BRAVO! One of the best essays on tariffs I have read. I have read every essay on tariffs from top economists since about January 2025. All the essays deliver the same message, but DF's essay delivers it in language anyone can understand --- if they want to. Some people just don't want to.
I don't have any academic training in economics, so if your response to this question is go read a book, that'll be fair. But maybe it's an easy thing to answer...
"while it increases jobs in import competing industries it reduces jobs in export industries"
If your export industries are all high-paying jobs, and you primarily import the products of low-paying jobs, and your constituency is seeking low-paying jobs, even if overall economically it all comes out in the wash of the exchange rate (or even overall leaves you economically poorer) would not tariffs still result in a promise fulfilled to your constituency, re-shoring the kinds of jobs they're looking for?
I ask because part of my moral calculus in evaluating these tariffs is that it might be worth a somewhat weaker overall economy if it increases job opportunities for folks lower on the totem pole. A lower tide prevents folks without boats from drowning, or something like that.
What you describe is possible. In that case the tariff is a costly mechanism for income transfer — costly because it does make the country on net poorer, something I didn't discuss in that post. The simplest way of seeing that is to view trade as a technology for producing things. Instead of building a car in Detroit you grow wheat in Iowa, put it on a ship and send it into the Pacific. It comes back with Hondas on it.
Trading wheat for Hondas is profitable only if it costs less US resources to produce cars that way than by building them. An auto tariff taxes the more efficient technology in order to shift production to the less efficient technology, so produces things as a higher cost in US resources.
In one of my earlier posts, on Vance not on tariffs, I speculate that your argument might be his position. But I have seen no evidence that either Vance or Trump has actually tried to design tariffs that way or tried to figure out how the labor force in export industries compares to that in import competing industries.
Possible? With constant returns to scale technology everywhere, it must be that in rich countries wages for sector specific skills rise more than prices in the short run, and that the same is true for unskilled labor unless it has weird tastes. In the long run, wages for the unskilled rise more than prices, full stop.
This is consistent with Trump's policies. This doesn't mean Trump is a sharp trade theorist, but it does reflect that since the 1920's this was the intuition of the US labor movement. At the time, most economists thought the intuition was wrong, but the correct answer got formalized starting around 1940.
There are some wrinkles with increasing returns to scale. One is that global free trade promotes competition to get a technological monopoly over the whole world, so that growth is promoted by free trade. But that's a long run phenomenon, after the next election for sure.
I don't see the argument for your "must be;" so far it is pure assertion. Wages rise in import competing industries, fall in export industries — more jobs for auto workers, fewer jobs for farm workers and people drilling for natural gas. You are now producing things in a less efficient way, so the total amount available to be distributed is less.
If the tariff reduces the trade deficit, possible but not guaranteed, that reduces the amount of foreign capital coming in, raising the return for domestic capital, a transfer from labor to capital.
Alas, this is the Heckscher-Ohlin Model. Assumes constant returns to scale in everything.
From free trade, the capital rich country imposes a uniform tariff. The relative price of the imported commodity, which is labor intensive, rises. This attracts capital and labor from the export industries, but in a different proportion than used in the import competing industry, more capital relative to labor. Hence, labor productivity rises in the import competing industry. Hence, the real wage for unskilled labor rises. Long run, of course.
In the short run, sector specific skills [capital] rise in the import competing industries. Unskilled labor is also attracted to increase output, and if the unskilled don't consume too much of the good whose price has increased, they are better off, too.
Apologies about assertions. It's difficult to know how much or how deeply to explain such processes. My experience -- not so much here, but everywhere else -- has been that nobody cares and they stick to their prejudices. So, first an assertion; then we see what happens.
Now, no one believes the constant returns to scale everywhere assumption anymore. So we get to computational methods, market by market, whose results are not necessarily general and I do not have in my head. My guess is that something of the constant returns results will be saved.
Wouldn’t this be around manufacturing, specifically?
I can buy the idea that if countries are atomic units, trade makes us all better off. But it seems to me the trade with China has been
- fantastic for American capital
- good for Chinese capital
- good for Chinese labor
- abysmal for American labor
I can happily believe that _on net_ the surplus is positive. But it seems the epitome of academic aloofness to ignore the declining quality of life for working class americans because look, we made our global adversary much richer, in exachange for enriching the Americans most likely to support restrictions on speech and contempt for the people who lost the most from the trade.
If you see tariffs as a form of income redistribution, you’re also ignoring the possibility of dignity from paid labor that isn’t present with a handout.
"If you see tariffs as a form of income redistribution, you’re also ignoring the possibility of dignity from paid labor that isn’t present with a handout."
It is a handout — taxing production by export industries via trade in order to subsidize production by import-competing industries.
You mean the dignity from hiding the fact that it is a handout?
Economics ignores things like conflict and war. Your theory works just as well whether we are trading with a country that shares our values or one that wants to dominate us. Once you assume certain boundary conditions, sure, trade equilibrates. Those boundary conditions don’t arise naturally and require investment. So to call tarriffs a hand out is to ignore the massive subsidies that exporters get, in the form of the U.S. Navy protecting their cost of shipping. If you don’t account for that cost and assume safe shipping lanes just happen, ok, then tariffs are a handout. But when you account for the fact that military power requires we manufacture our own weapons and weapon systems and logistics, you see that tariffs are charging exporters the appropriate cost of the services that taxpayers provide them.
That's a stretch. You may as well claim they get subsidized free advertising every time I tell someone how much fun I had on my ship in the Navy, and claim I am an externality every time I tell someone how much I hated the military side of it.
What makes it a stretch? Why should we ignore the cost of providing safe trade networks, given that exporters rely on these to run their business? How is that protection not a subsidy that benefits some industries moreso than others?
I do not buy this has been bad for American labor. Someone (Cafe Hayek? A quick search didn't find it) had a link in the last few days showing that manufacturing unemployment is less than general unemployment, and that manufacturers report more trouble finding workers.
Offshoring low skill low productivity jobs because high skill high productivity jobs crowd them out is good.
Farmers is a very broad category, like factory workers. Lots of jobs in food production are low skilled (“stoop labor”) and others are quite high skilled. You can’t think broadly about industries and the type of jobs they provide as they all tend to cover the gamut.
