Marc Dunkelman has a new book coming out next month called Why Nothing Works. His thesis is that America once did big things but now seems stuck—and much of it is the fault of progressives:
America is today the victim of a vetocracy that allows nearly anyone to stifle progress. While conservatives deserve some blame, progressives have overlooked an unlikely culprit: their own fears of “The Establishment.” A half-century ago, progressivism’s designs on getting stuff done were eclipsed by a desire to box in government.
This has become such a common critique that it's now practically conventional wisdom. But before we accept it, we need to know at least two things:
- Do we, in fact, not build things these days?
- Have all those progressive regulations been effective, or just pointless obstruction?
I'm going to try to assess both of those things as honestly as possible. What follows, I'm afraid, is immensely long. I got carried away. But the charts are all super simple and you should be able to scroll through them fairly quickly. Let's start with various categories of construction.
Total Construction
There are two ways of looking at construction: (a) total dollars adjusted for inflation or (b) as a percent of GDP.
In simple dollar terms, construction spending has grown steadily over the past 30 years. As a percent of GDP, it recovered from the Great Recession a few years ago and is now back to its pre-bubble average. Any way you look at it, we're building a lot of stuff.
But spending is a troublesome metric. If it's going up, does that mean we're building more stuff, or does it only mean that stuff has gotten more expensive? To answer that, we need to look at raw amounts of different kinds of construction.
Housing
Once again, there are two ways of looking at this: (a) the absolute number of units approved and built, and (b) the number of units after adjusting for population. The former gives a sense of whether the raw permitting process is working, while the latter tells us if we're building as many houses as we want to.
The top chart shows the number of housing units that managed to get built. It's gone up and down with the housing market and is about average today. The bottom chart is my favored metric: the number of units added each year as a ratio of the population increase each year. By that metric, residential building was fairly flattish up through the Great Recession and has since surged to historically high levels. Despite everything, we're still building a lot of housing.
Roads
When the interstate highway system was finished, the federal government naturally wound down its road-building program. Nonetheless, looking at all levels of government, road building has steadily increased over the past 40 years.
There's a dip during the pandemic, but it's already partly recovered. We're building a lot of roads.
Utility Solar
Big solar projects are infamous for their long permitting times, often due to opposition from environmental groups. Nonetheless, the construction of utility solar plants has skyrocketed:
I'd like to see this in acres of land used, but I couldn't find that. But this is a good proxy. One way or another, despite the opposition, we're building a lot of big solar plants.
Or so it seems. But the only way to know if this construction rate is impressive is to compare ourselves to the rest of the world. Since the US is a big country it naturally has a lot of solar capacity on an absolute level, which means it needs to be adjusted for population.
Among countries with more than 2 GW of solar capacity, the US is #3 in watts per capita. We're building a lot of utility solar.
Pipelines
Opposition to pipelines gets a lot of press, but that's limited to a few very high-profile projects. Here's how we're doing overall on pipeline construction.
Even averaged over five years this is a very volatile series. Still, it's been pretty steady over the years. We're building about the same pipeline mileage today as we did 40 years ago.
Bridges
We build thousands of bridges every year. However, most of them are tiny overpasses that are only technically bridges. For this chart I extracted just the bridges over a thousand feet long.
There has been a decline since 2016, but the overall trend is steadily upward. We're still building a lot of bridges.
Summary: total construction is up; housing is up; road building is up; utility solar is up; pipelines are flat; and bridge building is up. Given this, it doesn't appear that the progressive regulatory environment has slowed us down much—if at all. Still, there are a lot of regulations on the books these days. Let's take a whirlwind tour of whether they've been effective
What's the impact of all the regulations we've put in place?
Even if we're still building a lot of stuff, that doesn't mean regulation is good. After all, it obviously slows things down at least somewhat. It's only worthwhile if the cost (in amount of construction) is modest and the benefit (in human prosperity) is high. So let's look at the benefits of regulation.
The EPA has a lot of rules about what and where you can build things. Here's the impact of all those rules on air pollution and water pollution.
Air pollution levels are down substantially for all types of pollutants. Water pollution turns out to be surprisingly hard to measure, but two overall measures—safe for fishing and safe for drinking—both show huge improvements since the Clean Water Act was passed.
But that's not all the EPA has done for us. They've also had a huge impact on the crime rate by eliminating lead in gasoline.
The FAA is another big producer of rules. But it's paid off.
The NHTSA is a busy beaver at producing safety regulations for cars. Because of this, a lot fewer of us die every year.
