Gita Gopinath has spent the last several years inside the room where the world’s monetary plumbing gets fixed — or doesn’t. As first deputy managing director of the IMF and, before that, the Fund’s chief economist, she worked on sovereign debt restructuring, followed Argentina’s disinflation up close, and built her academic reputation on ideas like dollar dominance and the dominant currency paradigm. Now back at Harvard launching a new Global Economics Lab, her conversation with Tyler ranges from why exchange rates don’t adjust the way models predict, to what stablecoins really cost, to whether AI’s growth dividend will show up on schedule.
Tyler and Gita discuss why trade balances are only weakly related to real exchange rates, before turning to Argentina, Milei’s fiscal discipline, the peso, and dollarization. Then they discuss whether economists got the euro wrong; whether trade imbalances with China are really the problem; stablecoins, CBDCs, and the Treasury market; and whether AI will ease the debt burden, cause stagflation, or transform economics itself. Finally, they turn to her family’s Kerala roots: why the state keeps electing communists, the relationship between matriarchy and social indicators, single-sex education, and what she’s building next.
Watch the full conversation
Recorded August 27th, 2026.
TYLER COWEN: Hello, everyone, and welcome back to Conversations with Tyler. Today, I’m honored to be speaking with Gita Gopinath, who is professor of economics at Harvard, formerly first deputy director at the International Monetary Fund, which is number two in that institution; formerly chief economist at the International Monetary Fund; and one of the leading figures in international trade and finance and economics more generally. Gita, welcome.
GITA GOPINATH: Thank you, Tyler. Thanks so much for having me on your show.
COWEN: Today I have only easy questions. We’re economists, right? Relative prices matter, but it seems that trade balances are fairly weakly correlated with real exchange rates. How could you explain that?
GOPINATH: If you look at just what goes into a trade balance equation, it depends upon the relative price of your goods in international markets, but it also depends upon the relative level of demand in the two different countries for just overall consumption. For example, if in the US consumption is strong, while in China consumption is weak, even if relative prices were to favor, for example, US goods, just the overall level of demand being lower in China would reduce the overall level of exports from the US to China. Both those variables matters, not just the relative price, but also the relative levels of demand in countries.
COWEN: Why doesn’t, at some point, the real exchange rate keep on adjusting so that the substitution effect finally has some importance relative to the income effect?
GOPINATH: There is this notion that, if you were to write an economic model, that our models would predict necessarily that a country that’s running a current account deficit, for instance, would experience a depreciation of their real exchange rate, and a country that’s running a current account surplus would experience an appreciation of their real exchange rate. I think that’s an incorrect statement. This is the formal statement. It’s partly correct, but here’s the accurate statement: The only thing the models tell you is, as you know, we have what is a no-Ponzi condition, which is some sort of an intertemporal budget constraint.
What it tells you is that a country that is, for example, running deficits and therefore accumulating a lot of liabilities to the rest of the world will have to repay those liabilities because we have a condition, or either way, you default; one of the two. How do you repay those liabilities? You repay those liabilities in the model, and the equation will tell you that, at some point, you have to start running trade surpluses to repay those liabilities.
You’re assuming you don’t have an exorbitant privilege in the sense that the debt you’re showing to the world is always not much cheaper than what you owe to the rest of the world in terms of interest payments, is not always that much cheaper than what the rest of the world owes from you. There are periods in US history when that’s been true, that’s less true now, and I don’t think we can assume that’s going forward.
A country that’s accumulating a lot of debt to the rest of the world, or liabilities to the rest of the world, for instance, our model will say, at some point, have to run trade balance surpluses. Now, there are two ways that those trade balance surpluses can come about. One is through a real exchange rate depreciation, so that is true; that could be one channel. The other channel is if the country’s overall level of demand were to collapse, then that would also mean that they would import less, relative to the rest of the world, whose demand is higher, and they would import more.
Either of those channels, or both of those, can be in play, and if you look at the evidence in the data and you try to do some kind of a decomposition of how much of it is real exchange rate versus demand, there is no clear mapping from the fact that if you’re a country that’s running current account deficits or running trade deficits for a long period of time, that you necessarily should see a weakening currency.
COWEN: That puzzles me all the more. If we think of Australia, which does not have exorbitant privilege, as you know, they ran trade deficits for decades steadily. There was no reason to think Australian demand would suddenly collapse. Australia’s done fine, has strong fundamentals, yet the intertemporal constraint on that process, it never seems pushed into the present. Why is the inter-temporal constraint so weak, and how should that shape how we think about how economies actually work? Is everything just flashed period by period, like in the very simple Keynesian models, or is it something else?
GOPINATH: Well, it is a constraint that says that if you have net foreign liabilities to the rest of the world, then in a present discounted sense, at some point you have to be running surpluses to be able to pay that off. That’s the only thing that that constraint says, and that can happen through a combination of things. It can happen either because you have a discovery of a resource that you then export to the rest of the world and therefore your exports start booming because of that, or you have a recession in your country, and imports collapse.
The real exchange rate does play a role. To be clear, I am not saying that the real exchange rate never comes into play; it does. You have seen countries whose incomes have grown over time, have had their real exchange rates appreciate; it just takes a long period of time. To assume that on a year-to-year basis that if a country is running a trade balance deficit, it should necessarily have a weaker real exchange rate, I think making that tight link would be problematic not just from the theory, but from the theory side and from the empirical side.
COWEN: Now when exchange rates move, as you well know and have written plenty about, there’s very often not that much pass-through to the prices of imports and exports. That seems to be another case where relative prices, at least superficially, are not mattering in the way we would expect them to. What’s your best explanation for that?
GOPINATH: There are important distinctions here. The question is: What relative prices are we talking about? There is the question of imports and exports, and their link to the relative price of exports and imports. There you absolutely do have a pass-through, and you do have a quantity response. Then when you go to the real exchange rate, which is the nominal exchange rate adjusted by the consumer price indices, those consumer price indices include a whole bunch of stuff that’s nontraded, like healthcare, and that is less sensitive to an exchange rate movement or to the terms of trade itself.
Therefore, the pass-through to overall consumption from the real exchange rate can be much more muted as compared to what we see in terms of imports and exports. Now, the work that I did on the dominant currency paradigm was to point out that, unlike the typical assumption that was made, which was the Mundell-Fleming, Keynesian, Friedman assumption, that the relative price of your imports to exports moves almost one-to-one with your nominal exchange rate, because the idea is that, “Okay, I’m the US; when I sell to China, I’m pricing my goods in dollars, and that dollar price is relatively sticky.”
When China sells to the US, it’s pricing goods in the renminbi, and that renminbi price is relatively sticky. Therefore, when the dollar-renminbi price moves around, exchange rate moves around, that causes the relative price of my exports relative to my imports, which is the terms of trade, to move very much closely with the nominal exchange rate. If you look at the way the world works, is that when the US sells to China, it prices its goods in dollars, and that dollar price is relatively stable.
When China sells its goods to the US, it also prices in dollars; close to 90 percent of China’s exports to the US are priced in dollars, and it’s relatively stable in dollars. When the renminbi-US exchange rate moves around, the relative price of the terms of trade doesn’t move around that much. That particular channel works differently from the standard theory.
Now, I want to be clear that there is still expenditure switching here because, unlike the Friedman hypothesis, what you do not have is that when the US dollar, for example, depreciates, that means it’s going to reduce imports from China because, in the world, the dollar price hasn’t changed. That doesn’t have much of an effect on imports coming into the US.
On the other hand, US’s exports do expand because it’s priced in dollars, and in terms of Chinese renminbi, the cost or the price of that good has gone up. You get like half of the channel, which is when the dollar depreciates, you don’t get the import channel, but you get the export channel. Flipped around from the China perspective, because of this asymmetry, because of the dominance of the dollar, when the renminbi depreciates, they don’t really get the big export kick, but they get the adjustment through imports.
COWEN: Say if I look at the Trump tariffs, which you’ve written on in Journal of Economic Perspectives, there seems to be almost complete pass-through that’s well described by a simple model of more or less perfect competition. When we look at exchange rate movements, it’s all of a sudden a different model, even though they’re both changes in relative prices. What’s the meta model that explains how those two facts fit together?
GOPINATH: Yes. The meta model is, if you look at the simplest story, which I’m going to break down, if you were to think of the truly prices are sticky when China sells to the US, it sets a dollar price, and then the tariff is slapped onto it. Then that’s just mechanically full pass-through; US importers pay for it, inclusive of tariffs. Of course, that begs the question as to why is it that China is keeping that price in dollars relatively sticky.
COWEN: Sure.
GOPINATH: What we do know is, in the case of many countries in the world and many exporters, they also tend to be importers. The value-added component of trade has declined over time, and because of the dollar’s dominant role in the trading system, everybody is pricing their goods to each other in dollars. In a way, the way you want to think about it is that when China is exporting some certain goods to the rest of the world, if its imports are priced also in dollars, then those inputs that are going into its production function are priced in dollars and sticky in dollars, then you have an incentive to just price in dollars, and your dollar price is not going to move that much because the exchange rate adjustment is not doing much, because an important part of your cost of production is also in dollars.
That’s the reason why we see a fair amount of dollar pricing in the world. The sense that somehow China should be able to, or any other country should be able to, cut their dollar prices by a lot when their currency depreciates is not the case, because they are also importing inputs from the rest of the world that are priced in dollars, and they don’t have that much of a margin to squeeze.
Now, of course, there’s variation across goods. Some goods rely more on inputs that are dollar-priced, some depend less, and you do see in the data that that variation matters. You do see that the pricing behavior is a function of how much do these firms rely on imported inputs.
COWEN: It does seem to me odd that simply the unit of account, to some extent, determines whether firms behave competitively or as if they have some market power. That goes against all my economic intuitions, and it puzzles me.
GOPINATH: No, this is not the unit of account.
COWEN: The invoicing currency.
GOPINATH: I would say there’s a few things. One is there is the short term. There is the near term when the invoicing currency matters, but just the example I just gave you tells you why people have chosen to invoice in one currency versus another. The way you think about it is that, ideally, if you could flexibly adjust your price at every instant in time, then obviously invoicing currency is irrelevant, but companies do sign longer-term contracts.
One of the things I looked into when I was doing the research on this was to say, “If companies are signing long-term contracts, is this both about prices and about quantities?” It’s not just that I sign a contract that says I’m going to sell to you at this per-unit price, but it’s also that I’m only going to sell X amount of quantity, in which case then these prices are not allocative.
If you look at the contracts, that’s actually not the case. The way it works is that the companies say, “We’ll sell it to you at this price. The quantities, we have a range. Obviously, it’s not like we can send you any amount. There is a range of quantities at which we will send to you,” so there’s that much more flexibility. There is the short term, and then there is the medium term. I think what’s important to recognize is that the invoicing decision itself is not just some Calvo fairy. It’s not just some “Here, we’re going to assume that we wake up and somebody’s told me to price in dollars.”
There is a reason why the dollar is used as a pricing currency around the world. From a pure first principle, it’s not a decision about which currency to invoice. There’s evidence consistent with that.
COWEN: Now, as you know, the gravity equation is one of the most reliable regularities in international trade economics. Basically, the quantity of trade is inversely proportional to the distance of countries once you adjust for incomes and some other matters.
That seems to be another case where the relative price doesn’t matter that much because few people think the gravity equation is about transportation costs. Why is it, in your view, that the gravity equation holds so reliably and so consistently?
GOPINATH: Again, I’m not an expert on the gravity equation, but what we do see is it is the case that physical distance does matter. There’s a reason why trade between the US and Mexico and Canada should be as close as it should be, setting aside the tariffs that are happening right now. There’s a reason why it makes a lot of sense for countries that are physically closer that this should be the case. Distance does matter.
Over time, obviously, once you build up these networks and you’ve built up these relationships, they can persist over time. You can, of course, build pipelines and you can build other kinds of logistical networks to get goods around, but distance does matter. We also do know that the relative incomes of the countries do matter too. There’s more trade happening between the countries that are large, and that also plays a role.
COWEN: Distance matters because of networks. If we take US and India, which have pretty strong networks—Indians in the United States have earned very well, started lots of businesses, have tech ties back home—that then should violate the gravity equation if we consider the networks being strong?
GOPINATH: The biggest exports in terms of what it does to the US are service exports, which is the business process outsourcing. That’s the main form of export that happens. That, of course, is less constrained by physical distance in terms of literally shipping a product across the seas. Yes, again, depending upon the nature of the product that you’re exporting, the physical distance may or may not matter that much.
COWEN: I have some Argentina questions for you. Now, you’ve worked on Argentina. This is late August 2026. Why is it that right now inflation in Argentina is still so hard to bring down? It seems stuck, right? It’s gone down a lot, but it’s not obvious there’s a lot more progress to be had easily.
GOPINATH: Yes. That’s actually a lesson we learn every time with countries that are in this disinflation process where you’re starting off with inflations of three-digit numbers. Two years ago, I think Argentina’s inflation was around 150%, so coming down to 30 percent, that was tough. Bringing it down is impressive, but it’s always the last mile going from here to anything like a single-digit number, always takes much, much longer. This is nothing special about Argentina.
Now, in the case of Argentina, what would it take to bring it down much faster? Firstly, I think there’s got to be a lot more confidence in policy continuity. I think what the Milei administration has done very well compared to the previous administrations is recognize that the problem was fiscal, that as long as Argentina was running the kinds of deficits that they were doing and using monetary financing, basically money printing to pay for it, there was no possibility of getting out of this trap. He has been running primary surpluses since he came to power.
He’s absolutely completely committed to it. I remember I had conversations with him when I was at the IMF. This is something he absolutely will not budge from, but there are questions whether the rest of the political class has signed on to this. There’s question marks around that. There are elections coming around. There was a midterm last year. There is a national election in 2027 October. It’s very reasonable for everybody to pause and ask the question whether the reforms will continue or will there be a relapse, as we have seen many, many times in the past in Argentina.
COWEN: Say we knew that either he would be reelected or someone similar would be reelected, would then simple crude monetarism solve the inflation problem, if backed by enough will, or is there still some other thing that makes it hard to bring down the inflation?
GOPINATH: There is inertia in this process. The good news is that inflation expectations have been trending down in Argentina, so that’s great. It was at very high levels. Argentina’s inflation was completely deanchored at the time when he took over. Any news of the possibility of the exchange depreciating even by 1 percent or 2 percent would immediately show up in prices. This was an example of a case where just simply doing a nominal depreciation is not enough.
There were previous governments who were doing nominal depreciations, and in three months, the real exchange rate was basically back up to where it was before. Nothing changed because it just got priced into goods. That takes time. Anchoring inflation expectations is part science but also part art. Sticking the course and just showing you’re committed to keeping fiscal deficits low and having the central bank be independent, which is another step he’s taking right now, which is having central bank independence so that there is no monetization of deficits.
Argentina will also have to build up foreign exchange reserves, even though the goal is obviously to have floating exchange rates. We know that every country, pretty much, with the rare exceptions, can have disorderly market conditions. Given Argentina’s history with the currency, it will need to build up reserves much faster than it’s doing on foreign exchange reserves and what it’s doing right now. Also keeping in mind the elections that are coming around next year.
COWEN: In some recent times, it’s been quite expensive to fly down to Buenos Aires and buy a good steak because the real exchange rate was kept high. Now that may be a signal of credibility, but many economists criticize that decision that simply trying to peg exchange rates high has a pretty bad record historically. Was that the correct decision, and has reality vindicated it? That semi-peg hasn’t really been broken. Bessent backed it. At the time I thought that was a big American mistake. That too seems to have worked out okay. What’s your view on all of that?
GOPINATH: I do believe the Argentinian peso needs to adjust more and needs to be allowed to depreciate more than it is doing right now. They have a framework where they moved from what looked like an exchange rate peg in previous regimes to a crawling exchange rate to now a crawling band. That band has gotten pretty wide. In a sense, their framework should allow for more movement in the exchange rate.
I think they’re being far too hesitant in buying dollars to build up their foreign exchange reserves than they should. I think that’s something that they could do much faster because if they did that, yes, they would get some more depreciation, but it would be still well within the crawling band. Short answer, I do think that they need to let the nominal exchange rate move more. That, combined with fiscal policy and monetary policy that doesn’t have that depreciation feed directly into prices, that combination will generate some real exchange rate depreciation. I think that is valuable for them to do.
Now, the thinking behind why do you not want to leave the exchange rate to depreciate more, you can make an economic case for that when you say you worry about deanchored inflation expectations. This is not the world where our models work. It’s a world where when people wake up in the morning, see the newspaper, see that the peso is depreciated, immediately they want post higher prices, which are even higher than what the depreciation is, because they think that this is going to feed into all the prices that they have to deal with too.
When you have deanchored inflation expectations, there is this argument about, “Okay, we don’t want the nominal exchange rate to move around too much.” I think, over the past couple of years, thanks to the other policies, inflation expectations have come down, and that should give them some more confidence in letting the exchange rate move around.
COWEN: If further depreciation is required and indeed good, why isn’t the crawling peg just broken now? It’s not that thick a liquid market, right?
GOPINATH: Oh, there is fairly a wide band right now. Actually, because the crawl is now tied to inflation from two months ago, it’s actually opening, and it’s getting fairly wide. You’d have to do quite a bit of intervention to do something for it to happen, but at the same time, there are a lot of positives about the country. There’s a lot of reasons to actually want to invest in Argentina and to have a positive view of the country. The rest of the fundamentals would be consistent with attracting money into the country.
COWEN: How should they deal with debt deflation issues? People in Argentina took out plenty of loans under the expectation inflation would remain pretty high. Now inflation is much lower. They feel quite pinched. There are also people who vote. What should Milei do?
GOPINATH: Yes, the nonperforming loans have gone up. What Milei should do is put in place rapidly a resolution process for these loans, a market-driven resolution process. As I see it, Argentina is going through a major structural transformation, a major one. These will take years, not months. We would expect to see nonperforming loans also because sectors that were previously very protected have now been exposed to much more competition. Putting in place a resolution process that is market-driven—it’s not bailing out banks; I think that would be a mistake—I think that would be helpful.
You do need to deal with it. The idea that somehow you want to prevent this from happening, I think, would be problematic. What is needed—and that’s the challenge right now for Argentina—is that while inflation has come down to 30 percent—it’s interesting. It’s one of those countries that understands the second and the third derivative of the price levels very well. They look at what’s happening with inflation, and they think, “Okay, this is pretty good,” but what’s less good, of course, is what’s happening with jobs.
Growth is there, but it’s coming mainly from mining and agriculture, and much less from the job-intensive sectors—manufacturing—because they’ve also been open to competition now. There is going to be a difficult transition, and I don’t think the government should ignore that. The idea is to help workers and not necessarily protect the industry. The more it can do on that front, the better.
COWEN: Should we as economists be happy to see stablecoins introduced into Argentina becoming significant, possibly over the longer run, supplementing or even replacing the peso? Is that good or is it bad? It harms their state capacity too much? What do you think?
GOPINATH: The way I think about it is that some of these innovations serve as a disciplining device. The fact that people in the country have an option to hold their money in more stable forms of currency is a disciplining device on the government and what it can do, and therefore, puts pressure on having good policies in the country.
Right now, we know, but do we want to move to a situation where Argentina is dollarized? By the way, initially, actually, Milei campaigned on the grounds that he was going to move to a dollarized economy, and he hasn’t done that. I would say that I think that’s good because, first of all, there is no panacea in terms of saying just because you’re dollarized, you are safe, or you’re a well-managed country. The two main dollarized economies of the world—Ecuador, El Salvador—are all in programs with the IMF because ultimately, if you don’t have the right fiscal policies, you’re going to end up needing a bailout in any case. That’s not a panacea.
Of course, I am of the view that monetary policy is still very valuable for stabilizing economies. If you can get the credibility and the central bank independence, then having your own currency gives you much more ability to stabilize your economy.
COWEN: Take, say, the euro area. Both Milton Friedman and Paul Krugman thought the euro was a mistake. They disagree on many other things. There was a major euro crisis. What did Friedman and Krugman get wrong? If fixed exchange rates aren’t so great, why not just let the different European nations have floating rates to some extent? Maybe Benelux would be a fixed rate, but the others let them float. Same argument.
GOPINATH: As you know, the origins of the formation of the European Union and then the euro was a political contract.
COWEN: Sure, but we as economists, should we be happy about the euro or regret it?
GOPINATH: We could ask ourselves the same question. Let’s take the United States. It’s a large country. There’s states like California—very, very high income—and states with a much less high income. We are part of a currency union. Is there an argument to be made that it would be better off if everybody didn’t have? No. There is, again, this optimal currency area argument. In the case of Europe, the way I think about it is, if they can get increasingly to the goal of actually being truly integrated in terms of their product markets, have much less regulation that prevents trade across the borders of countries, then in that case, these are the benefits from having that.
At this point, I don’t think the debate of whether the euro is useful is really not that important anymore. I think that it is we are here, and there is a lot that can be gained if the European Union and the euro countries were to be truly integrated in the sense of having a lot more trade across their borders, a lot less restrictions. Right now, as we know that there is an issue with not having enough scale in their companies. All of that can be fixed. I think that’s the more interesting question than whether the euro makes sense.
COWEN: There’s countries that are not on the euro, right? They face decisions. Denmark has had a very strict peg, but they don’t actually formally use the euro, and that’s a choice they face. Iceland faces this choice, Armenia. We, as economists, I would be inclined to tell them not to enter the eurozone and to stay put. That seems like a pretty important choice. What would you tell them?
GOPINATH: Again, this is very country-specific because there are countries that could gain the credibility by actually tying their hands with their currencies. They get the access to the markets, and they get better terms for it, and they get funding from the union. That’s been very helpful for a lot of countries. They get transfers. From their own personal perspective, that can be helpful. If you’re a country otherwise that’s well managed, no need for resources from other countries or anything of that kind, then in that case, your flexible exchange rates are helpful.
COWEN: Now, Scott Sumner argues we shouldn’t be worried about trade imbalances at all. We don’t worry about them across American states. A lot of the EU has a pretty big current account surplus. Not many people are worried about that. Yet when it comes to China, the talk is all about trade imbalances. Scott argues that’s more of a political project than an actual economic argument. Do you agree with Scott or not?
GOPINATH: I would agree with Scott along the following lines, which is that the trade imbalance in and of itself is not something that we should be focused on. I think what we care about is welfare, and welfare involves jobs and consumption, what’s inflation, purchasing power, and so on. People do not wake up in the morning saying, “Okay, my current account deficit is too big, or my current account surplus is too big.” If all of your policies that were delivering good outcomes for your country were to bring along a deficit or a surplus, that’s perfectly fine. There are lots of good reasons to be running deficits and surpluses we know. There’s nothing that tells you that you shouldn’t.
The problem arises when you have policies that countries have in place that are inconsistent with any kind of a balanced growth model, and they manifest themselves in that trade deficits and surpluses. This is what I’m saying as an economist.
Now, of course, from a politician’s perspective, there are other arguments for why they would point to it. What we have seen historically over and over again is that whenever you had these increases in these deficits and surpluses, you’ve had trade wars or calls for protectionism. That’s what happened during Reagan’s time, 1980s. That’s what led to the Plaza Accord and then all the adjustments that followed after that.
Then you also do worry about the possibility of crises. The Great Financial Crisis was preceded by growing imbalances. There was a sense in which all this big savings glut. All this money flushing around. All these large surpluses and deficits were part of the problem. Now I would say where we are now in this third wave of concern about imbalances.
To be clear, it’s not the imbalance itself. As an economist, I would say that it’s not the imbalance in itself. You don’t wake up and say, “This is what I’m trying to prevent. I’m trying to prevent us from having a deficit. I want us to have balanced trade.” I think that’s bad economics.
COWEN: Say China is channeling what would have been wage income into investment, and that’s plausibly the case. Now, it may be politically unstable in the sense that we Americans object to it, but that’s not an argument per se, right? We have a choice as to whether or not we should object to it. It doesn’t seem that rigorous to say, well, this will cause another financial crisis like 2008. There’s just not real evidence for that. Scott Sumner would say, “Let’s just be happy we have cheaper goods. We send them paper. We get back stuff and go our merry way.” Why is that wrong?
GOPINATH: A few things. Firstly, I could go into these differences between trade deficits and surpluses versus what I would call sectoral imbalances, or the fact that China’s running a big manufacturing surplus, which is different from a trade surplus itself, or that it has big EV production. What we do recognize, again, is that we do a pretty bad job in moving or helping workers that have lost their jobs in certain sectors.
COWEN: We’re at full employment now or very close to it.
GOPINATH: Again, we’re at full employment right now, but we’ve had this period of time when we had communities that were deeply affected by—not just trade—automation was a big part of it too. We don’t do a great job in terms of getting people back into jobs, and that can affect, depending upon your ideal welfare function for the country, if you care a lot about those people, then obviously in that case, that’s something that should matter in your policy decisions.
In the case of China, I would say firstly that, in both the case of the US and China, we’re very far from the world where countries are doing good policies, and this is all about comparative advantage, and this is all about this would be the outcome we would have in a world where every country was doing the right policy. We’re very far from that.
China’s surpluses are a reflection of things going wrong in China. It’s not a reflection of strength in China. It’s a reflection of weak consumption. It’s a reflection of misallocated resources going into different sectors. They did that with their property markets. Now they have a huge property market problem that they haven’t been able to fix in five years. They have this now with other markets, including EVs and the other sectors. They have the problem with inflation being too low.
They’re trying to do so called anti-involution policies, bringing companies together and telling them, “You’ve got to keep prices higher than what you’re doing right now.” This is not the world where they’re playing good policies, and this is all comparative advantage, and this is the outcome that we see. I wouldn’t push the argument that this is, “Oh, we should be just happier with cheaper goods from China.”
COWEN: Should there be a digital euro? If so, how do we control or regulate access so there’s not too much disintermediation of private sector banking? It seems the more people can use the digital euro, the more disintermediation you get. If only a few parties can use it, maybe that’s fine, but there’s hardly any advantage to a digital euro. How do you think about those tradeoffs?
GOPINATH: I have a particular view on the whole discussion on digital money broadly. In this space, stablecoins comes in too, which is the closest to digital money because it’s basically backed one-to-one by a fiat currency and can be used for payments, unlike the others. There is the question, do we need this technology? The best argument I’ve heard from it is that, okay, well, it’s going to help cross-border payments because that’s where the true frictions are. If you look at the data in terms of the actual cost of doing a transfer using stablecoins, when you include the on-ramping and the off-ramping into fiat currency, it’s not clear at all that this is a cheaper way of doing it.
Then there is the argument that we need to be able to do 24 hours banking, and we need to tokenize all the assets so that we can move them at fractional amounts at infinite speed. Again, what’s the true value of that? It’s somewhat unclear. There is somehow this sense that we are in this equilibrium. There is private money in the form of private stablecoins coming out.
By the way, one thing I do like about stablecoins is that they’re finally putting pressure on the traditional banks to actually pay attention to the prices that they charge for their services, especially across the border. I love that. That’s a good competitive fringe argument for them. Over and above that, what is the benefit of all of this?
COWEN: It would be weird if money stopped evolving. Money evolves over centuries, millennia. To think that we’ll never have digital monies seems highly counterintuitive. If we’re going to have them, shouldn’t we have quality ones through central banks, like the European Central Bank, which at least talks about doing it? Thus, we should do it because the alternative is worse. We’ll get a worse digital money.
GOPINATH: Firstly, Tyler, we have digital money. All of the banking that happens is based on ledgers, and it’s all digital.
COWEN: Truly digital programmable money, right?
GOPINATH: You could do that with the ledgers that exist. It’s not as if the technology doesn’t exist. This is about whether you want to use blockchain technology versus another technology. We’ve come so far from using any kind of physical cash. It’s all digital right now.
COWEN: It’s also about giving more people direct access to central bank payment systems. That’s where the disintermediation would come in?
GOPINATH: What is the need for that? Why do they need to have accounts with central bank? If you want to go along and say, “Now it exists, this exists. Should there be central bank digital currencies?” I think the concern there is that if you are in a world where it’s all privately issued stablecoins, and for some reason we all gravitate to that world, and at some point, one of those stablecoin issuers decides that, “Well, we’re going to not be linked to the dollar, but linked to something else.” That’s the kind of difficult scenario that one would need to worry about. Then you are saying, “Okay. Well, again, the central bank digital currency works like a competitive fringe,” and says, “We are there. We can step in when there is any misuse of this kind of power that the private firms may have.”
Again, there’s a reason why central banks everywhere are still scratching their head about what exactly they should do about this. Look, if you think of the countries that have introduced central bank digital currencies, you think of China. Nigeria has experimented with it. There hasn’t been much take-up at all.
COWEN: Again, this is August 2026. Some of my friends are quite worried, whether correctly or not, that possibly the marginal buyer for US Treasuries today is both private and with some leverage. Some people say, “Oh, this is UK hedge funds.” I don’t know if that’s true or not, but if I want to set their minds at ease and tell them the Treasury market is going just fine, what argument should I make back to them? Alternatively, you might agree with them, but what’s your view?
GOPINATH: Well, I guess if I were really trying to be the advocate for saying, “Don’t worry too much about Treasuries,” I would just say that the alternatives could look worse. The question is, if you want safe assets, where would you put it? There is the US, which has the most liquid, largest Treasury market. Then if you look outside and you say, well, maybe one country that looks very good, Switzerland, but it is so tiny in terms of relative scale of its market that you’re going to lose money when you put your money in there. The markets are very small. If you look at France or Germany or any other country, Japan, it’s not as if you would look around and you’d say that there are great alternatives.
That said, that’s if I was trying to make an advocate for the Treasury versus others. I do think there is a problem. We have a problem in America with our fiscal situation, with what’s happening with debt, with Treasuries. I suspect more and more people are going to go into the short end in terms of what they hold, as opposed to the long end. You can see that with yield curves going up, because if you look at debt trajectories compared to other countries, including France, Germany, Japan, and the others, it’s the US debt trajectory going forward that is particularly concerning in addition to what the levels that we’re seeing right now.
COWEN: I have another group of friends, and they say that AI, or artificial general intelligence, it will raise US productivity maybe by half a percentage point a year. This will be like 1995 to 1998, and the budget won’t balance, but debt to GDP will converge at a ratio at something like 100 percent, 120 percent, and things will just be fine. Do you agree?
GOPINATH: I think what I would say is that if what you said is true and we do get growth going up persistently at half a percent or so, yes, that would absolutely help in terms of the debt-to-GDP trajectory, just in terms of the math. I think that’s very helpful.
COWEN: Do you think it will?
GOPINATH: At this point, I cannot say that that will necessarily be the case for the following reasons, because even if the technology seems wonderful—I use it a lot, I’m impressed with it—as you can see, there’s a lot of pushback against the use of AI. Whether there will be widespread adoption is to be seen. How much more can this technology penetrate? The way it gets used, I think there’s question marks around that.
Also, we could have a financial market correction that could happen, even if the technology is wonderful, as we know. You could have a correction, which then later on, you could have something better, but you could go through a correction. In this difficult time, given everything else that’s happening in the world economy, I don’t know whether we are going to be on this glide path to just having half a percent higher growth.
COWEN: Now, I have yet another group of friends who think that as AGI approaches, you probably end up with stagflation because real interest rates are quite high, because the private demand for capital is so extreme, and that may even crowd out government borrowing to some extent, and some people will be losing their jobs, and you end up with pretty high inflation and some degree of unemployment, and that’s an altogether new problem for central bankers. Agree or disagree?
GOPINATH: That seems a bit odd. If this is going to deliver the productivity boom—
COWEN: There’s an intermediate period where you get the demand for capital, which right now is very high, but the productivity gains from AI in this moment, we would agree, are quite small. For a while, you have stagflation, and then a radical deflation later on.
GOPINATH: We’re talking about now, where we could have this period where, with the build-out of all the AI infrastructure, we would get in demand-side effect, and I believe that is correct. I think that, as of now, we are seeing the effect of the build-out of AI which is more of a demand story, affecting prices and inflation, and there is. It’s not a huge amount, but you’re certainly seeing that show up, but you’re going to see it increasingly if the scales of investment that are projected actually materialize. There are some question marks around that too. Yes, in that case, the concern for if it’s a purely demand-driven shock, of course, the central banks, that tends to be an easier problem to manage in the following sense, which is worse is that there’s no tradeoff between inflation and output gaps or inflation and employment. That is true when you have a demand-driven increase in inflation.
The concern is if you have a supply shock, then obviously you do have a tradeoff in that case. If it’s a demand-driven inflation shock, then the right thing to do is to tame demand.
COWEN: It’s private sector stimulus. It’s hard to tame that demand. People want to build all these data centers and compute. Real interest rates are high, and maybe that’s bad for borrowing economies. We all know what the Volcker disinflation did to make the third world debt crisis worth back in the ’80s. Are we going to see a rerun of that, and then the IMF will just be super busy over the next 10 years?
GOPINATH: No. Firstly, I do hear about this fact that AI spend is insensitive to interest rates. On the margin argument, if it turns out, if Kevin Warsh at Jackson Hole were to say that we are going to now start raising interest rates going forward and that’s because we’re in an inflationary environment, I bet that that will change, if there will be an effect on AI spend. I don’t think that that’s the case. It’s a matter of degree of shift that’s going to happen.
Debt crises and so on—that’s a good question. It’s been very interesting because when I was at the IMF, I spent a lot of time on debt restructuring. The G20 created the Common Framework, which is basically how do we restructure a country’s debt, and do that with a platform that brings in official creditors, multilaterals, and the private sector and so on together. That seems like an efficient way to do it. Yes, that platform was created, the G20 was created, and I think what has been, in a way, surprising is that we haven’t had that many debt crises, despite everything we’ve seen up and down and sideways in terms of shocks.
That does not mean it won’t happen, with interest rates going up. Even more than we’ve seen, we could see more countries in debt crisis. Though I think the new question right now is what is the framework in a world where developed country bond markets are in trouble, real trouble? I don’t think we have that. When I see what Secretary Bessent is doing, it clearly seems like there’s not that framework. That could be an important area, maybe in terms of thinking about what would that look like.
COWEN: What do you think the future of academic economics looks like if, say, an advanced AI could write a top-five paper within two years with maybe modest human assistance, but nonetheless, could do most of the hard work? That seems quite plausible today.
GOPINATH: Have I seen such a paper? I have not.
COWEN: Certainly, it can write for third-tier journals today. It’s probably doing a lot of that. Again, two years, progress is extremely rapid. Two years ago it didn’t know how many R’s were in strawberry, and now they prove math theorems or disprove them. Things like differences and differences, they’re quite automatable, and have the human work for a week and set the agents on the problem and come back with a Review of Economic Studies piece. In fact, do 30 of those a year.
GOPINATH: Review of Economic Studies is a top five. In this case, what I would say is that as economists, we’re a social science. We’ve always, for decades, wanted to be associated with the science part of it because that makes us look more impressive, maybe. Maybe what’s going to happen now is that we’re going to be associated more with the social side of the social sciences. I see that.
Again, we’ll see what happens in two years and what the technology can do, but as of now, it’s very good at coding. Once you set up a problem and ask it to solve it, it’ll do it very well, but it doesn’t have a good sense of what are the right ingredients to throw into the model because there’s an art to it. Therefore, I’m waiting to see whether that’s the case, whether there is good, new insights that can come out of just an AI-written paper.
COWEN: Quite a few of the graduate students, especially from MIT, seem to be taking jobs at Anthropic or maybe OpenAI. What does this tell us about the future of the academic job market? Will there be an increasing brain drain of the smartest people into the private sector, which, of course, does pay a lot more?
GOPINATH: Well, I think there are actually two things that are also encouraging that shift, which is the other factor is what’s happening with funding, what’s happening with National Science Foundation grants, what’s happening with the tax on endowments. That is leading us everywhere, including at Harvard, to hire fewer faculty. There is less demand; there’s going to be fewer academic positions available because of what we’re seeing right now in terms of the policies coming out of this administration. Maybe that’s a good thing, who knows? That’s what the facts are. That’s what the market is.
Then, on the other hand, you have where the jobs are and where the money is much higher, which is in AI. I think we’re going through this period where, I can see why more good students and more PhD students and so on would be looking for jobs in the private sector.
COWEN: Immigration is tougher also, as you must know. Does that mean there’s a golden age of academic economics that lies immediately behind us? Now, more or less, at least for humans, it’s just going to get worse? I think that maybe is my own prognosis, but do you agree?
GOPINATH: There’s one sense in which this feels a bit like the golden age for economics because of the big transformations that are taking place. There’s AI, but there’s what we just talked about, the debt markets, there’s what’s happening with demographics, there’s numerous areas where it seems like the whole global economy is churning and shifting. Therefore, I think we all wake up, in a sense, very curious and excited about what we should be trying to help answer.
Yes, so this is not the great moderation period when we were, I think, writing papers on the derivatives. This is the period where we have big structural changes. I think this is a good period. I think it’s the responsibility of everybody, including economists, to engage on what we’re going through in terms of major transformation.
COWEN: Now, you’re from a family from Kerala. Why does Kerala keep on electing communists? How communist are they really at this point? It’s been a long time, right? Are they just another party, in essence, or is there something still communistic about them?
GOPINATH: Yes, so as you know, in every part of the world, whether you’re a liberal or not liberal, or whether you’re a Democrat or Republican, that definition of that just adjusts over time. First of all, in the case of Kerala, with the exception of one election, the government in Kerala has always switched between being communist for one term, and then it goes to the Congress Party for another term. The label of the Congress Party is, it’s Congress Party. It doesn’t say it’s communist. If you look at the policies of Congress, there are quite a lot to be left in the state of Kerala.
Just to clarify to what you said, which was that we do have, in Kerala, it switches between the Communist Party and Congress Party, with the exception of one election since independence. Why do they tend to be very left-leaning? This goes to the history of Kerala also, in terms of the maharajas and the way things were done, and the communists and levels of education were highest in Kerala for a long period of time. There are many factors behind it. Kerala does benefit from repatriated income from the Middle East. They get a lot of it. That also helps, but they are moving.
Kerala is actually moving toward doing more and more encouraging businesses and the private sector and so on. I think they are also moving in that direction, but what they do have is a very large social safety net.
COWEN: In both Kerala and many parts of Sri Lanka, they seem to have much higher social indicators than many other parts of South Asia. There’s better rights for women, however imperfect that may be. There’s higher literacy. Public health indicators are much better. Why is that? Since it’s Sri Lanka and Kerala, which are at least in parts ethnically or culturally similar, it doesn’t seem it would be the Communist Party or some very particular concrete explanation. It seems more general.
GOPINATH: Well, that’s a good question, and I can tell you what I think, but the honest truth is, and I haven’t exactly investigated this particular question. I just know that in the case of Kerala, the maharajas of Kerala were very big on educating women. There’s some reason to think that that filtered through. There’s also important large communities in Kerala where I come from, which is matriarchal and not patriarchal where who gets the property, it’s the women and not the men.
I suspect that that plays a role because in all of India it’s a patriarchy except in some of these communities in Kerala, though, like you, I would say that it’s an imperfect patriarchy. There is still a lot that doesn’t really seem fair to women. There are some of these unique characteristics about it. I think when I was growing up in India, I do remember having a lot of friends who made the point that their parents tried very hard to have a son, while I always grew up with the sense that girls were very important because, again, the matriarchy lineage. Maybe that plays a role too. I am, at this point, way outside my area of expertise.
COWEN: You grew up in Calcutta with the Keralite family. Is that correct?
GOPINATH: I was born there. I was very young when we moved out. I was like maybe three years old or four years old or something.
COWEN: You went to an all-girls Catholic school?
GOPINATH: Yes.
COWEN: What’s your view of unisex education? Do you have one? It worked for you, right?
GOPINATH: I’m in favor of it. My poor son, I was very excited to send him to Roxbury Latin School because it’s an all-boys school, and I thought that it was great as an education. My take on this, on the single sex one was, it was great because we did everything. There was no roles defined in what girls can do and boys do. When we put up a play, the girls were in the play. They acted, but they also built the stage. They did it all. I don’t know, I think that’s cool.
COWEN: Very last question. Now your IMF tenure ended in 2025. You’re back at Harvard, presumably recovering to some extent. For your future, what do you plan on doing next, working on next, writing next, researching next? What for you is coming up?
GOPINATH: Thank you. First of all, I’m really excited to be back and in the thick of all the discussions here, given the times that we live in. I think these are very interesting times to be an academic too. I’m working on digital money, stablecoins. Because of my work on dollar dominance, I’m curious to know how this can affect dollar dominance or not. That’s one big area, but also working on a project on AI.
COWEN: What’s your AI project?
GOPINATH: The AI project, which I’ve just kick-started, is actually looking at the impact of AI on jobs; everybody else is doing that. There is one difference I think that’s not coming up in the conversations, which is people are tracking what the effect is on what’s happening to jobs right now with the advent of AI, but what we’ve seen through past cycles of automation is that something like 85 percent of automation-related job losses happened in the first year of recessions. There is a sense in which there is a latent build-up of the fact that a pool of workers who, when there is a recession and you are in cost-cutting mode, companies lay off and then you have jobless recoveries. We saw that after the Great Financial Crisis. With my co-authors, I am working on trying to understand to see what we can do right now in terms of tracking what that latent pool could look like. I’m also interested in AI and financial markets and frictions and so on.
The last thing that I’m doing is I’m launching a Global Economics Lab at Harvard with the mission to help advance an international economic order that benefits people everywhere. The way I see it is that we’re at obviously a historic time in terms of the global economy: big transformations, shifts of protectionism, geopolitics, AI. Because now economics and politics are getting a bit intertwined, I think it’s harder for other institutions and multilateral institutions to be able to do full justice to the big questions that we face. I think independent researchers have a role to play here. That’s what I am investigating.
COWEN: Great. Gita Gopinath, thank you very much.
GOPINATH: Thank you, Tyler.