Austan Goolsbee is one of Tyler Cowen’s favorite economists—not because they always agree, but because Goolsbee embodies what it means to think like an economist. Whether he’s analyzing productivity slowdowns in the construction sector, exploring the impact of taxes on digital commerce, or poking holes in overconfident macro narratives, Goolsbee is consistently sharp, skeptical, and curious. A longtime professor at the University of Chicago’s Booth School and former chair of the Council of Economic Advisers under President Obama, Goolsbee now brings that intellectual discipline—and a healthy dose of humor—to his role as president of the Federal Reserve Bank of Chicago.
Tyler and Austan explore what theoretical frameworks Goolsbee uses for understanding inflation, why he’s skeptical of monetary policy rules, whether post-pandemic inflation was mostly from the demand or supply side, the proliferation of stablecoins and shadow banking, housing prices and construction productivity, how microeconomic principles apply to managing a regional Fed bank, whether the structure of the Federal Reserve system should change, AI’s role in banking supervision and economic forecasting, stablecoins and CBDCs, AI’s productivity potential over the coming decades, his secret to beating Ted Cruz in college debates, and more.
Watch the full conversation
Recorded March 3rd, 2025.
Read the full transcript
Thanks to an anonymous listener for sponsoring this transcript.
TYLER COWEN: Hello, everyone, and welcome back to Conversations with Tyler. Today, I’m very happy to be chatting with Austan Goolsbee. Austan is one of my favorite economists. He always thinks like an economist, is how I would put it. He has had a long-standing teaching post at the University of Chicago, served in the Obama administration, and now is president of the Chicago Fed. Austan, welcome.
AUSTAN GOOLSBEE: Tyler, thank you for having me. What a treat for me this is. I really appreciate it.
COWEN: What is it in academic macroeconomics, or just economics, that you found surprisingly useful being a Fed president?
GOOLSBEE: I was a data guy, as you know, in the field of economics. Soon as I got there, there’s all this pressure from the press and from others: “Are you a dove? Are you a hawk?” I used to say, “Look, I’m not one of the birds. I’m in the data dogs.” The first rule of the data dogs is, there’s a time for walking and a time for sniffing, and knowing the difference between those. I would say that discipline of academic economics — getting into the data — is super useful.
Then, we are used to thinking about causality and identification. I do think we could use a little more of that in a macro context.
COWEN: Say you’re trying to figure out the connection between the money supply and the rate of price inflation. What’s the first mental model you put on as a hat? You might disregard it when the data tells you otherwise, but where do you start?
GOOLSBEE: You sniff around. I’d say there are multiple models. There’re some people, and embodied in the machinery of the FRB/US — official Federal Reserve Bank model — is probably aKeynesian ‘inflation comes from overheating’ idea. I’d say the old-style ‘money supply is what’s correlated with inflation’ — a lot of those relationships between M2 and inflation, or that sort of thing, feel a little antiquated. They broke down in the data.
COWEN: Why are those wrong? What’s the theory in your mind? You’re trying to teach me. I’m in your class.
GOOLSBEE: Let me finish one thought, and then let’s come back to what’s wrong about it. I still like, most of all, the basic supply-and-demand framework, and before you can conclude anything, you’ve got to get a taste of, is this a supply shock or is this a demand shock?
In a way, a lot of the machinery of central banking and macro analysis, let’s call it, is oriented around demand. I’m not disputing that, in the past, that has been the source of the most frequent business cycle variations, but I’ve tried to caution everybody in weird moments, like when you’re getting major developments on the supply side — whether they’re labor supply, or supply chain, or productivity growth, or a number of things that are hitting the supply side.
Maybe all bets might not be off, but the training sample LLM version of being a central banker is going to be prone to hallucination problems because it’s going to give you things that are wrong because supply shocks might be driving inflation, not demand.
Then go back to your other question of, “Well, what’s wrong with taking M2? And why would it no longer be as correlated with price inflation?” I think that a lot of that is because of financial innovation that we have at great pride, and I love the cash vault at the Chicago Fed. They don’t like it if I say exactly how much money is in there. I’ll just say, many tens of billions of dollars of cash are in that vault, and we run hundreds of millions a day in and out.
