2025-06-05 14:41:59
In general, people don’t like to think about the economy. When the economy is running well, they can ignore it and think about their own personal situation — their business, their job, etc. It’s only when there’s a recession, or a surge in inflation, or a stock market crash, or some other calamity that regular people have to sit up and pay attention to the broader economy.
Back in April, everyone was paying attention. Trump’s “Liberation Day” tariffs, by far the highest in the country’s history, threatened to upend the American economy and way of life. Stock markets plunged. Consumer confidence dove. Interest rates spiked and the dollar fell, as investors moved money out of the country.
Two months later, the specter of looming economic catastrophe has all but vanished from the headlines. Stocks have bounced back, almost to where they were before Trump came to power:
Long-term interest rates are still elevated, but they’re no higher than they were at the beginning of the year:
The dollar is a bit down from where it was before the tariffs, but overall it’s about where it was in previous years:
Meanwhile, real economic data is a bit weak, but doesn’t look too catastrophic yet. Americans are cutting back on consumption of the kinds of goods that got hit hardest by Trump’s tariffs:
But remember that because imports don’t count in GDP one way or the other, this won’t show up as much in American production numbers — or employment, which is closely tied to production. The drop in motor vehicle parts purchases will eventually show up as decreased U.S. auto manufacturing, but that localized (and totally unnecessary) pain will probably be masked by whatever is going on in the broader economy. Meanwhile, despite some chaos in the logistics and trucking industries, employment rates were still near record highs as of April:
Nor does inflation look like it’s much above the 2% target:
Market expectations of future inflation over the next 5 years, meanwhile, are just a little over 2%.
What happened? Just a little over a month ago, I was warning that serious negative economic consequences were in the pipeline, scheduled to arrive in the next few months. That could still happen, but markets don’t look particularly perturbed anymore, and there’s no sign of major disruptions yet.
What happened was TACO. This is an acronym that a bunch of corny Wall Street traders invented for “Trump Always Chickens Out” — i.e., a bet that as soon as it looks like tariffs are hurting the markets, Trump will put them on pause. Indeed, most of Trump’s gargantuan “Liberation Day” tariffs are already on pause until next month, while the tariffs on China have been paused for even longer.
Wall Street is betting that the pauses will continue, or that the courts will end up blocking most of the tariffs (there have been several court rulings against tariffs in the past week; the administration is appealing these). That’s probably why the stock market has gone up. And the nation, perhaps sensing that Trump is just a bag of hot air rather than a madman, has become a little less unhappy about his presidency in recent weeks:
Trump obviously doesn’t like being told he’s a chicken. But broad tariffs of the Liberation Day variety are simply a terrible policy, and terrible policies hurt the markets. Trump doesn’t want to crash the markets or cause a big recession, because then he’ll be remembered as a disastrous failure of a President. So he’s in a bit of a bind here — either go through with his stupid, self-destructive plans, or get called a chicken.1
In any case, as a result of the blizzard of new tariff announcements and tariff pauses, the overall U.S. tariff rate keeps swinging wildly back and forth, but generally staying in the 10-20% range. Joey Politano has done a great job of tracking this:
(This overall rate is calculated by weighting the various tariff rates by the percent of U.S. imports that this rate would have applied to in 2024.)
So does TACO mean we’re saved? Are we destined to simply endure an endless series of tariff announcements followed by predictable pauses and pause extensions, all the way up until Trump leaves office? That’s a ridiculous way to run a country, but it’s better than actually making American businesses pay all those crazy tariffs. Is it time to breathe a sigh of relief and go back to our daily lives?
Well…maybe. But as I see it, there are still several reasons to worry. These are:
The return of big tariffs
Negative effects of the tariffs and tariff uncertainty that we already have
The GOP’s huge debt bonanza, and the rise in long-term interest rates that it might cause
Of these, the first is the biggest short-term worry, but the third is the biggest long-term worry. The U.S. government is showing that it does not have the political will to stop itself from borrowing at unsustainable rates.
2025-06-03 15:34:34
Let me tell you a story about World War 2. In 1940, before the entry of the U.S. and the USSR into the war, Britain was fighting alone against Germany and Italy. Despite being massively outnumbered and outgunned, the British managed to pull off a spectacular naval victory, using innovative new technology. They sent the HMS Illustrious, an aircraft carrier, to attack the Italian fleet in its harbor at Taranto. The British aircraft disabled three Italian battleships and several other ships, without the Italian navy even seeing their opponents’ ships, much less having a chance to fight back.
But that’s just the prelude to my story, which is not about a British victory, but a British defeat. Just a little over a year after the Battle of Taranto, Winston Churchill sent the battleship HMS Prince of Wales and the battlecruiser HMS Repulse to deter Japan from attacking Singapore. Despite their own crushing victory at Taranto, the British military leadership was skeptical that battleships moving under their own power at sea could be taken down by air attack alone. They placed their faith in the power of zigzag movement and anti-aircraft guns to deter attacking planes.
This was foolish. Japanese torpedo bombers found and sank the Prince of Wales and the Repulse quite easily. Here is an aerial photo of the British warships, taken from the cockpit of a Japanese plane, desperately trying to evade their doom:
The great battleships — the invincible masters of the sea in previous wars — were suddenly helpless against the swarm of tiny aircraft. Churchill reacted with shock and horror, and the British fleet withdrew, essentially leaving Southeast Asia to the Japanese.
