Numerous state highway programs have suffered cost overruns, say the Gannett papers (which include USA Today). What’s striking from the story, however, is how small and rare the cost overruns really are.
The papers found overruns in 19 states, but they focused on projects that actually had overruns and did not reveal how many projects had no overruns. Of the overruns they found, many were less than 2 percent, most were less than 5 percent, and only three–in New Jersey, New York, and Ohio–were more than 10 percent. The unweighted average was around 7 percent. Since Gannett did not discuss any projects that had zero overruns, the real average must be much less.
This contrasts sharply with rail transit cost overruns, which have steadily averaged around 40 percent. Nearly 10 years ago, Bent Flyvbjerg reported that transit cost overruns in the United States averaged 41 percent while highway overruns averaged 8 percent. More recent research has found similar rail overruns, but the Gannett analysis suggests that highway overruns remain well under 10 percent.
Why would rail overruns be so much greater than highway overruns? The most important answer is that highways are largely paid for out of user fees, and decisions are made by professional engineers overseen by appointed commissions. Rail projects are largely paid for out of taxes and decisions are made by politicians and, sometimes, voters who must be persuaded that the projects are reasonable. It would make sense for rail advocates to underestimate costs until they sell their ideas, then–after the decision has been made–suddenly “discover” higher costs.
The Gennett papers, and highway critics in general, often point to Boston’s Big Dig as an example of an outrageous highway cost overrun. But the Big Dig was not really a highway project; it was a city beautification project. The goal was not to expand highway capacity but to turn an elevated highway into a tunneled highway. Since the Federal Highway Administration would not pay for a city beautification project, the politicians–led by Tip O’Neill–had to step in, which meant the project had to be sold politically. (To placate the FHwA, the project ended up including a small capacity increase.) Perhaps not coincidentally, the original project cost estimate was made by Parsons Brinckerhoff, the company also responsible for many rail cost underestimates.
“Transportation costs too much,” argues a blogger who calls himself The Transportationist, which I suppose isn’t any more ridiculous than calling yourself the Antiplanner. The Transportationist lists several good reasons why transportation costs too much, including the fact that–despite spending hundreds of millions of dollars on transportation planning–no one really does benefit-cost analyses. But he misses the point about user fees: projects funded out of user fees are more likely to be efficient, partly because the agencies or private parties receiving those fees know the fees are limited and partly because they want to spend them in ways that will generate more fees (which means in ways that benefit users enough that the users are willing to pay for them).