When I lived in China, my non-CCP coworkers spoke of the "Iron Rice Bowl," hugely inefficient factories in the interior that produced nothing of note except air pollution and toxic waste to keep the unproductive middle of the country busy.
The CCP members called it something like "Our Outstanding Social Security System." Why you never read of this in the Western press is left as an exercise.
On the one hand, one would be hard-pressed to convince the UAW workers downstairs—interestingly, the executive offices are on the same floor—that on-shoring auto manufacturing at the expense of abstract improvements in efficiency is a Bad Thing.
The question that keeps recurring—the late Jerry Pournelle would occasionally blog on it—once we outsource every job that can be done by someone on the bottom half of the IQ distribution, what do you now do with half, by definition, of your population.
I don't have an answer, but I observe that I've never seen a politician from any country, even despotic dictators, who admit to any other model of international trade than "Trade Deficits: Bad" zero-sum mercantilism.
You don't outsource all such jobs unless you have a minimum wage law that makes low productivity workers unemployable. Besides, there are jobs, such as babysitting, security guard, housecleaning and yard work that can't be outsourced.
I usually tell people that dollars in and out have to equal, otherwise someone is using them to light cigars or as wallpaper. I usually get corrected by economists as to what "in" and "out" actually mean, but I can't think of a simpler and easier way to start a conversation and find out if they really want a conversation or are just ranting.
You’ve proposed a bunch of theories that claim trade always balances. Sure, if I accept the premises of your theory, I’ll agree. But why should I think your theories are true?
Is there evidence you can show which demonstrates that your claims accurately map back to empirical measurable reality?
The idea that some countries have cheaper labor than others is patently obvious to anyone with eyes to see. I’m sure you can come up with theoretical abstractions where this difference somehow doesn’t matter. What you’re doing here - not even trying to do - is explain why these tautological theories explain something that basic naked eye intuition is failing to see.
American workers clearly enjoy higher standards of living than their Chinese counterparts. I’m open to the possibility that somehow this doesn’t matter in terms of where labor goes. But you’ll need evidence , not just some clean and tidy set of tautologies where things balance out. Show me where my intuition departs from reality using actual evidence, and I’ll agree.
In summary this is like asking me to consider a spherical cow:
“ consider the simple case of two countries and no capital flows. “
Yeah but capital clearly does flow, that’s the whole point of it. For decades it’s flown to China, instead of being invested here. Why should I think this will somehow balance?
In explaining things it is often useful to start with a simple case, work through the logic of that, then add back what you eliminated to make it simple. Did you stop reading after I explained the logic without capital flows and went on to explain it with?
If you had bothered to try to understand what I was saying, you might have noticed that the trade deficit means Chinese capital flowing to the US, not the other way around.
Yes, I get that you’re starting with a toy system and building it up. I apologize for not investing in better understanding your claims. When you use examples like crystalline spheres and Miasmas, what I think is, you are imagining that economics has more sound empirical footing than it really does. I understand these ideas but suspect - and cannot prove - that “everything equilibrates over time” isn’t true, and what’s happening is you’re depending on some set of assumptions i can’t ferret out.
The closest I can get is this: yes, I understand this idea that trade balances are tautologically equivalent to capital flows. I understand this. But I don’t think capital flows are equal - Chinese “investment” in the U.S. seems to have the effect of subsidizing American government spending, which I think is a net waste. American investment in China, on the other hand, lead to the creation of a bunch of factories that produce real goods.
If I assume that all dollars are equal and prices measure value, I have to buy your conclusion. But I don’t think this is true.
I think not all investments actually produce long term value, and that taxing exporters to subsidize local manufacturing makes sense for the same reason that every cell in my body has walls around it. I’m sure you could make a fine economic argument that my body would operate far more efficiently if you got rid of all those barriers to nutrient transfer - but there’s a reason our physiologies make great effort to keep the outside world out, and I think the same applies to relationships between nations.
"what I think is, you are imagining that economics has more sound empirical footing than it really does."
I don't think any part of my argument depended on empirical claims. It was straight logic. You pretty clearly did not follow it, as demonstrated by getting the direction of the capital flow backwards.
There are doubtless things to criticize about economics, but replacing economic arguments with vague analogies to biology is not an improvement.
You are making claims about value, using a model that assumes a number of things, like “spending and investment are equally good, so long as the dollar numbers equal out.” Sure, I labeled an edge in your causal diagram backwards. I understand that you are claiming a toy system equilibrates, and i think your toy system ignores important parts of reality. You respond by claiming your toy system is just as solid as, say, radio astronomy- this is unearned credibility.
I can happily agree that tariffs on a toy system like you are sketching out don’t make it better. I don’t think we’ll agree on how well that toy system maps to reality. I think this is what you’re missing - people are rejecting your theory, and when you say you know the truth, they remember “subprime is contained,” and probably rely on some intuition like, “if economists knew all that much they’d all be rich from trading stocks, and we wouldn’t have giant crashes that somehow always seem to rebound in a way that helps the rich more.”
I do think tariffs are the least bad form of taxation, because the tax enforcement mechanism aligns incentives with the most essential function of government- securing the border. What do you think of tarriffs vs income taxes in terms of the incentives and externalities they create?
As it happens, economics is the one science that contains an answer to "if you are so smart, why are you not rich?"
I am not asking you to believe my conclusion because it is what economics says. I am showing you the relevant logic — and you are determinedly unwilling to make the effort to understand it.
"But I don’t think capital flows are equal - Chinese “investment” in the U.S. seems to have the effect of subsidizing American government spending, which I think is a net waste."
Then your complaint should be with the US government spending too much and borrowing, not those who loan them the money.
I have plenty of complaints there, for sure. But help me understand- wouldn’t my body operate far more efficiently without spending all that energy pumping ions across cell membranes? What makes nations different from biological structures, where lots of energy is spent creating barriers to the free flow of energy, even inside the physiology of the same organism?
Because competition. Because success breeds copycats. Because monopolies breed arrogance and sloppiness and caution.
And how can the US attract a trillion a year capital investment if capital has been flowing to China? You really need to show some actual numbers if you want people to believe your story.
In the 1950’s, a man could support a stay at home wife and three kids while driving a garbage truck for the city of Santa Clara. That’s not doable any more. No amount of numbers are going to convince voters who’ve seen their once healthy towns reduced to poverty and meth - with all the jobs having left for China - that the leaders know what they are doing.
from Don Boudreaux's Cafe Hayek, about growing up a little later than me, and his father's factory work. He has many more on the subject.