The world had changed, almost overnight. Air power had brought about a revolution in military affairs. Ironclad battleships went from the single most valuable piece of military hardware to being almost obsolete overnight. Yet people who had invested their countries’ treasure in battleship fleets, like Churchill, were painfully slow to realize the shift — even when it was their own technological innovations that rendered their old weapons useless.1
OK, so there’s your old WW2 parable, with a clear moral to the story: Don’t ignore technological revolutions. Now fast-forward to 2025. We may just have witnessed something akin to a modern Battle of Taranto. For years, Russia has used its strategic bombers — which can also carry nuclear weapons — to launch cruise missiles at Ukraine from a huge distance. The Ukrainians had attacked these bombers on the ground with drones, but the Russians simply moved them farther away, well out of reach of anything the Ukrainians could launch from their own territory.
So the Ukrainians got sneaky. They packed a bunch of drones — little plastic battery-powered quadcopters, not too different from a toy you would fly at the park — into trucks and (somehow) sent the trucks all the way across Russia. When the trucks got close to the air force bases where the Russians had parked their bombers, the Ukrainian drones popped out of the trucks and started blowing up the bombers — and other planes — on the ground. You can see the footage of the attack here:
And you can see some pictures of the drones used in the attack here:
It’s not clear how many Russian bombers the Ukrainians managed to take out, but everyone agrees it was a significant chunk of Russia’s bomber force. And these magnificent, enormously expensive, rare, highly prized machines of destruction were taken out battery-powered toys.
Again, the world has changed, almost overnight.
The American military is much better than the Russian military, but it’s ultimately not that different — it’s built around a bunch of big, expensive, heavy “platforms” like aircraft carriers, jet planes, and tanks. Each F-22 stealth fighter, still widely considered the best plane in the sky, cost about $350 million to build. A Ford-class aircraft carrier costs about $13 billion each. An M1A1 Abrams tank costs more than $4 million, and so on.
That’s the amount of value that will be destroyed every time a cheap plastic battery-powered Chinese drone takes out an expensive piece of American hardware in a war over Taiwan, or the South China Sea, or Xi Jinping waking up in a bad mood — not including, of course, the lives of whatever Americans happen to be inside the hardware when it gets destroyed. Except the true value lost will be much higher, since — like Japan in World War 2, or Russia now — the U.S. now has extremely limited defense manufacturing capacity, and thus won’t be able to easily replace what it loses.
As you read this, military planners all over the world are scrambling to come up with defenses against the kind of raid that Ukraine just carried out. Dozens of container ships arrive in American ports from China every day, each with thousands of containers. The containers on the ships then get unloaded and sent by road and rail to destinations all over the country. Imagine a hundred of those containers suddenly blossoming into swarms of drones, taking out huge chunks of America’s multi-trillion-dollar air force and navy in a few minutes.
That’s obviously a terrifying thought. How can the U.S. defend against that sort of attack? Possible countermeasures include hardened aircraft shelters and various forms of air defenses — guns, jammers, electromagnetic pulses, laser cannons, drone interceptors — along with improved surveillance of incoming container traffic. But whatever the eventual defenses are, the advent of cheap battery-powered drones has changed the game and made essentially the entire world into a battlefield.
The other question we need to be asking is: Why can’t the U.S. just do the same thing to China, in the event of a war? We have drones, right? Weren’t we the inventors of drone technology? Don’t we have innovative startups like Anduril, and Skydio, and lots of others racing to arm our military with the world’s best drones?
Well, OK. The U.S. did invent drone technology. But most of what we currently use are lumbering, expensive systems like the MQ-9 Reaper:
Each one of these giant drone planes costs $33 million. During the recent U.S. conflict with the Houthis — a conflict in which the U.S. was essentially defeated — the ragtag Yemeni militia shot down at least 7 of these Reaper drones, and possibly as many as 20. America in total has only a few hundred.
The kind of drones used in the Ukrainian raid, on the other hand, are “FPV” drones — that stands for “first person view”. These are small battery-powered plastic copters equipped with explosives. There are many types, but here’s one example:
These drones cost from a few hundred to a few thousand dollars each, depending on the type. Ukraine is currently producing thousands of these drones per day, and says it expects to be able to produce over 10,000, although either the base drone (before weapons and other military hardware are added) or the parts used to make the drone typically come from China.
Why so many? FPV drones aren’t just useful for the kind of long-range surprise attack that Ukraine just carried out. In fact, they’re steadily replacing every other type of weapon on the battlefield. FPV drones can take out tanks, including America’s best tanks. They are now estimated to cause 70% of the casualties on the battlefield — more than artillery, the traditional “god of war”. Here are some excerpts from a Bloomberg explainer:
Tens of thousands of the relatively cheap and expendable machines are now buzzing back and forth over the front lines, pinpointing Russian positions, gathering intelligence to anticipate impending assaults, colliding with enemy targets or dropping bombs on them.
By early 2025, drones were accounting for 60% to 70% of the damage and destruction caused to Russian equipment in the war, according to UK-based think tank the Royal United Services Institute…
Military commanders around the world are taking note. Taiwan is investing in mass-produced drones in anticipation of a possible conflict with China. Israel has recalibrated the Iron Dome air defense system in the war in Gaza to account for maneuverable drones — one of its biggest blind spots. European governments embarking on their largest rearmament since the Cold War have identified drones and counter-drone systems as an investment priority. The US Pentagon, which pioneered sophisticated and expensive drones sourced from big arms contractors, is looking to buy cheaper ones designed by startups and deployed en masse…
Small, light drones with multiple rotors have become the defining innovation of the war. Known as first-person view drones, they are typically controlled in real time via a video feed by an operator who can “see” through an onboard camera using electronic goggles so they can fly beyond the line of sight. Social media is full of videos showing the machines closing in on troops, armored personnel carriers, missile batteries and command posts until the moment of impact, when the picture turns to static…Other rotor drones are used to drop grenade-sized explosives on targets and can be reused if they make it back safely.