You've probably heard that the Hoover Dam took only five years to build. Why can't we do that anymore?
Part of the reason is that a hundred workers died during its construction, something we no longer tolerate. Nor is this just ancient history. Construction of the original World Trade Center took 60 lives. By contrast, its replacement, One World Trade, was built with zero loss of lives.
A lot of this is due to OSHA, which comes in for a lot of abuse over their seemingly arcane rules to protect workers. But OSHA has been around since 1934 and was given greater power (and its current name) in 1970. The impact has been substantial.
Workplace injuries are also way down.
This is not all due to OSHA, of course. Labor unions have also played a big role in pushing for safer workplaces.
How about banking regulations, which take up miles of paper from an alphabet soup of federal regulators? Banks may hate them, but they're good for the industry regardless.
After banks were regulated in 1934, failures plummeted and stayed low for 40 years. When S&Ls were deregulated, failures surged. And when derivatives were set free in 1999, within a few years they were adding rocket fuel to the Great Recession. But even when regulation was tight, bank profits have stayed strong. They aren't being hurt too badly.
It's not just banks, either. The Federal Reserve is responsible for making the entire economy safer.
After World War II the Fed began to take Keynes seriously and pushed back against cyclical booms and busts. The results have been pretty spectacular. Even the Great Recession produced only a tiny blip compared to half a dozen recessions and bank panics before the war.
How about the FDA? It was formed in 1906 alongside the Pure Food and Drug Act. Over the next half century salmonella food poisoning dropped to nearly nothing.
Public health regulations have an equally impressive track record.
In the Trump era the entire framework of civil rights and DEI has come under blistering attack. And maybe some things have gone too far. At the same time, the focus on civil rights from the Justice Department and the EEOC has paid off. Here's the share of managers and executives in private industry who are Black.
Governments are notorious for the failure of big software projects. And there have been some doozies. The project to upgrade air traffic control is infamous. The effort by the IRS to eliminate paperwork has gone on for years with no perceptible results. And remember the rollout of Obamacare?
There really are some bad IT projects out there—although it's worth remembering that Obamacare was up and running smoothly within a month of its initial failure. It's also the case that while the big failures get a lot of attention, they represent only a tiny portion of all projects. Here, for example, is an estimate of the status of all current federal IT projects.
I don't know how seriously I would take these self-reported figures, but even if they're optimistic they still suggest that we can build software just fine. Even a 10-15% failure rate wouldn't be very bad.
Also worth noting: the exact same thing happens in private industry. Big software projects blow up constantly. The difference is that they don't have inspector generals who snoop around and then write public reports about the failures. It's only government failures that produce dramatic congressional hearings and endless claims that "the government" can't do anything.
Building codes have gotten more and more intrusive over the years, but that's paid off too.
Finally, one more chart that's a bit of an oddball. It's about NEPA, the National Environmental Policy Act. It requires government agencies to write environment impact statements for large projects and to conduct public hearings about them. These EIS statements have gotten longer and more complicated since NEPA was founded in 1970, mostly because environmental groups have gotten more litigious.
But the production of an EIS is required for only a tiny fraction of the biggest projects, less than 1% in all. The rest either require a simpler Environmental Assessment or are simply cleared through as exempt. Here's the litigation rate of NEPA decisions.
Note the left-hand column. It's in tenths of a percent. And the litigation rate is going down.
(For some reason, the CEQ kept litigation figures for 2001-13 but not before or after. So this is all we have.)
Once again, the big lawsuits over controversial projects are the ones you hear about. But there are thousands of other projects that go through the planning process fairly smoothly and without any lawsuits.
But still, is NEPA worth it? That's a judgment call, and there's no chart I can show you about that. However, there's a mountain of anecdotal evidence that NEPA has improved the environment; given low-income communities a bigger voice in what's built nearby; and has often improved projects by forcing builders to plan more diligently before they start to build.
In Summary . . .
There are plenty of regulations that could use some pruning. But overall, the evidence is fairly clear that (a) we're still building a lot of stuff, and (b) a lot of regulations are there for good reasons even if they slow things down.
I'd also like to remind you of three axioms to keep front of mind whenever you hear that we can't do anything anymore:
- When the interstate highway system was finished, we didn't build a lot of highways anymore. When all the big rivers had been dammed, we stopped building big dams. After putting up several dozen skyscrapers over a thousand feet tall, developers realized that extreme height wasn't especially profitable and built a lot fewer of them.¹ In short, when all the big stuff has been built, you don't build anymore. That's common sense, not a loss of vision.