This bullshit nostalgia can't fade away fast enough. I don't know anyone who was alive then who thinks it was better. The only excuse I can imagine for being so deluded is that they were a child and had no responsibilities and only remember the fun of being a child with mommy and daddy taking care of all the dreary adult stuff.
This ignorant and bullshit historical favoritism really pisses me off.
* Medical care was miserable. The President's son died of an infected blister in 1924, just thirty years before, and medical car for ordinary people wasn't that far removed.
* Airplane travel was expensive and slow.
* Railroad was how most people traveled long distances, and it was expensive and slow.
* Crossing oceans was still a matter of ocean liners, expensive and slow.
* Telephones were expensive and limited to one per house unless you paid exorbitantly for extensions. Long distance calls were slow and really expensive. International calls were for the rich only, and slower than molasses.
* I can microwave food in my freezer and eat Chinese or Korean food in five minutes. I can drive to the supermarket and get other foreign food and eat it within an hour. Only a few city dwellers could get that even in the 1950s, and that required going to a restaurant and paying more.
* I have more varieties of food and drink in my pantry than John D. Rockefeller could have had for all his wealth. He died in 1937, just 13 years before your fake 1950s fantasy.
* The only thing John D Rockefeller had going for him was enough money to hire servants and boss people around.
* I would not swap my life for the 1950s no matter how much money was added in to the bargain.
ETA: I just asked Gemini to compare ship and airliner passenger traffic across the Atlantic. Airliner traffic didn't exceed ship traffic until 1958.
ETA: Polio was still a thing, with victims lucky to survive in an iron lung. There was a worldwide flu pandemic in 1957-58 which killed millions of people.
* It was a smaller crappier home and the kids shared bedrooms. Been there, done that.
* They had one crappy car. Been there, done that.
* The appliances sucked. TV sucked. There was no internet. Everything cost more in terms of hours worked, and they had less of it. Been there, done that.
* Who says the wife wanted to stay at home all day doing house work? My mom was there and did that.
Were you there, did you do that? If you enjoyed that primitive expensive life, more power to you, but butt out of everyone else's life. If you were not there and did not do that, is this all propaganda you've soaked up?
Then there's that propaganda about all the jobs moving to China. All the stats say otherwise. Do you think factory workers actually prefer factory work over office work? I frigging doubt most factory workers would prefer 1950s factory work, or even modern factory work, to an office job.
Your poverty and meth diatribe doesn't match statistics or reality. Even the fabled San Francisco streets full of shit and needles are a very small part of the city.
"In the long run, we are all dead" John Maynard Keynes
In the long run, and left to market forces alone, an asymmetry tends to accumulate. Setting aside international trade we might consider other examples: Farm jobs versus factory jobs, say. A slight job-hunters' preference for factory jobs leads to more employment in factories, more investment in new kinds of factories, more profits in the manufacturing sector of the national economy... In a not-so-very-long run a nation where the vast majority of work and wealth is farm-related can completely flip over to become majority manufacturing. Or manufacturing can give way to services. Consider too, imbalance between cash transactions and credit. Handling, storing, transporting, and just counting cash has costs. Offering credit has risks of loss, but methods of distributing those risks and actuarily accounting for such risks can bring about parity. And soon, in a not-so-long-run, the share of the economy associated with cash trades drops and the share involving (third-party) credit rises. We've seen it.
Now consider the asymmetry between traders, one of whom employs slaves, dumps toxic industrial wastes into the commons, and has imperial aspirations; and another who confers social benefits upon the citizenry via employers; has strict community/environmental protection rules, and politically wants mostly to just be left alone. Which has immediate advantage in a short term trading cycle? And what can we say of both, over the long run?
If you had followed the argument of the essay you were responding to, you would realize that none of the factors you listed affect "advantage" in a short-term trading cycle.
I do not understand this fascination with slavery as somehow conferring short term benefits. You aren't the first.
Slavery is only good for low skill (and not farm!) work where the chances for sabotage are minimal and the training costs are so low that killing slaves for low productivity scares the others more than the replacement cost. So it seems to my non-economist brain.
Even mining or digging ditches require some skill. You can't run factories on slave labor without having so many skilled overseers that you may as well skip the slaves and put the overseers to work instead.
Slavery on cotton plantations seems about as advanced as you can get. Nazi slave labor has enough stories about sabotage that I don't believe it was as efficient as it would have been to just pay them normally. I wonder if anyone has ever compared the training and overseer costs of Nazi slave labor to normal willing labor costs.
I've never seen the sabotage theory of slavery anywhere else except in my head. I think it's correct.
Slavery has been economically successful on a large scale only in cotton, sugar, and tobacco. It's hard to wreck those crops! [Slave owners were nice to house slaves, because house slaves could poison you.]
Without sabotage, one can use force. The physical productivity of cotton slavery seems to have rested on the "gang system" of planting and caring for the crops [not picking], which was enforced by the whip. When tried by free labor, wages had to be so high as to negate the physical productivity advantage of slavery.
ChatGPT says (after some prompting): If confusion about currency effects is a big driver of support for tariffs, as you suggest, then adopting a single global unit of account might change the politics. With no exchange rates to obscure the economics, protectionism gets harder to justify.
Yes, hard money implies deflation. But if that cost is smaller than the ongoing global drag from trade barriers, the tradeoff could be worth it.
With a single world currency — pretty nearly the situation with the 19th c. gold standard — the specie flow mechanism plays the same role as the changing exchange rate in the modern system. If a country has a positive balance of trade after accounting for capital flows that means that gold is flowing in, which pushes the value of gold down, prices up, which decreases exports, increases imports, restoring the balance.
Historically, foreign countries have used ownership of debtor nations obligations to exert control over the debtor nations. So there is a geopolitical consideration.
- Egypt in late 19th century accumulated debtor from modernization and sue canal, which led to Britain occupying Egypt.
- Ottomans also in 19th century defaulted on external debts, losing control to European powers including control over some domestic taxes.
- LatAm countries fell into debt and lost economic control in thr 1970s to the US led IMF.
- Delian League where alliance payments in silver shifted balance of power to Athens, who then used their silver reserves to turn their allies into defacto subjects.
