Navigating AI's Impact on U.S. Economic Growth

My fellow pro-growth/progress/abundance Up Wingers,

With tariff and immigration policies uncertain, and the emerging AI revolution continuing to emerge, there’s plenty to speculate about when it comes to the US economy. Today on Faster, Please! — The Podcast, I invite Joseph Politano to help us try and make sense of it all.

He is the author of the popular Apricitas Economics Substack newsletter. Politano previously worked as an analyst at the Bureau of Labor Statistics.

In This Episode

* Trade and immigration headwinds (1:03)

* Unpredictable trade policy (7:32)

* Tariffs as a political tool (12:10)

* The goal: higher tariffs (17:53)

* An AI tailwind (20:42)

Below is a lightly edited transcript of our conversation.

Trade and immigration headwinds (1:03)

You're going to have what is probably the largest one-year change in immigration in US history.

Pethokoukis: What are the main economic headwinds that you're tracking right now? Or is it just trade, trade, trade?

Politano: It’s hard for me to not say it’s trade, trade, trade because that's what my newsletter has been covering since the start of this administration and I think it's where the biggest change in longstanding policy is. If you look back on, say, the last 100 years of economic history in the United States, that's the kind of level you have to go to find a similar period where tariffs and trade restrictions were this high in the United States.

At the start of this year, we were at a high compared to the early 2000s, but it was not that large compared to the 1970s, 1960s, the early post-war era. Most of that, especially in Trump's first term, was concentrated in China, and then a couple of specific sectors like steel or cars from Mexico. Now we have one, you had the big jump in the baseline — there's ten percent tariffs on almost all goods that come to the United States, with some very important exceptions, but ten percent for most things that go into the US. Then, on top of that, you have very large tariffs on, say, cars are 25 percent, steel and aluminum right now are 50 percent. China was up to 20 percent then went to the crazy 150 percent tariffs we had for about a month, and now it's back down to only 30 percent. That's still the highest trade war in American history. I think that is a big headwind.

The headwind that I don't spend as much time covering, just because it's more consistent policy — even if it is, in my opinion, bad policy — is on the immigration stuff. You're going to have what is probably the largest one-year change in immigration in US history. So we're going to go from about 2.8 million net immigration to a year, to people like Stan Veuger projecting net-zero immigration this year in the United States, which would be not entirely unprecedented — but again, the biggest shift in modern American history. I think those are the two biggest headwinds for the US economy right now.

You’re highlighting two big drivers of the US economy: trade and immigration. But analyzing them is tricky because recent examples are limited. To understand the effects of these changes, you often have to look back 50 or 100 years, when the economic landscape was very different. I would think that would make drawing clear conclusions more difficult and pose a real challenge for you as an analyst.

Again, I'm going to start with trade because that's where I focused a lot of my energy here, but the key thing I’m trying to communicate to people — when people think of the protectionist era in US history, the number one thing people think about is Smoot-Hawley, which were the very large tariffs right before the Great Depression — in my opinion, obviously did not cause the Great Depression, but were part of the bad policy packages that exacerbated the Great Depression. That is an era in which one, the US is not a big net importer to the same degree; and two, trade was just a much smaller share of the economy, even though goods were a much larger share of the economy.

This is pre- the really big post-war globalization and pre- the now technology-era globalization. So if you're doing tariffs in 1930 or prior, you're hitting a more important sector. Manufacturing is a much larger share of the economy, construction is a larger share of the economy, but conversely, you're hitting it less hard. And now you have this change of going from a globalized world in which trade is a much larger share of GDP and hitting that with very large tariffs.

The immigration example is hard to find. I think the gap is America has not done . . . let's call it extensive interior enforcement in a long time. There's obviously been changes to immigration policy. Legally the tariffs have gone up. Legally, lot of immigration policy has not changed. We don't pass bills on immigration in the same way. We don't pass bills on tariffs, but we do pass bills on tax policy. So immigration has changed mostly through the enforcement mechanisms, primarily at the border, and then secondarily, but I think this is the bigger change, is the kind of aggressive interior enforcement.

The Steven Miller quote that was in the Wall Street Journal is what I think about, like, why aren't you going to Home Depot to try to deport people who are here undocumented? That's a really big change in economic policy from the first term where it was like, “Okay, we are going to restrict the flow of legal and undocumented immigrants at the border, and then mostly the people who are in the interior of the United States, we're only going to focus on people who've committed some other crime.” They got picked up by local law enforcement doing something else, and then we're going to deport them because of that.

This is very different, and I think also very different tonally. In the first term, there was a lot of, “People don't want refugees.” Refugee resettlement was cut a lot, but there was a rhetorical push for, “We should let some people in from Venezuela or Cuba, people who were fleeing socialist dictatorships.” That program [was] also very much torn up. So it's hard to find examples, in that case, where you’ve got to go back to 1924 immigration policy, you’ve got to go back to 1930 trade policy for the closest analogs.