Bloomberg says that the parts used to make Ukraine’s drone fleet are bought “online”, but that is a euphemism. They are made in China.
An FPV drone is basically:
some injection-molded plastic parts
some trailing edge computer chips (microcontrollers, sensors, etc.)
an electric motor made of rare earth permanent magnets
a lithium-ion battery
The U.S. can still make plenty of trailing-edge computer chips, but the rest of these items are all China, China, China.
China does a large fraction of the injection molding in the world — about 82%, according to one 2024 estimate.2 Currently, I know of no government plan to restore America’s lost capacity in injection molding. In fact, Trump’s tariffs — if they ever go into effect — are expected to severely damage the U.S. injection molding industry, by cutting American injection molding companies off from imports of the specialized equipment they need.
China also makes most of the electric motors in the world. This is because China makes most of the magnets, and an electric motor is basically just made out of magnets. The rest of the world is scrambling to add magnet production capacity, but for the rest of this decade, China will dominate:
But this will be hard to accomplish. The magnets for electric motors are made out of materials called “rare earths”, which are almost entirely mined and processed in China.
In fact, China recently slapped export controls on its sales of rare earths to the U.S., causing chaos in a number of U.S. industries, and probably contributing to Trump’s decision to pause his tariffs. So far, U.S. efforts to mine and refine rare earths have fallen short (which itself is a topic for another full post).
Finally, and most importantly, we have batteries. A battery is the essential component of an FPV drone — it holds the energy that makes the thing go. Larger drones can use combustion engines, but to get something as small and cheap as an FPV drone, you need a battery.3
China makes most of the batteries in the world. In 2022 it had 77% of global manufacturing capacity. Here’s a projection out to 2030:
Even this projection, which shows America catching up just a little bit, is probably way too rosy. It was made at a time when Joe Biden’s industrial policy — specifically, the Inflation Reduction Act — was dishing out huge subsidies for American battery factories. Here’s what that looked like:
This wouldn’t have put American battery-making capacity on par with China, but it would have given us a fighting chance.
Now, though, Donald Trump and the Republicans are canceling the policies that were promoting American battery manufacturing:
A tax and policy bill passed by House Republicans…would gut subsidies for battery manufacturing, incentives for purchases of electric vehicles by individuals and businesses, and money for charging stations that Congress passed during the Biden administration. And it would impose a new annual fee on owners of electric cars and trucks.
Electric vehicles are crucial for battery manufacturing capacity, because in peacetime, they’re the main source of demand for batteries. Pump up the EV industry, and you pump up the battery industry too — just as the chart above shows Biden doing. Kill the EV industry and you kill the battery industry too, just as Republicans now want to do. Harming the solar industry will also harm the battery industry, because some types of batteries are used to store solar energy for when the sun isn’t shining.
GOP policies are already mauling the American battery industry:
[M]ore [battery] projects were canceled in the first quarter of 2025 than in the previous two years combined. Those cancellations include a $1 billion factory in Georgia that would have made thermal barriers for batteries and a $1.2 billion lithium-ion battery factory in Arizona…“It’s hard at the moment to be a manufacturer in the U.S. given uncertainties on tariffs, tax credits and regulations,” said Tom Taylor, senior policy analyst at Atlas Public Policy. Hundreds of millions of dollars in additional investments appear to be stalled, he added, but haven’t been formally canceled yet.
In fact, the whole boom in American factory construction that happened under Biden appears to be halting and going into reverse under Trump, thanks to a combination of tariffs and the expected cancellation of industrial policies:
The Ukrainian attack on Russia’s nuclear bombers shows how insane and self-defeating the GOP’s attack on the battery industry is. Batteries were what powered the Ukrainian drones that destroyed the pride of Russia’s air fleet; if the U.S. refuses to make batteries, it will be unable to make similar drones in case of a war against China. Bereft of battery-powered FPV drones, America would be at a severe disadvantage in the new kind of war that Ukraine and Russia have pioneered.
Unfortunately, Trump and the GOP have decided to think of batteries as a culture-war issue instead of one of national security. They think they’re attacking hippie-dippy green energy, sticking it to the socialist environmentalist kids and standing up for good old red-blooded American oil and gas. Instead, what they’re actually doing is unilaterally disarming America’s future drone force and ceding the key weapon of the modern battlefield to China.
In any case, unless America’s leaders wake up very quickly to the military importance of batteries, magnets, injection molding, and drones themselves, the U.S. may end up looking like the British Navy in 1941 — or the Italian Navy in 1940. A revolution in military affairs is in process, and America is willfully missing the boat.
Ironically, Japan made a similar mistake, directing far too many of its scarce resources toward battleship production instead of aircraft carriers.
Actually, estimates for this number are all over the place. Some are lower.
Incidentally, this is why everyone who confidently tells you that batteries can’t replace fossil fuels because they have “lower energy density” doesn’t know what they’re talking about. Yes, if you measure just the gasoline or kerosene or diesel in a combustion engine, its energy density is higher than that of any battery. But open up a car hood, and you’ll see a huge array of heavy, bulky tanks and tubes and machinery — that’s the engine required for turning gasoline into kinetic energy. Batteries don’t need an engine to covert their energy into kinetic energy — they just need some magnets. This means that the true energy density of batteries, counting the extraction machinery, compares pretty favorably with combustion engines in many applications.
