One of the basic legitimations of capitalist growth and politics in capitalist democracies is productivity improvement. This can be disruptive. It can force social upheaval and creative destruction. But the triumph over scarcity – more for less – is hard to argue with. As Charles Maier pointed out, productivism was one of the keys to stabilization of capitalist political economy in the mid-century. The question is how are those gains and that political benefit distributed, if productivity gains are extremely uneven.

The forces that make for productivity improvement are complex and multiple. But the most important technical changes – electrification, IT etc – are very general in their application. At varying speeds most sectors benefit from productivity gains. Of course, some of these gains have hidden costs (environmental for instance), which if they were properly accounted for might actually negate the apparent increase in productivity. But even on the most conventional measures of output there are some sectors that exhibit not progress, but productivity decline. Of these, one of the most important and politically consequential is construction.

As the FT’s Cardiff Garcia remarked already in 2014: “For the nearly half-century through 2012, annual labour productivity growth in the US construction sector averaged close to zero, and it has been negative for the past two decades.”

This matters because it structures the entire debate about public infrastructure and the capacity for public action, which is so urgent in the US. Whilst Silicon Valley offers a triumphant story of private sector innovation, the public sector finds itself discredited by association with the chronic inefficiencies of the construction sector and megaproject management.

The technical obstacles to efficient construction have political effects. Noah Smith takes up the point from Matthew Yglesias: “the problem with high infrastructure costs is that they force us to debate the wrong things. If costs were reasonable, even skeptics would probably agree to fix roads and build better trains. But when the price of maintaining high-quality infrastructure is ridiculously high, the issue gets divided into two camps — a pro-building contingent that advocates biting the bullet and overspending to maintain transportation networks, and an anti-building group that throws up its hands at the price tag. When this is the debate, the country loses either way, because it ends up either spending too much money or living with potholed roads and trains that never arrive.”

Why construction in the US is so expensive, inefficient and prone to disastrous delay is genuinely mysterious.

There are two types of analysis that bear on the question. One is of the type already quoted which looks at developments within the construction sector over time.

Such comparisons are not simple, because measuring output is not simple. Nor is it easy to devise the appropriate measure of inflation to distinguish nominal from real increases in output. But, our best guesses all point in the same direction. Productivity in the US construction sector has been stagnating or even declining for decades. This is particularly true for housing construction, but the overall data for the construction sector do not suggest a radically different story in the rest of the industry.


Matthew Klein recently calculated this broad based index of construction productivity per hour.


This tendency is not confined to the US. Construction is a laggard everywhere. But in the US its performance is particularly bad. A recent FT piece highlighted the differences between Japan and the US: “the average Japanese construction worker currently produces about 37 per cent more new housing than the average American construction worker. Another way of putting all of this is that America built about the same number of housing units in 2016 as in 1992, but somehow required about 46 per cent more people to do it. Japan built 31 per cent fewer houses in 2016 than in 1992, but its construction workforce had fallen by 19 per cent. Productivity deteriorated in both countries, but productivity fell much further in America than in Japan.” Though employers often complain about a lack of construction labour, “(t)he data imply there is a glut of American construction labour, not a shortage. There are plenty of people, but they’re too inefficient.”

What can explain these overall trends? The reasons why productivity in construction has been less dynamic than in manufacturing as a whole are not too hard to fathom. As Daniel Gross comments, the construction labour and production process is inherently resistant to dramatic automation. It is “difficult to gain economies of scale — or to automate processes — when every job, or close to every job, is unique. If every T-shirt were made to order — different sizes, styles, cuts, fabric — it would be very hard to get a $3 T-shirt.” The sheer, overwhelming amount of choice in complex construction projects creates “huge variation from project to project, and within projects. In addition, there are important differences between home manufacturing and, say, car or appliance manufacturing. Factories create their own artificial environments that are conducive to the task at hand: The whole structure is designed for the optimal flow of people and materials, and these factories can run around the clock stamping out hundreds, thousands, or millions of units of the exact same product. As a result, operators can systematically experiment with and apply technology, tweaks, and improvements to make the process run more efficiently. Voila! Productivity gains!” In construction, “the manufacturer”, “often has very little control over the environment in which it operates. And so it has to operate slowly and carefully. …Because sites are typically small, construction has to take place in discrete, linear stages. … Each has to work sequentially to a large degree — so a single delay with a single trade, or poor coordination between trades, can throw the schedule off … Although some mechanized equipment is used, and although some builders are experimenting with prefabricated efforts, a significant part of the work is done on-site and with human hands.”

