High construction costs pose a major threat to the U.S. economy. Not only are highways and transit systems irreplaceable for most Americans, they enable the free movement of people and goods within and between cities — the glue that holds together networks of domestic production. Without smooth roads, solid bridges and well-functioning trains, supply chains break down, people can’t get to work and the whole economy gets gummed up. In the short term, the government can spend more, throwing money at the problem to keep infrastructure intact; but in the long run, high costs will make this option ever-less tenable. Eventually, local and state governments, and even the federal government, will balk at the price tag, and simply let infrastructure fall into further decay.
The Bureau of Labor Statistics finds that productivity in the construction of infrastructure like highways, streets and bridges has fallen in recent years:
Others find the situation even more dire. Economist Leah Brooks and law professor Zachary Liscow find that between 1960 and 1980, the inflation-adjusted cost of building one mile of highway tripled in the U.S. Rail systems are also expensive to build, especially when compared with other advanced countries.
Unfortunately, the cause of these high construction costs is a mystery. Many people, when confronted with the question, will quickly reply that strong unions, or the Davis-Bacon Act — which stipulates the wages federal contractors must be paid — are responsible. But the people who give this answer are incorrect. First of all, construction salaries simply aren't particularly high:
As for unionization, France and other countries that build infrastructure much more cheaply than the U.S. are even more heavily unionized or covered by collective-bargaining agreements. Furthermore, Brooks and Liscow examine a number of different data sources, and conclude that labor prices in highway construction aren't responsible for increased costs. Compensation has remained roughly constant in real terms, even as overall costs have exploded. They also find that the cost of materials hasn't gone up.
Another common suspect is land-acquisition costs. Unlike China, which kicks hapless peasants off of their land whenever it wants to build a new megaproject, the U.S.'s constitutional system forces government to acquire land from private owners (though it can use eminent domain to limit the cost of acquisition and compel a sale). But Brooks and Liscow find that land costs are a relatively minor piece of highway construction costs, and that this share has not increased over time. Changes in eminent domain law also don't seem to raise costs much. Nor, they find, do increases in planning costs explain much of the trend.
So if high U.S. costs aren't due to expensive labor, land, materials or planning, what explains the enormous expense of building roads and trains? With the easy answers ruled out, the economic detectives investigating the cost mystery are looking at subtler factors.
Brooks and Liscow investigate the impact of another feature of American democracy — the so-called citizen voice. In the U.S., wealthy homeowners are able to stop highway projects they don't like, often by forcing lengthy and expensive environmental reviews. This power dramatically increased after a number of legal changes, including the National Environmental Policy Act of 1970, the Historic Preservation Act of 1966 and various other environmental laws. When homeowners use these laws to block roads, it creates delays and uncertainty, forcing contractors to raise their costs. It can also force roads to take more expensive routes; Brooks and Liscow find a correlation between how "wiggly" a highway is and how much it costs per mile. They also find that construction costs tend to be higher in richer areas, suggesting that wealthy people use their power and legal knowledge to block noisy, unsightly roads.
Inefficient government may also be a big problem. Brooks and Liscow find huge state-to-state variation in the amount that highway costs have risen, suggesting that some governments are simply much worse at building roads cheaply. Transit blogger Alon Levy has documented problems in government procurement in states like California, which he alleges hires extremely inefficient but politically well-connected contractors who overcharge the taxpayer with impunity. A 2015 report by consultancy McKinsey & Co. suggested a number of ways that U.S. construction-procurement processes could be improved. Some reports blame inefficient union contracts for overstaffing on infrastructure projects — even if wages aren't high, cities may be forced to hire too many workers to do a job.
These issues don't lend themselves to easy fixes — construction costs won't be brought to reasonable levels by repealing Davis-Bacon or strengthening eminent domain. Instead, it seems like infrastructure in the U.S. is plagued by the same sort of cost disease that afflicts the country's education and health care — the country is simply so rich and complacent that it has allowed a thousand small inefficiencies to build up in the system. To make U.S. roads and rail transit functional again, these inefficiencies are going to have to be rooted out one by one, with consistent effort by policy makers and constant monitoring by economists and other outside observers. Many toes are going to get stepped on in the process, many local homeowners' wishes overridden, and many cozy contracting relationships disrupted. But the alternative — a country where infrastructure is just too expensive to build — is not acceptable.
Noah Smith is a Bloomberg Opinion columnist. He was an assistant professor of finance at Stony Brook University, and he blogs at Noahpinion.