2025-06-02 16:10:23
One thing I’ve learned, in my time as a pundit, is that many people claim to have read books that they have not actually read. This is true of giant dense tomes like Thomas Piketty’s Capital in the Twenty-First Century, but also — perhaps surprisingly — true of light, quick reads like Ezra Klein and Derek Thompson’s Abundance. The text of Abundance is only 222 pages long, and the font is large. And yet none of the book’s progressive critics seem to have any idea what Klein and Thompson actually wrote. For example, Aaron Regunberg, a lawyer who has been on a crusade against the abundance movement, writes:
[Y]es, abundance is about defeating progressives and remaking the Democratic Party as a libertarian, Never Trump Republican Party. Great, can we stop pretending it’s anything but that?
This is, of course, complete nonsense. To believe that Ezra Klein and Derek Thompson are libertarians or Republicans requires never having read anything that either of them wrote (or at least, pretending not to have done so). But most importantly, Abundance is all about how the government should have more power to build green energy, provide health care, and accomplish other progressive goals. You can argue with that idea, but you can’t call it libertarian in good faith.
But my favorite example comes from David Austin Walsh, America’s favorite angry history postdoc, who said all kinds of nasty things about Abundance before admitting, quite openly, that he hadn’t actually read it:
The Abundance critics continue to do a remarkably poor job of engaging the book or the abundance movement in general on an intellectual level. In general, they tend to be poorly informed, lazy, and sloppy.
And those critics aren’t winning. At the elite level, Democrats are embracing the abundance idea, and there’s plenty of grassroots interest too. I recommend the recent excellent writeup in the Wall Street Journal by Molly Ball. Some excerpts:
Democratic politicians are rushing to embrace the new [abundance] mantra. New York Gov. Kathy Hochul, Minnesota Gov. Tim Walz, Maryland Gov. Wes Moore and Colorado Gov. Jared Polis have all name-checked it publicly. New Jersey Sen. Cory Booker discussed it at length in his recent 25-hour Senate speech. Former Vice President Kamala Harris and the U.S. Senate’s Democratic caucus are among the many politicians who have recently sought the authors’ counsel. Not one but two congressional caucuses have recently formed to push legislation advancing the ideas laid out in the book…
It isn’t just party elders who have bought into the idea. Local Abundance clubs have formed in multiple cities and on college campuses…The book by Ezra Klein and Derek Thompson has been a surprise hit, with a sold-out national tour, hundreds of thousands of copies sold and two months on the bestseller list since its release in March.
But a recent poll suggests that although there’s lots of interest in the abundance idea, the message is still less appealing than populist red meat. Although you might think a poll by some pressure group calling itself “Demand Progress” might be biased toward populist causes, I found the wording in the questions to be reasonably fair:
Could I write a new phrasing for the “abundance” statement in this poll that would probably get more support? Sure, of course I could. It would go something like this:
2025-05-31 17:36:20
Oren Cass, the founder of the think tank American Compass, is probably the leading intellectual voice of MAGA economics. In a recent blog post, he discusses the question of whether economic forces are like gravity. He writes:
[Y]es, the physical world is governed by the laws of gravity. But it is not governed only by the laws of gravity. Indeed, anyone who thought he could reliably predict the motion of bodies with knowledge only of gravity would be something of a moron.
Oh, really? Would such a man be a moron? Then please explain to me how Edmond Halley, using only his knowledge of gravity, and having no understanding of any of the other forces of nature, was able to predict a solar eclipse in 1715 to an accuracy of four minutes:
In 1715…a total solar eclipse was visible across a broad band of England. It was the first to be predicted on the basis of the Newtonian theory of universal gravitation, its path mapped clearly and advertised widely in advance. Visible in locations such as London and Cambridge, both astronomical experts and the public were able to see the phenomena and be impressed by the predictive power of the new astronomy…Wikipedia will tell you that this is known as Halley’s Eclipse, after Edmond Halley, who produced accurate predictions of its timing and an easily-read map of the eclipse’s path. Halley did not live to see the confirmation of his predictions of a returning comet – a 1759 triumph for the Newtonian system – but he was able to enjoy his 1715 calculations, which were within 4 minutes[.]
Cass — who holds a Bachelor’s in political economy from Williams College and a law degree from Harvard, but has no apparent training in physics — confidently assures us that anyone who attempted what Edmond Halley did would be “something of a moron”.
This is the kind of insult that says less about the target than about the person doing the insulting. Before you make confident assertions about a field of study, you owe it to your readers to attempt to understand that field at least a little bit.
A botched physics analogy is harmless enough. But Cass’ main argument isn’t about gravity at all — it’s about economics. And it’s here where his willingness to make grand pronouncements about whole fields of study gets him into real trouble.
Cass’ post is a response to a Wall Street Journal op-ed by Matthew Hennessey. Hennessey, in turn, is responding to J.D. Vance’s declaration that markets are a “tool”. Hennessey argues that markets are more like a force of nature than a tool.1 Cass is trying to rebut Hennessey, criticizing market fundamentalism while also taking a swipe at the entire discipline of economics.
Now, I am no fan of market fundamentalism, and I spent my early years as a blogger bashing the field of (macro)economics — often with even more scorn than Oren Cass employs in this post. But I like to think that when I did this, I generally stuck to making specific criticisms about actual economic models and methods.