- Payments from first Punic war shifted balance of economic power to Rome, eventually leading in part to Carthages destruction.
- Roman Empire gave out loans similar to financial aid to client kings. Terms of these debts included military support and resource extraction.
- 13th century debts by crusaders to Venice resulted in sacking of Constantinople and Venetian trade monopolies in Eastern Mediterranean.
Lots more examples, but the point is (1) capital flows, depending on the circumstances, can shift the balance of power between states and (2) countries historically have intentionally accumulated capital with the purpose to shift balance of power. Economics smacks into geopolitics.
This only works if the creditor nation actually has the military ability to foreclose on the debt. As Daniel Webster pointed out during the controversy over the 2nd Bank of the United States, European investment in the US meant that, if there was a conflict, we had the canals and railroads they had paid to build, they only had claims.
Nationalizing foreign held property does impose costs on the country doing the seizure - not only the administration of the seizure, but a reduction in future foreign investment.
Generally coercive bargaining isn't on/off, since even a country that loses a conflict can still impose immense costs on the winner. Anything that shifts the relative military balance (historically national silver holdings for instance) shifts the expected outcome of a conflict, thus pushing one side to give up more than the other. Of course in reality countries aren't individuals, but the point still holds.
It's a pretty relevant because to build stateless security systems we need mechanisms for more efficiently resolving coercive conflicts. I've been through several coercive negotiations with hackers who stole funds from protocols, they usually will return at least some of the funds with the % being the likelihood of getting caught by police times the expected punishment. Lax enforcement plus underinvestment in international policing has put relative power into hands of black hat hackers over white hat ones.
> Historically, foreign countries have used ownership of debtor nations obligations to exert control over the debtor nations.
That generally only works if the creditor nation has the military force to invade the debtor nation, and then the debt is merely a legitimizing justification.
Wars tend to be short, so countries need a lot of capital stock (historically this was silver) to pay the armed forces. There's a direct corelation between capital stock and military power. In a lot of the examples I gave, accumulating capital while decreasing the capital stock of the adversary was an intentional and effective strategy.
Seems to me that the length of wars depends on available technology. Anybody heard of the 100 years war or the 30 years war or the seven years war [WW 0]?
WW I was conceived as a quickie, but the available technology damned the participants to four years of carnage. WW II was also conceived as a quickie and while it looked good at first, the technology damned the participants to six years of battering ram. Ukraine is a surprise in the sense that we are back to WW I attrition with WW III technology, drones.
Vietnam lasted 50 years from Ho Chi Minh's perspective, battling the French, Japanese, French, and Americans. Only, what, 8 years for Americans? Same as 8 years for Russia in Afghanistan, but 20 years for Americans in Afghanistant.
The EU tried that, and found it didn't work because each country's central bank was manipulating its own economy to its own (short term) advantage, and the single currency spread the pain, but not the benefits, to the other countries.
Or differently: exchange rates make all currencies equal. Metric and imperial measurements use different numbers to measure the same quantities. The French Revolutionary clock and calendar measured the same time with different numbers.
(One of my pet peeves is people who whine that metric makes calculations easier. My answer is they are just numbers, and people already have to deal with 60, 60, 24, 7, 28/29/30/31, 365, 366, 100, and 400, pi, e, and many many others.)
Or shorter: who would control this single world currency?
Does it need "controlling"? Could the world simply use gold or bitcoins?
I mean as a unit of account - we needn't actually move around physical gold, just the bits that say who it belongs to. And there are plenty of protocols in the works that allow blockchains to handle high transaction rates and small payments, but even if that doesn't work out, we could use existing banking systems to, again, just keep track of who owns the crypto. Just as we do now with dollars and yen.
Sure it would be deflationary, but would the damage from that be worse than the damage from tariffs and trade wars?
[BTW, if you want to see an example of how incredibly stupid AIs are on some subjects - I asked Google's Gemini about this - got a bunch of crazy excuses like "traders wouldn't be able to hedge exchange rate risk" and "market prices of gold/crypto/etc are unstable" - not realizing that there wouldn't *be* any exchange rate risk or "market price" since the currency is the unit of account! I'm amazed that an AI this dumb can write working code - yet it can! I guess it shows that coding doesn't really require some kinds of intelligence.]
"I mean as a unit of account - we needn't actually move around physical gold, just the bits that say who it belongs to"
Isn't that what was tried after FDR devalued the dollar, and foreign countries (France, I think) got suspicious and found the US had been lying about how much gold it held as the world's reserve currency? And that's how we got the US fiat currency as the world's reserve currency?
I am only dumping that from memory and do not swear it is accurate. And yes, audits could presumably be a check against that. But once governments get involved, it all goes to pieces.
With crypto (most of them) it's trivial to see how much each account holds. And account owners can cryptographically prove they own the account. No need for governments to get involved, and no need for settlement via physical movements of commodities.
Of course you could also have banks in trusted 3rd countries (Switzerland? Singapore?) hold gold and never settle by moving the gold at all - just keep track of who owns it. But crypto is probably easier.
BRAVO! One of the best essays on tariffs I have read. I have read every essay on tariffs from top economists since about January 2025. All the essays deliver the same message, but DF's essay delivers it in language anyone can understand --- if they want to. Some people just don't want to.
I don't have any academic training in economics, so if your response to this question is go read a book, that'll be fair. But maybe it's an easy thing to answer...
"while it increases jobs in import competing industries it reduces jobs in export industries"
If your export industries are all high-paying jobs, and you primarily import the products of low-paying jobs, and your constituency is seeking low-paying jobs, even if overall economically it all comes out in the wash of the exchange rate (or even overall leaves you economically poorer) would not tariffs still result in a promise fulfilled to your constituency, re-shoring the kinds of jobs they're looking for?
I ask because part of my moral calculus in evaluating these tariffs is that it might be worth a somewhat weaker overall economy if it increases job opportunities for folks lower on the totem pole. A lower tide prevents folks without boats from drowning, or something like that.
What you describe is possible. In that case the tariff is a costly mechanism for income transfer — costly because it does make the country on net poorer, something I didn't discuss in that post. The simplest way of seeing that is to view trade as a technology for producing things. Instead of building a car in Detroit you grow wheat in Iowa, put it on a ship and send it into the Pacific. It comes back with Hondas on it.
