World Economic Update

The World Economic Update highlights the quarter’s most important and emerging trends. Discussions cover changes in the global marketplace with special emphasis on current economic events and their implications for U.S. policy.

This series is presented by the Greenberg Center for Geoeconomics and is dedicated to the life and work of the distinguished economist Martin Feldstein.

MALLABY: So welcome, everybody, to today’s World Economic Update. This is a series, as you know, sponsored—dedicated to the life and work of the distinguished economist Martin Feldstein, who was a mentor to many of us.

We have, as usual, a great panel. Over there, Rebecca Patterson, my colleague at the Council on Relations. We have Adam Posen, president of the Peterson Institute for International Economics. Natasha Sarin next to me, professor of law at Yale Law School, a CFR term member. I mean, given your wisdom, the fact that you’re a term member is—(laughter)—I don’t want to be rude about term members, but I think of you as sort of senior to me intellectually, and all that.

Anyway. All right. So today we meet at a time when the bond markets have made it pretty easy to decide the focus of how to structure the conversation. The thirty-year bond yield in the U.S. was, yesterday, I think, the highest since 2002. So this is a pretty significant moment. I’m going to start with Rebecca and just talk about the immediate reason for this spike in rates. They’ve had this run up just in the, last, what three weeks? And a sense that the Iran war in particular did not have a likely, foreseeable, neat way to end the conflict. That Iranian proxies could hit oil infrastructure beyond the Strait of Hormuz in Saudi Arabia. This is what seems to have been the trigger for sending crude shooting up, and therefore inflation expectations. Talk a bit about that trigger.

PATTERSON: Yeah. First of all, it’s lovely to be here with all of you, and with all of you.

What’s so interesting to me about what’s happened with bond yields in the U.S. and, frankly, in other countries, since the Iran War began, is that very little of the rise in yields actually has to do with inflation expectations. If you kind of break down what’s driven the change in yields since the beginning of March, it’s about 5 percent inflation expectations and 95 percent everything else. And I think that’s really important when we think about what higher yields mean, right? Higher yields, higher borrowing costs are a drag on economic growth. It’s more expensive to get a mortgage, an auto loan, to borrow money to grow your business.

But what’s been driving yields is a repricing of the Federal Reserve, and how much it might need to raise rates, and how long rates could be elevated. It’s a rethinking of growth. And that’s tied in part to this huge AI infrastructure buildout. That’s not the only, but a large factor supporting economic growth. There’s supply concerns as the U.S. spends, in part on the war and in part on other things. There’s worries that we’re going to have more and more bond supply, and questions around whether there will be an equal rise in demand to keep yields steady.

And so all of these fundamental factors are pushing up bond yields. With inflation and the thought about the war, when you look at pricing of oil markets, if you look even at inflation expectations by businesses or consumers, it’s very up and down. So, yes, we’re going to have higher inflation in the next twelve months, but over the next five years it’ll come right back down. So there’s a sense that, whether it’s Iran or the U.S., someone’s going to blink and the war will end sooner rather than later. And I think you see that—even just yesterday we had some consumer confidence figures that included inflation expectations. Twelve-month inflation expectations higher, further out inflation expectations quite a bit lower.

And that, to me, is so important right now when it comes to not having this run away from us, because as long as inflation expectations over the medium and longer term can stay anchored that allows the Fed to have some really important credibility as it navigates all this. But even though oil prices, energy prices, diesel, refined products, all of that, it’s not what the Fed focuses on, right? They take out those volatile ingredients. But the longer the war lasts, and that was the fear, the more risk that these volatile components of inflation can feed through into the economy.

And diesel is a great example. You know, diesel prices just since the beginning of the war are up about 70 percent, if we just look at gas—diesel gasoline prices in the United States. And that’s going to feed through into food prices, not just this month but next year. That’s going to feed into transportation costs. And so if it can get brought more broadly into the inflation, that could affect the Fed, and that could affect yields, and push them up even higher than where they are now.

MALLABY: Adam, a complicating factor in this story recently, at least in the view of some commentators, is Treasury Secretary Bessent. So three weeks ago he intervened to support the yen. Then he boasted about his power over the markets, saying that, you know, “I am the house now.” And then he tried to push down the long-term interest rate by buying long Treasurys with issuance of short-term debt. And basically this seems to have backfired. Paul Krugman, on his Substack, put out a chart recently under the unflattering title The Beclowning of Scott Bessent. How far do you think that critique is fair?

POSEN: Let me—first off, Sebastian, thank you for having me back with distinguished colleagues.

Second, let me focus on the policy and not on the adjectives of individuals. (Laughter.) It is ironic that someone who took part in the so-called breaking of the Bank of England in 1992 suddenly believes that governments can go totally against fundamentals with small interventions. (Laughter.) He’d be a lot poorer now than he is if he hadn’t had that experience. And I say that not to score a point, but to make a point that, you know, for all its problems the U.S. Treasury market is the world’s largest, deepest, most liquid entity that has ever existed, in financial terms.

And the idea that you are not—you’ve passed a budget bill a year and a half ago that, as Natasha can talk about, is not good for the long-term fiscal sustainability of the economy. You’ve got no prospect of having a decent budget process between now and the 2028 elections, whatever happens in the November midterms. You have chosen to—and this is not just on the current Treasury secretary, but past Treasury secretaries—the U.S. government, has repeatedly passed up the chance to do a huge amount of financing at long-term when rates were extremely low, and now you’re putting yourself at huge rollover risk by financing short term, which is generally not a great idea.

