller but still expensive Mercer Mess fix, which consultants said would not improve congestion but has been embraced now by the City Council as an "urban renewal" project. It won't work as a transportation project, so the dreamers have quietly shifted the rationale. And how about the Port of Seattle with a staff that has demonstrably been willing to lie, break rules, and obfuscate to please the boss?
And talk about "optimism bias," the new New Deal is causing politicos to wrap everything in the banner of stimulus in a way that reflects all the self-deception Flyvbjerg says is par for the course. In essence, the New New Deal is itself the biggest of all megaprojects subject to the same dynamics. A kind of irrational exuberance seems to be at play. As Peter Callaghan notes in a skeptical column about it all in the Tacoma News Tribune, "Few politicians want to act as the killjoy, to toss the potato in the punch bowl...That is, should we solve a problem caused by too much borrowing by engaging in more borrowing?" Sounds like the classic Big Dig scenario.
There is a way out, but it's a process that may not give boosters the answers they want. It involves a detailed analysis of risk and the willingness to deliver "bad" news. It also requires the institutional character to look at the facts and say "no." A major component of analyzing megaprojects, says Flyvbjerg, should be to compare apples to apples. In other words, be hard-nosed in comparing your prospective Big Dig with all the other Big Digs.
Flyvbjerg uses the database he's assembled to run what's called a "reference-class forecast" which is a detailed assessment of a project's likelihood of success or failure and pinpointing where it might go wrong. The difficulty is that many bureaucrats and officials don't want to hear the truth. Another is that finding comparable projects can be difficult because of scale, location, government structures, and many other factors make the apples-to-apples comparisons difficult. Yet another is that a lot of after-the-fact project data is buried or unavailable. Gathering the forensic evidence on a high profile public failure can be hard to find as Jimmy Hoffa.
So the solution is to work harder to get comparative data, crunch the numbers, and make objective analysis of it a requirement of moving forward. "Flyvbjerg believes strong, confident bureaucrats, armed with the results of a reference-class forecast, can manage the process well enough to keep builders in check."
That is, presuming you can find the "strong, confident bureaucrats" who are willing to shatter politicians' dreams of leaving monuments that will stand (and cost) long after their terms have expired.