Trading wheat for Hondas is profitable only if it costs less US resources to produce cars that way than by building them. An auto tariff taxes the more efficient technology in order to shift production to the less efficient technology, so produces things as a higher cost in US resources.
In one of my earlier posts, on Vance not on tariffs, I speculate that your argument might be his position. But I have seen no evidence that either Vance or Trump has actually tried to design tariffs that way or tried to figure out how the labor force in export industries compares to that in import competing industries.
Possible? With constant returns to scale technology everywhere, it must be that in rich countries wages for sector specific skills rise more than prices in the short run, and that the same is true for unskilled labor unless it has weird tastes. In the long run, wages for the unskilled rise more than prices, full stop.
This is consistent with Trump's policies. This doesn't mean Trump is a sharp trade theorist, but it does reflect that since the 1920's this was the intuition of the US labor movement. At the time, most economists thought the intuition was wrong, but the correct answer got formalized starting around 1940.
There are some wrinkles with increasing returns to scale. One is that global free trade promotes competition to get a technological monopoly over the whole world, so that growth is promoted by free trade. But that's a long run phenomenon, after the next election for sure.
I don't see the argument for your "must be;" so far it is pure assertion. Wages rise in import competing industries, fall in export industries — more jobs for auto workers, fewer jobs for farm workers and people drilling for natural gas. You are now producing things in a less efficient way, so the total amount available to be distributed is less.
If the tariff reduces the trade deficit, possible but not guaranteed, that reduces the amount of foreign capital coming in, raising the return for domestic capital, a transfer from labor to capital.
Alas, this is the Heckscher-Ohlin Model. Assumes constant returns to scale in everything.
From free trade, the capital rich country imposes a uniform tariff. The relative price of the imported commodity, which is labor intensive, rises. This attracts capital and labor from the export industries, but in a different proportion than used in the import competing industry, more capital relative to labor. Hence, labor productivity rises in the import competing industry. Hence, the real wage for unskilled labor rises. Long run, of course.
In the short run, sector specific skills [capital] rise in the import competing industries. Unskilled labor is also attracted to increase output, and if the unskilled don't consume too much of the good whose price has increased, they are better off, too.
Apologies about assertions. It's difficult to know how much or how deeply to explain such processes. My experience -- not so much here, but everywhere else -- has been that nobody cares and they stick to their prejudices. So, first an assertion; then we see what happens.
Now, no one believes the constant returns to scale everywhere assumption anymore. So we get to computational methods, market by market, whose results are not necessarily general and I do not have in my head. My guess is that something of the constant returns results will be saved.
Wouldn’t this be around manufacturing, specifically?
I can buy the idea that if countries are atomic units, trade makes us all better off. But it seems to me the trade with China has been
- fantastic for American capital
- good for Chinese capital
- good for Chinese labor
- abysmal for American labor
I can happily believe that _on net_ the surplus is positive. But it seems the epitome of academic aloofness to ignore the declining quality of life for working class americans because look, we made our global adversary much richer, in exachange for enriching the Americans most likely to support restrictions on speech and contempt for the people who lost the most from the trade.
If you see tariffs as a form of income redistribution, you’re also ignoring the possibility of dignity from paid labor that isn’t present with a handout.
"If you see tariffs as a form of income redistribution, you’re also ignoring the possibility of dignity from paid labor that isn’t present with a handout."
It is a handout — taxing production by export industries via trade in order to subsidize production by import-competing industries.
You mean the dignity from hiding the fact that it is a handout?
Economics ignores things like conflict and war. Your theory works just as well whether we are trading with a country that shares our values or one that wants to dominate us. Once you assume certain boundary conditions, sure, trade equilibrates. Those boundary conditions don’t arise naturally and require investment. So to call tarriffs a hand out is to ignore the massive subsidies that exporters get, in the form of the U.S. Navy protecting their cost of shipping. If you don’t account for that cost and assume safe shipping lanes just happen, ok, then tariffs are a handout. But when you account for the fact that military power requires we manufacture our own weapons and weapon systems and logistics, you see that tariffs are charging exporters the appropriate cost of the services that taxpayers provide them.
That's a stretch. You may as well claim they get subsidized free advertising every time I tell someone how much fun I had on my ship in the Navy, and claim I am an externality every time I tell someone how much I hated the military side of it.
What makes it a stretch? Why should we ignore the cost of providing safe trade networks, given that exporters rely on these to run their business? How is that protection not a subsidy that benefits some industries moreso than others?
I do not buy this has been bad for American labor. Someone (Cafe Hayek? A quick search didn't find it) had a link in the last few days showing that manufacturing unemployment is less than general unemployment, and that manufacturers report more trouble finding workers.
Offshoring low skill low productivity jobs because high skill high productivity jobs crowd them out is good.
Possibly, but that is really hard to target. For example, the USA exports a lot of food, which supports a lot of low skill workers.
I hope you are not including farmers in your low skill category.
Farmers is a very broad category, like factory workers. Lots of jobs in food production are low skilled (“stoop labor”) and others are quite high skilled. You can’t think broadly about industries and the type of jobs they provide as they all tend to cover the gamut.
Then why did you specify "food" as supporting a lot of low skill workers? Do other generic industries not also support a lot of low skill workers?
Because the USA exports a lot of food, which is an example of an industry that supports a lot of low skill workers.
Did your reading comprehension break down as soon as you thought someone was talking about farmers? Usually you are much better than this.
You can't have it both ways.
Perhaps my reading comprehension can't cope with your writing comprehension.
When I lived in China, my non-CCP coworkers spoke of the "Iron Rice Bowl," hugely inefficient factories in the interior that produced nothing of note except air pollution and toxic waste to keep the unproductive middle of the country busy.
The CCP members called it something like "Our Outstanding Social Security System." Why you never read of this in the Western press is left as an exercise.
On the one hand, one would be hard-pressed to convince the UAW workers downstairs—interestingly, the executive offices are on the same floor—that on-shoring auto manufacturing at the expense of abstract improvements in efficiency is a Bad Thing.
The question that keeps recurring—the late Jerry Pournelle would occasionally blog on it—once we outsource every job that can be done by someone on the bottom half of the IQ distribution, what do you now do with half, by definition, of your population.
