Now that the wreckage of the I-35W bridge in Minneapolis is no longer all over our screens, it’s worth considering what the episode should teach us about infrastructure, public policy, politics, and engineering. Part of the answer is, not much: it’s very tricky to interpret rare and exciting events.
This one, recognizing that any accidental death is a terrible experience for victims, friends and families, was in any case not all that costly. It appears to have caused about fifteen deaths, about 3/10 of one percent of Minnesota’s annual highway total, and dozens of injuries. It trashed several dozen vehicles, at $20,000 each, maybe a million dollars’ worth. It interrupted navigation on a low-volume reach of the river, blocked an unimportant rail line, and will tie up road traffic in the Twin Cities for a couple of years (if history is any guide, much less than people now fear). This last will be quite costly, but hard to see because it sprinkles hours lost here and there thinly across thousands of people. It lost about the last fifth of the bridge’s planned service (it was slated for replacement in 2020), about $50m (the new bridge is being talked about as a $250m item).
Equally important, it’s indicative and salient, but not really data. Bridges almost never fail suddenly and by themselves; I can only find two such examples in the last twenty-five years. Usually they fail from events like earthquakes, floods, barge collisions, gas tankers catching fire under them, and such, or partly and slowly enough to close them before people or trains go in the drink. So while we should obviously do a good post-mortem on the pieces and history of this one to learn what we can, we will probably not find out how to avoid a lot of social cost in the future. There’s just not much payoff in having less of something that almost never happens. That said, there may be real payoff in designing for more extreme events when we change our probabilities of their occurrence: if we’re going to have more floods from global warming, we should spend more on bridge footings; as we come to recognize the seismic risk in Memphis and Boston, we crank up the building code. But any design conditions omit events so rare they aren’t worth building for.
Furthermore, big failures usually result from a confluence of unlikely things, like overloading combined with a construction or design error, not a single “cause” by itself, because structures are built with safety factors to meet extreme design conditions: we know materials have defects, people make mistakes, and the world is a stochastic place. Novelty is no guarantee against failures, either; the Boston Big Dig is leaking and famously dropping ceiling panels. And the guess of the week is that this bridge had a design error that wouldn’t have been corrected by any amount of painting and maintenance.
One really egregious example of bad inference from this event is its effect on Minnesota Republicans’ views about taxes; they are suddenly willing to increase gas taxes to pay for bridges and highways. The collapse may have illustrated what’s really at risk, or focused attention, but the infrastructure of Minnesota is no more dangerous this week than it was last month, and the failure added no real information to the previous stock by which that danger, and thus the net benefit of collecting and spending public money to reduce it, might be estimated. What we have learned is that Minnesota has a governor so stupid and stingy and ideological that he counts the shelf of engineering reports and data available before the accident less than this single extremely rare occurrence. That’s simply a humiliation of Minnesota voters. The job of a governor is to be ahead of the nightly news and on top of opportunities to invest resources for a good payoff, and to appoint a Secretary of Transportation who keeps him hip to stuff that isn’t in the news yet, not to be running around after CNN stories break saying “how was I supposed to know?”. “Less taxes!” is neither a moral principle nor a form of thought, it’s just cynical laziness and deliberate ignorance.
I think Pawlenty and his pals are now on the right side of the repair and maintenance issue, but not because their one bridge fell down. I take this position because the budget process and the session-to-session politics of maintenance bias the system against making good decisions (Minnesota, to its credit, at least has separate capital and operating budgets), and because the pervasive national deficit in infrastructure investment and maintenance has been well-known for years on the basis of real analysis, and is anyway obvious to anyone who just looks around. (Even here, it’s appropriate to proceed with caution. Among the most urgent advocates for more spending on infrastructure are the American Society of Civil Engineers and a bunch of contractor industry associations, that is, the people who sell us the stuff. They may be right, of course, and the engineers are professionals and thus obliged to put the public interest before their own financial gain, but wouldn’t you at least ask around before you let me and my colleagues tell you how much public policy education America should have?)
The big issues in infrastructure are no different than they were before the I-35W tragedy: we don’t take care of what we have, we don’t have enough, and we have the wrong kind. The costs of this fecklessness are not a wave of deaths: No-one died when the New York Subway pumps couldn’t keep up with the rain this morning for the fourth time in seven months; skip the tornado gee-whizzing and look at the tenth and eleventh paragraphs of this story. No-one is bored to death from spending hours in crawling commuter traffic instead of with her kids. No-one dies because there’s no community swimming pool, or because the high school has graffiti and falling plaster, or because there’s nothing for kids to do unless a parent drives them to it. The cost is a pervasive degradation of the quality of life, and the quality of life is a big deal, not a cherry on top of some more important sundae. It’s what public and private policy should be about.