Unpredictable trade policy (7:32)

People notice if the specific things that they associate with other countries go up in price, even if those aren't their most important export.

Trade policy seems especially difficult to analyze these days because it's been so mercurial and it’s constantly evolving. It's not like there's one or two clear policy shifts you can study — new announcements and reversals happen daily, or weekly. I think that unpredictability itself creates uncertainty, which many analysts see as a drag on growth, often as much as the tariffs themselves.

I think that's exactly right. I used to joke that there were three people in Washington, DC who know what the current tariff levels are, and I'm not sure any of them are in the White House, because they do change them extremely frequently. I'm going to give an example of the last 24 hours: We had the announced rate on imports from the Philippines from 20 percent to 19 percent, the rate on imports from Indonesia went from 32 to 19, the rate on Japan went from 25 to 15. None of those are legal changes. They've not published, “Here's the comprehensive list of exactly what we're changing, exactly when these are going to go into effect, yada, yada, yada.” It's just stuff that administration officials or Trump, in particular, said. So it's really hard to know with any certainty what's going on.

Even just this morning, the Financial Times had a good article basically saying that the US and the European Union are close to a quote-unquote “deal” where the tariffs on the EU would be at 15 percent. Then literally 30 minutes ago, Peter Navarro is on TV and he's like, “I would take that with a grain of salt.” So I don't know. Clearly some people internally know. This is actually the longest period of time that Trump has gone without legally changing the tariffs since he was inaugurated. 28 days was the previous record.

Normally — I'll give an example of the last Trump administration — what would happen is you'd have, “Hey, we are doing this Section 301 investigation against China. This is a legal procedure that you say that the Chinese government is doing ABC, XYZ unfair trade practices and we're going to retaliate by putting tariffs on these specific goods.” But you would have a very long list of goods at least a couple of months before the tariffs would take effect.

It wasn't quite to this degree, I don't want to make it sound like Trump won, everything was peachy keen, and there was no uncertainty. Trump would occasionally say something and then it would change the next week, but it was much more contained, and now it's like all facets of trade policy.

I think a really good example was when they did the tariffs on China going from 10 to 20 to then 145 percent, and then they had to come back a week later and be like, “We're exempting smartphones and certain types of computers.” And then they came back a week after that and were like, “We're exempting other types of electronics and electronic parts.” It does not take an expert to know that smartphones come from China. It's on the package that Apple sends you. And if you were very strategically planning this out, if you were like, “Well, are going to do 150 percent tariffs on China,” that would be one of the first questions someone would be like, “Well, people are going to notice if their iPhone prices go up. Have we thought about exempting them?”

During Trump's first term — again, you can take this as political or economic strategy — they mostly focused a lot of the tariffs on intermediate goods: computer parts, but not computers; brakes, not cars. That has more complicated economic costs. It, on balance, hurts manufacturing in the United States more and hurts consumers less, but it's clearly trying to set up a political salience. It's trying to solve a political salience problem. People notice if the specific things that they associate with other countries go up in price, even if those aren't their most important export. There's been much less of that this time around.

We're doing tariffs on coffee and bananas. I complain about that all the time, but I think it is useful symbolism because, in an administration that was less concerned about political blowback, you'd be like, “Oh yeah, give me a list of common grocery items to exempt.” This is much less concerned with that blowback and much more slap-dash.

Tariffs as a political tool (12:10)

. . . we're now in the process of sending out these quote-unquote “letters” to other countries threatening higher tariffs. It doesn't seem to me like there's a rhyme or reason why some countries are getting a letter or some countries aren't.

I think there’s a lot of uncertainty in interpreting administration statements, since they can change basically overnight. Even if the policy seems settled, unexpected events — like, oh, I don’t know, a there’s a trial of a politician who Trump likes in another country and all of a sudden there’s a tariff to nudge that country to let that politician go. If the president views tariffs as a universal tool, he may use them for unpredictable, non-economic reasons, making it even harder to analyze, I would think.

I think that's exactly right, and if you remember very early on in the Trump administration, the Columbian government did not want to take deportees on military aircraft. They viewed this as unjust treatment of Columbian nationals, and then Trump was like, “I'm going to do a 20, 30 percent tariff,” whatever the number was, and then that was resolved the next day, and then we stopped doing the military flights two weeks after that. I think that was a clear example . . . Columbia is an important US trading partner, but there's a lot more who are larger economies, unfortunately for Columbia.