I don't have an answer, but I observe that I've never seen a politician from any country, even despotic dictators, who admit to any other model of international trade than "Trade Deficits: Bad" zero-sum mercantilism.
You don't outsource all such jobs unless you have a minimum wage law that makes low productivity workers unemployable. Besides, there are jobs, such as babysitting, security guard, housecleaning and yard work that can't be outsourced.
I usually tell people that dollars in and out have to equal, otherwise someone is using them to light cigars or as wallpaper. I usually get corrected by economists as to what "in" and "out" actually mean, but I can't think of a simpler and easier way to start a conversation and find out if they really want a conversation or are just ranting.
Isn’t foreign investment or capital flowing in more important than a trade surplus?
“ I have just shown that trade always balances”
You’ve proposed a bunch of theories that claim trade always balances. Sure, if I accept the premises of your theory, I’ll agree. But why should I think your theories are true?
Is there evidence you can show which demonstrates that your claims accurately map back to empirical measurable reality?
The idea that some countries have cheaper labor than others is patently obvious to anyone with eyes to see. I’m sure you can come up with theoretical abstractions where this difference somehow doesn’t matter. What you’re doing here - not even trying to do - is explain why these tautological theories explain something that basic naked eye intuition is failing to see.
American workers clearly enjoy higher standards of living than their Chinese counterparts. I’m open to the possibility that somehow this doesn’t matter in terms of where labor goes. But you’ll need evidence , not just some clean and tidy set of tautologies where things balance out. Show me where my intuition departs from reality using actual evidence, and I’ll agree.
In summary this is like asking me to consider a spherical cow:
“ consider the simple case of two countries and no capital flows. “
Yeah but capital clearly does flow, that’s the whole point of it. For decades it’s flown to China, instead of being invested here. Why should I think this will somehow balance?
In explaining things it is often useful to start with a simple case, work through the logic of that, then add back what you eliminated to make it simple. Did you stop reading after I explained the logic without capital flows and went on to explain it with?
If you had bothered to try to understand what I was saying, you might have noticed that the trade deficit means Chinese capital flowing to the US, not the other way around.
Yes, I get that you’re starting with a toy system and building it up. I apologize for not investing in better understanding your claims. When you use examples like crystalline spheres and Miasmas, what I think is, you are imagining that economics has more sound empirical footing than it really does. I understand these ideas but suspect - and cannot prove - that “everything equilibrates over time” isn’t true, and what’s happening is you’re depending on some set of assumptions i can’t ferret out.
The closest I can get is this: yes, I understand this idea that trade balances are tautologically equivalent to capital flows. I understand this. But I don’t think capital flows are equal - Chinese “investment” in the U.S. seems to have the effect of subsidizing American government spending, which I think is a net waste. American investment in China, on the other hand, lead to the creation of a bunch of factories that produce real goods.
If I assume that all dollars are equal and prices measure value, I have to buy your conclusion. But I don’t think this is true.
I think not all investments actually produce long term value, and that taxing exporters to subsidize local manufacturing makes sense for the same reason that every cell in my body has walls around it. I’m sure you could make a fine economic argument that my body would operate far more efficiently if you got rid of all those barriers to nutrient transfer - but there’s a reason our physiologies make great effort to keep the outside world out, and I think the same applies to relationships between nations.
"what I think is, you are imagining that economics has more sound empirical footing than it really does."
I don't think any part of my argument depended on empirical claims. It was straight logic. You pretty clearly did not follow it, as demonstrated by getting the direction of the capital flow backwards.
There are doubtless things to criticize about economics, but replacing economic arguments with vague analogies to biology is not an improvement.
You are making claims about value, using a model that assumes a number of things, like “spending and investment are equally good, so long as the dollar numbers equal out.” Sure, I labeled an edge in your causal diagram backwards. I understand that you are claiming a toy system equilibrates, and i think your toy system ignores important parts of reality. You respond by claiming your toy system is just as solid as, say, radio astronomy- this is unearned credibility.
I can happily agree that tariffs on a toy system like you are sketching out don’t make it better. I don’t think we’ll agree on how well that toy system maps to reality. I think this is what you’re missing - people are rejecting your theory, and when you say you know the truth, they remember “subprime is contained,” and probably rely on some intuition like, “if economists knew all that much they’d all be rich from trading stocks, and we wouldn’t have giant crashes that somehow always seem to rebound in a way that helps the rich more.”
I do think tariffs are the least bad form of taxation, because the tax enforcement mechanism aligns incentives with the most essential function of government- securing the border. What do you think of tarriffs vs income taxes in terms of the incentives and externalities they create?
As it happens, economics is the one science that contains an answer to "if you are so smart, why are you not rich?"
I am not asking you to believe my conclusion because it is what economics says. I am showing you the relevant logic — and you are determinedly unwilling to make the effort to understand it.
"But I don’t think capital flows are equal - Chinese “investment” in the U.S. seems to have the effect of subsidizing American government spending, which I think is a net waste."
Then your complaint should be with the US government spending too much and borrowing, not those who loan them the money.
I have plenty of complaints there, for sure. But help me understand- wouldn’t my body operate far more efficiently without spending all that energy pumping ions across cell membranes? What makes nations different from biological structures, where lots of energy is spent creating barriers to the free flow of energy, even inside the physiology of the same organism?
Then let's just fall back on good old liberty and cut the cell membrane analogies. My trade, my *voluntary* trade, is none of your business. Butt out.
Because competition. Because success breeds copycats. Because monopolies breed arrogance and sloppiness and caution.
And how can the US attract a trillion a year capital investment if capital has been flowing to China? You really need to show some actual numbers if you want people to believe your story.
In the 1950’s, a man could support a stay at home wife and three kids while driving a garbage truck for the city of Santa Clara. That’s not doable any more. No amount of numbers are going to convince voters who’ve seen their once healthy towns reduced to poverty and meth - with all the jobs having left for China - that the leaders know what they are doing.
And my final reply is to suggest you read this:
https://cafehayek.com/2025/05/enough-with-the-nostalgia-for-the-manufacturing-economy-of-fifty-years-ago.html
from Don Boudreaux's Cafe Hayek, about growing up a little later than me, and his father's factory work. He has many more on the subject.
