Why America can’t build quickly anymore

If we have 12 years to fight climate change, we can't afford to take 17 years to build subway lines.
Why America can’t build quickly anymore
Photo by Tho-Ge.

Stripe CEO Patrick Collison has a hobby: he curates a list of examples of “people quickly accomplishing ambitious things together.” Sadly, of his examples from the physical world⁠—like ports and skyscrapers and railroads⁠—most come from before 1970.

For each impressive example of speedy construction from the early or mid-20th century, Collison includes a beleaguered project from today’s world for the sake of contrast: for example, the New York Subway system opened with 28 stations in 1904, just four and a half years after the first contract was awarded. By contrast, the 2017 Second Avenue subway opening, with just three stations, took seventeen years.

That might be a fine system if you’re already happy with the infrastructure you have, and living in a time of relative peace and stability. But we’re not, and we aren’t.

In recent years we have seen serious shocks to our economic order. Events like the COVID-19 pandemic and the Russian invasion of Ukraine⁠—or, on a longer time frame, climate change⁠—have forced us to speedily reassess how we do business, in substantial and material ways. Importantly, these shocks don’t just call for a response in the financial world, like shuffling about some numbers on a computer. They call for real, physical responses⁠—like redesigning city infrastructure or constructing new power plants.

I no longer have faith that we are capable of doing this on a reasonable timeline.

Sometime in living memory, the built environment of the U.S. began to freeze in place. I’d mark the time roughly at 1970, but it’s a process, not a single seminal event. It’s more like a prehistoric creature getting trapped in resin, losing strength fighting the sticky substance. If it doesn’t escape soon, that resin will polymerize and harden into amber, leaving its form preserved forever.

Today’s high gas prices are the perfect example. Americans are universally unhappy about the high short-run costs of energy and transportation. And in the longer run, they aspire to make our economic life more independent of foreign dictators and their wars.

This isn’t a pie-in-the-sky aspiration. We could start doing this now, with the technology we already have. We could break ground on power plants to supply more energy. We could reduce energy demand by building public transportation projects, or build homes more densely to shorten commutes and save on heating.

But these things are easier said than done. Not because we lack the construction workers or the engineering expertise, or even readily drawn-up plans, but because there are legal and political structures that stop each of those ideas from happening through procedural delays.

At best, this is an annoyance that wastes billions of dollars and years worth of time. At worst⁠—at least, if you agree with climate change hawks⁠—it’s the end of the world. Rep. Alexandria Ocasio-Cortez (D-NY) believes in an urgent need for greener infrastructure, at any cost.

“The world is going to end in 12 years if we don’t address climate change,” she lashed out at critics in 2019, “and your biggest issue is how are we gonna pay for it?”

Forget how we pay for it, for a moment. The bigger problem is that urgency just isn’t there, money or no money. Ocasio-Cortez represents parts of Queens and the Bronx, where you might imagine public transit would be part of the green infrastructure mix. But new subway stops take decades, and are scarcely built anymore. Queens hasn’t gotten a new subway station since October 1989, the month she was born. The Bronx hasn’t gotten one since May 1941.

If all you get out of twelve years is two thirds of a Second Avenue subway, then what chance do you have, really? The deadline might as well be next Friday.

The rise of citizen voice

When someone’s actions have a negative impact on uninvolved third parties, economists call this a negative externality. There are a few ways to deal with negative externalities. English economist Arthur Pigou recommends a tax, ideally in the amount of the harm. Sometimes⁠—as Ronald Coase showed⁠—the parties can negotiate a side payment to efficiently handle the harm.

But they were aware of the limits of their solutions. Sometimes you don’t know how much the harm is, or the parties can’t negotiate efficiently among themselves. At this point, you need to make policy choices, one way or another.

And this is where I feel that lawmakers of the 1970s made a huge mistake. Rather than accept the need for general rules, or choices by accountable elected officials, the lawmakers built a dispersed power structure filled with veto points that lends itself to analysis paralysis.

This style of thinking is present especially in environmental laws like the National Environmental Policy Act (NEPA) at the federal level, or the California Environmental Quality Act (CEQA) at the state level. These laws both require the government to conduct an exhaustive review of government projects—sometimes even permitting decisions on private projects—that might have negative environmental impacts. But more broadly, it’s also present in any political environment where politicians solicit community input on a specific project before going forward.

The theory goes, if you talk to everyone and collect information on all the possible effects of a decision, you have a great chance to get the best outcome. In practice, these “citizen voice” systems end up rigged heavily towards preservation of the status quo, they contain an incentive structure that makes them more expensive, and some of the information they act on is unquantifiable at best.

How environmental review misses a key tradeoff

Review systems are supposed to weigh costs and benefits, and then help you make the best possible choice. They aren’t binding; they don’t force the government to act one way or another. Let’s say, for the sake of argument, that they get their object-level cost-benefit analysis generally correct.

There’s still a meta-level tradeoff, between acquiring more information and the costs of acquiring that information. The reports aren’t free. And more importantly, time isn’t free. The longer a project takes, the more it’s likely to cost, and the more land, efforts, and resources are tied up in a partially-completed project not yet bearing fruit. The longer a project takes, the more likely some aspect of the plan becomes obsolete or unfeasible midstream. And the longer a project takes, the later it begins to benefit people.