The example you're giving about Brazil is one of the funnier ones because . . . on April 2nd, Trump comes out and says, “We're doing reciprocal tariffs.” If you take that idea seriously, we should do tariffs against countries that employ unfair trade practices against US exports. You take that idea seriously, Brazil should be in your top offender categories. They have very high trade barriers, they have very high tariffs, they have domestic industrial policy that's not super successful, but does clearly hurt US exports to the region. They got one of the lowest tariff rates because they didn't actually do it by trade barriers, they did it by a formula, and Brazil happens to export some oil, and coffee, and cashews, and orange juice to the United States more than they buy from us. That was the bad formula they did looking at the bilateral trade deficit.

So you come back, and we're now in the process of sending out these quote-unquote “letters” to other countries threatening higher tariffs. It doesn't seem to me like there's a rhyme or reason why some countries are getting a letter or some countries aren't. We sent one to Libya, which is not an important trading partner, and we sent one to the Philippines, which is. But the letter to Brazil is half, “Okay, now we remembered that we have these unfair trade practices that we're complaining about,” and then it's half, “You have to let Jair Bolsonaro go and stop prosecuting him for the attempt to stay in power when he lost the election.”

It's really hard to say, okay, what is Lula supposed to do? It's one thing to be like, economically, a country like Brazil could lower its tariffs and then the United States would lower its tariff threat. You'd still be worse off than you were at the start of the year. Tariffs would still be higher, trade barriers would still be higher, but they'd at least not be as bad as they could be. But tying it up in this political process makes it much less clear and it's much harder to find an internally consistent push on the political thing. There are out-and-out dictatorships that we have very normal trade relationships with. I think you could say we should just trade with everybody regardless their internal politics, or you could say trade is a tool of specific political grievances that we have, but neither of those principles are being applied consistently.

As a business owner, totally separate from the political considerations, is it safe to import something from Mexico? Is Trump going to get upset at Claudia Sheinbaum over internal political matters? I don't know. He was upset with Justin Trudeau for a long period of time. Trudeau got replaced with Mark Carney, who is not exactly the same political figure, but they're in the same party, they're very similar people, and the complaints from Trump have dropped off a cliff. So it's hard to tell what the actual impulse is. I follow this stuff every day, and I have been wrong so many times, it is hard to count. I'll give an example: I thought Trump, last month, was like, “We're going to do 50 percent tariffs on the European Union.” And in my head I was like, “Oh, this makes sense.”

With every other major trading partner, we go from a baseline level, we raise to a very large level, we keep that on for a very short amount of time, and then we lower back down to a level that is much higher than what we started at, but much lower than what was in practice. We went from average 20 percent-ish tariffs on China, we went from that to average 40 percent-ish tariffs, and then we went into the mid-100s, and now we're back down to average 50 percent-ish tariffs on China if you count stuff from Trump's first term.

So I was like, “Oh, they paused this for 90 days, they're going to come back and they're going to say, ‘Well, everyone except the European Union, everyone except Japan, everyone except Brazil is doing really well in negotiations. We're going to raise tariffs on Brazil to 50 percent for a week and then we're going to lower them back.’” And that was obviously just wrong. They just kicked the can down the road unceremoniously.

The goal: higher tariffs (17:53)

It's not as though Donald Trump has a specific vision of what he wants the tariff rates to look like in five years, at a number level, per country per good. It's that he wants them to be higher.

Do you feel that you have a good understanding, at this point, about what the president wants, ultimately, out of his trade policy?

I do. In one word, he wants tariffs to be higher. Beyond that, all of the secondary goals are fungible. Recently, the White House has been saying, “Oh, tariffs don't raise prices,” which is an economic conjecture I think is empirically wrong. You can look at pre- and post-tariff import prices, post-tariff prices are up. It's not a 100 percent being passed through to consumers, but you can see some of that passed through in stuff like toys, and audio equipment, and coffee, and yada, yada.

Point being, if you believe that conjecture, then it really can't industrialize the nation because it's implying that foreigners are just absorbing the costs to continue passing products that they make in Japan, or China, or Canada, into the United States. And then inversely, they'll say, “Well, it is industrializing the nation. Look at this investment, this factory that's being built, and we think it's because of the tariffs.”

Well, if that's happening, it can't raise revenue. And then they'll come back and say, “Well, actually, it's fixing the budget deficit.” If that's happening, then you're in the worst of both worlds because it’s raising prices and you're still importing stuff. So it's hard to find an internally consistent justification.

Part of my mental model of how this White House works is that there's different camps on every issue, and it's very much not a consensus institution on policy, but it's also not a top-down institution. It's not as though Donald Trump has a specific vision of what he wants the tariff rates to look like in five years, at a number level, per country per good. It's that he wants them to be higher.

He has this general impulse that he wants to reduce trade openness, and then somebody comes up to Trump and goes, “Hey, Mr. President, we should do 25 percent tariffs on cars. Remember where they come from?” And he goes, “That's a good idea.”

And then somebody comes up to him and goes, “Hey, Mr. President, we should do a 10 percent baseline tariff on everything that comes into the United States.” And he goes, “That's a good idea.”