This bullshit nostalgia can't fade away fast enough. I don't know anyone who was alive then who thinks it was better. The only excuse I can imagine for being so deluded is that they were a child and had no responsibilities and only remember the fun of being a child with mommy and daddy taking care of all the dreary adult stuff.
This ignorant and bullshit historical favoritism really pisses me off.
* Medical care was miserable. The President's son died of an infected blister in 1924, just thirty years before, and medical car for ordinary people wasn't that far removed.
* Airplane travel was expensive and slow.
* Railroad was how most people traveled long distances, and it was expensive and slow.
* Crossing oceans was still a matter of ocean liners, expensive and slow.
* Telephones were expensive and limited to one per house unless you paid exorbitantly for extensions. Long distance calls were slow and really expensive. International calls were for the rich only, and slower than molasses.
* I can microwave food in my freezer and eat Chinese or Korean food in five minutes. I can drive to the supermarket and get other foreign food and eat it within an hour. Only a few city dwellers could get that even in the 1950s, and that required going to a restaurant and paying more.
* I have more varieties of food and drink in my pantry than John D. Rockefeller could have had for all his wealth. He died in 1937, just 13 years before your fake 1950s fantasy.
* The only thing John D Rockefeller had going for him was enough money to hire servants and boss people around.
* I would not swap my life for the 1950s no matter how much money was added in to the bargain.
ETA: I just asked Gemini to compare ship and airliner passenger traffic across the Atlantic. Airliner traffic didn't exceed ship traffic until 1958.
ETA: Polio was still a thing, with victims lucky to survive in an iron lung. There was a worldwide flu pandemic in 1957-58 which killed millions of people.
"* Crossing oceans was still a matter of ocean liners, expensive and slow."
But, man, was it fun!
And the ice in your drinks was free!
* It was a smaller crappier home and the kids shared bedrooms. Been there, done that.
* They had one crappy car. Been there, done that.
* The appliances sucked. TV sucked. There was no internet. Everything cost more in terms of hours worked, and they had less of it. Been there, done that.
* Who says the wife wanted to stay at home all day doing house work? My mom was there and did that.
Were you there, did you do that? If you enjoyed that primitive expensive life, more power to you, but butt out of everyone else's life. If you were not there and did not do that, is this all propaganda you've soaked up?
Then there's that propaganda about all the jobs moving to China. All the stats say otherwise. Do you think factory workers actually prefer factory work over office work? I frigging doubt most factory workers would prefer 1950s factory work, or even modern factory work, to an office job.
Your poverty and meth diatribe doesn't match statistics or reality. Even the fabled San Francisco streets full of shit and needles are a very small part of the city.
Your are full of nonsense.
I agree with many of the commenters - this was a perfect encapsulation of the misconceptions about tariffs and how best to actually think about them.
It was an honour to meet you on Thursday, Mr. Friedman. Thank you for your talk in London.
"In the long run, we are all dead" John Maynard Keynes
In the long run, and left to market forces alone, an asymmetry tends to accumulate. Setting aside international trade we might consider other examples: Farm jobs versus factory jobs, say. A slight job-hunters' preference for factory jobs leads to more employment in factories, more investment in new kinds of factories, more profits in the manufacturing sector of the national economy... In a not-so-very-long run a nation where the vast majority of work and wealth is farm-related can completely flip over to become majority manufacturing. Or manufacturing can give way to services. Consider too, imbalance between cash transactions and credit. Handling, storing, transporting, and just counting cash has costs. Offering credit has risks of loss, but methods of distributing those risks and actuarily accounting for such risks can bring about parity. And soon, in a not-so-long-run, the share of the economy associated with cash trades drops and the share involving (third-party) credit rises. We've seen it.
Now consider the asymmetry between traders, one of whom employs slaves, dumps toxic industrial wastes into the commons, and has imperial aspirations; and another who confers social benefits upon the citizenry via employers; has strict community/environmental protection rules, and politically wants mostly to just be left alone. Which has immediate advantage in a short term trading cycle? And what can we say of both, over the long run?
"In the long run, we are all dead"
If you had followed the argument of the essay you were responding to, you would realize that none of the factors you listed affect "advantage" in a short-term trading cycle.
"In the long run, we are all dead" John Maynard Keynes.
And now Keynes is dead and we're in the long run.
I do not understand this fascination with slavery as somehow conferring short term benefits. You aren't the first.
Slavery is only good for low skill (and not farm!) work where the chances for sabotage are minimal and the training costs are so low that killing slaves for low productivity scares the others more than the replacement cost. So it seems to my non-economist brain.
Even mining or digging ditches require some skill. You can't run factories on slave labor without having so many skilled overseers that you may as well skip the slaves and put the overseers to work instead.
Slavery on cotton plantations seems about as advanced as you can get. Nazi slave labor has enough stories about sabotage that I don't believe it was as efficient as it would have been to just pay them normally. I wonder if anyone has ever compared the training and overseer costs of Nazi slave labor to normal willing labor costs.
I've never seen the sabotage theory of slavery anywhere else except in my head. I think it's correct.
Slavery has been economically successful on a large scale only in cotton, sugar, and tobacco. It's hard to wreck those crops! [Slave owners were nice to house slaves, because house slaves could poison you.]
Without sabotage, one can use force. The physical productivity of cotton slavery seems to have rested on the "gang system" of planting and caring for the crops [not picking], which was enforced by the whip. When tried by free labor, wages had to be so high as to negate the physical productivity advantage of slavery.
Penal slavery also worked for rowing galleys.
Probably worked really well if they were chained to their benches and defeat meant drowning.
ChatGPT says (after some prompting): If confusion about currency effects is a big driver of support for tariffs, as you suggest, then adopting a single global unit of account might change the politics. With no exchange rates to obscure the economics, protectionism gets harder to justify.
Yes, hard money implies deflation. But if that cost is smaller than the ongoing global drag from trade barriers, the tradeoff could be worth it.
With a single world currency — pretty nearly the situation with the 19th c. gold standard — the specie flow mechanism plays the same role as the changing exchange rate in the modern system. If a country has a positive balance of trade after accounting for capital flows that means that gold is flowing in, which pushes the value of gold down, prices up, which decreases exports, increases imports, restoring the balance.
Historically, foreign countries have used ownership of debtor nations obligations to exert control over the debtor nations. So there is a geopolitical consideration.
