Paul Krugman Podcast

Lunch Money with Paul Krugman and Heather Cox Richardson

I’m posting our Wednesday conversation as this week’s video. Transcript below. . . . TRANSCRIPT: Paul Krugman in Conversation with Heather Cox Richardson (recorded 5/20/26) Heather Cox Richardson: How are you doing, Professor Krugman? I know you’re on vacation. Paul Krugman: Yeah. As I wrote the other day, I’m in Europe, which means I don’t have to think about Trump 100% of the time, only about 90%. So that’s a little bit of release psychologically. HCR: It’s really astonishing, isn’t it? But hopefully we don’t talk entirely about him today. I’m actually interested and would love to hear what you have to say about artificial intelligence, not itself as an entity, but as a factor in the economy. Because boy, it sure looks to me like we are way overinvested in AI. I think the growth on the stock market is basically AI companies. We know now that there’s more construction in AI data centers than there is in commercial real estate. And I’m wondering, can we just talk about that and you walk us through what this looks like? Because everybody keeps saying, “Oh, it’s a bubble like the housing bubble or like the dot-com bubble.” And I’m looking at it and saying… PK: Obviously, history is mostly what we have to go on. There have been many bubbles like this. There’s some broad similarities to dot-com, which was also a telecommunication thing. It also looks like the canal bubble in England, which was earlier. Most of the bubbles were pretty clearly bubbles at the time and that was certainly true for dot-com which I sort of still remember in real time. But with AI, I’m finding that the contrasts with the late 90s bubble are really illuminating. Obviously it’s again technology with lots of investment. There’s an enormous enthusiasm of a kind, but in other ways, it’s quite different. HCR: Well, let’s start with this. What exactly is a bubble? PK: Yeah, it’s always a question, but a bubble more or less means that people are investing in something that has no realistic chance of paying off—not socially but just commercially, to an extent that justifies the amount of money being thrown at it. Crucially, a bubble is something that people do because everyone else is doing it. So, Robert Shiller, the great bubble theorist of modern economics, said that a bubble is a natural Ponzi scheme. It’s something where you get in and you make money because other people get in, and people keep on coming in because everybody before them made money. But in the end, it’s a game where the money isn’t really there. It all depends on fresh crops of suckers coming in. And at some point you run out of suckers. So that is a Ponzi scheme, especially when someone like a Bernie Madoff does it deliberately in a bubble. It also happens naturally. Nobody is orchestrating it but nonetheless the logic of it is the same as a Ponzi scheme. So basically, it’s a lot like pornography where you know it when you see it. But it’s not just the fact that people are wrong but that people are wrong in a way that should have been predictable and where it’s really something that is sustained by the momentum, by the fact that other people keep on coming in until they don’t. HCR: Okay, so when historians talk about this, they example they often use is tulips. It’s something that you can explain to people as a reference because it’s kind of a cool story. When you take it out of the economic system that we understand now, it’s easier to see. PK: Yeah, I mean, I’m not really fond of the tulips analogy but sort of the first thing that people think of as being something like a modern bubble was the tulip mania in the Netherlands. 17th century Netherlands was not quite the first modern economy because they weren’t quite modern, but they were on the way. They were commercialized. They were banking. And people were speculating in tulip bulbs, which were in fact valuable investments, but it got crazy. The prices went up because people were buying and buying and then prices went up further. And so, you can see the financial logic there, but I’m not really fond of this example because there wasn’t a whole lot of real investment. People weren’t building tulip infrastructure. But I guess in terms of the psychology, the market logic, it was not that different from railroad shares or dot com shares. So, yeah. And it is telling you, the fact that this is the Holland of Rembrandt and not only wasn’t there an internet, there weren’t even telephones, and yet the psychological logic was the same. And that’s kind of telling you that in some ways there’s a kind of universality about bubbles. HCR: So when we look at AI now, am I correct that there are two super companies in which the majority of AI money is invested? PK: Yeah. There’s OpenAI and there’s Anthropic and who are the big players but it’s an industry. It’s not just that these are the two biggest AI models. So you’re either talking to ChatGPT or to Claude which are the two leaders but then Google has its own model which is Gemini and then Elon Musk has a really bad one, Grok. And then there’s a bunch of Chinese versions, where they’ve taken a very different strategy. So it’s a little bit more complicated than that. And then there’s this network. So in a lot of ways, you want to think of this whole AI boom bubble as being a little bit like the California gold rush, another historical parallel. The people who are selling Anthropic and OpenAI are like miners, prospectors looking for gold. And what we know in California in the 1840s was that the people looking for gold mostly ended up bust but the people who made money were basically the Levi Strausses who didn’t make money by finding gold. They made money by selling equipment, by selling jeans and picks and shovels and also brothels and liquor to the prospectors. The equivalents of that now are companies like Nvidia which is selling the specialized chips that go into AI and there’s a bunch of other companies making a lot of money basically renting out computational capacity. So now we’re starting to see at least a little bit of money being made by Anthropic. All of my friends are playing with Claude and I just can’t get myself to do it. The big thing seems to be vibe coding, which lets you do programming without knowing how to program. And so Anthropic is actually making some money because people are subscribing to that service. But at this point, most of the money being made is from people basically selling equipment, selling the suppliers to this thing. And so the question from a kind financial economic point of view is whether there will ever be enough revenue, whether people actually end up paying enough for AI, this thing that we call AI, to justify all of the money being thrown at the industry. And history would suggest there’s a very good chance that the most likely outcome is no. The most likely outcome is that it will end up being a waste. But again, history doesn’t always repeat so maybe this pays off but I don’t think that explains the enthusiasm. HCR: Well, it’s interesting because one of the things that you’re seeing lately is the changing model for paying AI. That is, most of the use of AI currently is subsidized really quite heavily for every dollar of computing power that people use. It’s subsidized between $3 and $25 at the minimum. And the idea that people are actually going to pay the extraordinary costs that certainly right now it would warrant…it doesn’t seem like it’s going to happen. PK: Well there’s a question. Let me play devil’s advocate here for a second. When the dot-com bubble happened and people were offering all these services on the Internet where people weren’t willing to pay remotely enough to justify the money that was being thrown at it. But what eventually happened was that a few companies managed to create walled gardens. They managed to create enclaves. Essentially, Facebook is a walled garden where people pay for ads or watch ads or whatever. Google basically ended up being a kind of walled garden. The search was free, but Google was making money out of pushing targeted ads. We used to joke about Amazon. I’m old enough to remember when Amazon was famously unprofitable and was never going to be profitable. But it turns out that, well, in the end, Jeff Bezos built a moat with all of the infrastructure, the distribution centers. And so now Amazon is a huge moneymaker and evil. But that’s another story. And what’s happening with AI is, to a certain extent, they’re building walled gardens from the beginning. So I know people who’ve been using Claude or have been playing with Claude, I think would be a better description, and the results have been terrible. And it turns out that the results are terrible unless you pay and buy a higher tier of service. Now even there it’s not remotely enough to justify the expense [of investments] but clearly Anthropic is trying to create a situation in which people get hooked on vibe coding and then end up addicted and they’re going to end up shelling out large amounts of money to have the the version of Claude that works. And with something like that you can already see the outlines, at least, of how the industry intends to make money. Now, history suggests that usually there are only a few winners. Although one thing that’s also different from the dot-com bubble, is that in the dot-com bubble, there were hundreds of players trying to succeed, and in the end, just a few highly profitable corporations survived. This is not like that. This industry, at least on the U.S. side, is just a handful of players. So the chance that one or two or maybe three big AI models will end up becoming highly profitable monopolies, it’s not that remote. So, as I say, things tend to be somewhat different. I mean, we don’t want to start talking about what AI is exactly, but I think there are inherent weaknesses of it. I mean, it’s a technology where you cannot predict exactly what the tools will do, and you cannot know when they’re going to betray you; when they’re going to deliver hallucinations instead of actual-actual true results. That’s weird. I don’t know if there’s anything like that and you have to wonder, just how much will our society be willing to rely on technology that every once in a while just decides to go crazy or basically turn into Frankenstein’s monster on you. So that would be my guess, but it’s not as if there’s no possible way these guys could make money. HCR: Well, but there is something interesting in it as well, and I think you’ve identified that many of the things that we’re identifying as bubbles actually start with a product that people want. They don’t have to create their own markets. And the other piece of that is I certainly have heard people say exactly what you’re saying, that there will be a fallout where we’ll get a few good ideas out of where we are. And then you can have your walled gardens around those things. But it’s rare. I mean, I can think of an occasion for it when we got the Union Pacific Railroad in the 1860s, because Congress recognizes that people actually would like to get to California. But if you actually wait for there to be enough of a market in the plains to get those railroads going all the way to California, you’re going to be waiting a very long time. So they put the money up to create a market for those railroads. But then very quickly you get all these branch roads that lead to nowhere and end up feeding that railroad boom in the 1870s that collapses. So it does feel to me like this is something different. You’re not getting those walled gardens right now where people say, “Yeah, I really want to get into that and I’m willing to pay for it,” the way we were with iPhones, for example, or the way we were with the internet. I remember the first time I turned on the internet I was teaching at MIT and they made us take seminars so that we understood this new technology and I can still remember going home and saying, “Oh my god. My world just changed because I can do all this research.” This is the very early days but you look at the AI stuff and, I started using it pretty heavily just to see what it would do and I have become completely against it because so far I haven’t seen anything that isn’t crap. And I was agnostic. I’m usually pro-technology. Now, I am willing to admit that there are places where it is probably a good thing, like checking engineering plans in construction plans, for example. We know that there are ways in which mixing cement can be much more efficient if you use AI [for calculations]. But right now, I don’t see it taking off. PK: Well, you and I are not typical, of course. I think there’s an important distinction here but what I actually am using a little bit of AI for is actually producing transcripts of videos. You run a video through AI to produce a transcript which is often hilarious in detail but you can fix that. You wouldn’t believe what AI was making of the words, “vibecession.” But anyway, it can do certain things. I also find that with economic history, often there are a lot of papers that have tables and charts and I can feed them into a sort of low grade AI model as a PDF and get the numbers out instead of having to type the numbers from the old papers. So there are uses even for someone like me. I mean, in a lot of ways AI is kind of awesome in how much it manages to produce intelligible if sometimes dishonest responses to plain language questions. That is awesome given where we used to be, even if it’s not totally reliable. But the main thing is that a lot of AI—and certainly what is likely to be the paying uses of AI—is not coming from individuals. It’s not coming from me or you or some middle manager deciding, “Hey, maybe I can use AI to do this better, or maybe I’m just going to have some fun with it.” (Slightly scary but I do know people who are developing relationships with Chat GPT.) But it’s mostly coming from people working at businesses and large organizations who are being told, “You must use AI.” And this is something I’ve never seen before. This is kind of coercive technology adoption where the big money is telling workers that you must use this technology. And one thing you’ll remember from the early days of the internet, it was joyful. People loved the internet. People hate AI. We’re now having a regular pattern at college commencements of speakers who start talking about AI and all of the students start booing because everybody hates this. And the question is, how far can you go with a technology that everybody hates? So that’s one of the things that is unprecedented. You think of the people whose jobs were displaced by power looms, the Luddites. Okay, they hated the technology because they didn’t like what it was doing to their jobs but people hate using AI and they hate the fact that other people are using it. But they are to a large extent being dragooned into doing it and I’m not sure that I can think of a historical example like that. It doesn’t seem like it’s a very sustainable path forward. HCR: So, Henry Ford would have something to say about trying to force people to take on new technologies. I actually saw an Edsel a few years ago. I’d never seen one. I’d always just heard about how much they were rejected. And I saw it and I’m like, “That’s it? They just didn’t like the front grill?” And yeah, people just didn’t like the front grill and they wouldn’t go with it. But that brings up another question for me. You’re hearing a lot and there were stories out just today about companies cutting thousands of jobs because people were being replaced by AI. And I have a question for you about that. I actually then want to end with, what does this look like for the entire society? But it certainly looks to me that as the economy slows down, that it’s certainly possible that companies are letting workers go saying it’s AI. And what they’re really doing is they’re reducing their forces. Is it right that AI is possibly simply being a cover for people who wanted to downsize anyway? PK: Well, there’s some of both. I mean, if you’re a company that wants to, in effect, increase the workload on a smaller number of workers, then AI is a great cover story. You can say, “Oh, we’re doing this because of the wonders of modern technology.” And by the way, we expect you to, in effect, put in 10 hour days. We keep getting stories of companies that lay off a lot of workers saying that AI can do it better and then it turns out it can’t. And I don’t think these are just stories. If we’re saying that AI is just doing routine stuff. Some of my friends who actually work on this, like Henry Farrell, say that AI is a social technology. It’s basically agglomerating what lots of people have said. And it’s delivering back to you what a lot of people who know something about a subject would say if asked the question you asked. And it’s not understanding. There’s no mind there. But it is delivering a kind of aggregated, standard response. And a lot of jobs are like that. If you’re talking to the help desk at a call center somewhere thousands of miles away, the person that you’re talking to, if it is an actual person, may very well be there with a three-ring binder looking for what they’re supposed to say. And AI can replace that job. AI is basically doing much the same. A lot of people are doing fairly routinized, standardized work. It’s just the common opinion of common opinion responses to things as a way of doing their jobs. So that’s real. So it’s not just that AI is an excuse, but again, it’s an excuse. I mean, we always see this, right? To the extent that businesses care either what their workers feel or what their customers feel, stuff happening provides external excuses. This is the story of greedflation, that companies may raise prices when there’s an energy crisis, not because their actual costs have gone up, but because with everybody raising prices, who will notice if I get greedy? So there’s something like that on AI and jobs as well. But I don’t know. I mean, again there are enough stories now of companies that have laid off all of their experienced professionals for AI, and it turns out, well, the experienced professionals could actually deal with questions that were not routine, and they didn’t hallucinate, and so they’re finding that they made a mistake. But it’s amazing how little we know about how this works. I don’t remember there being so much uncertainty about what you could actually do with the internet. And of course, I don’t have any memory of what people thought you could do with railroads. But I think this is kind of unprecedented as this massive technology that we’re investing trillions of dollars in and still nobody quite knows how it works or what it will do. HCR: Well, I want to end with my real question. That is, if I’m correct, and these people I’m reading are correct about it looking like a real bubble, what does it look like if that bubble bursts? This is the reason I use the comparison of the 19th century railroads, or you could do the 1920s with cars, and the investment in AI in data centers, in hiring practices, certainly in investments, certainly in NVIDIA and all these different places that are tied into that specific technology. Now, I’ve heard from a lot of people, you included, that we’re going to get some good technologies out of it no matter what happens. And I agree with that. We always do. But with all the pressures that are on the American economy right now, I’m worried. And should I be? PK: Yeah. Let me give you sort of good news and definitely bad news. The good news, and I say sort of for a reason, is that on the face of it, if you just look at the scale of the AI investment, it looks like that’s driving all of our economic growth. But it turns out that an awful lot of the AI spending is actually imported tech gear. It’s actually imported chips and computer equipment and so on. So if the AI bubble bursts, a large part of the burst would be falling imports. It would be a big shock to the domestic economy but not nearly as much as you might think. There’s been a back and forth about how much economic growth has been AI and how much the high import intensity of the stuff. So in some ways this is a shock to the world economy and not so much to the U.S. economy, specifically. So I guess that’s kind of good news, though not so good for other countries. But, you know, Taiwan has experienced an enormous economic growth because of all the chips they’re selling to U.S. AI companies. So a lot of the bad news will end up showing up in Taiwan rather than in the U.S. The bad news: this would have been true of railroads, as well, but the dot-com bubble in terms of the actual really big money laid out was telecoms rather than dot-coms. It was the telecommunications companies investing especially in fiber optics, laying down tremendous amounts of fiber optic cable which stayed unused for a long time. There was lots of dark fiber after the dot-com bubble burst but it was still there. Fiber optic cable doesn’t depreciate rapidly. It was still there in the ground and eventually got used. So it was a lot of useful investments. As I understand it, these data centers that are being built, the investment in chips, the investment in software, this stuff will depreciate physically pretty fast. It will become outmoded pretty fast. So I think there’s likely to be a much higher proportion of just wasted investment that never finds a use out of this boom than there was out of the last tech boom. So, not so great. And by the way, the Chinese are taking a very different approach. They’re building much more limited models that just don’t use as much information but get a high fraction of the performance and use a lot less energy. If the world ends up going to that model of AI instead of the all-encompassing ones then we will have just wasted the money. We will have spent a lot of money on building super impressive stuff that nobody actually wants to use. Obviously the railroads still had railroads. You could use the tracks later on. You could use dark fiber. I think the original boom that looks something like this was, in fact, the British Canal boom around 1800. All of those left usable legacies. And this one might not. HCR: Wow. Fascinating. Absolutely fascinating. Thank you. It’s always a pleasure to talk to you. PK: Good to be on. Let’s do this again. Get full access to Paul Krugman at paulkrugman.substack.com/subscribe [https://paulkrugman.substack.com/subscribe?utm_medium=podcast&utm_campaign=CTA_4]

A Whiff of Stagflation

For more videos, visit my YouTube channel [https://www.youtube.com/@PKrugman]. Transcript Hi, Paul Krugman here. Different city, different country, still not home. Unfortunately, couldn’t manage to do this one in a cafe, but we have been sitting in cafes a fair bit. I just want to weigh in on a really kind of alarming report on consumer confidence that came out today. This is the long-running University of Michigan survey of consumer sentiment. It is kind of time hallowed. I don’t know that it’s necessarily the gold standard — there are other surveys — but this is the one that people really do focus on most. The numbers are terrible, people. We’re hitting a record low on consumer sentiment which fits in with the general picture. We know that people are very upset about prices; they’re very upset about economic management; they just don’t feel that there’s anyone making any sense who’s in charge of things; which is all true. I mean we can argue that objectively things are not as bad as all that. We have consumer sentiment that’s worse than at the depths of the financial crisis. We have consumer sentiment that is worse than during the stagflation circa 1980. And it’s hard to say that that’s really justified. But OK, the customer is always right. If people are feeling this down then we need to take that seriously. But that is actually not the big issue. The really big issue is inflation expectations. Now why do we focus on that? Inflation for a short period of time is not good but it’s tolerable. If we have a year of elevated inflation — even if you do something stupid, if you impose tariffs and raise consumer prices, or you start a war and mishandle it and you drive up oil prices that is not good. But it only turns into a really, really serious problem if it gets “entrenched” in the economy. That is usually the term that people at the Federal Reserve use. And what they mean is this. If you think about how wages and prices are set, think about the process of inflation. Not all prices are set at the same time. There’s a kind of a leapfrogging in which each individual company, each individual employer is setting prices based both on inflation in the past and on inflation that they expect in the future. They’re looking over their shoulders at what they think competitors are going to be charging. They’re looking over their shoulders at what they think is going to happen to their costs. And they need to do that because for many prices, it’s impractical and costly and disruptive to change them too frequently. So you set prices for a year in advance, something like that. You set prices for a while, which means that a lot of what’s happening to prices now is determined by what people think is going to be happening to prices in the future. We don’t have great measures of what’s in the minds of people who are setting prices, but we have pretty good, or at least consistent over time, measures of what consumers expect. And, you know, we’re all living in the same society. So that’s telling you something about where we are in terms of expected inflation. If you have a spike in inflation, if inflation comes and goes, but it doesn’t get built into expectations of higher inflation for a long time, then okay, you ride through it. Maybe people vote the bums out, but you ride through it. If it gets built into expectations, then it’s a much a much more difficult situation. Then you have to somehow wring those expectations of high inflation out of the economy because if you don’t, inflation will just feed on itself. Prices will rise because everybody expects prices to rise and those expectations will be confirmed and it just goes on. So if you want to return to an acceptably low rate of inflation and if people are expecting a high rate of inflation, then while there may be other ways, normally what we do is we put the economy through a wringer. which is what happened at the beginning of the 1980s. After the inflation of the 1970s, inflation was eventually brought under control, but that would happen through years of extremely high punishing unemployment. Some people looking at inflation four years ago, looking at the inflation of 2021-2022 predicted that we’d have to do the same thing, that having seen a burst of inflation after decades of low inflation, that we were going to have to go through something like the end of the 70s stagflation, that we’d have to go through a severe recession with high unemployment for years to get inflation back down. Something I called right — we all get things wrong, but something I called right — was that I said no, that that’s not going to happen, that it’s a false analogy. And the reason I said it was a false analogy was because medium-term expected inflation had not gone up very much. Now, we go for medium term because we know that for short-term inflation, well, people’s expectations about that bounce around a lot, often driven by fluctuations in gasoline prices. But medium-term expectations are normally more stable, so if they rise that’s an indicator that you are starting to get entrenched inflation and things will be really bad. In 2022 — sorry, let’s go back to 1980 — medium-term inflation expectations as measured by the Michigan survey were about nine percent — expected inflation over the next five to ten years was 9%. That was really bad. That said that people had basically internalized the inflation of the 70s and expected it to continue indefinitely. This meant that actually getting inflation back down to tolerable levels was very costly and very painful. In 2022, well, expected inflation over the next five years had crept up by a fraction of a percentage point, but it was still quite low. People were not at all building in anything like the expected inflation that prevailed before the great painful disinflation of the 1980s. And so I was quite confident that the dire predictions about what it would take to bring that inflation back down were wrong. Well, guess what? Especially in the last two months, expected inflation over the next five years has gone up a lot. It’s 3.9 percent in the latest Michigan release. That is, it’s not 1980 but it’s really bad. It’s the worst we’ve seen on that number since the early 1980s. It is saying that the person on the street is starting to believe after the tariff shock and now the Iran shock that we’re in a higher inflation environment. And we have to suspect that people making decisions about prices are thinking the same way. They’re going to start building those expectations into pricing. So we’re starting to get the thing that everyone in the economics biz fears, which is entrenched inflation. And if that’s happening, then the costs of the policy failures, the policy foolishness of the past year and a half are going to be a lot bigger than anyone is now reckoning. This is going to be an extremely painful situation that we have. It looks, at least according to these preliminary indications, as if Donald Trump has managed to create the kind of environment that we had at the end of the 1970s stagflation, which means that this is going to be really, really ugly and that we are going to be paying the price for these misadventures for years to come. Happy thought. Have a nice day. Get full access to Paul Krugman at paulkrugman.substack.com/subscribe [https://paulkrugman.substack.com/subscribe?utm_medium=podcast&utm_campaign=CTA_4]

G. Elliott Morris on Vibes and the Midterms

For more videos, visit my YouTube channel [https://www.youtube.com/@PKrugman]. I’m away but alas staying in touch with political news at home, and thought I would check in with one of the best public opinion quants about where we stand right now … TRANSCRIPT: Paul Krugman in Conversation with G. Elliott Morris (recorded 5/13/26) Paul Krugman: Hi everyone. Returning today to G. Elliott Morris [https://www.gelliottmorris.com/], my favorite polling and public opinion analyst. We’ve had an eventful time with redistricting and there’s a lot of stuff going on, so we’ll see where this goes. The big news is, of course, the Court leaving Democrats stunned by overruling the referendum with Virginia redistricting, which now gives Republicans a substantial lead. You’ve been doing some analysis. How should we think about how this changes November? G. Elliott Morris: Yeah. Big picture is: as long as Democrats are still winning the popular vote by four points, they’re still taking back the House of Representatives. A lot has changed over the last three weeks. First, the Supreme Court has invalidated section two of the Voting Rights Act. This was the portion of the law that prevented state legislatures or other state bodies from diluting the power of black voters. Krugman: Right. Morris: This, of course, matters for our partisan calculations, because black representatives in the South tend to be Democrats. Now the Supreme Court has said states can divvy up their votes. Republican-led states in the South, including Tennessee, Alabama, and Louisiana, have since passed, or are about to pass, maps that will take away three Democrats at least, and potentially five. So that is quite a few seats. On top of that, there has been other redistricting news. Virginia voters had passed a constitutional amendment to adopt a Democratic gerrymander that has been struck down. So Democrats in Virginia are going back to their old map, and they will lose two seats because of that—two seats that they would have otherwise gained. So, if you’re catching up on the math here, that’s three seats lost from the Democrats. It would have only been one; now it’s three. But then we have redistricting in Texas, Florida, North Carolina, Ohio, and Missouri. If you add up all of those Republican states, they have taken away about 13 seats from the Democrats, and Democrats have gained only five or so out of California, the only state they have really redistricted. If you add up all this, then Democrats are down about six seats from the gerrymandering wars that Donald Trump started last year. And that could be potentially decisive in a close race. Krugman: My very informal impression was that prior to all of this stuff, the kind of Republican bias of the voting was largely gone, and that the House majorities tended to more or less reflect the popular vote. But now we’re in a situation where we’re back with probably the biggest ever Republican lean there. Morris: Yeah. So the 2024 congressional map was technically still biased toward the Republicans. If in a perfectly average year with perfectly average candidates—and that is the big “if”—if 2024 had been rerun, we would have expected Democrats to lose the majority of seats, even if they had won the popular vote by about a point. The big benefit in 2024 was that recruitment by Democrats in close seats was really, really good. So they beat expectations. But if you rerun it, it still would have been slightly biased towards Republicans. Now we’re at around a Republican bias of four points, which is close to the bias right after the 2010 redistricting. What happened in 2012? So this is pretty bad. We’re getting to the point where the structural bias in basically every electoral institution at the federal level is significantly overweighting Republican votes, just by the fact of where they live or who’s in charge of drawing the maps. Krugman: A few weeks ago, I talked with Kim Lane Scheppele, my old friend from Princeton, about Hungary. And, you know, a lot of what Orbán did was, in fact, basically whatever the Hungarian word for gerrymandering is, but on a heroic scale. She said that they basically weighted rural voters by about 3 to 1 over urban voters. But of course, that was overtopped by a huge wave election. And you would still think that the most likely scenario, given the current polling, is probably still that the Democrats are going to probably crest. Morris: I think the 2026 election will be significantly pro-Democratic, and that the gerrymandering won’t matter. It won’t matter in terms of who wins the majority of the seats. Democrats will still be down six seats, at least, from where they should be. But if they’re gaining twelve, then, you know, they’re still managing to recapture the House because it was so close last time. Republicans only had three extra seats at the last election. So it’s a pretty easy wave election for the Democrats. But they’ll still be down seats, like they’re still deprived of representation in the South. And more importantly, in 2028, when we’re not expecting Democrats to have such a large wave—unless the country comes to its senses. I know you talk a lot about tariffs here. That’s a big example. Then we’re expecting a much closer election. And in that 2028 scenario, this gerrymandering could give Republicans the majority, even if Democrats win the popular vote. Krugman: Just a quick, amateur question on this stuff: to what extent is there the possibility of a “dummy-mander”? I was just thinking about the Hispanic vote—that the Republicans may have drawn these districts on the belief that the 2024 Hispanic vote was going to remain. And they seem to have really lost that, at least if the polling is at all right. Does this mean there’s a possibility that the Republicans have essentially diluted their own support in order to wipe out Democratic districts, and that they’ve opened the possibility of losing a lot of normally red seats? Morris: So, it’s a great question. I’ve done some math on this. My own simulations of election outcomes, where I assume rationality. But Republicans basically went after five districts in Texas. Maybe two or three of those are highly susceptible to a dummy-mander. In which case, if you do the math and Latinos move 20 points toward the Democrats, and everyone else only moves ten points to the Democrats—assuming Latinos are moving twice as much as everyone else, which is pretty close to what happened in the 2025 elections—then Republicans only gain two seats out of Texas, but they’re still gaining seats. So there is a possibility that they have drawn themselves too thin in the case of a big Latino backlash. But they’re just subtracting some seats that they could have otherwise gained. So it’s not the fact that they’re going to lose overall in terms of the overall gerrymandering. In other words, they’re still coming out ahead. Krugman: Okay, that’s slightly depressing, but I’ll take it. I find myself wondering: if we really have a very clear, massive, public backlash against Republicans, but these maneuvers keep them in control of the House, how much damage does this do to legitimacy and feelings about the government? Morris: I don’t know how much worse feelings of legitimacy or approval of the government could get. I mean, approval of Congress is 20%, SCOTUS is 20%, and Trump’s approval rating is 35%—only by virtue of that question being really partisan-polarized. If you actually ask Americans how they approve of Donald Trump’s handling of stuff like prices and tariffs, then it’s closer to 25%. So, it would be very striking to have lower confidence in the US government to solve the problems of everyday people. Basically, this might make an impact on how Americans view the functioning of their democracy or what have you. And actually, from my point of view, that type of education could be useful for stuff like electoral reform or proportional representation. But we don’t have to get into that for now. But it’s pretty bad out there. It’s pretty bleak out there, Paul. Krugman: The unpopularity of Donald Trump is really extraordinary, and the unpopularity of the policies. Things have really gone downhill. Things were really going downhill, I think, even before the Iran war. Morris: Yeah. If I’m telling the story of the Trump administration, I’m looking at five main events besides his inauguration, which is itself a sort of negative signal to the American people. I’m looking at the Liberation Day tariffs in April of last year, which caused a drop in Donald Trump’s approval rating and then mostly trickled along, slightly dipping as every day people are realizing what the administration is doing. They tend to react negatively to the president regardless of what he does. This was true for Biden as well, by the way. It’s just a sort of weird factor of political psychology here. And then the next event I’m looking at is a sort of confluence of immigration events that happened from May to June of 2025. So you have the deportation of Kilmar Abrego Garcia. Donald Trump sends the National Guard to LA and to Chicago. And this creates a lot of negative press attention for him. And you see his approval rating on the economy and deportations overall drop again by about 5 or 10 points. And then he trickles around; he’s losing support. And then over the next six months, really not much happens in his approval rating. And then the government shutdown happens, and a lot of Americans come around to the news that Donald Trump has basically defunded a lot of Medicaid and premiums are going to increase. That, of course, sets in in January of this year. And then I would be looking at the Iran war. You would also want to add the killings in Minnesota as well. That was a big signal to Americans about the negative outcomes of Trump’s deportation agenda and militarization of U.S. cities, essentially. So those events are about immigration and prices. The health care thing is still kind of a price anxiety. And you can see his approval rating dropping. His approval rating was positive when he started. He’s at -25 or so now in our polling; he’s closer to -30. So it is worth emphasizing: very few people like Donald Trump. If you walk down the average American street, you will encounter between three out of ten or four out of ten people who actively support what he is doing. And that’s very low for a president of the United States historically, and in absolute terms, it is just worth emphasizing that people do not like this period. Krugman: Yeah, I had forgotten that he did have a positive approval briefly. Morris: He did have a bit of a honeymoon when he started. But it really deflated very fast. Krugman: And it’s extraordinary, actually. It feels like much longer ago than it was. Morris: There seems to be an awful lot of enabling of Donald Trump on the part of congressional Republicans. A lot more than you would expect based on his approval ratings overall and his approval ratings in their districts. Now, in my opinion Trump reacts to public opinion only from a narcissistic point of view. He shares the polls when they’re good and he calls them fake news when they’re bad. He’s constantly talking about how much the public loves him. That is the type of thing a narcissist would do. But in terms of reining in political actions from the White House, I do think we’ve seen rather little evidence that the polls are meaningfully moving him. Now, there are a couple of cases. The big one is the retreat from Minneapolis after the killing of Rene Goode and Alex Pretti there. There was a dramatic increase in support for abolishing ICE and a dramatic decrease in approval ratings for ICE and the president’s immigration and deportations agenda right after that. So it seemed to matter in terms of public opinion. But look, I think we’re in an environment where most legislators, especially on the right, are insulated from general electorate opinions, especially the opinions of the average person who might not turn out to vote. And that is enabling an awful lot of bad behavior on the part of the president. And partisanship is really an overwhelming force for bad on the right, given the president’s proclivities. So I think you are right here to say, you know, the polls are the polls. And it is important to say that people don’t like this. But Trump is not necessarily the type of actor you would expect given that information. Krugman: I just wonder, because the papers are full almost every day with some scandalous or just outrageous behavior. Kash Patel’s personal brand of bourbon and all of that stuff. But one of the things I learned from you about swing voters, and you have a very straightforward definition of being badly or poorly informed, which is just: do they know who controls Congress? But I wonder whether any of this stuff even reaches a lot of voters. Morris: Yeah, I doubt the average person knows about the Kash Patel whiskey—the “cash money” whiskey. By the way—I’m a big fan of whiskey, and that seems like a real betrayal to all the whiskey fans. Yeah, there’s a problem here, which is that most legislators just really don’t care what the public thinks, including in their district and including overall. And really, nowhere is that clearer than in the Republican Party when Donald Trump is passing tariffs that will cause inflation or asking for $1 billion for his ballroom, etc. And the voters who aren’t really paying attention to the news might not hear about that stuff, but it doesn’t mean it doesn’t matter. And they are getting signals of the president’s incompetence from stuff like gas prices going up. And just a general news environment being bad about the President of the United States. So some of this does filter through to them. Krugman: Okay. Let’s go on to the vibecession. There’s a lot of payoff in the economics and business punditry world for turns of phrase. And Kyla Scanlon with the vibecession, as I think she said. Morris: Yeah, she should get a Nobel Prize for vibecession. Krugman: Yeah, she’s set for life on just that one term, although she’s actually very good on other things. But it is quite amazing, right? Just before we recorded this, the latest survey of household economic dynamics from the Fed came out, and so even leaving aside the approval ratings and so on, public views about how they’re doing—most people still say we’re doing okay—but the views about the state of the economy have just fallen off a cliff. As we might expect, people are incredibly negative. The first question is: do we think that, in some sense, people are more negative than the reality? But do you have a different take on that? Morris: I will just respond directly to the last thing, which is yes, there is a vibecession. The vibes are still lower than you would expect. Even fundamental indicators and even this one that Jared Bernstein has proposed—that I’ve sort of back-tested in some modeling: the excess inflation number. Even if you account for excess inflation, or just price levels being higher than people expect, consumer sentiment as measured by the University of Michigan is still about 10 to 15 index points lower than you would expect. So there is still some level of anxiety out there that is breaking from our historical understanding of economic anxiety. Krugman: Okay. So you think that there is, in fact, still a mystery component, at least based on the consumer confidence index. Morris: Yeah, I guess the other way to say it is that those historical models that predict consumer sentiment are still missing something. Maybe they’re missing that people are reacting more to inflation now than unemployment or other structural variables than they were in the past. And you have to find some way to account for that. I mean, I’ve tried every way possible. Even if you P-hack it, you really can’t get there. Krugman: People may not know, but P-hacking is essentially playing with variables until you get something that is statistically significant. Morris: Except that it isn’t really, because we’ve tried all the alternatives to find the thing that seems to work. By random chance, you would have arrived at an answer. But what I’m saying here is, even by random chance, you cannot arrive at a prediction of consumer sentiment that is perfect. There’s some fundamental break around two years ago in the vibes about the economy. And it’s lower even if you account for stuff like excess prices. And that’s got to be an important part of our story. There’s a subset of internet commentators, mainly on Bluesky, who insist that the economy is actually good and the vibes are just wrong for no reason. I don’t think that is right, but I wouldn’t go so far as to say there’s no vibecession. I think it’s somewhere in the middle. Krugman: Okay, so I guess there should have been a break two years ago. But you think that things are worse now relative to the fundamentals than they were in 2023? Morris: Yes. If you predict consumer sentiment with excess inflation via the S&P 500, economic growth, and such—you can get a very good prediction—almost perfect out-of-sample until December 2024 or January 2025. Krugman: Which is an awfully convenient result if you are focusing on Trump-related sources of economic pressure. Morris: But it also is around the time when the president started passing inflationary policies. So it could just be that people, after January 2025, were hyper-aware of inflationary policies like tariffs, or just the “horse in the hospital” aspect of this presidency. Maybe they’re mapping that onto their economic sentiment. I am still searching for answers for the last year or so. But if you include excess prices in your model of consumer sentiment, this basically fixes, I think, the original vibecession aspect through the end of 2024. But now we’re in a sort of different vibecession environment, perhaps related to Trump. I’m not sure. Krugman: Okay, so “Vibecession II,” which is to go along with “Trump II.” So, you really are saying there is basically a “Vibecession II,” which is interesting, and that there’s something that goes beyond all the solutions that we’ve tried to find to explain why people were so depressed in 2024. Now, even with all of that, something else has happened now. Morris: Yeah. I think I can explain the vibecession of 2022, 2023, and 2024 very well as an excess price shock. Krugman: Alright. Morris: I don’t think we have a good explanation, economically or otherwise, for the 2025-2026 vibe session—”Vibcession 2: Electric Boogaloo.” Krugman: Yeah. Well, of course, people are feeling really bad because we have crazy tariffs, we have cuts to health care. And you know Trump is so terrible. So, of course, people are feeling something terrible. But I don’t think that’s what’s going on in the minds of the average American. So, there is something going on there. Morris: If I were putting on my political scientist—maybe political psychologist—hat, it is very possible that the amount of coverage about the Biden economy in derogatory terms, and inflation and the blame of the president for inflation in 2022 and 2023, caused voters to think about the president more and then think about the direction of the national economy. And therefore, if that is true, then getting a figure like Trump into office would cause a pretty negative backlash in overall economic sentiment, even if it’s not causing this negative backlash in their personal financial situation, as you know. Krugman: Yeah. I mean, it probably doesn’t matter for the public views, but in my view, Biden bore very little responsibility for that inflation. It was supply chain disruptions in the aftermath of COVID, and European Union inflation was basically identical to U.S. inflation. But this time around, you really have to say, well, the 3.8% inflation that we just got is— Morris: Yeah, you could definitely put a “Trump-flation” label on it, at least for the time being. Jerome Powell said so. “Daddy” said so. So you gotta listen. Krugman: Yeah. No, it’s pretty amazing. So for listeners, in all of these discussions, economists like to talk about inflation, which is the rate at which prices are rising. And the conventional approach to understand consumer sentiment is to talk about inflation and then unemployment and maybe some other things as well. But if you talk to actual people, they talk a lot about what things cost. And so there’s this argument that says people are upset because even though inflation came way down from its peak in 2022, prices didn’t. The level of prices leveled off rather than coming down. But this split raises a lot of problems. So why don’t we talk about the excess inflation? Morris: So, this work is based on the theory that voters react negatively to a shock in prices, or really a shock in inflation. Especially if inflation had been low for some amount of time—20 or 30 years in the most recent case. To measure excess prices, I’ve built on economist Jared Bernstein [https://paulkrugman.substack.com/p/talking-vibes-with-jared-bernstein]‘s work. So our model for this is to predict what prices would have been today using inflation from the last 20 years or so. And then we measure the residual between actual nominal prices and the prediction of prices. And at least in my work, when I say “prices,” I mean the index price of vehicles, shelter, and of food. But the results are actually the same if you use PCE. Krugman: Personal Consumption Expenditure, as the Federal Reserve calls it. Morris: Which is like the CPI, Consumer Price Index, but it’s arguably a little bit better, and the Federal Reserve relies on the PCE. Krugman: Yeah. Morris: So it really doesn’t matter what goods you’re looking at. Today, prices are about 15% higher than they would have been given 2% inflation over the last six years or so, basically since COVID. And if you add that variable to some model of consumer sentiment that has traditional measures of economic activity like inflation, the S&P 500, and unemployment, then you do get a much better prediction of consumer sentiment over time, including in the 1970s when the change in the price level was even worse than it was over COVID. Krugman: There’s what I think of as the “Morning in America” problem. You may think people are upset now because things cost a lot more than they did before COVID, but, well, that was also true in 1984 under Ronald Reagan. It turns out that the increase in consumer prices in Ronald Reagan’s first term was almost identical in percentage terms to the increase in prices under Biden. But of course, Ronald Reagan ran as a triumphant rescuer of the U.S. economy. “It’s morning in America”. And Biden was deeply unpopular. And the explanation, which I think all of us working on this have come to, is that at the beginning of the 1980s, people were expecting lots more inflation. And at the beginning of 2021, they were not. And that’s kind of what you’re measuring. Morris: Yeah. And this isn’t just me talking, either. If you look at the political scientists’ voter psychology work on what they call “retrospective economic perceptions” and predict those ratings based on changes in economic indicators, then inflation causes a much more negative impact on economic evaluations after a period of what they call “good times” when inflation is low. So psychologically, this works, too. If people are primed to see increases in prices of 10-15% for a decade and then they see it again, they react less negatively than they do in, say, your COVID-era price spike after 30 years of low inflation. Krugman: Yeah. Although what is kind of interesting—and I know that you’ve been doing statistical modeling and I’m just pulling stuff out of my— Morris: Well, I’m not an economist and I don’t have a Nobel laureate. Krugman: Well, yeah, but that was a long time ago. But anyway, in the mid-’70s, people were still completely shocked. I mean, I’m also an old guy, and I remember the ‘70s, and we were all really, really shocked. And yet, consumer sentiment, even in the Ford administration, was not as negative as it has been lately. And still, times were really good in the ‘60s and up through about ‘73. I’m still kind of shocked at just how bad perceptions are now. But your models seem to track the ‘70s okay. Morris: They do. Yeah. And they do because of this adjustment for the good times versus the bad times. So if you take our excess price measure, which again is just the percent difference between expected prices and actual prices in nominal terms, and you adjust for the average inflation in the CPI over the last decade, then you essentially decrease the excess price measure for the ‘70s and hold it about constant for the post-COVID period, mainly 2023 being the peak. And you get a much better fit in the model. So this is built on our voter psychological theory that people react more negatively to higher prices after a period of good times than bad times. So things are being triangulated here in our overall story of the impact of excess prices, even if, as I said at the beginning, this isn’t a complete explanation for the vibe session here in 2026, which is somewhat different somehow. Krugman: Just an interjection—I’m a garrulous old guy here—but I associate stagflation with the taste of Hamburger Helper because I was working summers as an undergraduate as a research assistant, and my friends and I, in our dreadful shared apartment, were using a lot of Hamburger Helper because we didn’t know how to cook. And also meat was really expensive, or seemed so at the time. So, yeah. But it’s interesting that people were not as depressed. And I think that maybe they had already kind of internalized that the economy can be tough or something. Morris: That’s, in effect, what we’re saying here. They weren’t as surprised. They’d internalized high prices as something that could happen in their lifetimes. You know, I was but a twinkle in my daddy’s eye in 1970. But you can do a lot worse than Hamburger Helper. Hamburger Helper is a good staple food for your working-class person. Krugman: Well, I had some roommates who insisted on soybeans with everything, and that I could have done without. But anyway, it was the ‘70s. So this question of what do we think are the prices that people expected—and you’ve been basically fitting a trend to recent price movements, right? And projecting forward? I think you’re using something like the average inflation rate over the past five years to project forward? Or how are you getting that? Morris: For excess prices? I mean, it’s the trend in prices. So that is mathematically equivalent to the average inflation from the 15 years prior to whatever date you are predicting on. 15 years prior to the five years prior. So the idea is that people have formed their expectations for inflation over some period of the last ten years on average. Krugman: So, we have direct measures, supposedly, of what inflation people expect. There are surveys. There’s University of Michigan. And some surveys, but especially University of Michigan, do ask people what they expect the inflation rate to be over the next 5 to 10 years, which kind of gives you a medium-term expected inflation. And you can get an implied inflation forecast out of the bond market—the TIPS spread, the break-even inflation, whatever jargony stuff. But there is an implied inflation forecast. So those are not necessarily congruent with lagged— Morris: Just the excess price measure. Krugman: Yeah. So here’s my question: let’s say consumer expectations of inflation over the next five years are somewhat elevated now. They’re higher than they were. This is not, I think, the way it comes out in your analysis, but I would have thought that would make it easier to end the excess price stuff. Because if you want to have prices lower than what people are expecting, given that they’re expecting higher inflation at this point, then they’ll be pleasantly surprised if we only have 2% inflation. But I think that is not how you’re seeing it, right? Morris: No, I’m not using the survey measure of what you would expect your inflation to be over the next five years. Krugman: So what you’re doing is sort of saying that people’s expectation of inflation is something like inflation over the last five or ten years. And you have actually used the expected inflation of the survey, which says, “What if inflation is actually that high, and then it’s going to be really bad ?” But I would have turned that around and said, “Well, people are already expecting pretty high inflation, so they’ll be pleasantly surprised if it’s lower than that. And that should make it easier to get back to a price level that people find acceptable.” But I don’t know if I’m making sense. Morris: Yeah, you’re making some sense. But we are not using a psychological measure of excessive prices, and that is different from a survey measure. We are using an actual mechanical level of excess prices from the residual of the trend. So one way to reconcile the fact that the objective measure of excess prices, rather than the survey-based measure, is more explanatory—you could say people are bad at predicting prices in the future, just in general, which would be true. Or that the survey isn’t picking up on anxiety about the price level with that variable as well as you would expect. And one thing to mention here is that the University of Michigan’s measure of what I’ve called “price anxiety”—which is just the percent of people who have a bad opinion of the economy—the percent of those people who attribute it to worse personal finances is at an all-time high. And it surged in 2021 and 2022 and stayed there; it never came back down. Which is similar, you’ll notice, to the trend on consumer sentiment through the University of Michigan. I don’t have the Conference Board data in front of me or memorized. So it is possible that people are bad at predicting what prices should be. One idea would be if we took the expected change in prices over the next year and divided it by average CPI—overall inflation—over the ten-year period preceding. I wonder if that number would be at an all-time high. That’s a very easy check after the fact. I bet it would be near an all-time high. Krugman: Probably getting too meta, but what we’re trying to predict is a variable that is consumer sentiment, which is not a behavioral thing. It’s like asking, how do people answer a questionnaire? And this is a question: should we also be using questionnaire-type answers to predict it? Obviously, at some level we’re interested in objective economic stuff, but I wonder whether we should inherently prefer the objective economic stuff as a way of predicting. I’m not making a whole lot of sense here but— Morris: No, this is making sense to me because I spend a lot of my time thinking about the difference between our perceptions of objective reality and these survey-based measures which, for whatever reason, can deviate from that. And my argument would be that we should be using the survey-based measurement of anxiety in addition to our “economic fundamentals”—our structural variables—because our models and our job as modelers is fallible. And we can’t rely on the people when they tell us in surveys that they are anxious for whatever reason, instead of pouring cold water on it because our models don’t line up. Maybe this is just my opinion, but you are right. Of course, we want to know how people are reacting to objective conditions on the ground. And the only way we can really do that is by looking at the match between executive positions lying around and some other outcome variable. So, I’m acknowledging it’s tricky. There’s no clear answer, I guess. Krugman: So, two questions left. S&P 500, and again, if people don’t know, that is the broad index of the stock market—that really shows up as something that explains how people feel. Morris: Yeah, the annual change in the S&P 500 is pretty direct to consumer sentiment, even after controlling for stuff like your annual change in CPI, PCE, etc. Krugman: So that’s really kind of interesting, because the vast majority of Americans own very little stock, so the impact of the S&P 500 on most people’s economic position is really kind of small. I’m wondering whether that’s more like a signal. People like me are always saying the stock market is not the economy. But it’s not clear that’s how people see it. Morris: Yeah. The S&P 500 impacts media coverage quite a lot. And in our models, we try to control for negative media sentiment. But again, our empirical analysis of media sentiment is often different from how people are interpreting this. So I tend to really land on one answer here, which is, if you look at the polling on price anxiety—the percent of people who are saying their situation is worse because of personal finance issues—that’s at an all-time high. And if you trust the people, that is pretty explanatory of consumer sentiment on its own. But it requires some hurdles to get there. Chart 7A in the University of Michigan shows the percent of people whose finances are worse and who say personal finances is the reason why. Krugman: Yeah. So I mean, at some level, if our numbers say that personal finances are actually better, but people say they’re worse, at some level, customers are always right. Morris: Yeah. Exactly. But that leaves us at a loss for an explanation. It does leave us putting our shoulders up. Krugman: Last thing. And again, I’m just throwing stuff out there because I’m puzzling over this stuff myself. So a lot of these issues are in some ways harking back. I still always think that “Morning in America” in 1984 is in many ways a crucible for making sense of all this stuff. But 1984 as a year was closer to the end of World War II than it is to today. And it was a very, very different country then. And I always wonder, are we trying to get a model that fits a society that has changed immeasurably over time? Morris: Yeah, absolutely. I mean, so much of the vibe session discourse—not necessarily Kyla Scanlon, but I believe Nate Silver wrote this article for the New York Times Opinion Page that was about how the consumer sentiment index broke down. This was after The Economist had done something similar, I believe. Much of the discussion of that was based on the idea that you could build a model of consumer sentiment historically. And now it’s breaking down. And one conclusion from that is that people aren’t thinking about the economy rationally anymore. But another conclusion is that they’re thinking about the economy differently than they have been previously. And that seems entirely legitimate to me. And if that is the case, then we should be looking at the polling data and the perceptions data more and the fundamentals indicator less to explain consumer sentiment. Krugman: Okay, but the big news to me is that we really are seeing sort of a second downward leg in the vibecession. Morris: “Vibecession 2.0,” yeah. Krugman: Which is really quite remarkable. It’s going to matter enormously in many ways, obviously in the elections. So that’s news to me and actually worth highlighting. Morris: Well, now you have a headline. Krugman: Now I do. Hey, gotta feed the beast on Substack, as you know. Morris: Yeah. Right. Well, look, in terms of consequences, and maybe to go back to where we started: Donald Trump’s approval rating on prices is like 30% or less, and from 70%—it’s down -40 or so. And that was the last time I looked at this, which was a week and a half or two weeks ago. And he’s been losing ground very fast. That is congruent with an electorate that is very upset about prices, even if the objective data don’t explain why to us. So there’s some triangulation of the anxiety in terms of evaluations of the president. And if that number stays as low as it is, then we should expect the type of rout in the midterms that is large enough to overcome, basically, effectively, the Republican cheating through gerrymandering over the last decade or so. That might be where I would leave it. Krugman: Yeah. Get full access to Paul Krugman at paulkrugman.substack.com/subscribe [https://paulkrugman.substack.com/subscribe?utm_medium=podcast&utm_campaign=CTA_4]

My President Went to China, and All I Got Was Even More Expensive Gasoline

Transcript My president went to China, and all I got was even higher gas prices. Hi, Paul Krugman here. Recording a video today rather than one of my usual posts,.I just feel like I’ve been too immersed in charts and wanted to do something shorter, simpler, whatever. It’s Friday. The summit between Trump and Xi Jinping has wrapped up. Hard to say what exactly was accomplished, but at least Trump is saying that China is going to buy more US oil. That’s the headline on at least some of the financial sources that I look at. And in fact, crude oil prices in the United States have risen at least a little bit in response to that announcement. The odd thing is that Trump seems to think that China buying more US oil is a good thing, which it is probably not from the point of view of the United States and definitely not for him. So let’s walk through what all of this means. Background: The United States, which used to import a large part of its oil, is now a net exporter of oil. That’s what shale oil did. Since about 2020, we’ve been selling more oil to the rest of the world than we buy. Now that’s a net number. It’s actually a fairly complicated picture because there are different kinds of oil and the geography of oil creates some complications. So we have the upper Midwest importing oil from Canada, Texas exporting oil to Europe and it’s a bit of a mix of of where the oil is produced and where it goes so that there are actually quite large gross flows underneath that net transaction. But, okay, that’s not all that relevant here. What is relevant, you might think, is that the United States overall produces more oil than it consumes. And you might think, in fact, some people still seem to imagine, that this insulates the United States, that when the Strait of Hormuz is closed and global oil prices skyrocket, well, America, which is actually selling more oil than it buys, should be a net winner, which arguably is true for the economy as a whole, although as I’ll say in a minute, even that’s not entirely clear. But it’s definitely not true for most of the U.S. public. Most of us have little or no stake in the oil industry, but we buy gasoline and diesel and we buy products whose price ultimately includes the cost of gasoline and diesel. So higher oil prices hurt most Americans. And it’s the global situation that has led to higher prices in the United States. If you look at net exports of oil and petroleum products before the Iran war started the United States was exporting about 2.9 million barrels a day on net of oil. Now we’re exporting about 5.8 million barrels a day. That’s a response to the very high prices that buyers in Europe and Asia are willing to pay for oil now that the Strait of Hormuz is closed. That’s how markets work. That’s a significant contributor by the way to easing the impact of Hormuz on the world economy outside the United States. There were something like 20 million barrels a day of oil being shipped through the strait. Some of that is being routed around through pipelines across Arabia to the Red Sea. But a significant part of it, something like 15% of that shortfall is being made up by increased exports from the United States. We still have a looming crunch because a significant amount of oil demand is being met by drawing down inventories and we’re kind of getting to crunch point there but that’s a whole other issue. Now the downside for the United States is that more oil shipped abroad, unless we have a large increase in US production— which is not happening and won’t happen any time quickly — that means more oil being shipped abroad means less oil for the US market so prices have risen. And so here we have the United States which is self-sufficient and more than self-sufficient in oil, we’re nonetheless seeing gasoline prices $1.50 or $1.60 a gallon higher than they were before the war started and diesel prices up even more than that. So this is not a great thing for, this is a bad thing for most Americans and having the Chinese buy some more oil — hard to know what if anything is going to come of this — but having the Chinese buy even more US oil is going to enhance the negative effect. It’s going to drive gas prices even higher. Now there are beneficiaries. Basically, oil producing companies are getting a windfall. They’re getting a much higher price for their product. West Texas Intermediate, the benchmark price of the United States, was around $65 a barrel before the war started. It’s around $102 as I record this. So that’s a pretty big benefit in terms of profits for a select group of oil companies. Who benefits from that? Because in the end, corporations are not people, they are ultimately owned by people. Well, okay, people who own stock in oil companies are the ultimate beneficiaries of these higher profits. So who are those people? We don’t know exactly. If you look at it, it turns out that oil industry stuff is largely owned by institutional investors who in turn have other investors and it’s a more than a little bit not transparent exactly who the beneficiaries are here. But in general what we know about US stocks is two things: A significant fraction are owned by foreigners. There’s a little bit of dispute on this, but I’ve seen estimates that run as high as 40% of U.S. equities being foreign owned. If that’s the case then of these excess profits something like 40 cents on the dollar might be going to foreigners. Not totally sure about that number but what we do know is that among U.S. investors, among the U.S. public, stocks are basically held by a small fraction of the population, about half by the richest 1%, another 37%, according to the latest numbers I’ve seen, by the next 9%. So 10% of the US population, and this by the way includes mutual funds, it includes pension, it includes your TIAA — sorry TIAA, that’s only what academics have — it includes your 401ks. So US stocks are overwhelmingly held by a small fraction of the population. The great bulk of the U.S. population has very little stake in the stock market. For all the talk about it, it really is not something that’s terribly relevant to most people. On the other hand, almost everybody has to fill up their tanks and even if you don’t, even if you are carless in New York City, which is not very many people but anyway, even so, the price of almost everything you buy is affected by the price of fuel. And it’s affected by the price of fertilizer, which also is very much petroleum related. So, on balance, certainly 80, 85% or more of the US public is a net loser from higher oil prices and hence a net loser from increased US sales of oil abroad. Okay, you can think of a couple of ways that you might be able to change that conclusion. It would be more beneficial to the US public at large if oil companies paid a lot of taxes on their profits. Well, I can stop right there. Obviously the oil industry has historically been famous for not paying very much in taxes. It could be a good thing for the American public if wealthy investors who have capital gains as a result of this surge in oil prices pay a lot more in taxes. But again we can stop right there. The U.S. system in general gives people who derive their standard of living, their wealth, their income from capital gains a much, much lower burden than ordinary people. I mean, the income tax system is progressive. The income tax rate on — my favorite line from the movie Wall Street, $400,000 a year working Wall Street stiffs — they pay quite high personal tax rates, especially of course if they live in New York City. But the people who are getting their money from stocks and from gains in stock prices pay much lower tax rates. So this is not going to be a significant source of revenue and therefore it’s not going to ease the burden of paying for government on the rest of us. So very hard to see how you can treat increased Chinese purchases of oil as a win for America. It’s a win for people who benefit from higher oil prices, but that’s a small group of people, and it’s a loss for people who are hurt by higher oil prices, which is almost everybody. Why we should think of this as a positive outcome, well, obviously, I’m tired of pointing out things that Trump doesn’t understand, but what you would think is a little peculiar is that this is bad politically. I mean the price of gasoline has become a real flashpoint in the US political debate. You could argue that it’s looming larger than the actual share of gasoline in people’s budgets can justify. But in this case: historically, presidents have had very little impact on the price of gasoline. It’s always been a kind of a standing complaint among political observers that this price that presidents really don’t control should play such a large role in politics. Except this time around the price of gasoline is higher because Donald Trump decided to start a war. End of story. So in this case to the extent that it’s a negative — and the approval of Trump on prices in general and gas prices in particular is incredibly negative — you would think that he would know that getting China to buy more US oil is not something that you want to do now. It’s certainly not an achievement that you want to trumpet, but here we are. In the end I I actually don’t think this is going to happen. I think that the Chinese will in practice do what they’ve done on previous trade agreements, which is just say that they’re going to do stuff and not do it and it’ll all get kind of lost in the shuffle. But to the extent it happens, this is not a gain. If this was the major consequence of the summit, the United States scored another own goal. On that note, have a great day. Get full access to Paul Krugman at paulkrugman.substack.com/subscribe [https://paulkrugman.substack.com/subscribe?utm_medium=podcast&utm_campaign=CTA_4]

Why Did Trump Take Elon Musk to China?

Transcript What’s good for Elon Musk is not necessarily good for America. In fact, it may go the other way around. So why did Donald Trump take Musk and a bunch of other top executives to China with him? Hi, Paul Krugman, again from a cafe, a little noisy behind me, but I hope it’ll be tolerable. So Donald Trump has gone to Beijing. I wrote something about it earlier today, about the economics and about the generally pathetic state of the United States in geopolitics right now. But I want to focus for this video on the remarkable decision of Trump to bring a bunch of wealthy executives, in the case of some of them, like Musk, extremely wealthy executives, with him on a trip that is supposed to be something about serving the interests of the United States. America’s corporations are not America. They really have very distinctive differences in interest from those of the general public. You may have heard the old line that what’s good for General Motors is good for America. That’s not exactly what the CEO of General Motors said. What he said is that what’s good for America is good for General Motors and vice versa. But in any case he said that a very very long time ago, when corporations’ role in American life was not what it is now. General Motors at the time was a “stakeholder” corporation. That is, it did not see itself as solely serving the interests of stockholders. It viewed itself as having multiple groups that had a stake in the company. There was the workers who were represented by a powerful union. There were customers who were considered to be part of the story. They played a role in the wider community. Today corporations ruthlessly maximize value for stockholders, unless they do it for the founder who is considered to be the owner. (It’s not entirely clear that Tesla is run in the interest of Tesla stockholders. To a large extent, it’s run just in Elon Musk’s interest, but it’s certainly not run in the interest of U.S. workers or U.S. national security or anything like that.) Why then should we care? It’s probably worth knowing that to the extent that corporations are run in the interest of their stockholders, the stockholders of an “American” corporation are by no means necessarily American. We think that something like 40% of US equities are owned by foreigners. So anything that enhances the profits of corporations, you should think of 40 cents on the dollar of that gain actually going to other countries. And among Americans, stock ownership in the United States is extremely concentrated in the hands of the top 10% of the population, a large fraction just in the hands of the 1% or less. and most Americans have very little stake in stock prices. They may have some stake in the success of business in the United States, but that doesn’t have to be what we consider American corporations. It’s not really right to think of Tesla or NVIDIA, whose Jensen Huang also went to China, as being somehow America going to China. These are corporations that serve stockholders around the world, serve some tech bros who have a special control over them. What they want is profits . What they want is access to the Chinese market, including being able to sell China stuff that from the US national point of view maybe we shouldn’t be allowing them to sell — you know, highly sophisticated equipment that on national security grounds we should actually try to restrict the access of fundamentally unfriendly powers. Anyway, we know that’s what’s good for Nvidia is definitely not good for America. What’s good for Elon Musk is more problematic but there’s very little reason to think that any business advantages that Tesla might gain out of this, or xAI, or whatever whatever enterprise is he’s hoping will realize some gain, that this is going to redound significantly to the benefit of US workers. T,o the extent that it benefit redounds the benefit of these guys the people who are on the plane, why should we care? An extra billion dollars in the hands of Elon Musk or Jensen Huang doesn’t do anything for the great majority of Americans. And yeah, it does something for them, but not very much, right? When you have that much money, a billion here, a billion there, and what’s the difference? So this is a really peculiar group to be taking. unless you try to think about what does Donald Trump want? Well, from Trump’s point of view, his son Eric, who runs the family business, was on the plane. They claim it’s just it’s just a family thing — yeah, right. He might as well have been walking around Beijing with a sign that says — in block capitals, of course, this is Trump — BRIBE ME. That’s very clearly what that’s about and as for the rest, well, you know, these corporations are in a way Trump’s base or at least they gave him a lot of money both in campaign funds and directly in one way or another. I’m still wondering, by the way, why do we need a billion dollars for that ballroom? I thought the corporations were were paying for the ballroom by bribing Trump. But maybe I don’t know where that money is going. Anywa,y whatever the story, these are not U.S. national interests being represented here. The whole visit — aside from the fact that it’s humiliating, that it’s really a pathetic display of U.S. weakness and Chinese strength — the whole visit is also yet another spectacular example of the corruption that now pervades everything about U.S. governance. And we should be angry. We should be outraged. We certainly shouldn’t allow Trump and company to spin whatever comes out of this as a victory. We mostly defeated ourselves here, but we certainly aren’t getting anything for us. Maybe something for Elon Musk comes out of this, but there’s nothing for the rest of us coming out of this essentially tributary visit to China. Take care Get full access to Paul Krugman at paulkrugman.substack.com/subscribe [https://paulkrugman.substack.com/subscribe?utm_medium=podcast&utm_campaign=CTA_4]