And then somebody goes and says, “Hey, Mr. President, we should do a tariff that's reciprocal that's based on other countries trade barriers.” And he goes, “That's actually a good idea.”

Those are very, very wildly different goals that are conflicting, even in just that area. But it's not that there's one vision that's being spread across all these policies, it's that there's multiple competing visions that are all getting partially implemented.

An AI tailwind (20:42)

This is the one area where it's only American companies that dominate, and the depth is so high that [other countries] feel like they're not even competing.

I see AI as a potential tailwind toward productivity gains, but my concern is that any positive impact may only cancel out the headwinds of current trade and immigration policies, rather than accelerating growth. Is it a big enough tailwind?

I do think it's a tailwind, and the US has several distinct advantages specific to AI. The first being that most of the companies that are major players, both from a software-development and from an infrastructure-development point of view, are in the United States. We are here in the DMV, and this is the largest data center cluster on planet Earth, which is kind of crazy that it's in Loudoun County. But that kind of stuff is actually very important. Secondarily, that we have the depth of financing and the expertise that exists in Silicon Valley that is so rare across the rest of the world. So I am optimistic that it will increase GDP growth, increase productivity, maybe not show up as a growth in productivity growth immediately, if that makes sense. Not quite an acceleration, but definitely a positive tailwind and a tailwind that is more beneficial in the United States than it is in other countries.

The counter to that is that the AI stuff is obviously not constrained by borders to even a nominal degree, at this point. The fact that everyone talks about DeepSeek, for obvious reasons, but there are tons of models in the Gulf States, in Western Europe, in Australia, and you can access them all from anywhere. The fact that you can access ChatGPT from Europe means that not all the benefits are just captured in the narrow area around open AI headquarters in San Francisco.

The secondary thing is that, in my opinion, one of the most important reasons why the United States continues to benefit from this high-tech economy that most other high-income countries are extremely jealous of — you talk to people from Europe, and Japan, and even places like Canada, the prize that they're jealous of is the stuff in Silicon Valley, because they feel like, reasonably, they can make cars and do finance just as well as the Americans. This is the one area where it's only American companies that dominate, and the depth is so high that they feel like they're not even competing. Anyone who wants to found a company moves to San Francisco immediately, but that relies on both a big research ecosystem and also a big immigration ecosystem. I don't know if you saw the Facebook superstars that they're paying, but I believe it was 50 percent non-American-born talent. That's a really big advantage in the United States’ case that lots of people want to move to the US to found a company to work for some of these big companies. I don't think that's demolished, but it's clearly partially under threat by a lot of these immigration restrictions.

The other important thing to remember is that even though the president's most controversial immigration policies are all about undocumented immigrants, and then to a lesser extent, people who are documented asylees, people who are coming from Haiti, and El Salvador, Venezuela, et cetera, the biggest direct power that they have is over legal immigration, just from a raw numerical standpoint. So the idea that they want to cut back on student visas, they want to cut back on OPT, which is the way that student visas basically start working in the United States, they want to add more intensive restrictions to the H-1B program, those are all going to undermine the benefits that the US will get from having this lead in artificial intelligence.

The last thing that I'll say to wrap a big bow around this: We talked about it before, I think that when Trump was like, “We're doing infinity tariffs April 2nd,” there were so many bits of the computer ecosystem that were still tariffed. You would've had a very large tariff on Taiwanese computer parts, which mostly is very expensive TSMC equipment that goes into US data centers. I think that Jensen Huang — I don’t know if he personally did this . . . or it was the coalition of tech people, but I am using him as a representative here — I think Jensen Huang went in and was like, “We really badly need this,” and they got their exemption. The Trump administration had been talking about doing tariffs on semiconductors at some point, I'm sure they will come up with something, but in the meantime, right now, we are importing absolute record amounts of large computers. It's at a run-rate of close to $150 billion a year.

This is not all computers, this is specific to the kind of large computers that go into data centers and are not for personal or normal business use. I don't know what happens to that, let's say a year and a half from now, if the tariffs are 25 percent, considering how much of the cost of a data center is in the semiconductors. If you're going to have to then say, “Well, we would really like to put this somewhere in Virginia, somewhere in Pennsylvania, somewhere in Arizona, but you have a 25 percent premium on all this stuff, we're going to put it in Vancouver. We're going to put it in somewhere in the Gulf States,” or what I think the administration is very worried about is, “We're going to put it somewhere in China.” That chart of US computer imports, in trade policy, it’s really rare to get a chart that is just a straight line up, and this is just a straight line up.

On sale everywhere The Conservative Futurist: How To Create the Sci-Fi World We Were Promised

Micro Reads

Please check out the website or Substack app for the latest Up Wing economic, business, and tech news contained in this new edition of the newsletter. Lots of great stuff!

Faster, Please! is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit fasterplease.substack.com/subscribe