- Egypt in late 19th century accumulated debtor from modernization and sue canal, which led to Britain occupying Egypt.
- Ottomans also in 19th century defaulted on external debts, losing control to European powers including control over some domestic taxes.
- LatAm countries fell into debt and lost economic control in thr 1970s to the US led IMF.
- Delian League where alliance payments in silver shifted balance of power to Athens, who then used their silver reserves to turn their allies into defacto subjects.
- Payments from first Punic war shifted balance of economic power to Rome, eventually leading in part to Carthages destruction.
- Roman Empire gave out loans similar to financial aid to client kings. Terms of these debts included military support and resource extraction.
- 13th century debts by crusaders to Venice resulted in sacking of Constantinople and Venetian trade monopolies in Eastern Mediterranean.
Lots more examples, but the point is (1) capital flows, depending on the circumstances, can shift the balance of power between states and (2) countries historically have intentionally accumulated capital with the purpose to shift balance of power. Economics smacks into geopolitics.
This only works if the creditor nation actually has the military ability to foreclose on the debt. As Daniel Webster pointed out during the controversy over the 2nd Bank of the United States, European investment in the US meant that, if there was a conflict, we had the canals and railroads they had paid to build, they only had claims.
I'll read Webster on this, thank you.
Nationalizing foreign held property does impose costs on the country doing the seizure - not only the administration of the seizure, but a reduction in future foreign investment.
Generally coercive bargaining isn't on/off, since even a country that loses a conflict can still impose immense costs on the winner. Anything that shifts the relative military balance (historically national silver holdings for instance) shifts the expected outcome of a conflict, thus pushing one side to give up more than the other. Of course in reality countries aren't individuals, but the point still holds.
It's a pretty relevant because to build stateless security systems we need mechanisms for more efficiently resolving coercive conflicts. I've been through several coercive negotiations with hackers who stole funds from protocols, they usually will return at least some of the funds with the % being the likelihood of getting caught by police times the expected punishment. Lax enforcement plus underinvestment in international policing has put relative power into hands of black hat hackers over white hat ones.
> Historically, foreign countries have used ownership of debtor nations obligations to exert control over the debtor nations.
That generally only works if the creditor nation has the military force to invade the debtor nation, and then the debt is merely a legitimizing justification.
Wars tend to be short, so countries need a lot of capital stock (historically this was silver) to pay the armed forces. There's a direct corelation between capital stock and military power. In a lot of the examples I gave, accumulating capital while decreasing the capital stock of the adversary was an intentional and effective strategy.
What do you consider "short"? The Gaza war is a year and a half old by now. Ukraine and Russia have been at it over 3 years.
Seems to me that the length of wars depends on available technology. Anybody heard of the 100 years war or the 30 years war or the seven years war [WW 0]?
WW I was conceived as a quickie, but the available technology damned the participants to four years of carnage. WW II was also conceived as a quickie and while it looked good at first, the technology damned the participants to six years of battering ram. Ukraine is a surprise in the sense that we are back to WW I attrition with WW III technology, drones.
Careful what you wish for!
Vietnam lasted 50 years from Ho Chi Minh's perspective, battling the French, Japanese, French, and Americans. Only, what, 8 years for Americans? Same as 8 years for Russia in Afghanistan, but 20 years for Americans in Afghanistant.
You're a good man Charlie Brown! I also forgot about the Napoleonic Wars, 1803 - 1815, admittedly a mere 12 years.
One can make a case for short wars in the long 19th century, 1815 -1914. But now we do it continuously if not discretely.
These are unusual. Most conflicts are more similar to the India Pakistan situation.
Most long conflicts are lots of short conflicts all strung together. Very few consist of, say, a Leningrad siege and nothing else.
It depends how you measure it, but the average is about 4-5 months. This article explains the average and why Ukraine is unusual. https://www.foreignaffairs.com/ukraine/hard-truth-about-long-wars
So why doesn't everyone just use the same currency and make this problem go away?
We used to. It was called "gold".
The EU tried that, and found it didn't work because each country's central bank was manipulating its own economy to its own (short term) advantage, and the single currency spread the pain, but not the benefits, to the other countries.
Or differently: exchange rates make all currencies equal. Metric and imperial measurements use different numbers to measure the same quantities. The French Revolutionary clock and calendar measured the same time with different numbers.
(One of my pet peeves is people who whine that metric makes calculations easier. My answer is they are just numbers, and people already have to deal with 60, 60, 24, 7, 28/29/30/31, 365, 366, 100, and 400, pi, e, and many many others.)
Or shorter: who would control this single world currency?
Does it need "controlling"? Could the world simply use gold or bitcoins?
I mean as a unit of account - we needn't actually move around physical gold, just the bits that say who it belongs to. And there are plenty of protocols in the works that allow blockchains to handle high transaction rates and small payments, but even if that doesn't work out, we could use existing banking systems to, again, just keep track of who owns the crypto. Just as we do now with dollars and yen.
Sure it would be deflationary, but would the damage from that be worse than the damage from tariffs and trade wars?
[BTW, if you want to see an example of how incredibly stupid AIs are on some subjects - I asked Google's Gemini about this - got a bunch of crazy excuses like "traders wouldn't be able to hedge exchange rate risk" and "market prices of gold/crypto/etc are unstable" - not realizing that there wouldn't *be* any exchange rate risk or "market price" since the currency is the unit of account! I'm amazed that an AI this dumb can write working code - yet it can! I guess it shows that coding doesn't really require some kinds of intelligence.]
"I mean as a unit of account - we needn't actually move around physical gold, just the bits that say who it belongs to"
Isn't that what was tried after FDR devalued the dollar, and foreign countries (France, I think) got suspicious and found the US had been lying about how much gold it held as the world's reserve currency? And that's how we got the US fiat currency as the world's reserve currency?
I am only dumping that from memory and do not swear it is accurate. And yes, audits could presumably be a check against that. But once governments get involved, it all goes to pieces.
With crypto (most of them) it's trivial to see how much each account holds. And account owners can cryptographically prove they own the account. No need for governments to get involved, and no need for settlement via physical movements of commodities.
Of course you could also have banks in trusted 3rd countries (Switzerland? Singapore?) hold gold and never settle by moving the gold at all - just keep track of who owns it. But crypto is probably easier.
There's still a need for audits to physically count the gold, even sample and assay it.