- Always remember that California and New York City are outliers when it comes to building. And within California, the Bay Area is simply insane. Californians genuinely resist further growth, and construction in NYC is genuinely eye-poppingly expensive. When you read about our ability to build stuff, always be suspicious if the examples are all from San Francisco, Los Angeles, and New York. They aren't representative of the country.
- The fact that you've heard of something is almost a guarantee that it's unusual. Nobody writes a newspaper article headlined "New light rail project coming along normally." So while it's true, for example, that California's high-speed rail project borders on the absurd; New York's water tunnel #3 will take 60 years and $6 billion to complete; and a project to create a virtual border fence failed, then its successor failed, and then its successor failed—the fact that you've heard of these things is because they're epic fumbles. It doesn't mean that government is incompetent any more than New Coke means the Coca Cola Company is incompetent.
It shouldn't be necessary at this point to mention that of course there are some burdensome regulations out there. NEPA should probably be modernized, the Pentagon needs to fix its acquisition process, and our price support system for farmers is just short of insane.
But be careful before you declare a regulation pointless. Do we really need dozens of pages of rules about the sinks in restaurant bathrooms? It might seem silly until you realize that it's mostly to make them accessible to the disabled. If you still don't think it's worth it, that's fine. But you should at least know why the regulation is there and what it's accomplished.
¹In Europe they never built them in the first place. Outside of Russia, there are only two buildings over a thousand feet on the entire continent. Australia also has two, Africa has one, and South America has none. Asia has 176.
"Given this, it doesn't appear that the progressive regulatory environment has slowed us down much—if at all"
The problem is value. Building in the US is much more expensive than most of the OECD.
https://www.vox.com/22534714/rail-roads-infrastructure-costs-america
Again, where are they comparing it to, and is that actually compared to wages and gdp (ability to pay) and purchasing power (how much do they get for those wages)?
If Marc Dunkelman did any real research and told the truth, instead of going for sensationalism, who would buy his book?
The formula for selling cheap and easy BS is so simple. Reader's Digest perfected it long ago. I still remember their "That's Outrageous!" column. That formula made the owners of that rag very rich.
Apropos of nothing, young Tucker Carlson once wrote frequently for the Readers
Digest...
That's the ticket! Bury him in graphs!!
... and now, I will go back and actually read the posting.
People like nice roads, but hate road repairs (construction season), and certainly don't want to pay for it. Why does construction take soo long? Why does it cost soooo much? Blame "regulations"--much easier than why you have to do soil tests to make sure it can handle the load.
Most people's direct interaction with "regulations" typically deal with parking tickets, getting permits for house repairs, and finding out they now need to bring things up to code. This all costs money.
You are a gem, Kevin, & this article exemplifies why. This country badly needs your calm, fact-based, deeply researched approach, and I just wish you had the much bigger platform you deserve. (But only if it entailed continuing coverage of Hilbert and Charlie, of course!)
Happy New Year.
Mr. KD, Bluesky could use your insights. I cross-posted this there and noticed you haven't engaged yet. Come on in, the water's cool and inviting.
Yeah, much better than Twitland.
To the downward trend which began circa 2013.
the Marc Dunkelman who was the DLC's vice president for strategy and communication in the aughts?
blaming "progressives"?
shocker
People should stop claiming that the introduction of new coke was a failure, let alone a spectacular failure. It was one of the biggest successes in commercial history. (It's more an example of turning using bad publicity into a massive advantage.)
Facts that people overlook in claiming the introduction of new coke was a disaster.
1) Few people actually wanted to drink the original Coke at the time of new coke's introduction. It is true that the original coke (and yes original should be in quotes since it obviously wan't original since it didn't contain either cocaine or Kola drug) was still number one in market share, but that was because it had a virtual monopoly on sales where people couldn't choose between different colas. In venues like grocery stores where people could chose Pepsi or Coke, Coke had fallen to 2nd place and market share was still plummeting with no end in site.
2) Coke did not return to the previous flavor of coke with coke classic, but the Canadian version which was half way between the original and new coke. (See number 1 above, almost no one liked the original flavor.)
3) Diet coke, Coke's number one seller today is the same flavor as new coke.
Yes, new coke was not successful. But it's introduction and coke's response to the push back of it's consumer base stopped the market share hemorrhaging it was experiencing cold and got it back to dominating the market where consumers had a choice.. The only thing keeping Coke (the drink) alive prior to the role out was it's near 100% control of venues that didn't give the consumer choice. It was going to lose 100% of that 100% over the next 5 years because almost no one wanted to drink it. This made the rollout quite possibly the biggest successful marketing campaign in history. It just wasn't the marketing campaign Coke originally intended.