A lot of econ critics don’t do this. Back when I was at Bloomberg, I used to have fun poking at the grandiose broadsides against economics that periodically appear in British publications like The Guardian or The Telegraph. These critiques tend to repeat the same old nostrums over and over — economics isn’t a science, it doesn’t do controlled experiments, its assumptions are bad, its theories don’t work, people can’t be predicted like particles, etc. etc.
There are grains of truth to these boilerplate critiques, but the people who write them generally haven’t bothered to pay much attention to what modern economists actually do. Here’s what I wrote in a Bloomberg post back in 2017:
[E]conomists have developed some theories that really work. A good scientific theory makes testable predictions that apply to situations other than those that motivated the creation of the theory. Slowly, econ is building up a repertoire of these gems. One of them is auction theory, which predicts how buyers will bid for things like online ads or spectrum rights -- Google’s profits are powered by econ theory as much as by search algorithms. Another example is matching theory, which has made it a lot easier to get an organ transplant. A third is random-utility discrete choice theory, which is used in everything from marketing to transportation planning to disaster preparedness.
Nor are econ’s successful theories limited to microeconomics. Gravity models of trade, though fairly simple in nature, have proven very successful at predicting the flow of international trade.
These and other successful economics theories can be used confidently in a wide-variety of real-world situations, by policy makers, engineers and businesses. They prove that anyone who claims that econ theories will never be reliable, because they deal with human beings instead of atoms, is simply incorrect.
Yes, studying mass human behavior is different than studying the motions of the planets, in a number of important ways. But the intellectuals who loftily declare that economics “isn’t a science” don’t seem like they’ve bothered to think very hard about what those differences are, or when and why they matter.
For example, what do the people who write that “economics isn’t a science” think about natural experiments — the empirical technique that has taken over much of econ research in the last three decades? Do they think that these are always less informative than lab experiments in the natural sciences? And if so, why? What do they think are the strengths and weaknesses of natural experiments relative to lab experiments, and how much can they help us test theories and derive general principles about how economies work?
I suspect that very few of the econ critics have thought seriously about these questions, and that far too few have even heard of natural experiments. Certainly, if they have, they must have some good reason for never mentioning them. And certainly they must have good reasons for never talking about auction theory, matching theory, discrete choice models, gravity models, or any of the other economics theories that prove themselves in the real world day in and day out. Right?
But the screeds in The Guardian or The Telegraph are downright erudite compared to what Oren Cass serves up in his own criticism of econ. Here’s what he writes:
Economics is nothing like physics. Its principles are not generated from repeatable experiments, nor do they hold consistently across space and time. Trusting otherwise is a quite literal example of the blind faith and fundamentalism at issue.
That’s it. That’s literally Cass’ entire criticism of the field of economics in this post. He spends the rest of the post pulling quotes from conservative political thinkers — G.K. Chesterton, Robert Nisbet, Yuval Levin, Roger Scruton, etc. — who urge us to value things like community, tradition, etc., or who assert that markets can’t work without a robust social fabric. Those are interesting things to think about, to be sure, but they don’t bear on the question of what, exactly, Cass thinks is so inadequate about economics.
Cass does not name or criticize any specific economic theories in this post. He cites zero papers and names zero researchers. I looked through a bunch of his other posts about economics, and I almost never found him naming or criticizing any specific theories in those posts, either. In one post, I did find him criticizing the theory of comparative advantage, and he did make one useful, substantive point about it — that comparative advantage can’t explain trade deficits and surpluses. He’s right about that. That’s by far the most substantive, knowledgeable criticism of economics that I could find on his blog.
If Cass is aware of any economic models other than comparative advantage and the basic Econ 101 supply-and-demand model, he plays his cards close to his chest. He doesn’t mention gravity models of trade, which some economists use to try to predict the effects of tariffs (with some success). Nor does he mention Paul Krugman’s New Trade Theory, which implies that countries can sometimes benefit from targeted tariffs against other countries’ national champions (but which wouldn’t recommend the kind of broad, sweeping tariffs Trump has tried to implement).
Neither of those theories is outside the mainstream; both were invented by economists who went on to win the Nobel. Why doesn’t Oren Cass mention them, or grapple with their implications, or use their existence to inform his criticisms of the field of economics? My guess — and this is only a guess — is that Cass is completely unaware that these theories exist, that he has no interest in discovering whether such theories exist, that if he did discover them he would have no idea how to evaluate them, and that even if he did know how to evaluate them he would have no interest in doing so.
A sophisticated understanding of what mainstream economics actually says and does is not useful to Oren Cass’ project, which is to denounce intellectual rivals within the conservative movement. If you want to actually figure out how trade works and what tariffs do, it would help to look at the research literature. If you didn’t get the training needed to understand that literature, it would help to ask some people who did get that training, or at least read a little Wikipedia and ask ChatGPT a few questions.2 That won’t give you all the answers — the world’s best economists don’t even have all the answers — but it would leave you a lot more knowledgeable than you started out, and it would give you a much better idea of where economists are on solid ground and where there are gaps in their understanding.
On the other hand, if all you want is to dunk on Wall Street Journal writers, perhaps all you need is some hand-waving rhetoric about “market fundamentalism” and a vague half-knowledge of one simple trade theory developed 200 years ago.
The real problem with these econ critics — both the lefty writers in The Guardian and the new crop of MAGA defenders — is that their project is fundamentally political. The lefty writers think that if everyone accepts that econ isn’t a science, and that the econ Nobel isn’t a real Nobel, and people aren’t like particles, and so on and so forth, then some sort of lefty ideology — Marxism, or degrowth, or whatever — will flow in to fill the hole left when economics vanishes. The MAGA writers think that it will be Trump’s economic ideas that fill that void instead.
