Want More Infrastructure? Make It Cheaper to Build


Early on in Donald Trump’s presidency, increased infrastructure investment seemed like it might be a big winner for him, a populist policy with bipartisan appeal. Instead, Trump’s “infrastructure weeks” became a running joke and federal infrastructure spending fell. As a share of gross domestic product, federal outlays for major non-defense physical capital investments in fiscal year 2019 were the lowest they’ve been since 1957.

Want More Infrastructure? Make It Cheaper to Build

There are reasons to hope this will change under President-elect Joe Biden, who has big plans for investment in roads, transit and clean energy, among other things, and a better record than Trump of actually getting infrastructure built (as vice president he was in charge of overseeing spending from the $800 billion stimulus bill enacted in 2009).

There are reasons to doubt it, too. If Republicans retain control of the Senate, big spending increases may be off the table. Also, it is state and local governments that finance most infrastructure spending nowadays, meaning the federal role will likely be limited in any case.

Want More Infrastructure? Make It Cheaper to Build

Then there’s what may be the biggest challenge facing Biden and anyone else trying to upgrade the nation’s roads, bridges, transit systems and other infrastructure. Building such things in the U.S. has gotten to be extremely expensive, relative both to the past and to other wealthy countries. Without some success in bringing down the cost of building here, not all that much is going to get built.

This realization that U.S. infrastructure costs have gotten way out of line went mainstream just a few years ago, mainly because of a couple of spectacularly pricey transit projects in New York City: the $2.5-billion per mile Second Avenue subway and $3.5-billion per mile “East Side Access” tunnel connecting the Long Island Railroad to Grand Central Station. Subway lines in Continental Europe and Japan generally cost $200 million to $500 million per mile to build, according to transportation researcher Alon Levy.

A bunch of attempts to explain the differences ensued, notably Tracy Gordon and David Schleicher’s “High Costs May Explain Crumbling Support for U.S. Infrastructure” at Real Clear Policy, Brian Rosenthal’s “The Most Expensive Mile of Subway Track on Earth” in the New York Times, Josh Barro’s “Why New York Can’t Have Nice Things” in New York magazine and Levy’s “Why American Costs Are So High” at his Pedestrian Observations blog, where he has been writing about such matters since 2011. (Oh, and I wrote one too.)

There are many important insights in these pieces, but almost too many to make sense of. Yes the union staffing demands, political infighting, overuse of consultants, ignorance of overseas practices, poor procurement systems and tendency to build subway stations with gigantic mezzanines detailed in these accounts all seem bad. But is there a simpler explanation, and one that doesn’t apply mainly to New York City?

Well, in 2016 two economists began taking on the subject in the systematic way economists do, by gathering a bunch of data and running regressions on it in hopes of coming up with an overarching explanation. In two papers now nearing their final form, Leah Brooks of George Washington University’s Trachtenberg School of Public Policy and Public Affairs and Zachary D. Liscow of Yale Law School use data on interstate-highway construction spending from the late 1950s through the early 1990s, when the system was completed, to try to suss out what the key drivers of higher costs might be.

In one paper they compare per-mile spending across states (controlling for things like hills, swamps and dense populations that make building more expensive) and conclude somewhat unsatisfyingly that it has something to do with average incomes and differences in culture — “which could be either the underlying preferences of state citizens or the institutions that aggregate those preferences, or both.”  In the other they look at the increase in spending over time and find a related but more compelling story.

First, though, they dismiss some factors that don’t seem to have played a role in the quadrupling of the real cost of building a mile of interstate highway from the late 1950s and early 1960s: labor and materials, the cost of which didn’t budge much in inflation-adjusted terms.

Want More Infrastructure? Make It Cheaper to Build

What did change, seemingly beginning in the 1970s? Well, Americans had become a lot more affluent over the course of the 1950s and 1960s, and while growth in median incomes slowed after that, growth in top incomes accelerated. At the same time, the mechanisms for people with money and free time to take action against projects they saw as harming the environment, their community or their interests grew more numerous.