The Bureau of Labour Statistics in a substantial study dug deep into the problem of measurement and productivity calculation.

Source: p. 20

Interestingly, the BLS showed that the pattern of productivity development was not the same across housing and road building. In road building In the US in the last decade there was actually some productivity improvement. The billion dollar question is from what level and how that path of improvement compares with other countries.

International comparisons of the cost of infrastructure projects produce surprising, puzzling and dismaying results from a US point of view. Megaproject cost comparisons is a cottage industry in which the blogger Alon Levy is the most important voice. He offers regular lists of per km costs for comparable urban infrastructure projects which range from $ 4 bn plus per km for the East Side Access route, to $ 40 m per km for a recent tunneling effort in Barcelona.

Citing the extraordinary data compiled by Alon Levy one study reports: “New York’s Second Avenue Subway will cost roughly eight times more than Tokyo’s Koto Waterfront line and 36 times more than Madrid’s Metrosur tunnels on a per-kilometer, purchasing power parity (PPP) basis. But this is not strictly a New York problem. Outside of New York, there are three more U.S. projects in the top 12: Boston’s proposed Red-Blue Line Connector, San Francisco’s Central Subway, and Los Angeles’s Westside Subway Extension.”

How to explain this gap? Noah smith canvasses the obvious culprits: “One popular villain is union labor. The Davis-Bacon Act, passed in 1931, mandates that infrastructure workers get paid locally prevailing wages, which usually means the wages that union members would receive. Some studies have claimed that this law and other union-friendly policies drive up costs in the U.S. But unions probably don’t help explain the yawning gap between the U.S. and other rich countries. The reason is that places like France have some of the strongest unions in the world. Strikes by rail workers are commonplace. Yet France’s trains cost much less. Japan is another counterexample. The median salary for a Japanese construction worker in 2014 was about 4 million yen a year, which at current exchange rates is roughly $36,000. Construction workers in the U.S. make about the same — an average of $37,890 in 2016. Now, it’s possible that the average Japanese worker is capable of building much more in an hour than the average American worker, meaning U.S. laborers could still be overpaid in the relative sense. But it seems unlikely that the difference is that huge. The numbers are pretty clear — high wages aren’t the big culprit in U.S. costs. Another bogeyman is land-acquisition costs. People think of China’s authoritarian government forcing millions of people to move in order to build dams and highways, and assume this must be why it can get things done so much more cheaply than in the democratic U.S.”

But as Alon Levy points out: “land-acquisition costs are much higher in Japan, where eminent domain laws are weaker. … And yet, somehow, Japan still lays train track much more cheaply. Explanations based on geography — the U.S. is too spread out, or New York City is too dense — also fail to stand up to scrutiny. Cumbersome environmental impact reviews are a possible culprit, but it’s hard to believe that countries such as France would be so willing to pave over their natural beauty and slaughter endangered species that their trains would cost only half as much as America’s as a result.”

Every aspect of US project construction is worse. “In China, a 30-story building can be completed in only 15 days. In Japan, giant sinkholes get fully repaired in one week. Even in the U.S. of a century ago, construction was pretty fast — the Empire State Building went up in 410 days. Yet today, it takes the U.S. many years to spend the money that Congress allocates for infrastructure. New buildings seem to linger half-built for months or years, with construction workers often nowhere to be found. Subways can take decades.”

What this suggests Smith argues, is that the key to higher US costs is “general inefficiency — inefficient project management, an inefficient government contracting process, and inefficient regulation. It suggests that construction, like health care or asset management or education, is an area where Americans have simply ponied up more and more cash over the years while ignoring the fact that they were getting less and less for their money.”

Stephen Smith a Brooklyn-based analyst focusing on New York projects highlights some key issues: “A huge part of the problem is that agencies can’t keep their private contractors in check. Starved of funds and expertise for in-house planning, officials contract out the project management and early design concepts to private companies that have little incentive to keep costs down and quality up. And even when they know better, agencies are often forced by legislation, courts and politicians to make decisions that they know aren’t in the public interest.”