But Matthew Hennessey, the Wall Street Journal writer, got one big thing right: Simply replacing academic theories with your own ideology can win you power, but there are important things it can’t do. It can’t change the nature of what tariffs actually do to the economy. Even if you and your friends and your political allies all shout very loudly that tariffs will restore American manufacturing, and act very scornful toward nerdy academics who tell you it doesn’t work like that, the economic headwinds that tariffs actually create for American manufacturers won’t change one iota.
However limited mainstream academic economics is as a tool for understanding the consequences of your policies, ideology is even worse.
Who’s right, Vance or Hennesey? Both are right. Market forces are, quite literally, forces of nature. And markets themselves are, quite literally, a tool. Tools work by harnessing forces of nature. A pendulum clock works by harnessing the force of gravity. A market works by harnessing the forces that drive people to buy and sell things. Vance is right that markets should be shaped to serve our desired ends, rather than being an end in and of themselves. Hennesey is right that market forces can’t be denied, ignored, or wished away.
Just to see how AI is doing, I asked ChatGPT o3 about the papers on gravity models. Its characterization of the papers’ results was oversimplified and omitted crucial nuance about the ex ante predictive accuracy of Fejgelbaum et al. (2020). But overall its explanations weren’t too bad!
2025-05-30 15:03:18
In the 2010s, economics discussions went mainstream. Blogs and (especially) social media meant that data-driven, high-level debates about economic issues were no longer confined to the op-ed pages of the Wall Street Journal and the talking heads on CNBC. This was a good thing, but I think it had some unintended consequences. One is that a lot of the narratives that made sense in 2012 or 2015 got embedded into the popular consciousness, and have stayed there even though the underlying reality has changed.
One big example of this is wage stagnation. In the early 2010s, lots of people “knew” that American middle-class wages had stagnated since the early 1970s, with only a brief temporary revival in the late 90s. And so naturally, a lot of discussions revolved around how to fix that. But even as we were having those arguments, wages accelerated again. Now, when we look back with the benefit of ten more years of hindsight, the years since the mid-1990s look like a bumpy but substantial increase:
Another big narrative, related to the wage stagnation idea, is that service costs are relentlessly rising in the U.S. economy. You’ve probably seen some version of Mark Perry’s famous “price changes” chart:
This chart tells a simple, powerful story: Services get more expensive, while physical goods get cheaper. Health care, education, and child care went up in price faster than wages, while cars, clothing, electronics, and toys got more affordable.
And the story was compelling because it came with a simple theory to explain it. This was the notion that manufacturing productivity naturally increases faster than service productivity. Conceptually, it seems easier to figure out how to rearrange production processes in a factory, and apply new machine tools, than to figure out new ways to educate kids or take care of Grandma.
And during the 20th century, this principle held true. For example, from 1987 to 2011, manufacturing productivity more than doubled, while overall productivity increased by only 70%:
In fact, if you look on Wikipedia’s page for “Baumol’s Effect”, also known as Baumol Cost Disease, it lists this productivity divergence as an explanation for diverging costs.
If service costs rise relentlessly while manufacturing costs fall, it portends a grim future — one where we have cheap gadgets, but where the big necessities of modern middle-class life are increasingly out of reach. And in fact, that was the story a lot of people were telling in the mid-2010s.
That story led to a certain techno-pessimism. If technology could give us cheap gadgets, but couldn’t make the basics of modern life any cheaper, what good was it? The failure of educational technologies like online education only seemed to drive the point home — it seemed like we’d always just be stuck with a teacher giving lectures on a board to 20 or 30 kids. Though some people dreamed of robot nurses and nannies, these seemed very far from materializing.
It also fed into cynicism about trade. If China gave us cheap gadgets but we were still struggling to pay for health care and education, wasn’t globalization sort of a booby prize? I think we see echoes of that sentiment in the GOP’s willingness to embrace Trump’s tariffs.
On the policy side, rising service costs led to two different big ideas.
First, if technology couldn’t make services cheaper, some people (including myself) reasoned, we should use policy to make them cheaper. Perhaps deregulation was the key — many libertarians blamed rising service costs on a policy of “cost disease socialism” that restricted supply while subsidizing demand, implying that we just had to eliminate the supply restrictions and cut the subsidies and the market would do its thing. Progressives advocated a more statist approach — after all, most other countries have some form of national health insurance, and they all have cheaper health costs than America does, so perhaps government bargaining power was the key to driving costs down. (Personally, I supported doing both approaches in parallel.)
The second big policy idea said that what we want isn’t necessarily for services to cost less to produce; what we want is for services to cost less to the people who consume them. Productivity improvements in the service sector would just throw human beings out of work, as they had done in manufacturing; instead, a future where everyone gets a high-paying job taking care of everyone else sounded pretty good to some progressives.
So how do we make services cheap to consume, but keep using “care jobs” as the future of good-paying middle-class work? Taxes and subsidies were how progressives planned to square that circle. If rich people could be taxed at much higher rates, that money could be used to create higher wages and more jobs for care workers and lower prices for consumers. For example, here’s what Heather Boushey, one of Biden’s most influential economic advisors, told the Aspen Institute in 2021:
“This year has taught us all, across the nation, how the care economy is the foundation of our economy,” Boushey remarked…Boushey reflects on the twin crises of a lack of good care and a lack of good jobs and notes that investing in the care economy creates a positive feedback loop that allows family caregivers to find quality care and care workers to find quality jobs.
