Are we cooked? We’re going to be revisiting this question quarterly now, it seems. To discuss, we have Peter Harrell, former Biden official and host of the excellent new Security Economics podcast, Kevin Xu, who writes the Interconnected newsletter, and Matt Klein, author of Trade Wars Are Class Wars and The Overshoot substack.
We discuss…
Short-term positive indicators, including the stock market rebound, the continued independence of the Federal Reserve, and the administration’s compliance with court rulings,
Why talent is the ultimate bottleneck to AI progress, and how the attacks on foreign students, universities, and scientific research will impact America long term,
Whether Trump’s policies will cause mass divestment from dollar-denominated assets,
The state of U.S. alliances, from the apparent thawing of relations with NATO and Zelenskyy, to signs of friction with Japan,
Next steps for Trump’s industrial policy.
Listen now on iTunes, Spotify, or your favorite podcast app.
Dollar Depreciation and Self-Inflicted Brain Drain
Jordan Schneider: Peter, kick us off.
Peter Harrell: We did this last quarter in late April, and looking around the short term over the last three months, the indicators are decent. We’re looking less cooked. Trump TACO’d out on a lot of the tariff wars, so the stock market rebounded. He can post memes about bombing Iran and get away with it — the short-term indicators actually suggest the last couple of months have gone decently well for the United States from a big picture perspective.
However, I’m pretty concerned about the long-term trajectories. Looking at some of the core things that have made the US what it is today — substantial investment in scientific R&D, attracting talent from around the world to do science and innovation, business leadership here in the United States, and the social fabric we have — I remain worried about whether we’re cooked long term, even though I have to give some credit to the last three months.
Kevin Xu: To add onto what you just said, Peter, in preparation for this podcast, I looked at the value of the dollar chart because I remember we brought it up the last time we talked about this. It went down by another 5 to 6 percent from our last recording to today. That was one of the biggest conundrums at the time. Right after Liberation Day, everyone who thinks about tariffs in perhaps the most conventional way would expect the dollar to go up, everything else being equal.
But the erosion of the dollar, no matter how you want to slice it, continues slowly but surely going down. It’s no longer a big story because there have been so many other attention-grabbing geopolitical and economic stories since then. But for me, that’s something I’m going to watch from time to time as this collective “how does the world vote with their feet” question regarding how much credibility they place in the United States government and leadership through how they treat the dollar.
Matt Klein: The overall level of “cookedness” has not actually changed much one way or another. Partly because I discounted the damage from the tariffs. Not that the tariffs announced on April 2nd were good, but that was never the main issue. The attacks on scientific research, the attacks on foreigners coming into the country — and we’re not just talking about changes in Southern border policy, but the way European tourists are treated — across the board restrictions on foreign students coming into universities, the attacks on the rule of law — these things have not abated in any way.
There has been ongoing pushback from courts, but many things remain problematic. The changes to scientific research policy, the changes to approaches on medical research funding in particular and vaccinations — that has not improved in the past few months. If anything, it’s been entrenched.
If I wanted to be more pessimistic, I would say the fact that immediate concerns about things such as tariffs have subsided and the stock market has recovered might actually reduce the pressure to unwind those longer-term, more damaging policy changes that have occurred over the past six months, making them more likely to persist. Even if the immediate impact is not necessarily substantial, it could be a very long-term drag, not just on GDP growth, but on the quality of life, power, and well-being in this country.
The analogy is Brexit. When the UK was voting about whether to leave or remain in the European Union, there were concerns — people called this “project fear” — about an immediate catastrophe if they voted to leave. That didn’t happen, so people said the case was overstated. But the more realistic and correct argument was that everything would be worse — not catastrophically worse, but persistently worse over time. That has been borne out. I fear we’re ending up in a similar, if not worse, situation as a result of all this.
Jordan Schneider: I’m going to take the other side of this argument. Many of the worst-case scenarios we were contemplating in April are much less likely to manifest now. I’m taking Brexit as a win, if that is our analogy, as opposed to something even more dramatic.
Elon is done. DOGE’s energy is abating for the most part, which I think is positive for governance and civil service. We have Congress — maybe they’ll pass a bill — it doesn’t look great, but it’s also not republic-wrecking. There’s going to be an election not that long from now.
On the immigration side, yes, the direction of travel is not great, but we have not banned all foreign students. The country bans that we’ve seen and will continue to see are directionally problematic, but we could have had much more aggressive paths that would have presaged an even more aggressive four years, which have not happened. You’ve even seen Trump at one point tweeting about protecting the farmers, saying we can’t kick out all of our undocumented immigrants.
It’s not just on trade where this administration has blinked. It seems there are other topics — science accepted, we’re in a pretty poor timeline on that front — but at least on the immigration front, it seems we could be in a worse place than where we are now.
Peter Harrell: I’ll add two points, one on each side of the ledger, building on what you’ve both said.
On the positive side, Trump appears to be stepping back from the brink, at least somewhat, on whether he’s going to outright defy the courts. That felt more real to me a couple of months ago. He did bring home the individual who was sent to El Salvador and charged with crimes — a complicated issue. But he does seem to be stepping back from his showdown with the judiciary, though time will tell. I’d put that in the positive ledger.
On the other side are things that are eroding U.S. centrality in the international economic system, at least long term. I want to pick up where Kevin left off on the decline in the dollar’s value. We’re seeing several other data points where foreign governments and private parties have quadrupled or quintupled their efforts to create financial networks substantially outside U.S. jurisdiction.
You see this in the drop in the dollar’s value — investors are clearly beginning to pull back from the United States. You also see it in the large run-up in gold prices. This 3,000-year-old store of value has had an explosive couple of months, building on a strong year or two, driven very much by central bank purchases of gold as a way of hedging and reducing their exposure to the dollar.
But perhaps moving to gold isn’t even enough. There’s now a debate in Europe, in Germany and Italy, about whether they need to repatriate their gold reserves — which are currently stored deep in the rock underneath Manhattan, in the custody of the New York Fed — because they don’t even trust us as custodians of their gold anymore.
This is a significant data point. Now, it’s not easy to develop large-scale financial networks outside the U.S. — it’s very hard and doesn’t happen overnight. We’re still seeing foreign investment here, and we’ll still attract the capital levels we need. But this foreign government distrust about the long-term stability of the United States, whether right or wrong, is driving something that you’re beginning to see in the numbers.
Matt Klein: I might push back on that a little because while that intuitively makes sense, the numbers aren’t necessarily showing something as extreme as one might have expected.
The dollar depreciation was sharp around April 2nd, but if we zoom out, the dollar is still quite expensive on a trade-weighted basis relative to the rest of the world. It’s still up something like 20 percent compared to early 2021, up 30 percent compared to 2014. It’s still quite expensive — maybe not quite as expensive as before.
Foreign investors had been massively overweight in U.S. assets. On a market cap basis, they were equally weighted, but 70 percent of stock market capitalization being American feels odd given that the U.S. economy is a quarter of the world economy. Yes, American companies may be more profitable, but still, there were plenty of reasons why, with more benign catalysts, you would have seen some of these market moves, particularly on exchange rates or foreign allocations.
In some ways, the story is that there hasn’t been as much of a move yet. I agree that if the Germans and Italians move their gold, that’s symbolically interesting, although what that means in practice is debatable. But it’s striking what hasn’t happened, which is honestly surprising.
The same thing applies on the U.S. side — have Americans been moving a lot of money out of the U.S. into foreign assets? Maybe some, but it’s not super clear in the data that it’s been enormous volumes of the kind you would have expected.
We have other examples of countries with sharp moves in response to political instability — Turkey in March, for example, after the arrest of the mayor of Istanbul. There was a very sharp move, actually a similar-sized move in the exchange rate for the Turkish lira as for the dollar. But in terms of the quantities of what was actually happening, it was on a much different scale. Turkey’s central bank only has $40 billion in reserves to keep the currency stable. We didn’t see that here.
As much as it would be logical to say that’s what’s happening, I don’t think we’ve seen it yet. Maybe we will; maybe it’ll be more gradual. The gold price move is certainly interesting, although that really dates back to the sanctions on Russia’s central bank when that trend began. Otherwise, it’s hard to see it as a dollar-specific story.
Jordan Schneider: Matt, we’re doing better on the Fed, right? That was a news story that completely burned itself out. I think we’ll have a Fed come 2028.
Matt Klein: Maybe. This is the other thing — I talk about the independent Fed, the independent judiciary, but these are Senate-confirmed positions. It’s not clear to me what kind of pushback the Senate would provide if you have a Trump-supporting majority there.
The judiciary we have now is not necessarily the judiciary we’ll have four years from now. That could change. The Fed situation could change. If I remember correctly, there’s a passage in the big, beautiful bill they’re putting through that basically says administration officials would be immune from contempt of Congress. I’ll defer to the lawyers on the exact terminology, but there are things being done by the other branch of government to support the potential for further executive overreach, which should be concerning even if it’s not necessarily generating the same kind of headlines and market responses as tariffs.
Peter Harrell: In fairness, I said a couple of minutes ago that Trump is stepping back from his confrontation with the judiciary in a noticeable way, though the long-term impact remains to be determined. Part of that is because he’s winning. He’s making the assessment that he can win at least a large enough percentage of his cases in front of the judiciary.
We saw that today, at least temporarily, with what the Supreme Court did on nationwide injunctions and birthright citizenship. He’s essentially saying, “If I’m winning two-thirds of my big cases in the courts, maybe I don’t need to outright confront or defy them over the third that I’m losing."
As a lawyer, I’ll take compliance with judicial rulings. That’s all I’m saying.
Jordan Schneider: Peter, give us the “tariffs are actually illegal” update.
Peter Harrell: This is one near and dear to my heart. Most of Trump’s tariffs — certainly all of the Liberation Day tariffs from April that caused significant market turmoil — weren’t imposed under a trade law statute. He imposed these tariffs under a 1977 emergency powers statute that previous presidents had used for sanctions on Iran and Russia. It had never previously been used for tariffs.
This is progressing through the court system. There’s a specialty district court in New York called the Court of International Trade. Plaintiffs challenging these tariffs won at the specialty district court, but that victory is on hold while the government and plaintiffs deal with an appeal before the Federal Circuit, one of the federal appellate courts.
The Federal Circuit will have its hearings on this case on July 31st. We’ll likely get an opinion from the Federal Circuit by the end of August, and then this will go up to the Supreme Court in the fall.
If I were betting, I would bet Trump loses on the merits here. At the end of the day, these tariffs are going to be found unlawful. That doesn’t mean the end of all tariffs — it means a mad scramble across the U.S. government to see what tariffs they can recreate under other legal authorities.
Matt Klein: They’ve been doing that with Section 232. The 50 percent tariff on steel isn’t being challenged. The pharma tariffs in the pipeline are now separate.
Kevin Xu: There’s your alpha for the episode.
Peter Harrell: For those thinking about this from a market perspective, one thing that’s going to happen is that companies paying the tariffs this year will get a rebate. They’ll be entitled to a rebate of paid unlawful tariffs, which would probably happen next year. From a market perspective, you’ll see an earnings shift from this year to next year as they get rebates of tariffs paid this year, which they’ll then have to pay going forward again. Retroactively, 2025 might look better on earnings than it does right now.
Matt Klein: Are the tariffs tax-deductible? Are tariff payments deductible against income tax? Is the rebate going to be taxed?
Peter Harrell: I’m not a tax lawyer — that’s a good question. It’s certainly a business expense, so you can claim it as such.
Kevin Xu: They’re not going to tax tips.
Peter Harrell: How to tax this is full employment for lawyers. That’s Trump trade policy 101 — full employment for lawyers.
Talent > Transistors
Jordan Schneider: Kevin, how about an AI update?
Kevin Xu: One thing in the AI world — whether it’s investment or technological progress — speaks to how equipped the U.S. is. I’m not necessarily saying China will lead or run the table — I’m analyzing our country specifically.
CapEx is growing for at least another one or two years. The AI trade is on. But everyone’s realizing that GPUs are no longer scarce. That was more of a temporary thing. If you have the money, you can get whatever latest NVIDIA GPU you want for a data center.
The pace of energy coming online in the U.S. is still relatively slow compared to other places. That’s why the Middle East swing by the president and all the AI executives is so consequential. There’s an executive order coming from the White House to make that easier in the U.S. — we’ll see what happens in real life.
All of this really comes down to people. I’m talking about both the supply of general contractors to build and dig data centers and wire all the power sources to make data centers come online as quickly as possible, as well as the top-end research talent that’s becoming more of a bottleneck — not just for which country could advance, but which company within which country will advance.
The most interesting position right now is probably Mark Zuckerberg’s spending spree to hire whoever from OpenAI is willing to jump ship. He’s willing to pay whatever he needs to pay. It’s like a classic “buy the free agent market” way to win the championship.
It’s an interesting comparison, keeping the sports analogy going, to how DeepSeek got to where it is — a pure scouting and drafting way of building their team. Their team got built within China (maybe they got some secret offer from Meta too, we don’t know), but these are very different ways of accumulating AI talent, which we know is in real scarcity — not the hardware anymore.
How do we extrapolate that to which country or company will be first to reach the promised land? This makes me think about the immigration issues we discussed at the beginning. Our country never had a great immigration policy for any kind of immigrant, to be perfectly honest. We had enough prestige and resources to attract the brightest in the world. But when that chilling effect sets in, it could be a problem for the long haul, while the scouting ecosystem or drafting source on the China side seems sufficient to put together a DeepSeek, if nothing else.
Peter Harrell: Can I ask you a question, Kevin? It’s not an area I know much about. Is Europe anywhere on the map with AI these days, or is this really remaining a U.S.-China competition?
Kevin Xu: This is very much a U.S.-China competition, particularly because of the talent question — not so much about the money. Jensen went through a Europe swing just a couple of weeks ago and announced a bunch of sovereign AI projects in the UK, France, Germany, Slovenia, and some smaller countries. The governments are generally pretty good at building hard infrastructure from the ground up and signing really big checks when everyone comes together politically. That’s what governments are good at.
But when it comes to organic talent development, most European top-end talent either moves to the U.S. if they can — because they find the European system generally stifling if you want to move fast — or they work locally at the labs of American companies that are building locally to recruit there. The three most recent successful Zuckerberg recruits from OpenAI were from the Sora Lab of OpenAI. Is that European talent or American talent? You tell me.
Matt Klein: DeepMind is UK, right? They’re owned by Google, but —
Kevin Xu: That’s the one shiny example.
Jordan Schneider: Does it count anymore? They were a subsidiary as of 10 years ago. It’s an interesting question, all this sovereign AI buildout. It almost reminds me of Chinese provinces that spin up their own data centers and proudly buy some hardware chips. What are we doing here?
Another aspect I wanted your take on, Kevin: there are rumors that Meta was going to bow out of the open source game. In that case, we’d have no 10-figure companies in the U.S. really pushing on open source, basically seeding the entire ground to Alibaba and DeepSeek. What does that mean from a 10-year perspective?
Kevin Xu: First, Meta’s version of open source — LLaMA — since we do have a lawyer in the room here, is actually from a legal perspective less open than all the Chinese open source models currently, from a legal licensing perspective. Meta came up with its own version of an open source license that was hotly debated within the tiny community of open source licensing lawyers that I’m part of.
Part of those restrictions is that once you use LLaMA’s model up to, I believe, 700 or 750 million monthly active users, then you need to engage with LLaMA’s parent company to have a commercial licensing agreement. That’s an important nuance to start with.
From that perspective, and I’ve heard this from other sources too, there was at least an expectation from LLaMA at some point to still make money directly from its open source models. If it gets competitive enough — whether it embeds into another coding agent like OpenAI and Anthropic, or some other B2B or enterprise product — they had some expectation of that. They weren’t just going to throw a ton of money and talent into this theoretical financial dark hole and do something good for the world, even though they were falling behind.
The fact that this was even leaked as an internal discussion point — I don’t know if they’re really going to do it — but the fact that serious people inside Meta have considered abandoning the open source part of LLaMA and actually abandoning LLaMA entirely in favor of just the best model in the market has so many layers of irony. We have the open society here in America, the open this and that, freedom and whatnot, but we chose this most singularly transformative technology direction to be as closed as possible among the companies.
You can imagine a lot of siloing going on. A lot of things that probably could have been shared but aren’t shared for one reason or another. That actually blunts progress on the long-term horizon, as well as the diffusion pace of certain technology, where open source is just the best way to diffuse a lot of technology — even if it’s the worst way to make money off that technology. That might be the reason why a lot of U.S. companies are choosing the closed source path, while the Chinese side is right now leading in all the open source model benchmarks.
Peter Harrell: Jordan, it’s a question for you. When we think about “Is America cooked,” we often think about this in the U.S.-China context.
Peter Harrell: Has China done anything right over the last three months?
Kevin Xu: I’ll jump in here real quick. This is actually related to the dollar again — speculative because it’s very recent. They did have a recent high-level finance summit in Shanghai, the Lujiazui Forum 陆家嘴论坛, where the governor of the People’s Bank of China and all the relevant high-level financial regulators came and gave speeches.
What came out of this specific forum this year is a concerted willingness to really promote the RMB in ways and places where the dollar is falling short for one reason or another — not to necessarily replace the dollar as the global reserve currency, because that’s a whole other can of worms that I don’t think anybody in China really wants to be. But in countries or trade directions where having the RMB be the settlement currency in lieu of the dollar, or being able to buy global commodities in large chunks using the RMB versus the dollar, is advantageous, there is now a policy appetite.
This is very much a reaction to Liberation Day, to the way that the U.S. instigated global trade war has unfolded. That prompted this. How far will this particular policy direction go? Nobody knows at this point — it’s way too early. But the appetite is there now to be a little more proactive about making whatever reforms necessary to make the RMB more attractive in certain situational settings when the dollar is becoming a little less appealing.
On Allies and Immigration
Jordan Schneider: How bad is Trump going to end up being to the allies? Q1 was probably pretty close to a worst-case scenario with those really obnoxious speeches in Munich. But since then, we had Haley do a trip to Asia where she basically said things that a Nikki Haley administration would have said.
What does the Iran bombing mean for America’s commitment to Asian allies? On one hand, it shows that Trump’s willing to bomb stuff, but on the other, it’s bringing us back to the Middle East. Maybe that’s a bit of a wash.
There was also talk of some crazy DOGE cuts to the Pentagon, but there are basically going to be no defense funding cuts. Whether they’re actually going to reform defense acquisitions and buy stuff we need to buy is an open question.
Even though we’ve had some serious U.S.-Japan weirdness — with Japan being sick enough of the Americans to cancel a two-plus-two dialogue — I still think we could be on a worse timeline than the one we’re on now with the way Trump has treated Asian allies.
Matt Klein: If you really want to be optimistic, you just say the way that the administration is now treating European allies — that’s where you’re seeing the biggest shift, maybe. We’re friends with Zelenskyy all of a sudden. Who knows how long that will last, but that is notable. The European NATO countries seem to have managed to figure out how — with the exception of Spain — to put together a deal that looks like it’s getting everyone on the same page. Who knows how that will play out longer term, but if you want to be encouraged, that’s definitely somewhere to look.
Kevin Xu: Isn’t that just giving Trump what he wanted?
Peter Harrell: That is part of it.
Matt Klein: You reframe it. He said he wanted five percent, and they said three and a half percent plus one and a half percent other, which is fine. Everyone agreed on it. Arguably, it’s good for them too. But the shift with Zelenskyy is interesting, as Jordan said. Trump met him at the NATO summit and said they had a great conversation. Now he’s saying, “Supposedly, Putin called and I need help with mediating with Israel, and I need help with you.” Who knows?
Jordan Schneider: It’s hard to know what’s really going on there. But if you want to be optimistic, the thing with this stuff is that it was unclear if Trump was going to be doctrinaire about these things or just vibrate to weirder places than any American president had gone to, then vibrate back to a more mainstream place and keep swinging. But him staying at “No, we’re leaving NATO, this is a terrible thing" — that’s not something that’s happened. He has shown the capability to get pissed off and fed up by people like Putin, which is something you may not have entirely priced in April of this year. I consider that positive.
Although I’m the Trump optimist here.
Matt Klein: To be fair, it’s also not as if we’ve actually adjusted our Russia policy in a way to imply that he’s actually gotten upset with them. So far, it’s just words. But we’ll see.
Jordan Schneider: I’ll take words.
Matt Klein: Better than the other words.
Jordan Schneider: They’re a start.
Peter Harrell: I very much agree with the point Matt made. Obviously, the words coming out of the NATO summit and the words around Zelenskyy are much better. The lived experience over the last six months is that words change. The thing I’d note in the other direction: two or three weeks ago, Trump was saying nice things about Canadian Prime Minister Carney. Right before we got on this call, he’s now back to saying Canada’s the worst and we’re going to tariff them all over again. Nothing really stays stuck with our current president.
Jordan Schneider: But if things don’t stay stuck on the dumbest timeline, then that is — again, it’s more the Brexit arc, less the actual America self-owned catastrophe one, which is okay.
Peter Harrell: Maybe I should be worried here. The other thing I’m watching to see how it plays out — conceptually related to immigration but also distinct — is where we’re going with our universities. To pick up on a point Kevin made earlier, it’s not that we always had the best immigration policies. American immigration policy has been totally dysfunctional for my entire life — I’m 45 — but we have had, against a backdrop of dysfunctional immigration policy, some very important bright spots. The ability to attract truly great talent to some of the world’s best universities who then stay here and found and run businesses and contribute to scientific innovation.
Trump’s war on the universities has now spread, at least with Harvard, beyond just cutting off funding to trying to cut off their ability to enroll foreign students. It appears that he may be trying to get the University of Virginia to fire its president. I’m not going to defend all kinds of crazy things universities have done over the last five to eight years — there have been plenty of challenges with American universities — but they are a huge source of long-term American strength from a scientific innovation perspective, from a talent attraction perspective, from educating people who then go out and do great things.
It’s looking like a fairly full-out war and not just something that he’s looking for a détente on. There’s a ton of institutional strength in American higher education. I have bias here because both my parents were full-time academics all my life growing up, so I’m biased toward American educational institutions. I worry that if we structurally harm our higher education system, that would be another thing — cutting scientific R&D — that helps us be cooked longer term.
Matt Klein: That’s part of how we’re hurting it. They get a lot of that scientific R&D funding, and if that’s cut off — and also, this is not just a Trump obsession — in the bill that currently passed was this punitive tax on endowment income for universities. For universities that have enough endowment income for that to matter, that pays a very large chunk of their operating expenses. Most of the rest comes from things like government grants for research.
If you cut those things at the same time, they’re going to go out of business or they’ll have to drastically change what they do. Again, this is not something that’s really getting a lot of pushback because people pay a lot more attention to other things — the tariff on this is going up by this amount or whatever — but these other things are still steadily going on despite the fact that people’s attention isn’t focused on that. As you said, it could be a relatively benign Brexit outcome, but it adds up and would be very harmful.
Kevin Xu: That will never show up in these more short-term twitchy indicators, whether it’s the market, the dollar, or any of that stuff. Two points I want to add to what Peter said — which probably makes the best case for long-term cookedness if there’s no really forceful reversal of what’s happening right now, but hopefully some actual tangible improvement on immigration — not just keeping together whatever we had left prior to this presidency.
First, the way in which a lot of these scientific research grants are cut isn’t just a problem because they’re cut from an absolute value perspective, but they’re cut from a “Control-F” perspective. You go through the grants, you Control-F for words that the administration doesn’t like for entirely ideological reasons, and you cut them. That is the reason a lot of the top academics and researchers who came to the United States from other countries specifically came to escape from.
Matt Klein: Right, like “neurotransmitters” because it has “trans” in there. Now you’re saying you’re not looking at Alzheimer’s research.
Kevin Xu: Exactly. The word “diversity” just trips up a bunch of alerts from this Control-F way of thinking about scientific research when the word “variance” could have replaced “diversity” and that grant could have been safe. All these very blunt instrument ways defeat not just the fact that we get to have this innovation happen in our country, but that the people who come to this country to do the research and want to stay — they didn’t just come for the money. Nobody in academia, even at Harvard or Princeton, really celebrates the amount of salary they get to make, even at the highest level. They came here because they don’t have to be worried about being Control-F’d in their own country of origin.
The second point that builds off of that — which is really the long-term cook scenario — every time we have this American declinism conversation (this is not the first time we’ve had this conversation as broader public discourse) the most common way that we patted ourselves on the back or were able to go back to sleep after the conversation was immigration. We still had immigration. All the smartest people in the world are going to save our ass by coming here. That may not be an answer we can just default to anymore. That’s probably the one thing that would put you on the cook ledger in the 10-year timeframe — certainly not a quarter-by-quarter timeframe, because who knows what the market is going to do.
To come back to immigration today — we’re actually selling Trump cards. We’re selling a spot in the best country in the world, quote-unquote, to the highest bidder, not the smartest people. What kind of people do you attract from that? I’ll let the listeners have their own imagination over what those people could be.
Matt Klein: It’s worth noting that the U.S. university system has been so dominant and so good for so long that we assume it has always been that way. But that’s not actually true. Before 1933 — which is a significant date to think about — the U.S. university system was not particularly competitive globally. After that, it was. English wasn’t even the dominant language among academics — it was German. It’s not inherent that it has to always be this way. You’d have to work hard at it, but if you’re committed enough to self-harm for ideological reasons, then you can get rid of these long-term advantages.
Peter Harrell: Another topic that the Trump administration needs to prove it can handle — and it’s an open question — is supporting U.S. manufacturing and some of these strategic industries and products it talks about. There’s also a lot of bipartisan support for the idea that we should have at least some onshore or ally-shored critical minerals production, shipbuilding, and these strategic lower-tech industries.
The administration has certainly talked a good game about some of this. They’ve put out executive orders, they’ve said they want to reduce permitting hurdles to make it easier. It seems entirely up in the air whether they’re actually going to be able to achieve any of this manufacturing renaissance that they’re talking about.
This is going to require a more integrated approach than what we’ve seen from them to date. That’s a big X factor for me: Are they going to be able to actually put this together or not? Will it be in the industries that actually matter to us? All due respect to Howard Lutnick, I’m not sure “screwing tiny, tiny screws into iPhones” is really the industry we want to be focusing on in terms of our manufacturing.
Jordan Schneider: Are there any data points that are encouraging on this line? We’ve seen various efforts to gut the DOE loans program. The CHIPS Act has been exploded.
Peter Harrell: If you look at what they’ve actually done with the money, the DOE loans program was focused on clean energy, and they don’t want clean energy. That seems to be fizzling. There’s anecdotal evidence — Matt, you may have plugged into the data more — that some of the big announced manufacturing investments in those sectors have been put on pause. There’s a slowdown there.
While Trump’s talking about things like shipbuilding, we’re not actually seeing shipbuilding. Let’s put aside the climate impacts of oil and gas and current energy prices — I’m not sure you’re going to see a huge oil and gas boom.
Matt Klein: We haven’t. The joke about “now we’re going to unleash drilling” — you should talk to the shale drillers. They don’t want to do that. Briefly, when Brent went close to $80 because of fears of Iranian activity in the Strait of Hormuz, if it had stayed there long enough, maybe you would have seen more shale. But between the fact that their costs have gone up tremendously because of tariffs on steel — what do you think the pipes are made of? — and the fact that the price of the product they sell has gone down, plus labor costs are still going up over time, it’s not super compelling.
Investment in manufacturing peaked in the middle of 2024. A lot of the growth in investment and manufacturing capacity in this country had been directly attributable to CHIPS and IRA — semiconductors, batteries, things like that. If that’s what the incremental investment is coming from, and then those sources of incentives are being gutted as quickly as possible, it’s not surprising that you’re going to see that move in reverse.
Some of the money may already be out the door enough that it won’t be pulled back on projects that have already started. But it’s not super encouraging. The Loan Programs Office was talking about clean energy, but they were also talking about doing it for critical minerals. Those minerals have green energy-related uses, so you could say, “Oh, we’re going to get rid of it.” But to the extent that they say those things are strategic — whether it’s magnets or whatnot — if you cut the funding, it doesn’t matter whether you call it green energy or not. You still are not going to have the resources.
It’s going to be very difficult to have a domestic industry in those sectors be competitive without some kind of government support because they’re so cyclically volatile on price and because the main producers right now are in China, where market-based constraints are not as severe. Because the capacity is there, they can basically flood the market when they want to. If any seemingly viable competitor shows up — which is why you’ve had such problems with lithium in the U.S. in the past — you need some kind of government backstop to prevent that. That was being done out of DOE.
It could change, but it’s not clear to me there’s strong interest in doing that. I don’t know why a manufacturing renaissance would occur. On top of which, if part of the rationale for a manufacturing renaissance is that these tariffs are going to make U.S. production for the U.S. market more competitive, but then no one knows what the tariffs are going to be — either because they’re illegal or because they’re going to get negotiated away for unclear gains — why would you make a commitment now? It doesn’t pencil out. You want to use a pencil, not a pen, because you can just change.
I don’t see it. It’s not showing up in the data that we’re having any kind of manufacturing renaissance. Far, it’s the opposite. It’s not like it crashed or anything, but it’s not clear to me why we would see a big change given what we know so far.
Kevin Xu: This is a question for Matt, but obviously anybody can jump in. What are the things that got some play but did not get nearly enough play during the DOGE stuff? The cutting of staffing or investment in places like the Bureau of Labor Statistics and BEA. Basically the only neutral arbiter of actually accounting for how cooked we are or not as a country — hard data, not soft data, not live data, not sentiment data.
That’s another thing that just confounds me. If we were to actually answer this question from a statistics perspective or in a rigorous way a year from now, two years from now, whatever — I don’t even know if I could really trust any of the GDP numbers going forward, any of the non-farm payrolls going forward. The government pumps out so much good data and it takes so much time to build up that infrastructure, but you need one bad cut to completely get rid of that discipline.
Matt Klein: One thing I would say is some of that actually precedes even the election. It had to do with some of the budget deals that were being done in 2023-24. I don’t know why these were agreed to, but basically — I’ve talked to the BEA people and seen they announced tables have been discontinued because they’re not in the budget for it. This was over a year ago. That’s unfortunately been an ongoing problem.
The BLS, which compiles things on inflation and labor market, has been having issues for a while — not necessarily because of underfunding, though that’s probably related to it, but just because people don’t respond to surveys the way they used to, for who knows why. You take a survey of employers every month, figure out how many people are working, and then aggregate it up to give an estimate. But if the employers are not answering your call for whatever reason, or fewer of them are proportionally, that creates potential problems. That was already an issue.
It’s compounded by cutting people arbitrarily. There was an announcement in the past few months — the Commerce Department, which runs the BEA and the census (not BLS, that’s in the Labor Department), had this outside body of advisors, mostly academics, to judge the quality of federal statistics. They just decided to cancel the meetings. Who knows? Maybe they just decided we don’t need the meeting. But it’s the kind of thing where if you were to say, “Okay, this would be the first step if you wanted to fudge the numbers."
I don’t see any evidence that’s what’s going on so far. I don’t want to impugn any of the numbers or any of the work that’s being done now. I don’t see any evidence of that yet. I don’t even know what it would look like, but I agree it’s something to be worried about. Just the fact that we’re talking about it is indicative of the kinds of people who are in charge and what they’ve generally shown in their attitude toward things — that this is a possibility, that it could be happening, even if it hasn’t actually happened yet.
Jordan Schneider: It would be pretty remarkable if we’re in a world where we start treating American numbers like Chinese numbers. There’s the intel community stuff of fudging judgments and people leaking out what the real numbers are. I can imagine that happening also in BLS pretty quickly, where there’s enough of an institutional allergy to this that you hear pretty fast within that org of any of these types of shenanigans happening.
Matt Klein: I had two things. One is that far there are independent ways of verifying inflation stuff. The Billion Prices Project was invented by an Argentine working in the United States as an academic economist — very appropriately — to figure out alternative measures of tracking inflation. They just look at online listings. Far for the U.S., you say, “Okay, what are the prices of a basket of goods that matches the CPI basket of goods based on online listings?” And it’s tracked perfectly so far. That’s encouraging.
On the Chinese side, I guess you might know more than I do. My understanding is the headline GDP numbers may or may not be massaged. The underlying stuff is probably the good faith best estimate. But if they don’t like something, they just stop publishing it. It’s not even that they make it up — they just stop publishing the number. That seems like that actually could be an issue, especially given you have budget cuts anyway.
Jordan Schneider: There’s enough money in getting good numbers on U.S. employment that people will just invest in getting this right.
Matt Klein: It might not be public. There’s value in it being publicly available.
Kevin Xu: It’s locked up in Jane Street’s data center somewhere.
Peter Harrell: Bloomberg will add it to the terminal for an extra $500 a month.
Jordan Schneider: Does anyone have recommendations? Anything you’re reading or enjoying?
Peter Harrell: I just finished reading a book from 15 to 20 years ago by a Wall Street Journal reporter about the rise of the Texas oil families from the 1930s to the 1950s — the wildcatters like the Hunt family and similar figures. It’s fascinating because there are so many parallels, both culturally and psychologically, to the people who have made fortunes in tech over the last 20 years.
These were individuals doing innovative things with new technology who made tremendous amounts of money and then started converting that wealth into political influence. They began buying up media — literally purchasing radio stations at the time. It was quite striking to see how these people started out drilling oil wells in the 1920s and eventually came to build or at least fund the modern conservative political movement in the 1950s. It’s a very interesting piece of history.
Kevin Xu: I started reading Inside the House of Money, which is a collection of interviews with macro hedge fund managers, conducted by another macro hedge fund professional. The interviews are very intimate and revealing.
One of the interviews features our current Treasury Secretary, Scott Bessent, from when he was running his own fund right after leaving Soros. Apparently, going into this interview, he was extremely sleep-deprived — he hadn’t slept for days — and was worried about something that might happen in China. He didn’t specify what, but that was the context for his interview.
This book was written between 2004 and 2005, and I still haven’t figured out what was happening in China at that time that kept our future Treasury Secretary awake at night. If anyone can help me solve that mystery, I’d appreciate it. Regardless, the book is a really engaging read if you enjoy understanding the mindset of global macro traders.
Matt Klein: I don’t have anything quite as sophisticated, but I recently started watching Your Friends & Neighbors, and I’ve been enjoying it tremendously. I also bought The Party’s Interests Come First on your recommendation, though I haven’t started reading it yet. It sounds excellent.
Jordan Schneider: I have another recommendation that will preview a future China Talk episode — To the Success of Our Hopeless Cause: The Many Lives of the Soviet Dissident Movement. It just won the Pulitzer Prize.
It was actually amusing because I booked the author before he won the Pulitzer, and then we had to postpone our interview because he suddenly had major media obligations to attend to. The book is beautiful and literary, with incredible sources. You have all these transcripts from court cases where these dissidents are essentially mocking the prosecutors and judges, taking them to task at every turn.
While it doesn’t necessarily tell you much about this moment in American politics, it provides a fascinating window into the world of these very brave and frankly eccentric people who made remarkable choices in the 1950s, 1960s, and 1970s in the Soviet Union. This was after the Stalin era, so they weren’t going to be executed, but they were still putting an enormous amount on the line to confront the regime.
Amazing read. The answer is "mega cooked." I'll just talk about the AI talent race in particular, most of the AI researchers recently hired by Meta are Chinese, with Chinese bachelor's (graduate degrees not yet) from top Chinese universities. We are defunding our education and excluding even highly intellectual immigrants. China might not have the same immigration history as we have and a huge language and political barrier to even get there, that being said they are investing in their own population in a way we are not. While China has diversified their investments the US is singularly focused on become more militarized and isolated. We still have the advantage of our financial and capital systems but as the US dollar hegemony comes into question. It leaves a lot to be wondered about the future of our economy and politics.
Given the OBBB was passed today, we're cooked. We're ceding the technologies of the future to China and ruining our fiscal situation so our ability to do things in the future is even more constrained. We're cooked for a generation.