Brooks and Liscow cite the National Environmental Protection Act of 1970, with its requirement for environmental-impact reviews of projects getting significant federal funding, as the most important such mechanism. The 1966 National Historic Preservation Act, 1972 Clean Water Act, 1973 Endangered Species Act and the Supreme Court’s 1971 decision in Citizens to Protect Overton Park v. Volpe were among the other landmarks in the rise of what they call “citizen voice.”

As a result, building interstate highways became more complicated. One clever way Brooks and Liscow devised to measure this was by calculating the “wiggliness” of interstate highways (again adjusting for geographical factors). They found it increased markedly after the early 1970s. 

Want More Infrastructure? Make It Cheaper to Build

Wiggliness is not necessarily bad. Early interstate highways that plowed straight through cities with little regard for those who lived there may have caused economic and environmental damage that far outweighed the construction-cost savings.

Brooks and Liscow tell the story of Interstate 696 in the suburbs north of Detroit, the first two legs of which were relatively straight, ground-level freeways. The third, after years of opposition from community activists making ample use of federal environmental laws, was built as a wiggly, sunken roadway bordered with noise walls and covered in several spots with parkland. Completed in 1989, it cost seven times as much per mile, adjusted for inflation, as the first leg (finished in 1964) and twice as much as the second leg (1979).

Maybe that was worth all the extra cost. But when, say, the Beverly Hills Unified School District uses similar tactics and more than $15 million in bond proceeds in a seemingly futile but-construction-cost-increasing attempt to stop the Los Angeles County Metropolitan Transportation Authority from digging a subway tunnel under Beverly Hills High School because of fears the digging might release some fumes from old oil wells, one has to wonder whether “citizen voice” has gotten a little out of hand. As Brooks and Liscow write:

The most benign interpretation of these findings is that the rise of citizen voice causes government to internalize what were previously external costs of highway construction. The malign interpretation is  that this increased spending delivers little of value but serves to enrich favored parties. At this point we cannot say which type of interpretation more closely fits the data.

Which brings us back to the prospects for future infrastructure weeks. Politicians from both major parties actually have been making modest attempts lately to restrain citizen voice in order to make it easier to get things built. This summer the Trump administration finalized a regulation aimed at speeding up environmental reviews for federal infrastructure projects, while “yes-in-my-backyard” Democrats in California, Oregon and elsewhere have been trying for several years to sidestep local zoning rules and other regulations to allow the construction of more, denser, cheaper housing.

But the politics of telling people to shut up are of course complicated. Secretary of Housing and Urban Development Ben Carson seemed to embrace the Yimby cause in 2018, but then his boss decided to campaign this year on “saving our suburbs” from apartment buildings. And given that the environmental groups that are a key part of the Democratic coalition all seem to have hated Trump’s executive order, the Biden administration might try to quickly repeal it (although that will be hard if Republicans hold the Senate) or amend it. “The thing about citizen voice is that both the left and right like to use it,” says Brooks.

The rise of citizen voice also seems to have interacted in the U.S. with existing institutions such as the common-law legal system and the extremely fragmented nature of local government in ways that make it much harder to get infrastructure projects built than in other affluent counties. “The U.S. has always been a common law country, but you have a common-law country with new legislation, new social movements and new judicial doctrine,” says Liscow.

We’ve always been quite fragmented, but you have fragmentation along with these same things, and they’re activated differently. It’s just this massive problem and we’re just beginning — I wouldn’t even say we’re beginning to get a grip on it. We’re beginning to begin to get a grip.

There you have it. A U.S. infrastructure push probably can’t succeed in a big way unless construction costs are brought down, and we’re only beginning to begin to get a grip on why they’re so high.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Justin Fox is a Bloomberg Opinion columnist covering business. He was the editorial director of Harvard Business Review and wrote for Time, Fortune and American Banker. He is the author of “The Myth of the Rational Market.”

©2020 Bloomberg L.P.

BQ Install

Bloomberg Quint

Add BloombergQuint App to Home screen.