This is a fool’s errand. If you’re a tiny, small, open economy, you get more of an excuse for this kind of intervention for two reasons. First is, the market in your government bonds and your currency, whatever, is smaller, so if you somehow can get enough money together you can have a bigger impact, and also through regulatory measures can have a bigger impact. But also you have less choice, because you’re more subject to the whims of capital markets and the exigencies of the things Rebecca was talking about.

Ultimately, he’s not the house. You know how HGTV discovered several years ago that you could get males—American males to watch the design shows if you spend a quarter of the show doing demo? (Laughter.) You can always be credible doing demo. You can always be credible doing demo, flex your arms doing demo. In other words, if your goal is to drive down the value of the U.S. dollar, if your goal is to drive people out of U.S. Treasury markets, you can always achieve that.

MALLABY: So just—the demo is demolition. It’s not demonstration.

POSEN: Right. Sorry. Demolition.

MALLABY: I was a bit confused for a second.

POSEN: Sorry, sorry, sorry. You don’t watch enough HGTV. (Laughter.) Demolitio, not demonstration. But I mean—I mean, the point is, it’s asymmetric. If he wants—if the Treasury secretary, he or she, wants to drive down the value of the dollar against the yen, or whatever, they can do it. If they want to shove everything into short-term rollover bonds at some risk, they can do it. But if the goal is to minimize the long-term funding cost of the U.S., or the goal is to strengthen the dollar when it’s under attack, you can’t do that. That needs fundamentals. These kinds of tactics won’t work.

MALLABY: Right. Right, right. So, Natasha, I think Adam referenced this, but, you know, the structural issue, in terms of worries about the bond market, is obviously the fiscal outlook. Clearly the government debt is growing. But then those of us who think about this just intermittently and superficially, well, it’s always been growing. Explain to us what it is about now that explains why the market is suddenly worried.

SARIN: It’s a sort of an eternally hard question. And I want to actually start by piggybacking on versions of what Rebecca and Adam have already described. In that—and I think you can kind of think about what we’re seeing with respect to the Treasury market at the moment, is I can tell you a pretty optimistic story about what we’re observing. In that, what is fundamentally driving the fact that yields are rising is the fact that we have massive AI capital expenditures that are competing for dollars. And given these very valuable private investment opportunities, the government just has to pay more to people if it wants people to hand them their dollars instead. And, in fact, this is, like, the Kevin Warsh thesis of what we are observing at the moment. It is actually a story of fundamentals. It is a positive story of fundamentals and of the fundamental dynamism and growth that we’re about to see in the American economy.

The question is, how much of what we are observing is that, which is, like, good, relative to how much of what we are observing is that the world, and investors in particular, are finally waking up to the fact that, for a host of reasons, like, kind of, like, our credibility is on the line. And that has to do with the fact that, you know, we haven’t been at the 2 percent inflation target since 2021. There have been real questions raised by this administration about the nature of central bank independence and the extent to which the Federal Reserve is going to be focused on its sort of mandate about employment and about inflation. There are really challenging fiscal dynamics that we are making worse, not better, as we do things like last year’s reconciliation package, which added $3.4 trillion to primary deficits, and if you include the cost of interest over $4 trillion to the nature of our fiscal challenges.

And, by the way, the next president is going to inherit the expiration of the Social Security and Medicare Trust Funds. And what’s really interesting about that, is that the Social Security situation is kind of, like, baked into CBO’s estimates of the budget trajectory, because statute requires it. So what statute requires CBO do is it requires that it say that we are going to continue to pay out beneficiaries and make them whole, even though there are no dollars to do that once we get past 2032. And I think you should think about that not as a statement about the likelihood that the deficit picture is actually going to look better, because we’re going to see across the board benefit cuts in the range of 28 percent, but actually that the most likely outcome at this moment is that the way in which we are going to continue to make good on promises we’ve made to Social Security beneficiaries is by rolling over into general revenues, which is going to make our deficit picture worse, not better.

And so all of that has me kind of trying to disentangle the extent to which what we are observing today is a, you know, fundamental shift in the nature of whether or not the U.S. is going to be able to find its way to fiscal sustainability over any horizon that is meaningful, or whether it’s this sort of more positive story, or whether if what I speculate is that it’s actually both of the things happening at the same time. And I think what that means, frankly, for those of us who are really interested in trying to advocate for better economic policy, is that we all should hope that—for a long time people said, you know, what the U.S. needs is like a mini fiscal crisis, because you want to create the sort of exigency to actually do something from the policy perspective, and actually address the fact both with a combination of thinking seriously about how we spend dollars but also about thinking seriously about how we raise them. But there just hasn’t been the political will in recent history to do much of that.

And maybe if you get the sort of bond market doing what it’s supposed to do, which is discipline policymaking and encourage them to think seriously about these problems, that’s, like, for the best. And in that sense, the Bessent stuff is actually kind of, again, a little optimistic, because it shows that the bond market still is able to play this disciplining device. They feel like they have to do something. These, like, tiny interventions on the margins with respect to buybacks or what they’re doing in Japan, that shows me that it matters and that the yields can actually drive action in a way that I worry everything else can’t.

MALLABY: Go ahead, Adam.

POSEN: I basically agree with Natasha that it’s both. But I want to broaden the focus a little bit beyond the U.S., because what I think is important is that it’s not just the U.S. Japan, the U.K. France in particular, Germany, Canada are all having major fiscal issues, to varying degrees. We’re in the odd situation where in the G-7 I think the country with the largest primary surplus is Italy, by a large measure. (Laughter.) And that tells you something. And so part of this, in line with what both Natasha and Rebecca have said, is we are competing for funds. Each government is competing for funds in a global market. And if returns on capital invested in AI are going up, you have to compete more for funds.

But additionally, it’s that we’re seeing less flows across countries, particularly from Asia into these—into the bond markets in the U.S. and in the West, particularly from China. We’re seeing demographics, obviously, playing a role, which Natasha, and colleagues of mine at the Peterson Institute, and others have been talking for a long time. But the part I want to emphasize is, in terms of Natasha talking about a helpful small crisis, we’ve got an additional aspect here. Which is, let’s go back to the euro crisis a dozen years ago, or fifteen years ago. When it was just, quote/unquote, Greece and Spain and Portugal and Ireland, but not Germany, U.S., whatever, the bond market vigilantes could go whole hog after those countries.

PATTERSON: Whole hog? That was good.

POSEN: Sorry?

PATTERSON: Well, the pigs crisis.

POSEN: Ah, right. Sorry, I didn’t think of that. (Laughter.) You’re better than I am, subliminal. They could go after these, and you could really have an effect. But when it’s six or seven of the G-7 simultaneously, and people don’t want to put their money into China for a bunch of reasons, there’s no place to go. So, in a sense, you’re going to have the higher interest rates, but you may not get the same bond vigilante benefits as if it was one country getting targeted, or a set of smaller countries getting targeted. They’re just—as Rebecca was saying in a meeting earlier, there’s only so many Singaporean and Norwegian bonds you can buy. (Laughter.) And so this may, perversely, lengthen the time till the sort of useful crisis occurs, because there’s no place for the money to get out.

MALLABY: Rebecca, the British thirty-year gilt is the highest since 1998. When I looked yesterday, yields on ten-year JGBs higher since 1996. So, you know, as Adam rightly says, this is something in all of the big markets. Do you want to pick up on, that and talk about this international dimension and where you see the fragility?

PATTERSON: Sure. Sure. And I’m just delighted—we just had Adam on our Spillover podcast that CFR kindly produces. That’ll be coming out shortly.

POSEN: Yay.

PATTERSON: And Natasha and I had fun a few weeks ago talking about all things fiscal. And we started our podcast back in February, if I remember. And one of our first, maybe the first episode, was what we called The Fragile Four: the U.S., U.K., France, and Italy. And we were worried that there were some commonalities in these countries, populations, rightly or wrongly, who increasingly expect the governments will come with some sort of help when there are affordability concerns, or a pandemic, or some shock. And because of various political issues in the different countries, the politicians don’t want to disappoint.

And we’ve seen in France—I mean, I actually had to go get Claude to double check for me because I’ve lost track. France, if you count—one person was in twice, so we’ll only count that person once—they’ve had five prime ministers in five years. Because every time they say, well, actually, we need to bring the budget deficit down because we’re out of the growth and stability pact guidelines, you know, we need—you all are living longer. Like, we can start your pensions a year later. No, no, no, out. You’re gone. And I feel like a version of that is happening in all of these countries. It becomes politically very difficult. It’s almost political suicide to embrace fiscal austerity.

And so the problem is continuing in all of these places. Again, slight differences, but a lot of commonality. And when I look at Europe, you know, we got data just this week that you can see the extent of the energy crisis there. So the headline inflation number was 3.3 percent in August. That was the highest since 2024. But core was down—core inflation, which takes out energy and food, was down to 2.4 percent. So that’s a very big gap. And a lot of that is energy. And that’s tied to the war in Iran and the continuing war in Ukraine. The ECB is raising rates. They raised rates twice this summer. So when you look at what’s happening in their bond yields, some of this is feeding into inflation and expectations that the central banks will have to keep raising rates. Because what happens with the policy rate affects the entire yield curve? Some of it is the supply, the fiscal dynamics that we talked about.

And, to me, France is the one to keep an eye on, which is interesting. I spent some years earlier in my life living in France and Italy. And I would always pick on Italy. I love Italy, but I would pick on it. (Laughter.) Now I have to pick on France. You know, aside from the political instability and the credibility issues—and you do have an election coming up there next April. And right now if you were to look at the polls, you would think that the most likely outcome will be more fiscal spending from the parties that right now, at least, are in power. But what’s interesting there is that the French ten-year yield is higher than the Italian yield, and has been now for a while. And it’s significantly higher than its German ten-year equivalent, the German Bund. So you are seeing a discount, if you will, on French bonds, tied to supply and political concerns.

And what I worry about with France is that, unlike the U.K., Japan, or the U.S., they don’t print their own currency, right? They have the euro area. They’re not in charge of that. So they can’t simply inflate their way out of this mess. So that’s—France is probably—I worry about France a lot, in terms of where our mini crisis could come from. And importantly too, remember these markets are all correlated. Most investors today own global bond portfolios, whereas ten or twenty years ago they tended to be very home biased. And that means if you have a crisis in one, you can have contagion very quickly to the others. So a crisis in France doesn’t stay in France.

And then in terms of the U.K., again, not dissimilar. A lot of different leaders in recent years. I might not have the number right. You’re going to get a new budget from this government on October 28. They’re talking about some major welfare reform. They are talking about a few tax hikes, but they’re also talking about more spending. So we’ll see where that goes. And the Bank of England is trying to figure out how to navigate this. They held rates steady at their last meeting. They’re probably going to hike before year end. If they hike, that pushes up yields. But the issue there is the supply, et cetera.

And the Bank of England just today came out and said, we’re really worried about the hedge fund exposure in in our market, and how that could cause volatility and contagion if positions get unwound. And they’re still—they’re still having PTSD from 2022, when a fiscally loose budget proposal caused a crisis for about a month that forced Bank of England intervention to calm down the market. And it did create contagion, including to the United States. So the U.K., to me, the bigger worry there is the market is just big enough to attract a lot of speculative investors. And those investors, if things start smelling bad, could get out quickly, and that could cause a bit of a tidal wave into all these other markets. So Europe is not a place to hide right now.

MALLABY: I love the conclusion on the bad-smelling investors. I’m going to ask one more question for Natasha, and then I’m going to open it up. So please get your questions ready.

Maybe one thing you could talk a bit more, Natasha, about is AI. So you mentioned earlier that the huge demand for capex generated by AI is one of the background factors to rising rates. I think that’s worth double-clicking on because there’s a story out there that says, look, AI will drive better productivity. That will drive better growth. So our debt to GDP, maybe we can just grow out of it with faster GDP growth. But what the bond markets are telling us is what AI giveth fiscally, it also taketh away, right? That you may have a faster growing economy, but your debt stock is going to be more expensive to service. And given the size of that debt stock, that’s a problem.

So talk a bit about the plausibility. And, by the way, the other thing is, of course, the tax base in most advanced economies, and especially the U.S., very geared towards labor. And so if AI starts contributing more to GDP, and therefore labor contributes less, we’re not taxing the machines that are generating the output. And that’s a problem, too. So give us your view on how AI is going to impact the medium-term, long-term fiscal outlook.

SARIN: I liked when I got to be optimistic. And now I have to be pessimistic. (Laughs.) In that, it’s actually—and I have to sort of highlight the work of my colleagues at the Budget Lab, because we’ve been asking ourselves this question about, can you get enough of a productivity boost from AI that you relax some of the fiscal challenges and find your way closer to fiscal sustainability? And in some sense, the answer might be at least partially, yes. In that, if you take the median economist’s forecast of productivity growth that they expect over the course of the next five to seven years, they expect labor productivity growth in the range of 2 ½ to 3 percent And, by the way, that’s much lower than what you hear from technologists or the people running these labs who are saying kind of crazy things like 10 percent or 30 percent or whatever it is.

But even at this 2 ½-3 percent, that’s very substantial of an uptick relative to where we’ve been over the course of the last decade, which is in the maybe high ones. And what that means is that you are going to get a boost to growth from the technology. But what is also happening simultaneously is that, again, the median economist forecast, is that you’re going to see labor force participation rates decline by between 2 to 5 percent. And, again, if you talk to people at leading labs, they’ll tell you those numbers are way low, but they at least give you a sense of where the economic consensus is on some of these dimensions.

The challenge, as Sebastian is highlighting, that the bond market seems to realize—which I think all of us, as we start to grapple with the question can also realize—is if what is happening is that you are getting massive productivity growth that comes at the heels of potential massive disruptions to the labor force, that is going to actually put pressure on the need to find solutions geared—that cost money. So things like—that politicians are talking about right now, things like universal basic income, or massive job retraining programs, or whatever your solution is, is going to be something that is going to take away from the potential growth benefit that comes from the fact that the economy is just larger, because you also have to invest more fiscally in order to be able to support the various sort of segments of the economy that are potentially struggling with the transition to this new world.

And also, you are doing so at a moment—and it’s also the case that—this is also related to what Rebecca was saying. If you look at inflation expectations, they are somewhat—actually, if you look at inflation expectations, but also if you look at the nature of what is happening in the bond market right now, you can look at what’s happening to ten-year yields, but you can also kind of break those into two pieces and look at what’s happening to the five-year and then the five-year five years out. And what you see is that the five-year is quite elevated, and the five-year five years out is not. And the reason for that is that the market kind of expects that the Warsh Fed is going to do what it appears that they have committed, and that the chair kind of committed to do, which is to be laser focused on getting inflation back to expectations.

What that means is it is going to mean a higher interest rate environment in the near term. And what that means is that servicing our existing stock of debt is going to be expensive. And that is kind of putting a fair bit of pressure fiscally on a situation that already seems like it might have a need for another type of investment in order to deal with what’s happening in the labor market. And thirdly, as Sebastian already started to describe, we are quite good at taxing labor in this country, and, like, pretty bad at taxing capital. And we’ve gotten worse, frankly, over time, because of a lot of tax preferences. There’s a great new book by Owen Zidar and Eric Zwick that everyone should read called The Everywhere Millionaire, which talks a lot about many things, but one of them is the rise of passthroughs, and the ways in which we’re not really doing that good of a job at taxing the massive stock of wealth that’s accumulating in the form of capital income in the economy.

If what AI is doing is it is shifting the composition—it’s achieving this productivity boon by shifting the composition of the economy more in the direction of capital income and further from labor income, then actually the boon to the fisc is less than a lot of our models at Budget Lab suggest, or basic models suggest, we’ve now added in this component. Because you’re expecting to be able to reap the types of rewards that you’re reaping today from a structure of an economy that’s heavily reliant on labor. And as we transition away from that, that actually has consequences fiscally.

My sort of optimistic hope that I will leave you with is there is a lot of stuff—and we can debate the nature of whether or not we should do wealth taxes or the mark-to-market taxation of capital gains—but there is a lot of what I call, like, plain vanilla capital tax reform, which—and corporate tax reform—which is in the direction of having a slightly higher corporate tax rate for multinationals, getting rid of stepped-up basis, getting rid of the 199A deduction for passthroughs. Just, like, stuff that’s been out in the policy ether for quite some time, and has some—certainly it’s what I think those on the left would—at least some of them—would like to see happen in a 2029 tax bill.

But I also think they’re not particularly—if you look at what happened in the first Trump tax package, Business Roundtable, which is this collection of sort of CEOs of leading companies, what they wanted was a corporate tax rate at 25 percent, and what they got was a corporate tax rate of 21. So I think there is, like, space to do something here. And I actually think that—my hope for AI, from, like, a tax perspective, is that it almost gives you the capacity to call this broad suite of things that we should have done a long time ago, and definitely should do now as we shift towards a more capital-dependent economy, we call that our AI tax plan. And that’s a way that it actually gets over the finish line when it hasn’t been able to in times past.

MALLABY: OK. So let’s open it up to members to join the conversation. Remember, it’s on the record. I’ll go to Tara in the front first.

Q: Thank you so much. My name is Tara Hariharan. I’m co-CIO of NWI, a hedge fund here in New York.

Based on everything our esteemed panelists have said, I would love them to further comment on where you think the Sell America trade will go, the one that everyone was getting excited about during the liberation day tariffs last year. The concern went away. But now if you think about it, the three legs of the buy America stool, I would argue two of them are crippled. If you think about bonds, we are if not seeing outright outflows from Treasurys, we’re certainly seeing concerns about Treasurys. The dollar, we are seeing foreigners now starting to increase their hedging ratios after earlier having found that it was not the best trade. But now, given the fiscal risks, they are increasing. And, frankly, the only leg of that three-legged stool that’s still working is the interest in U.S. equities, primarily based on AI.

But I’d also argue that the move ahead for the AI trade is entirely dependent on AI pricing power and enterprise take up. And if AI right now is the primary driver of growth, capex, et cetera in the economy, and everything else is falling behind, to expect that the rest of the economy is going to provide that strong demand for AI might be a little bit of wishful thinking. So would love to hear the views from the panelists. And to press further on, you know, the outlook for long yields, I appreciate Natasha’s optimism about ways we could find fiscal solutions, but right now it doesn’t seem like that’s going to be an issue. It seems like the hyperscalers are going to continue to issue at the long end, and they’re fine with higher and higher yields.

MALLABY: OK, let’s focus on the question about where does—where does the outflow go, if there is an outflow, if we buy the premise. Do you want to start, Rebecca?

PATTERSON: Sure. You know, there’s still a tremendous amount of money coming into the U.S., despite the worries around the U.S. from a variety of sources. And part of it is growth, right? The U.S. economy is continuing to be very robust, despite a number of shocks hitting it. We can talk about some of the dynamics of the labor market under the hood, but the headline unemployment rate certainly is low. Weekly jobless claims have fallen since the end of last year, and now are quite low. So people generally have incomes. Their confidence is destroyed. Consumer confidence surveys are pretty horrible. But they’re still spending. They’re dipping in their savings more, and they’re using credit cards more, but they’re still spending. And so that’s supporting earnings.

And so even though you have higher borrowing costs for these government bond yields, and competition from corporate bonds, you aren’t seeing that as a huge detriment to equities yet. Yet, I would say. The question in my mind is, when—you know I’m going to use a very, very well known, very technical term from the Ghostbusters movie. When do the streams cross? (Laughter.) You know, at what point are yields so high that the borrowing costs become a dominant factor over their earnings? And, to me, that’s something that makes me want to continue watching the consumer confidence and consumer behaviors very, very closely. And so if the war goes on longer and the Fed maybe has to raise rates more because of passthrough from energy prices into broader inflation measures, that could be a source of risk. If there’s some disappointment within the AI ecosystem, that could be a source of risk.

But if you have the borrowing costs becoming a dominant worry and something becomes a little squishy within the AI ecosystem, either of those things could cause this Jenga tower to fall down. And what I worry, going back to the points we were talking about before, I think you were making this beautifully, Adam, where do you go, right? So if money leaves the U.S., everyone said, well, we’re fine. We’re globally diversified. I have Korea. I have Taiwan. I have Latin America. Well, guess what? They’re all AI trades. Either they’re making the inputs for AI, or they’re making the commodities that fuel the AI. So diversification truly, where you go to be safe right now—aside from cash, which is being inflated away a little—it’s a hard one. And the joke about Singapore bonds is not a joke. It’s a very small market because they’ve had budget surpluses for a very long time. I remember dealing with this in 2008, and again in 2012 as an investor. Maybe we go back to gold.

MALLABY: Yeah.

POSEN: Or, we are going back to gold.

PATTERSON: Yes.

MALLABY: OK. There was a question in the front. Yes, yeah, right here.

Q: The panel has presented a pretty gloomy picture.

MALLABY: Could you identify yourself, sir?

Q: My name is Mark Rosen. I’m from Advection Growth Capital.

A gloomy picture of the economy, particularly even the U.S. economy—although you mentioned that there’s—growth is still relatively robust. You’re the only member of the panel who’s said anything positive today. I mean, are there not a group of economists—Aghion, the Nobel Prize-winning economist last year, Autor from MIT, Cochrane from Stanford, and a number of others—who’ve said that, in fact, AI is going to produce tremendous productivity growth, that it’s not going to impact the labor market, and that we predicted since the advent of the printing press that technology was going to destroy jobs, and that we were wrong. And this time is no different. So I think you said that there was a consensus among economists that the labor market was going to be negatively impacted. I would say, is that right, given what these other economists are saying that directly contradicts what you and others have said?

POSEN: Can I respond?

MALLABY: Sure. Natasha, do you want to go first, and then we’ll come to that?

SARIN: Yeah, I should say that that I am not a pessimist on the labor force impacts of AI, in that I think if you listen to Dario Amodei, for example, he was giving versions of you’re going to see 50 percent displacement of white collar work over some horizon that’s very short. I think that is very unlikely. I also think that in moments of technological change, and this is the—this is related to the Autor stuff—in moments of technological change historically, take the Industrial Revolution, what you saw is that, in fact, new jobs did appear. And I suspect again—and maybe this makes me a sort of overly optimistic on AI because some people think it’s going to be the end of the industrial age—but I suspect, again, new jobs will appear.

The challenge is that in the transition—and we know in the U.S. we don’t actually manage these transitions particularly well—what you are going to see is you are going to see particular types of people or particular types of communities potentially displaced by the fact that as the—there is no way to accomplish the productivity growth without it being sort of related to the fact that what AI is enabling is the automation of certain types of tasks. And so that is going to come on some horizon with a labor force effect. And I speculate that is going to require fiscal investment. That does not make me pessimistic about the technology. I’m hugely optimistic about it. And I’m very pleased, in fact, that we are the house and the home of the leading developers of it.

POSEN: I’m going to give you a slightly different perspective, but first I basically agree with Natasha. I made a slight face when you listed those three very distinguished economists because their reasons for optimism are completely orthogonal to each other. John Cochrane has a blind faith in the ability of markets to reallocate. Philippe Aghion has a very fundamental insight into creative destruction. And David Autor has an empirically based insight that overlaps with what Natasha said about how labor markets respond. So if you just want to select on who’s positive, you can put those three together, but they’re actually three very different assessments of what’s at work.

And so when we talk about labor, and we just held an event at the Peterson Institute a couple days ago on this with Luis Garicano and Anna Stansbury, two other not-yet Nobel Prize but quite good economists—you might want to look at it—we discussed the idea of Garicano’s concept of “messy jobs,” which is a very profound insight that you’re probably not going to see anywhere near as much job destruction as Amodei, or some of the other AI CEOs have said, because most jobs are not single tasks. They’re embedded in networks of specified knowledge, of relationships, of management, of particular skills and assessments. And so there will come a point for certain tasks like, simple coding or point-to-point trucking with automated trucking, t jobs will go away. But these enormous lists of AI-exposed jobs are probably wrong, because they overemphasize the idea that a job is simply a task. And most jobs—not all—most jobs are much more bundles of tasks and relationships.

So, and then Anna Stansbury makes the point, and other economic historians are making this point—there’s a brilliant paper—God, I forget her name. There’s a new postdoc had a paper out going back to the Industrial Revolution, looking at bootmakers in nineteenth-century U.K. That was essentially automated overnight. You didn’t see mass displacement. What you saw was very few new people, young people, went into boot making. They went into other jobs. And the industry slowly declined. And so Anna Stansbury makes the point about even if we can make it so that most of the displacement is—as we’re seeing so far in the data, that fewer young people get hired in certain places, but it’s not we’re displacing in en masse in Ohio a bunch of fifty-five year olds. That the labor market disruption is less.

So I think you’re leaping ahead by saying, oh, there are these optimists out there in the labor market. You can be quite robust on the labor market and still be more pessimistic. And I’ll give you three reasons why.

First one we already mentioned is if you’re competing up the cost of capital by the returns on AI, you’re going to have all kinds of distortionary effects. It may be efficient, but you’re going to have all kinds of interesting effects where there’s one sector of the economy that’s getting very high returns and is driving up the cost of capital for everything else. And we’ve seen that before in certain environments. And that doesn’t necessarily lead to stable or sustainable growth.

The second reason I would emphasize that you’re pointing, and you said about U.S. growth, is that I think people vastly overestimate the extent to which current leads and cutting-edge AGI attempts will confer lasting competitive advantage on U.S. versus, say, Chinese or foreign producers. Not just of AI straight up, but of robotics and applications. And if the U.S. politically continues to Galapagos-ize itself, like it’s doing with EVs, you may end up with a world where Chinese open models or European hybrid models with GDPR protections get much more adopted around the world than whatever the U.S. firms are doing. And you got to have that in a way.

The third reason I would be pessimistic, at least not just map from a positive view of U.S. AI productivity to wonderful outcomes, is political economy. We already had a technology to replace radiologists. It was remotely hire radiologists in India and Nigeria. And politics completely killed that. Do not underestimate the ability of people who are very scared in democratic societies to resist technological change. So we don’t have to be unrealistic pessimistics about the labor market to not be glowing about AI solving all our problems.

MALLABY: Another question? Yes, over here on the aisle.

Q: Thanks. Robyn Meredith from Morgan Stanley.

Just in light of the points that you made on the economy moving more towards a capital-dependent economy versus labor-dependent, and also the point you made earlier about Social Security needing to reset, and that being inflation-indexed, therefore the cost of it will go up every year, does that sort of limit the ability of us to simply inflate our way out of the problems we’ve all been describing? If so, how? Or is it in fact the inevitable solution to all of these things we’ve been talking about with growing deficit, et cetera?

SARIN: So I don’t think inflating your way out of debt is a great strategy in general. (Laughs.) Unsurprising. But I am actually, I think, more optimistic than you are on the nature of where fiscal issues are going to situate themselves in our political discourse. And maybe that puts me at odds some with Adam, who sounds pessimistic on politics and political economy at the moment, though maybe not because maybe on this issue we agree. I think what the expiration of the Social Security Trust Fund in 2032 is going to do is it is going to force every candidate who is running for president, on both sides of the aisle, to have a strategy, or at least acknowledge the existence of the need to think seriously about entitlements.

And by the way, part of the challenge of DOGE—which you all remember Elon was trumpeting all of these, like, we’re going to cut deficits by this, or we’re going to achieve hundreds of billions of dollars of budgetary cuts—is that, like, employees of the federal government are not actually that substantial of a line item, from the perspective of what the government spends money on. These sort of mandatory spending items, like Social Security and Medicare and Medicaid, they really are. And so the whole ball game with respect to the spending side of this conversation is around entitlements. And we are going to be forced to engage with the question of what to do about entitlements in just a couple of years. And you’re already starting to see, like, bipartisan sort of working groups forming around ideas to think about fiscal sustainability and deficit reduction. And so I am just a little hopeful that the mini crisis might be on the horizon, and it might come in the form of the fact that the trust fund is about to expire.

PATTERSON: I mean, one thing we know is in an age of social media where content is generated by the bucket loads—not saying it’s useful content—but this will be a headline. This will be in everyone’s feed every single day. And what shapes voters’ views becomes a political campaign becomes policy. They’re linked. So I would count on that.

POSEN: I guess I’m going to be briefly more pessimistic. I mean, think of the description that Rebecca gave earlier of the French situation. My colleague Olivier Blanchard from Peterson now spends basically all his time on some combination of academic work, policy advising, and social media, trying to say we’re hitting the wall in France. And they are hitting the wall in France. And the politics are not moving forward. It’s nice—I hope Natasha’s right. I don’t view it as that way.

Just one more point, and Sebastian’s heard me make this point in past meetings. Twenty years ago I pointed out that the U.S. had a dynamic for a lot of years, where roughly every ten years or so, sometimes twelve, there would be a sense of mini crisis, like the Social Security Commission in the eighties, where a bipartisan presidential commission or a bipartisan Congress would do something to get the train back on the tracks for ten years or so. Wouldn’t fix the whole problem permanently, but make a material course change. Later, Alan Auerbach, who’s a superb professor at UC Berkeley, econometrically showed in a Brookings paper that that was the pattern.

That pattern broke around 2005. We’re now twenty years in since the last time this happened. And again, I hope this is right. I want us to lobby for this being right. We have to act as though our democracy is capable of making these decisions. But I don’t think we should totally discount the experience of the last twenty years being different from the experience of the previous forty or fifty years in terms of the self-correcting nature of—

MALLABY: I believe that 2005 was the year in which Facebook started, but I presume that’s not the full explanation. What does Auerbach say is the reason for the—

POSEN: I don’t—

MALLABY: Or does he not—

POSEN: I think he speculates. He doesn’t go there. Natasha may remember the paper better than I do. But I think—and this goes to something you and I and Rebecca were talking about before this meeting—I think history unfolds, right? So you have the 2008 financial crisis. Nobody’s going to do fiscal austerity then. Then you have backlash from that, and the countries that did austerity got into trouble. Then you have COVID. Then you have political polarization. So, sequence of events.

SARIN: Can I just be optimistic for one second, because I know Sebastian wants to do more questions. But part of what I wonder is if we are approaching a moment, this is way outside my field, but of generational change in our politics, and if that is going to be a driving force behind some of the—sort of, my hopeful version of this. In that we spend, like, $9 on every senior for every $1 we spend on kids in this country. People are increasingly—a big chunk of what people are telling you when they say they’re worried about affordability is they’re worried about big things like the cost of housing, or the cost of health care, or the cost of childcare—where even in low-income neighborhoods in Manhattan it costs over $2,000 a month to be able to get full-time childcare for your kids. And that is why people are increasingly not having kids.

So I wonder if part of what we are going to see in our politics is—I’ve been thinking about, like, calling it, like, farewell to, like, the OK Boomer economy, and a bit of a push towards thinking seriously about the fact that investing in one generation inherently means taking from younger generations.

MALLABY: Interesting. Joe, in the front here.

POSEN: The self-selection of the front row people being the assertive people seems to be working. So this is good. (Laughter.)

Q: Joe Gasparro, Royal Bank of Canada. Thank you again for this incredible panel.

PATTERSON: The mic’s not there. I think that—

Q: All right. Joe Gasparro, Royal Bank of Canada. Thank you again for this incredible panel.

Natasha, you bring up the great point that the U.S. is better at taxing labor than capital. What’s the cleanest way to increase our tax revenues without discouraging capital or the investment that’s actually responsible for generating all the productivity from AI? Thank you.

SARIN: I’m going to be super fast and give you, like, three things, OK? One is sort of table stakes, which is one of the things that I think our tax system should just do better is collect taxes that are owed already. And we don’t have to do any law changes for that. But we are currently missing out on 3 percent of GDP annually. That’s about $800 billion in tax revenue that would otherwise be paid, but isn’t for a combination of people make mistakes and corporations make mistakes, and also there is just a lot of tax evasion. We should invest in the IRS to deal with that. And that’s going to become even more important in a world with AI, because AI-enabled tax evasion—finding every loophole and pushing to the edge of what is legal and past that—is just going to be the norm.

The second piece of this is, I think, thinking seriously about—I’ve seen a bunch of policymakers who are keen for ideas about how do we tax the labs, or how do we tax the tokens, or how do we tax stuff that is in the AI space? And my point to them, that I don’t super succeed on because I’m no good at messaging—but I’m hopefully OK at economic policy—is that, like, what you actually want to do is tax the winners in the economy who are benefiting and profiting from this technological moment. And we already have a tool that does that in the corporate tax. And we can keep corporate tax rates what they are today for 98 percent of corporations, and just raise them for the largest 2 percent just a little, and we are going to generate a ton of revenue—trillions of dollars over the course of the next decade. If you do a bunch of sort of—if we also deal with the multinational tax challenges and have some sort of global minimum tax rate, like the Biden administration was pushing for.

The third thing I would do is just think about undoing some of the preferences that exist today for capital income relative to labor income. And some of that is, like, you know, why should it be that Mark Zuckerberg is going to be able to pass his entire accumulated wealth, which is disproportionately in his shares of Meta, to his kids, essentially untaxed? And we have, again, the tools that we would need to tighten up to do something about that. We might want a whole new system to do something about that, like inheritance taxes. I’m not entirely sure. But dealing with that problem seriously and doing small things. Raise the capital gains tax rate a little bit. The idea that Mark Zuckerberg was sitting in his Harvard dorm room and would have been deterred from starting Meta if the capital gains tax rate was slightly closer to the ordinary income tax rate strikes me as, like, categorically absurd. (Laughter.)

MALLABY: Do we have one more question? OK, quick, last one.

POSEN: Oh, second row.

Q: Yeah. It’s Bruce Churchill, representing the slackers in the back. (Laughter.)

Like, no one’s mentioned China? Is China irrelevant? Less important. What’s up?

MALLABY: Doesn’t have a bond market meltdown.

PATTERSON: Well, China has—it has gone in a completely different direction. They just cut interest rates. Their bond yields are around 1.7 percent or so. The challenge owning a lot of Chinese bonds is that you’re taking some geopolitical risk and you’re taking some liquidity risk. And that could change in a day, right? And so that isn’t your safe haven. China and the U.S. are still, as they were, interestingly, twenty years ago, 2006—global imbalances, when we think about current accounts and we think about savings and investment, it peaked in 2006, driven primarily by China and the U.S. As of today, we are about two-thirds back to that peak. And guess who’s driving it? China and the U.S. Those are the similarities. Everything else has changed.

China today is exporting its excess capacity. Consumers aren’t spending. That’s why the bond yields are so low. And they’re just saying, OK, we’re going to grow our way out of this through exports. And that has a limit to how far that can go. The U.S. has moved its issues from the private sector—2006, we had a lot of consumer leverage and corporate leverage. Today, we have government leverage. The big difference today versus 2006 is that the capital has changed. The market structure has changed. You know, in the past China was putting a lot of money into U.S. Treasurys. Japan was putting a lot of money in U.S. Treasurys.

Today foreigners are putting money in U.S. equities, but not necessarily in Treasurys. And the capital isn’t coming from sticky central banks that are price insensitive. It’s coming from shorter-term players like hedge funds, not to pick on hedge funds, but that capital can move much faster. And so, China is struggling economically, but it’s still contributing, just in new ways, to imbalances. And those imbalances today don’t necessitate a crisis, but it creates a higher level of vulnerability that if one of these things breaks, whether it’s the French bond or something in the AI ecosystem or whatever, you know, it could be more than a mini crisis.

MALLABY: OK. Well, with that cheerful note, we should wrap up. (Laughter.) Thank you for coming. Thank you to Rebecca, Natasha, and Adam. (Applause.)

(END)

This is an uncorrected transcript.