Paraphrasing, from memory, one of the Coke executives at the time, he said something to the effect of 'Some people will say we had no idea what we were doing; some other people will say we planned this the whole time. We're neither that stupid, nor that smart.'
+++
😄😄😄
Solar power graph but no wind power one? Or is wind included because it's really solar power too? (Probably not.) There's a little geothermal and other such technologies too.
It's not, it'd be its own graph also showing about the same information.
"A desire to box in Government" is precisely what conservatives have had for generations.
As Kevin notes, it's normal to start with a massively biased sample and leap to conclusions from it (the most famous ought to be lotteries, where you can read announcements about how much winners won, but never see how many losers lost or how much money they lost). But it's worth noting the following as well:
1. Regulations create extra work and trouble for whoever has to follow them, whether it's the extra expense of engineering and constructing a motor vehicle to be safer, or workplace safety measures and training which reduce workplace accidents. And the costs associated with these extra expenses aren't always reflected in the marketplace: if Ford vehicles were massively unsafe, Ford would lose money, but Ford has to spend on safety while the costs from preventable deaths will not fall on the company (aside from possible legal settlements or fines). The Ford CEO is going to hear constantly about the costs and issues with regulation, but rarely or never hear about the benefits of unleaded gasoline, or the general benefits of saving drivers' lives. The first will always be reflected on the corporate bottom line; the others are intangible.
2. A subset of the business community has gone all-in on "first quarter thinking" and vulture capitol approaches. All of the benefits from regulations are future based--for example, getting lead out of gasoline takes decades to pay off--and corporate incentives (or, for that matter, many American tax incentives) being focused on rapid short-term gains over long-term thinking mean corporate internal pressure against regulations will shoot up. If you want number to go up in three months, something that causes a six month delay hurts a lot more than if you're operating on a ten-year planning cycle.
Of course, that short-term model of investment is, if anything, interested in unloading debts and costs onto others, so the idea of taking a quick profit and causing long-term harm that isn't registered on the investors' own account books is kind of the point, if not the sole purpose. Regulations are designed specifically to prevent some behaviors of the sort that short-term thinking incentivizes. These investors, then, and the politicians who represent their interests, will want to dump regulation and its benefits in their own self-interests regardless of the larger costs to citizens/society/the nation.
But note that short-term investor behavior ALSO causes the problems that the anti-regulation folks claim regulations cause. Why invest in long-term construction and infrastructure projects? If you even still own this business in ten years, it'll be time to move elsewhere to cash in big state incentives; why spend $150 million on a permanent corporate headquarters expansion when you're likely to be gone before construction is complete? Kevin's shown that regulations haven't impeded things as much as Dunkelman and others claim, but they also ignore the ways in which the approaches they support (and the financial interests they speak on behalf of) ALSO damage the sorts of projects they claim deregulation will help to spur.
When big investor behavior drives markets, and big investors have undue influence over governmental tax structures, and when maximizing profit as an investor gets increasingly decoupled from actual production and services, then we would expect to see the big investors trying to change financial and free market structures in ways which minimize their own personal risks and maximize their rewards; further, they'd seek to reduce the relationship between their own performance and their profit. Making a poor or dangerous product is disincentivized in a free market because people will just purchase a better/safer alternative, but given that cutting costs is a way of increasing profits under CEO control, while consumer behavior is not under direct CEO control, an alternative is to restrict or eliminate effective competition so that the quality of your product no longer costs you money because nothing better and cheaper exists to compete.
And indeed, this accounts both for the desire to pursue monopolies and for the kinds of collusion between big providers which ensure uniform pricing structures for uniformly poor necessities.
Assessing regulations and their benefits requires careful attention to who benefits. The broader argument about government regulation has as much to do with a central disagreement about who government serves, and why.
Accessing regulation is key, but bad regulation it is never an excuse for lax regulation for problems that are predictable and can be solved with that minimum vigilance.
Also +1
Just because something is bad doesn't mean its lack would be better.
+1
When I began doing business writing, I had lots of people go on and on in my ear about 'regularity uncertainty.' Never had too many actual examples of it, but the theory was that fears of a changing regulatory environment was enough to stifle business investment.
Through my years of reporting on both business and land use issues, I began to feel that the opposite happens: good regulation creates environments where businesses can flourish and investors can invest because they don't have to worry about catastrophic failures. The lack of regulations the led to the BP Gulf of Mexico disaster brought this concept into focus. Whenever I mention it to MBAs, the scratch their heads and try not to hear what I'm saying, often they try really hard.