Additional pages of environmental paperwork are only valuable to the extent that they actually help change material decisions for the better, and only if the improvements they facilitate are more valuable than the time lost and the resources spent on the report.

It is obviously fair for authorities to take some time to plan things out and weigh the costs and benefits. But they spend, well, an inordinate amount of time weighing the costs and benefits. In a 2018 study of environmental impact statements under NEPA, the mean statement took 4.5 years to complete—about as long as it took to complete the original 28-station New York City subway back in 1904⁠—and ran 575 pages.

Here’s a thought experiment for you, one in what economists call thinking at the margin: take a 575 page report, and rank the pages from most valuable to least. How likely is it that the 500th- or 540th- or 572nd-most valuable page makes a material impact on your choices?

I imagine that the first six months and fifty pages of study yielded some important results, or even the next six months and the next fifty pages. But it is quite hard to imagine that government decisions are best served by spending the entire length of World War II, from Pearl Harbor to VJ day, doing cost-benefit analysis.

It would seem as if environmental review processes miss the meta-level tradeoff. The cost-benefit analysis never measures the cost of the analysis itself.

But the more likely explanation is that some participants understand the meta-level tradeoff all too well.

You don’t have to win to win

The problem with processes where you spend a long time gathering information and hearing arguments before you take action⁠ is that they are heavily biased towards maintaining the status quo.

Let’s say you think something should be done, and I don’t. While we work on our disagreement, nothing gets done. This isn’t a good structure for speedy resolution of disagreements, because I am incentivized to make the process longer.

The key tool that opponents of action have available is lawsuits. They can sue the state for permitting projects they don’t like⁠—not because they can prove a decision was wrong on the merits, but simply because the environmental review didn’t sufficiently evaluate their concerns. This forces the government to go back and write an even longer environmental impact statement. Such a move might not change the ultimate decision, but it drags out the process for a few more months or years.

Consider, for example, the Berkeley lawsuit under CEQA, California’s state-level version of NEPA, which Tim wrote about last month. Activists argued that expanding enrollment at Berkeley was an action subject to environmental review, and that the state had not sufficiently studied the impact of Berkeley enrollment. Critically, they didn’t need to prove that raising Berkeley’s enrollment would actually harm the environment. They simply argued that Berkeley enrollment hadn’t been adequately studied, and that was enough—at least until the legislature overrode the court decision.

Our civic utility monsters

The current rules to challenge environmental reviews do not require opponents’ concerns to be especially plausible, or even sincere. Instead, challenges often remind me of Robert Nozick’s Utility Monster, a creature whose preferences are so strong that it gains tons of happiness any time it wins a concession and becomes violently sad any time it loses at anything. By expressing its dramatically strong preferences, the utility monster hijacks utilitarian systems of decision making.

For example, it is difficult to believe that shadows⁠—an issue used frequently to block new buildings in San Francisco⁠—are actually deeply concerning to people living there. After all, they spend quite a lot of time under the shade of the already-existing buildings, often deliberately through features like porches.

Instead, I believe that complaints about shadows are a pro forma ritual. San Francisco’s community input process more or less gives NIMBY groups a veto over development without formally enumerating the veto into a property right. Instead, neighbors just have to demonstrate sufficient anguish to produce the utilitarian justification for the veto.

And this is where I think the advocates of externalities-based thinking have gotten too far over their skis. It is well and good to have the government handle large externalities, or quantifiable ones. I should pay for the wear and tear my car puts on the road. I should pay for the carbon impact of my emissions. But by expanding government’s scope to handle small externalities, and unquantifiable ones, they’ve created a system where few things are doable as-of-right, everything is up for debate, and people are incentivized to exaggerate or even fake grievances.

For a particularly egregious example of fake grievances, consider the Vineyard Wind project, an attempt to build wind farms off of the coast of Massachusetts. This is a project desired by both the state and federal government, and both accountable levels of government were ready to give it the green light.

However, because the wind farms were in the water, they were subject to permitting review by the Bureau of Ocean Energy Management (BOEM), and therefore, required an environmental impact statement under NEPA. Although the environmental impact statement was more than 2,000 pages long, a rival energy producer nonetheless sued BOEM in 2021 to argue that the statement did not sufficiently consider various concerns, like whether 260-meter turbines might have a different impact on migratory birds than 200-meter turbines.

It is, I suppose, plausible that the rival energy producer has deep and sincere concerns about the accuracy of BOEM’s study with respect to birds that fly between 200 and 260 meters of height, specifically. But I think it much more likely that this is an affectation: utility monster behavior by an energy company that wants less competition.

Much of the change aversion built into our political systems reflects people’s sincere preferences, especially at the local levels. But I think reforms can be made at the macro level. I doubt people ever intended for our infrastructure-building process to become so frozen as it is, in aggregate. But we can fix it some,  by reining in lawsuit-based case-by-case input processes and returning towards clearly-defined laws and executive decisions by elected officials.


Only members can comment.
Please subscribe to a free or paid plan or sign in to join the conversation.