Natalie Foster and Amanda Newman wrote that “policy [should] contribute to building a care economy that dignifies the work of caregivers and expands access to quality, affordable care”. Felicia Wong of the Roosevelt Institute endorsed the idea of universal child care as part of a new New Deal.
This idea represented a substantial portion of Biden’s economic plans. The Build Back Better bill — which never made it into law — would have created a national family and medical leave program and a universal pre-kindergarten education, while heavily subsidizing child care and various in-home services. The total price tag would have been in the hundreds of billions of dollars. This big push for service subsidies was mostly defeated in Congress, though some health care subsidies did make it into the (misnamed) Inflation Reduction Act.
In other words, the trend of increasing service costs defined many of our economic debates for a decade. There was just one small problem — by the time we started talking about how to address this trend, the trend had changed.
I’ve been meaning to write this post for a while, but what prompted me to finally do it was when Matt Yglesias posted a very nice chart of health spending as a percentage of GDP. Here’s a version of that chart with two different measures of health spending:
Note that “personal consumption expenditures” includes spending on behalf of consumers by the government and by nonprofits, so this isn’t just household spending on health care — it’s all of it.
We also see the same pattern when we look only at hospital services — the item whose cost went up the most on the Mark Perry chart:
So what we see is that until around 1990, health spending rapidly ate up a bigger and bigger portion of our national income. Then the increase slowed down, but it did go up some more until around 2009. But after that, it leveled off; in 2024, Americans didn’t spend a greater percent of their income on health care than they did in 2009. And in fact, the increase since 1990 has been pretty modest — if you look only at the service portion of health care (the blue line), it’s gone up by about 1.5% of GDP over 34 years.
OK, so, this is total spending, not the price of health care. Is America spending less because we’re getting less care? No. In cost-adjusted terms, Americans have been getting more and more health care services over the years:
What’s happening is that although health care prices have still been going up, they’ve been going up more slowly than before. Here are health service prices relative to the overall price level, as measured with the PCE:
You can see that costs level off in 2009, and then actually drop after the pandemic relative to other costs. This doesn’t mean health care got cheaper after 2020; what it means is that health care prices rose more slowly than other prices did.
If you use the CPI (as Mark Perry’s chart does), things look a little different — medical costs keep rising faster than overall costs during the 2010s, but then slow down quite a bit after the pandemic:
Why the difference between CPI and PCE? Two main reasons. First of all, health care itself is a bigger component of PCE than of CPI, while CPI weights housing more heavily instead. So the fact that health care prices have gone up faster than housing prices means that CPI will show a greater increase of health prices relative to the total. Second, PCE tries harder to adjust for quality improvements. Health care has gone up a lot in quality over the years, and this has been hard to capture with traditional price measures. This is from the conclusion of a Brookings literature review from 2016:
Traditional measures of productivity growth in the health sector most likely understate it, because they don’t adjust prices for substitution from higher to lower cost inputs and because they don’t take account of changes in quality over time…The evidence to date suggests that adjusting health care expenditures for changes in quality leads to a significant reduction in the rise in health prices over time, and, indeed, these prices may even have declined relative to other prices.
PCE has this problem less than CPI does, so it shows more of a cost slowdown.
OK but anyway, what we really care about at the end of the day is affordability — i.e., how much health care an average America can buy. A good way of measuring affordability is to look at median income divided by an index of health care prices — in other words, how much health care the typical American can buy with their annual income.
When we look at this, we see that health care got steadily less affordable until around 1990, then leveled off, and now has been getting more affordable since around 2012:
So overall, health care is probably now more affordable for the average American than it was in 2000 — in fact, it’s now about as affordable as it was in the early 1980s. That doesn’t mean that every type of care is more affordable, of course. But the narrative that U.S. health costs just go up and up relentlessly hasn’t reflected reality for a while now.
Why have health care costs stopped going up rapidly? Some credit the cost control provisions of the Affordable Care Act of 2009 (aka Obamacare). This is from 2019:
March 23, the ninth anniversary of the ACA’s passage, presents a good opportunity to examine its legacy on cost control…Fast forward to December 2018, when [the Office of the Actuary of the Department of Health and Human Services] released the official tabulation of health care spending in 2017. The bottom line: cumulatively from 2010 to 2017 the ACA reduced health care spending a total of $2.3 trillion…[H]ealth care spending in 2017 was $2,000 less per person than it was projected to be. And for the 176 million Americans who have private employer-sponsored insurance, their lower premiums averaged just under $1,000 per person.
Another possibility is that Americans just got tired of paying more and more for health care, and started balking at higher prices. People who made excuses for high American health costs claimed that it was just an income effect — that as nations get richer, they simply spend a higher and higher percent of their income on health care. But as I pointed out, at some point this explanation has to fail, because it can’t be true that health care eventually consumes all of a rich nation’s income, to the exclusion of everything else.
It could just be that Americans were willing to pay more for health care as they got richer, up to a point, but that at some point they said “OK, that’s enough.”
It’s not just health care whose costs have leveled off. College tuition is no longer getting more and more expensive every year. In fact, in inflation adjusted terms, tuition has actually fallen since the pandemic:
Here’s another way of visualizing that same data, and you can see that adding in housing and food doesn’t really change the story:
Of course this doesn’t include financial aid (nor does Mark Perry’s chart, nor do official inflation numbers). Financial aid has been going up, especially at private schools. When you include that, it turns out that private four-year nonprofits are actually less expensive in inflation-adjusted terms than they were in the mid-2000s, even without accounting for rising incomes:
For public schools it’s about the same:
In other words, higher education has been getting more affordable for years, and the decrease in affordability in the late 2000s and 2010s was significantly overstated. The popular narrative that college is getting less and less affordable is wrong, and it needs to change.