But this nice set of charts documents the concept nicely. Thank you.
The earliest graphs in this post don't go back far enough, since the 70s is when the change is typically dated.
It's laughable to date "the Federal Reserve era" as beginning in the 1950s when it was created in 1913. Ben Bernanke even admitted to Milton Friedman that the Great Depression was indeed the Fed's fault, but if you only consider the Fed responsible for the Great Moderation era you can handwave away their mistakes. Why don't you compare Canada, which didn't have a central bank during the Great Depression?
Saying that California & New York are outliers/exceptions is granting too much. Those are some of the largest states in country, who have lately been losing population to other states where building is legal, but still have huge populations. They also contain some of the biggest wealth-generating industries. Logically you would be expecting people in regions where pay is worse to be moving to those states to take (comparative) advantage of those spillovers, but they're not because building restrictions are so onerous.
Housefires would have declined over time regardless as technology improved and houses were built out of less flammable material.
Yeah, having retirees move out of the state is great for those other states.
Yes, Kevin, please do get on Bluesky. I really appreciated this post, and didn’t think it was too long. One quibble, that I’ve brought up before: I think Seattle belongs on your list of cities that are infected with NIMBYism and resist building, and they have the housing prices and huge homeless problem to prove it. (Full disclosure: I’m from Tacoma, and we are building here.). If you have data that says otherwise, please share.
It depends on which specific municipality; Seattle is good, Ballard is not.
And if you compare King County to San Francisco, Santa Clara, or especially Marin or Santa Cruz, you can see how much better your metro is doing - even if it's not enough.
There are two main reasons why the economy has been less volatile since around 1950. First there have been no world wars. But aside from that it has been regulation and control of finance that has prevented another Great Depression. The danger of excessive leverage in the stock and other markets has been recognized and partially controlled, for example by the SEC, and the danger of massive bank runs as in the Depression is controlled with deposit insurance. The recent bank failures involved uninsured deposits.
The Fed Maestro Alan Greenspan was always an advocate of deregulation and he was instrumental in the failure to regulate derivatives, which was one of the problems in 2008. The Fed failed to act to rein in the housing bubble. The Fed may have prevented a worse collapse of banking at that time by throwing money at the banks but next time there may not be a bailout (most Republicans voted against TARP).
The Trump administration threatens to do a number of things to cripple important regulation of banking and finance, and bubbles will probably develop - they always have. We could be entering another period of volatility.
Speaking of the Interstate Highway System, something to consider: Construction began in 1956 and was completed in 1992. That's 36 years. Ike had been out of office for 32 of those years, and dead for 23 of them. There's an argument brewing in Dem party circles that projects are not worth doing unless you can get one enacted, planned, and built in 4 years, so it helps the D president's reelection efforts. Otherwise, why bother? l'd suggest that idea is complete and utter nonsense. It's an idea that has lost faith in the future. We need at least one party in this country to believe big things can and need to be done.
It's also ignorant of our past. It's hard to find examples of historic projects that didn't require multiple terms, and multiple presidencies, to get done. Hoover Dam, going to the moon, Obamacare, and yes, building infrastructure to support transition from fossil fuels to a green economy.
Well, we're dealing with a media who seems to think its job is clicks and false balance instead of objective facts.
It means people forget.
It also doesn't help that there are those who delay, delay, delay instead of actually blocking.
All those fancy graphs and words are meaningless in the face of the fact that the US has not built a new battleship in decades!!!!!! 1944!!!! The US Missouri!!!!
This post is an instant classic. Thanks for writing it.
I do have a question (a bit off topic). WHY are things so much more costly in NYC than in all the other major U.S. cities that also have lots of the factors that I would think make things so expensive in NYC?
My one issue with KD's article is that the short version is "things don't take longer to build or cost more except in places where building takes longer and costs more". Which is true but we don't really need to build major highways or transit infrastructure in places where not very many people live.
A lot of these charts are fatally flawed by the fact that they start in 1980, and the key regulations that have strangled building were put in place in the 1970s. To prove what you claim they prove, they need to go back to at least 1960 and preferably 1950.
There is a lot of quibbling about start dates which I agree with but one always has to make choices when presenting data. I would argue that the only way to look at this in any meaningful way is to chart each one at two scales. The maximum for which we have data and the last 20 years. It is easy to see the difference between dropping from say 20 to 2 but the subsequent drop from 2 to 0.5 is hard to discern yet both are significant.