The contrast between the high cost American system and Spain’s low cost methods of construction is striking. Spain’s approach to station design is strictly functional. “American politicians have different priorities. The Port Authority of New York and New Jersey is spending $3.8 billion on a single subway station at the World Trade Center designed by Santiago Calatrava, a Spanish architect known for his costly projects. If New York could build subways at the prices that Paris and Tokyo pay, $3.8 billion would be enough to build the entire Second Avenue subway, from Harlem to the Financial District.”

Spain cuts costs by minimizing the number of consultants employed on the project. “Not so in the U.S. Parsons Brinckerhoff, perhaps the biggest name in the nation’s transit construction industry, is both the lead-design contractor and project manager for California’s planned high-speed rail line, and the company stands a good chance of winning construction contracts for its own designs. As if that conflict of interest wasn’t bad enough, the California High-Speed Rail Authority’s new CEO, Jeff Morales, arrived at the agency after a stint as senior vice present at Parsons Brinckerhoff, where he worked on the authority’s business plan.”

In projects that it completes abroad Parsons Brinckerhoff works at affordable rates and reasonable cost. “But absent the right incentives and oversight, even the best private companies will resort to rent seeking.”

New York’s legal system and procurement rules are not conducive to delivering high quality low cost projects on time. As one experienced observer put it: “In the private sector, if you rob your customer, you will suffer a hit to your reputation and possible losses in the courts,” he said in an interview. “Not so if you rob an agency like the MTA. Then it’s all rights and no responsibilities.”

As a result the MTA is forced into ““writing longer and longer and longer contracts, expressly prohibiting every way it has been ripped off in the past.” The byzantine contracts that come out of this process drive entrants away, limiting competition and pushing up costs.”

The impasse prompts Smith to language that is positively Bolshevik in its intensity: “To fix the problems choking U.S. construction, reformers are going to have to go through the system and rip out the inefficiencies root and branch. Unfortunately, this is going to be hard, given all the vested interests and institutional inertia blocking deep reform of the construction sector.”

As Yglesias reports, a study by the Government Accountability Office looking into the problem of high train-construction costs was killed on Capitol Hill without explanation.

The result is a vicious circle of underperformance, indignation, stigma and an increasingly irrational debate. As Alon Levy remarks: “In the transit-related forums I participate in, people know that the US builds subways at higher costs than all other countries, because I talk about it often. This feeds into various stereotypes Americans have of government effectiveness; Americans of many political stripes understand that there are serious problems with US governance, and compare the US negatively with certain countries that are famous for getting things done. Thomas Friedman periodically raves about China’s massive infrastructure investment; when he was secretary of transportation, Ray LaHood made the same praise, and connected this to Chinese authoritarianism, while saying that American democracy was still overall a better system.”

Faced with the kind of shambles that New York has witnessed on 2nd avenue it seems logical to connect “low construction costs with authoritarianism” or with other types of cultural difference. But as Levy remarks: “Cultures are far too diverse to be reduced to these oppositions, and this is especially true on the level of political subcultures, such as transit investment. The reality is that the places with the lowest construction costs do not really match the stereotypes.”

In the US meanwhile the debate has stark partisan implications. As Yglesias remarks, costs and effectiveness matter and they ought to matter for progressive politics. The disastrous website launch for Obama’s Affordable Care Act is one instance of a wider phenomenon: “That embarrassing governance failure undermined the president’s signature policy initiative in serious ways, with crucial long-term repercussions. And any big ambitious new progressive initiative is potentially vulnerable to this kind of failure. It’s hard to pass a law providing the money to create a new guarantee of universal access to high-quality pre-K. But it’s even harder to actually build and manage preschools that deliver high quality education. Until places like New York and California — the bluest jurisdictions that are most open to the idea of taxing and spending to improve public services — get better at actually delivering those services in a cost-effective way, it’s going to be difficult to persuade residents of more skeptical jurisdictions that it makes sense to take the same agenda national.”

To end on a positive note this gives strategic significance to LA, which is driving the majority of mass transit infrastructure spending in the US and is also doing a relatively good job at containing costs, at least by US standards. Its new tracks are coming in around $ 400-450 m per km, a fraction of the New York disaster stories. These are way above comparable projects in the high performing European and Asian builders. But, if they set a new standard for the US it will be a major breakthrough in the politics of US infrastructure.