Why are college costs going down now? It’s basically a story about demand. College enrollment peaked in 2010 and began to fall:
This enrollment decline happened for every age group of students. It wasn’t just because the number of young people in America plateaued, either; enrollment fell as a proportion of young people.
As for why enrollment rates declined, that’s an open question. My own hypothesis is that college costs rose and the college wage premium fell until a market equilibrium was reached.
It’s not just true that service costs have stopped exploding in America. The narrative that service productivity is stagnant, while manufacturing experiences fast productivity growth, has been flipped on its head since 2008. Manufacturing productivity abruptly flatlined around the time of the Great Recession, while overall productivity has risen:
This implies that service productivity growth has accelerated. And indeed, if you look at specific industries, you see a lot of services where productivity is going up quickly in America:
Now, these changing trends don’t mean that services are cheap and we can stop thinking about service costs. First of all, there are still some services that are getting less affordable over time — most notably, child care. Second, the recent mild increases in affordability for health care and higher education haven’t erased the big cost increases that happened in the 1980s, 1990s, and early 2000s; Americans still pay a lot more for these things than Europeans or Asians do, relative to their incomes. So there’s still probably scope to bring down the costs of health care and college.
But with all that said, the change in the trends in service costs and service productivity mean that our debates about these topics need to change.
First of all, services don’t look as resistant to technological improvement as we once thought. The increased adoption of IT in health care is probably having a big effect. We can probably apply that insight to child care and K-12 education as well. And the introduction of AI into education — and everything else — will probably accelerate the trend even more.
The possibility of technology-driven productivity improvements in service industries also means that we don’t have to resort to expensive subsidies in order to make services both better-paying for workers and cheaper for consumers. Instead, we can encourage technology adoption, and enjoy higher wages and lower prices, just like we did with manufacturing in the 20th century.
(As for where the jobs of the future will come from, it’s looking more likely that professional jobs, rather than care jobs, are the next big thing.)
Falling service costs should also make us more optimistic about capitalism, and about the future of the middle class. For the last decade and a half, a lot of our economic debates have been premised on the idea that the American economy is fundamentally broken, and that life just keeps getting more and more expensive for regular people. That’s looking a lot less true than it did in 2012.
We spent a long time talking about how service costs were eating the world, and I don’t expect that narrative to disappear overnight just because I posted one blog post with a lot of charts. But it’s well past time for our discussions on this topic to start shifting in response to new information.
2025-05-29 15:43:37
The Trump administration has just paused all visa interviews for all foreign students seeking to study in the United States. This means that until and unless that pause is lifted, the number of new students who can apply to come study in the U.S. is zero. Ostensibly, the pause is a short-term measure that will give the Trump administration time to set up a procedure to vet prospective students’ social media accounts for ideas it doesn’t like; interview scheduling could resume within days, or it could take longer.
This is important because it’s part of a more general attack on foreign students. Trump already forbade Harvard University from enrolling students from overseas, although the action was blocked by a court. This was part of a high-profile confrontation with Harvard over anti-white racial discrimination, but it also reflected Trump’s belief that international students are taking spots that ought to go to Americans. Bloomberg reports:
President Donald Trump said Harvard University should cap foreign student enrollment at 15%, ratcheting up his effort to force policy changes at the elite institution.
“I think they should have a cap of maybe around 15%, not 31%,” Trump said Wednesday at the White House. “We have people who want to go to Harvard and other schools, they can’t get in because we have foreign students there.”
Trump has the same attitude toward college spots that he has toward immigration and imports. To him, everything is just a lump of fixed size — a pie to be divided. In his mind, if you kick out immigrants, the number of jobs doesn’t go down — the jobs just get parceled out to native-born Americans instead. If you ban imports, Americans don’t consume less — they just buy American-made products instead. And if you kick out foreign students, the number of college spots doesn’t go down; American kids just get more.
Of course, Trump is wrong about that, as I’ll explain. But kicking out international students dovetails with a number of other Trump administration priorities. It allows them to strike out at left-leaning universities. It allows them to implement ideological tests for immigrants — something they want to do in general.1 It allows them to reduce the number of foreigners in the country, which they feel strengthens Western civilization.
And for the China hawks in the administration, it allows them to kick out the Chinese, whom they suspect of being spies:
The U.S. will “aggressively revoke” visas for Chinese students, Secretary of State Marco Rubio said Wednesday…“Under President Trump’s leadership, the U.S. State Department will work with the Department of Homeland Security to aggressively revoke visas for Chinese students, including those with connections to the Chinese Communist Party or studying in critical fields,” he said in a statement. “We will also revise visa criteria to enhance scrutiny of all future visa applications from the People’s Republic of China and Hong Kong.”
So kicking out foreign students is, in many ways, the perfect policy from the Trump administration’s point of view. The problem is, it’s a terrible policy. But like tariffs, this policy will immediately and needlessly hurt some parts of America, while also leading to a long-term erosion of American competitiveness, innovation, and general economic strength.
There are three basic ways that international students are important for the U.S. economy:
Their high tuition payments subsidize the education of native-born American students.
The money they spend in America helps keep the economies of small towns and cities going.
They are crucial for American innovation and technological strength.
American universities fund themselves five main ways: 1) government money, 2) alumni contributions, 3) investment income from endowments, 4) revenue from hospitals and other businesses, and 5) tuition. Tuition currently provides about 17.5% of the total overall, and almost a third for private nonprofit universities. Here’s data from 2020:
