Within the Trump administration there is a vigorous debate between two camps. One group, headed by Peter Navarro, might be called the “true believers”. They favor mercantilist economic policies of the sort that Argentina implemented during the 1940s and 1950s. Another group, headed by Elon Musk, might be called the free traders. In the middle are people like Scott Bessent and Howard Lutnick.
Donald Trump is definitely a true believer, and indeed has favored tariffs since at least the 1980s. His “Liberation Day” tariff proposal reflected the fairly extreme views of Peter Navarro. When the negative market reaction created fears of an economic crisis, Bessent and Lutnick went to Trump and encouraged him to delay the “reciprocal tariffs” (which have nothing to do with reciprocity) by 90 days.
Meanwhile, the debate continues to rage within the Trump administration:
Billionaire presidential adviser Elon Musk attacked White House trade counselor Peter Navarro as “dumber than a sack of bricks” as a fight over President Donald Trump’s tariff regime spilled onto social media on Tuesday.
Navarro has a Harvard PhD in economics, which generally suggests a high level of intelligence. In 1984, Navarro wrote a book mocking the views of protectionists. But then something happened. By 2016, he was making some extremely odd claims. This is from a post I wrote in December 2016:
Navarro and Ross are making the EC101 mistake of drawing causal implications from the famous GDP identity:
GDP = C + I + G + (X – M)
Students often assume that trade deficits subtract from GDP, because there is a minus sign attached to imports. What they forget is that the goods imported then show up as a positive in either the consumption
ofor investment category. Navarro also seems to have forgotten this fact.
Elsewhere, Navarro made a number of other elementary errors in economic reasoning. He even made up imaginary experts to buttress his case. In 2024, he served 4 months in prison. Today he is the architect of perhaps the most disruptive policy initiative of the 21st century.
So how has Elon Musk been able to advocate free trade, and remain in the good graces of Donald Trump? Musk is quite clever. He understands that the Trump/Navarro approach to trade is based on a misconception—the idea that our trade deficit is caused by unfair trade practices in our trading partners. This is not true; the main causes of the deficit are factors that generate low saving and high investment in the US economy. But that myth provided an opening for Musk. If it really were true that unfair trade practices were the problem, then a reasonable solution would be negotiations where both sides reduced their trade barriers:
“At the end of the day, I hope it’s agreed that both Europe and the United States should move ideally, in my view, to a zero tariff situation, effectively creating a free trade zone between Europe and North America,” the tech billionaire [Musk] told Matteo Salvini, the leader of Italy’s right-wing League Party.
I suspect that Trump will not accept Musk’s argument, but he might end up somewhere between the extremes of Navarro and Musk, if only to prevent a stock market meltdown. Time will tell.
It is often said that “truth is the first casualty of war”, and economic truth is certainly the first casualty of trade wars. To convince the public to sacrifice by paying higher prices for imports, it was necessary to create a myth that nefarious foreigners are stealing our jobs.
In my December 2016 post, I added this postscript:
PS. It’s not at all clear that Navarro’s ideas will actually be implemented. Some people believe that Trump is more likely to govern as a traditional Republican. The next four years will provide a test of the “Great Man” theory of history. I’m in the camp that believes presidents are far less consequential than most people assume.
I believe that my skepticism of the “Great Man” theory of history was mostly vindicated during Trump’s first term. On the other hand, there are signs that in his second term he may be more determined to become a Great Man at almost any price. So perhaps in the long run I’ll end up being wrong. 
(Let’s hope it’s Austerlitz, not Waterloo.)
READER COMMENTS
Billy Kaubashine
Apr 21 2025 at 9:44pm
Navarro is like a novice boxer. He is very confident in his strategy, but fails to think about what punches his opponent might throw>
David Seltzer
Apr 22 2025 at 10:27am
Billy: Yeah! The great street philosopher, Mike Tyson, said, “Everybody has a plan until they get punched in the mouth!”
nobody.really
Apr 22 2025 at 2:02am
Typo? Perhaps of should be or?
Feel free to delete this message.
Scott Sumner
Apr 22 2025 at 12:43pm
Thanks, I fixed it.
Warren Platts
Apr 22 2025 at 12:44pm
Scott, this is a little unfair. As a true believer myself, I can say they are not pursuing mercantilist policies; they are pursuing protectionist policies. Big difference. Under mercantilism, the goal is to run big trade surpluses; under protectionism, trade happens, but it’s not emphasized. Under mercantilism, wages are suppressed in order to force up savings; under protectionism, preserving high wages is the main idea. Mercantilism boosts economic growth through beggar-thy-neighbor policies that import foreign demand; protectionism boosts growth by boosting domestic demand by making sure people got money in their pockets.
As for GDP = C + I + G + (X – M), I understand that M doesn’t affect GDP in the superficial sense that GDP is supposed to measure domestic production, and so M is merely an accounting device to back out the foreign production in C+I+G. That is, the full equation is:
GDP = Cd + Cf + Id + If + Gd + Gf + X – M
where M = Cf + If + Gf
But consider: if we’re running a big trade deficit in one year, then M >> X. But what would happen if the next year we magically made M = X and we magically assume X remains unchanged. What happens? If M really has no effect on GDP, then GDP must also remain the same. And since GDP only measures domestic production, that entails that Cd+Id+Gd must also remain unchanged. Therefore, C+I+G must go down. Thus GDP remains the same, but we get to consume less. That’s no fun!
But why must that necessarily the case? C is not a mistake because it contains Cf that must be corrected by M. That is, if Cd is a measure of domestic supply, then C [Cd+Cf] is a measure of domestic demand. And why should we expect demand to change? People won’t cut back their spending for no good reason. Therefore, it’s more reasonable to expect that it is C+I+G that will remain the same.
In that case, since Cf+If+Gf must go down if M goes down, and if C+G+I remains the same, then Cd+Id+Gd must go up by an equivalent amount. But that’s precisely what GDP measures, thus GDP must also go up. And if we let X go down because of retaliation, all that entails — assuming trade remains balanced — is that M will decline by the same amount as X, further entailing that Cf+If+Gf will go down by the same amount, thus entailing that Cd+Id+Gd must go up by the corresponding amount because C+I+G remains unchanged. Therefore, a decline in X would not affect GDP at all.
Travis Allison
Apr 22 2025 at 4:22pm
@Warren: See my reply below. I hope it helps.
Student
Apr 22 2025 at 6:03pm
Oy vey.
Warren Platts
Apr 22 2025 at 7:49pm
I’m merely trying to explain Navarro’s thinking on the GDP identity so he can have a fair hearing. The economics shibboleth that M has no effect on Y seemingly implies that C is misleading mistake, that must be corrected by M because it’s easier to calculate M based Customs information — as if that’s the only reason M is in the GDP equation. C represents total consumption that is equal to total production. That’s not a mistake. That’s an important bit of information. The consumptions is all domestic: it represents American spending on consumption. But the production is split between domestic producers and foreign producers. So if C stays the same and Cf goes down, the Cd must go up by the same amount.
Student
Apr 22 2025 at 7:53pm
It’s astrology.
Warren Platts
Apr 22 2025 at 11:49pm
Thank you for your well thought out responses.
Jon Murphy
Apr 23 2025 at 9:32am
Student-
I don’t think your being fair to astrology. Astrology at least has the virtue of being consistent and non-contradictory in its reasoning. The Trump Administration and the True Believers change their reasoning on a day-to-day, sometimes minute-to-minute basis, often adopting contradictory stories.
Warren Platts
Apr 23 2025 at 11:49am
Mere gaslighting. If there were contradictions, you’d be able to point them out. Yet you don’t. Meantime, here’s one contradiction on you: you say trade deficits don’t matter, yet you must admit that large negative NIIP must matter at some point whether its -60%, -100%, 1000%, 10,000%. If one of them matters, then both matter. If one of them doesn’t matter, then both must not matter. But it cannot be the case that one matters and the other doesn’t matter. That’s your contradiction.
Jon Murphy
Apr 23 2025 at 4:19pm
I have on multiple occasions (most recently on the 6th). Your response was to acknowledge the contradictions and say “But you know what? NOBODY CARES”
I don’t get the rest of your comment. I have explicitly denied that an NIIP being negative matters. You’re the only one making that claim (which actually does to another contradiction, this one tied directly to you. Just the other day you said that it is not per se bad. Now you are implying it is).
nobody.really
Apr 24 2025 at 11:14am
“The only function of economic forecasting is to make astrology look respectable.” Ezra Solomon, Burmese-born American economist, The Bulletin (1984), Reader’s Digest (1985) (NOT J. K. Galbraith).
In fairness, I picked that campaign slogan less for accuracy than ‘cuz it fit well on buttons and bumperstickers. Today I guess I’d pick a slogan based on whatever a consultant thought would look good on a chyron or Instagram post.
Scott Sumner
Apr 23 2025 at 12:11pm
Warren, You are mixing up identities and causal arguments. It is not true that the GDP identity shows that imports reduce GDP. If you wish to argue that reducing imports raises GDP, feel free to do so. But you cannot start the argument by postulating that M goes down for no good reason, and then criticize people that suggest that C would also go down “for no good reason”.
You need to explain the shock that causes M to decline, and then consider how that shock might affect other components of the GDP equation. For instance, Keynes once argued that a decline in M might boost aggregate demand (I disagree.)
Navarro’s silliness is not the claim that imports reduce GDP (It’s probably wrong, but an argument can be made), rather that the GDP equation demonstrates this causal connection.
Warren Platts
Apr 23 2025 at 1:53pm
Scott, I actually agree with everything you just said. I’m just looking at the identity itself. If we can forget about the real world for a moment and think of the GDP equation as a game. It has two non-negotiable rules:
Y = Cd+Id+Gd + X
M = Cf+If+Gf
Now the question is whether we should impose a third rule:
3a. Any change in M must have no effect on Y; or
3b. Any change in M must have no effect on C+I+G (where C+I+G = Cd+Id+Gd + Cf+If+Gf
Both 3rd rules seem equally arbitrary to me. But if the starting setup is M >> X, and then the next move is M = X where X stays the same, then the 3rd move, what happens to the other variables, is quite different under 3a compared to 3b. Under 3a, Y stays the same, but C+I+G must go down by ΔM. Under 3b, Y must go up by ΔM.
What this means for the real world, I dunno. I guess it depends on the behavior of consumers: if imports are reduced because of tariffs or other trade restrictions, will they reduce their consumption? If so, then Y will stay the same. But if they maintain the same level of consumption, then Y will go up.
Jon Murphy
Apr 23 2025 at 4:21pm
And this is why merely relying on an accounting identity and not understanding the theory behind it is extraordinarily dangerous.
Any intro econ student can answer that question.
Warren Platts
Apr 24 2025 at 12:14pm
OK Professor. This blog exists to educate people, right? Why don’t you tell us? No, I’m exactly right. In Y = C+I+G+(X-M) when going from M >> X to M = X, it is of course true that Y need not necessarily change: but then something else must change. If Y stays the same, then C+I+G must go down by ΔM. Whether this happens in the real world depends on what households, businesses, and what the government does. They would have to reduce their consumption and/or investment by ΔM. However, it is at least as plausible they will maintain the same level of spending by substituting domestic goods. In that case, Y will indeed go up by ΔM. And of course, those two examples represent extreme ends of a spectrum of possibilities: Y might go up some and C+I+G might go down some. Yes, there would probably be some short-term hysteresis.
Scott Sumner
Apr 24 2025 at 2:23pm
You asked what would happen if M goes down. Suppose M goes down because an asteroid destroys planet Earth. In that case, GDP falls to zero. Or M might go down because we are in a depression, in which case GDP would fall. Or M might go down due to trade barriers, in which case GDP probably goes down. There are many reasons why M might go down, and each has a different impact on GDP.
Staring at that equation won’t help you to answer these questions.
Travis Allison
Apr 22 2025 at 4:21pm
@Warren. I totally get that protectionism seems appealing. Trade theory is subtle. Suppose you cut M. If C and X were somehow to stay the same (unlikely), then Id would decline. You aren’t going to increase GDP by changing the names of the buckets that you assign to different spending. Do you think we’d increase GDP by increasing government spending to the moon?
In the US, we basically have full employment. If you apply tariffs to goods, people in the US will gradually shift to manufacturing goods and away from performing services. GDP would decline, because we would be doing things that we don’t have a comparative advantage doing. We’d consume (overall) fewer manufacturing goods and fewer services.
Before clicking the first link below, ask yourself whether you think Japan had been largely running trade deficits or a trade surpluses for the past *10* years. (Click 10Y on the graph.)
https://tradingeconomics.com/japan/balance-of-trade
https://www.nippon.com/en/japan-data/h02370/
Warren Platts
Apr 22 2025 at 7:38pm
It depends. There are two extremes: (1) you require that Y (GDP) stays the same (because M has no effect on Y); (2) you require that C+I+G stay the same and allow Y to float (because people won’t change their spending habits). So when you go from M >> X to M = X, under (1) C+I+G must go down; or under (2) Y must go up. There could be a combination of (1) and (2).
Nobody’s changing the names of any buckets. lol!
Isn’t that called “fiscal stimulus”? Yes, if you increase G, you increase Y, other things being equal.
Not at all. The service sector, on average, has a lower labor productivity than the manufacturing sector. Not all service workers are doctors and lawyers and such. Some service workers work at McDonalds for a living. So when you increase manufacturing employment, you’re not sucking financiers from Wall Street to work the assembly line; you’re pulling McDonalds burger flippers and Dollar Store clerks and sticking them on the assembly line. This increases overall labor productivity and should increase GDP.
Are you kidding? I checked back to 1970. Over the ensuing 55 years, Japan ran an annual current account deficit exactly once, and that was in 2014. Because of the Fukushima meltdown, they had to import a lot of extra energy that year.
Travis Allison
Apr 23 2025 at 5:20pm
At full employment, increasing G is unlikely to increase Y, other than having people work overtime, etc. Eventually this would result in inflation. The Fed, if it’s doing its job, would fight against inflation. Remember 2013(?), when the government reduced unemployment benefits. Krugman thought the economy would decline. Instead, it grew. The Fed is the last mover when performing correctly. At full employment, increasing G crowds out the private sector. Overally GDP will not increase.
Click on the links I gave you. It shows that Japan has had a *trade deficit* for the past 10 years. It still has a current account surplus.
If you put tariffs on manufacturing goods, the average price of manufacturing goods will go up. Some people who used to produce services transition to producing manufacturing goods. The the new manufacturing employees will produce fewer goods than were consumed at the lower price because demand curves slope downward. Thus manufacturing consumption drops. Those new manufacturing employees are no longer producing services, thus service consumption has to drop. In sum, total consumption of goods and services drops due to tariffs. There will be government revenue from the tariffs, but it isn’t enough to make up for the drop in overall consumption.
Jon Murphy
Apr 23 2025 at 6:01pm
Good stuff Travis. One thing I’d add is that tariffs also reduce exports, so consumption of those goods fall as well
Rajat
Apr 22 2025 at 8:52pm
It seems clear that Trump is being a Great Man right now. The best case scenario is that he ceases being a Great Man some time within the next 19 months or so, either due to a reality-check from markets or others, or due to a monumental mid-term backlash. The worst case scenario is that Trump remains a Great Man for the rest of his life, followed by a procession of more Great Men. My money (and hope) is on the first option, as I suspect is the market consensus.