There was a time when that cash use was central to the payments of the United States, and bank accounts, and checking accounts, and writing a physical check — again, central to the function of the financial system. As we’ve spread to electronic payments and credit cards and debit cards and that sort of thing, it has made . . . I guess in the old model, you would say it’s radically changed the velocity of money.
It wasn’t just M equals PY. It was M times V. If M is moving around, and V is moving around at the same time, you’re going to get a little mixed-up interpreting to over-indexing on that theory.
COWEN: Okay, if the instability comes from the velocity side, that means that we should favor a monetary-growth rule to target the growth path of a nominal GDP, M times V, right?
GOOLSBEE: [laughs] Yes, and now you’re going to get me in trouble, Tyler. Here’s the thing I’ve known —
COWEN: You can just say yes. You’re not in trouble with me.
GOOLSBEE: I’m not going to say yes because, remember, I don’t like making policy off accounting identities. There’s no economic content in accounting identity. If you are trying to design a rule, that rule may work if the shocks are the same as what they always were in previous business cycles. I called it the golden path.
When we came into 2023, you’ll recall the Bloomberg economists said there was a 100 percent chance of recession in 2023. They announced it at the end of 2022. That’s when I came into the Fed system, the beginning of ’23.
That argument was rooted in the past. There had never been a drop of inflation of a significant degree without a very serious recession. Yet in 2023, there was. Inflation fell almost as much as it ever fell in one year without a recession. If you over-index too much on a rule that implicitly is premised on that everything is driven by demand shocks, I just think you want to be careful over-committing.
COWEN: I’m a little confused at the theoretical level. On one hand, you’re saying M times V is an identity, but on the other hand, it drives inflation dynamics.
GOOLSBEE: It’s why I started back from the — I bring a micro sentiment to the thinking about causality and supply and demand. I sense that you want to bring us to agreeing on a monetary policy rule, and I’m inherently a little uncomfortable. I want to see what the rules say, but I fundamentally don’t want us to pre-commit to any given rule in a way that’s not robust to shocks.
COWEN: Now, you mentioned the post-pandemic inflation and the role of the supply side. When I look at that inflation, I see prices really haven’t come back down. They’ve stayed up, and I see service prices are also quite high and went up a lot, so I tend to think it was mostly demand side. Now, why is that wrong?
GOOLSBEE: There’re two parts to that. I won’t say why it’s wrong, but here are my questions. If you’re firmly a ‘this-all-came-from-demand’ guy, (A) you’ve got to answer, why did inflation begin soaring in the US when the unemployment rate is over 6 percent? Or we could turn it into potential output terms if you want, but output is below our estimate of potential. Unemployment is way higher than what we think of as the natural rate, and inflation is soaring. That already should make you a little questioning.
COWEN: I can cite M2. You may not like it. M2 went up 40 percent over a few-year period, right?
GOOLSBEE: Two, the fact that the inflation is taking place simultaneously in a bunch of countries of similar magnitudes that did not have the kind of aggregate demand, fiscal or monetary stimulus that we had in the US is also a little bit of a puzzle.
Then the third is, if you don’t think it was supply, then you need to have an explanation for why, when the stimulus rolls off, everything about the stimulus is delta from last year. We pass a big fiscal stimulus, we have substantial monetary stimulus that rolls off, the inflation doesn’t come down. Then in ’23, when the supply chain begins to heal, you see inflation come down. Those three things suggest there’s a little bit of a puzzle if you think it was all demand.
COWEN: No, I don’t think it was all demand, but you mentioned other countries. Switzerland and Japan — they import a lot. They were more restrained on the demand side. They had much lower rates of price inflation. That seems to me strong evidence for being more demand than supply.
GOOLSBEE: Wait a minute.
COWEN: I’m waiting.
GOOLSBEE: You’re going to bring in Japan?
COWEN: Yes.
GOOLSBEE: And you’re going to try to claim that Japan’s low inflation is the result of something in COVID? Japan had lower inflation all along, for decades before. They were going through deflation.
COWEN: But if it was mostly supply, a supply shock would’ve gotten them out of the earlier deflation, right? A demand shock would not have.
GOOLSBEE: Let’s back up and try to look at big countries that are not export-oriented, that are primarily domestically driven like the United States. I think imports are what, 10 percent, 12 percent of personal consumption. You take Nigeria, you take India, you take a series of big economies that are primarily domestically driven. Why was inflation high in all of these places simultaneously?
I do think there’s a puzzle. I don’t think it’s 100 percent supply. I do think there was a serious demand component, but you’ve still got to explain when the demand stimulus rolls off, inflation doesn’t go down, and then when the negative supply shock rolls off, inflation does go down. I’m still curious about it, but I find the argument that this was predominantly or entirely demand unpersuasive for some of those reasons.
COWEN: Let me try a question from the other extreme. Why does money matter at all? Isn’t it a very close substitute for T-bills? You’re paying interest on reserves. You balance that rate with other interest rates. There are slight differences in terms of maturity and liquidity profiles, but it could be like the proverbial swap of two nickels for a dime. Maybe money just doesn’t matter. Could you teach me why that’s wrong?
GOOLSBEE: I need to ask you to teach me why it’s wrong. I know you’ve raised this before, and it’s not really an answer to say the market seems to feel like they’re not the same, that there’s some yield to safety that we don’t quite understand, I’d say, and it becomes more relevant perhaps in a world where people are creating money-like deposits all over the place. We’ve got stablecoins as well as the use of credit cards, debit cards, a whole bunch of things that look like old-fashioned bank accounts or checking accounts.
I guess the long answer is the definition of a money-like deposit — why is it different from explicit money? I don’t totally know, but part of me wants to caution you back from going back to M2 in a world like that, where there are imperfect substitutes for money.
COWEN: It’s weird that we economists can’t explain it, right?
GOOLSBEE: Yes, probably. It is probably weird. But look, you’re going to find yourself back into the data dog caucus and out of the pure theory.
COWEN: Do stablecoins increase the true money supply or the rate of price inflation? It is a kind of intermediation. Say they’re backed fully by treasuries. Let’s say it’s for real; it’s not fraud.
GOOLSBEE: Now, you just sent a chill down my spine when you said it’s backed by treasuries. They have the treasuries in there?
COWEN: Yes, 100 percent reserves, and someone creates a stablecoin.
GOOLSBEE: I’m liking the sound of that. I’m not anti-crypto. There’s a lot of anti-crypto, but I am nervous. If you tell me you’re going to create money-like deposits, that people are going to put their money in there, and they can take it out whenever they want, I feel like either (a) there better be some deposit insurance, or (b) there better be some pretty serious restrictions on the assets backing that or restrictions on what the entity can do with the money, or else that thing ends in bank runs and tears. We have 500 years of financial history that tell us that.
COWEN: I’ll come back to the stablecoin.
GOOLSBEE: I’m just a little nervous about that.
COWEN: You ought to be very nervous, though. We’ll come back to stablecoins, but if you look at banking as a whole, I’ve seen figures that formal legal banks are about one-fifth of the lending total. Private equity is more important.
GOOLSBEE: It’s that low.
COWEN: The real sector of the economy, and that’s not FDIC insured, and a lot of that is, in fact, runs-prone, as we learned in 2008, so you’ve just got to be hyper, super nervous.
GOOLSBEE: Yes. Look, that’s the very essence of financial stability. The conundrum of financial stability that we face as a society and as central banks around the world are exactly this — that there is an official banking sector that we have oversight of. Then there’s a shadow banking sector, which is getting bigger and bigger every year and can be prone to runs, and that runs. What is the role of a central bank as a lender of last resort in an environment where it doesn’t play?
COWEN: By trying to keep banks special, haven’t we created a world where 80 percent of the lending is outside banks, which is much harder for the Fed to deal with, even if you should sometimes step in, and that at the current margin, it’s counterproductive to keep banks legally special because we drive more and more business out of banks.
GOOLSBEE: That was two different statements. One is, how is that a problem? The second was the normative — isn’t it useless, and so we shouldn’t do it?
Jeremy Stein is an old friend, and I’m a massive admirer of his work. When he was a governor at the Fed, he highlighted exactly this, that we have to strike a balance in financial stability. If you squeeze really, really hard on one side, there can be regulatory arbitrage, and it can lead to activity to shift to the area where we don’t control, and oftentimes, don’t even have the information. I agree with you on that part.
I don’t know if I would go so far as your second statement, that therefore it’s counterproductive oversight. Raising capital ratios on official banks is not effective because it leads to shadow banks.
Very early, I arrived at the Fed, and then the Silicon Valley Bank events happened. I was there back in 2009 when the big banks were the centerpiece of the problem. The fact that in March of ’23, the big banks were not the problem, I think a lot has to do with all of the efforts that were put in place to raise capital at those larger banks. We can argue about whether it was done right, but you’ve got to admit it, it made you feel better that it wasn’t the biggest official banks that were facing critical deficiencies in capital.
COWEN: On the Great Financial Crisis, I have a question. I’m curious how you frame this. In 2009, I was convinced we had had a housing bubble. Today, I look at real estate prices, and I think most of the country didn’t. Maybe suburban Orlando did. The prices were right, and we had some anti-bubble panic, and that was the problem.
GOOLSBEE: That’s a fascinating point.
COWEN: What’s your view?
GOOLSBEE: I hadn’t thought of it that way. My view has been more, part of the job of the Reserve Bank presidents is we have a district, and our district in Chicago is heart of the Midwest — most of Wisconsin, Iowa, Illinois, Indiana, Michigan. I’m out talking to business people. I’m talking to individuals, and overwhelmingly, what you hear is despair — I would even call it despair — about the cost of housing. That housing — they can’t move.
This is not just in cities where you could argue a lot of it maybe has to do with building codes and zoning. We went out to the Iowa Farm Bureau, and in rural Iowa, I asked them, “What’s the biggest problem?” They said, “Attracting workers.” I said, “Why is it so hard to attract workers?” They said, “Because they can’t afford to buy housing.”
I’ve spent a long time trying to think that through. It’s not wrong that it’s just more extreme. The last couple of years, house price inflation has been radically higher than goods price inflation. If you just compare buying a house to buying stuff at Costco or Target, there’d been a big differential, but what’s important is that’s not new in the last three years. That’s been going on literally for decades.
If you take the 12 years before COVID, house price inflation was 3.5 percent or 4 percent a year, and goods price inflation was actually deflation of around 1 percent a year. The relative price of housing has been rising 4 percent, 5 percent per year for a decade and a half. It doesn’t take a PhD by any means to recognize that something compounding at 5 percent a year is going to add up to a big number. I think it goes to your question of, “Well, maybe it wasn’t a bubble.” I don’t fully understand why the relative price of housing has been trending upward like this. I find it hard to explain.
I have a paper, you might’ve seen, with Chad Syverson, that’s about negative productivity growth in the construction industry over long periods of time, which is itself a puzzle. Maybe that’s part of it. Some component of it may be regulatory in nature, but as I say, you see it in rural areas, too, where the land use regulation is not as prevalent. I think that’s a real puzzle.
COWEN: On the construction productivity puzzle, what do you think is the main reason for that? That it’s negative. If it were zero — that would be a little easier to understand, but we’re forgetting how to build homes?
GOOLSBEE: Yes, somehow, we’re forgetting. We’re doing it worse. We tried every way we could think of to analyze this. Maybe it’s just mismeasurement, so we got evidence on the physical number of homes and the value of those homes that you could deflate by the very localized price deflators. At the best, it’s zero, and it mostly looks like we’re getting worse at it.
Ed Glaeser, who’s an old friend of mine — and I know you’re a fan of his work as well — he believes that it is very much tied to land use regulation and other regulations on the construction industry, that if you go back to the 1970s or ’60s, they were moving in a modular direction. Think of it as a higher-brow prefab housing, turning construction more into manufactured good, and that the regulations, in some sense, have forced an inefficient scale on our producers, but I don’t know. We left it. We documented it, we showed what it was not, but we weren’t able to find a smoking gun for what it was.
COWEN: Now, in your role in running the Chicago Fed, which is like running a corporation, right?
GOOLSBEE: Yes.
COWEN: What is it from microeconomics that you have found especially useful or helpful? Because you face a lot of managerial problems that have nothing to do with headlines about the Fed, right?
GOOLSBEE: Yes, it’s true. The first thing to know about the Reserve Banks is there is a professional COO, called the first vice president. Chicago’s first vice president is a woman named Ellen Bromagen who’s the best. Everyone says amazing.
I was at an early meeting. I’d been at the Fed for — I don’t know — maybe five months, and one of the other presidents said — it seemed like he was being nice — he said, “It seems like it’s going great. You’ve been really doing a good job.” I said, “Thank you.” He said, “Well, that’s what I was thinking. Then I realized Ellen is your first vice president, and Anna is your research director, Anna Paulson. How big of an idiot would you have to be not to look like you were doing a good job?” There’s something to that.
The micro theory of delegation, I think, is important. And things that have a marginal cost of zero — it’s fine if you do more of those things. That’s the idea of what is a cost, what is an opportunity cost, and what is a marginal cost, are some of the most important microeconomic ideas that I think we can apply to management.
COWEN: What would be an example of something with a marginal cost of zero?
GOOLSBEE: [laughs] Well, I don’t want to reveal anything about our operations and get myself in trouble about the Federal Reserve operations. If you look at marginal cost of zero things, opening meetings to include others and having folks work together, sharing of information can often have very low cost — if not literally zero — and strong benefits. In the same way in our growth models, knowledge is a public good, information can have a very low marginal cost and can help us facilitate working together, and so I’ve tried to emphasize those internally.
COWEN: Now, we live in an age of fiscal pressures, as you know, and recent experience with DOGE has shown that even if the amounts of money at stake are small, people will go after highly visible targets. Do you think, say, 10 years from now, that we’ll have so many different Federal Reserve branches that each have their own research staff in any large numbers?
GOOLSBEE: You mean, might they kill that system? That’s a system we have.
COWEN: Or someone. Right, that’s a system we have. Someone might ask, why does the Federal Reserve Bank of Dallas need as many research staff as it has?
GOOLSBEE: Then let’s back up to a little bit to how the Fed is created. It was made in 1913, so it’s a little bit kludgy, and like every politically created thing, it involves some compromises, but there are a couple of pieces, I think are genius, or at least very durable, important contributions about how the Fed is built that we should not lose.
The first thing to note is, we have a hybrid federal system in which, yes, there is a chair, and there are seven governors that are political appointees named by the president and confirmed by the Senate. In 1913, as today, people were deeply uncomfortable with the idea that either Washington, DC, or New York City, or just a combination of those two, would control the US financial system with no input from the rest of the country.
So, they didn’t set it up that way. They added 12 Reserve Banks from around the nation to be part of the FOMC, and we go and we sit around the table. There’re 19 people; 7 of them are political appointees, and 12 of the people sitting around the table are not political appointees. They’re chosen by boards of directors out in the flyover states like us, where business leaders, civic leaders, and people from the region choose to have representation.
I think it’s critically important that we maintain that kind of monetary policy, independent thought, that it’s not just New York City and Washington, DC, that are coming with one perspective. You’ve probably seen some of the analyses over the years, a lot of the new ideas about monetary policy, about banking and supervision that came out of the Reserve Banks.
That’s why we have our own research departments, and that’s why we take very seriously the idea that when we go to the FOMC meeting, I love hearing what the other presidents and the governors have to say. I said with no irony, “I consider the FOMC to be the world’s greatest deliberative body at this point.” No offense to the US Senate or to anyone else, it’s an amazing group.
If you’re an econ nerd, you go into that room, and it is just about the coolest thing there is on this planet. The shades come down. There’s a giant table, and they go around the table and Jay Powell is going to say, “Here’s what I see in the economy.” And then it’s going to go to Chris Waller, and it’s going to go to President Barkin and President Bostic. “What do you think?”
I think that’s really important. If we put ourselves on a path that we’re going to go chip away at that, and that somehow it would be more efficient to hammer away and get rid of the national representation on the FOMC, I think that’d be a terrible mistake.
COWEN: Now, I like the structure of the committee. I think that’s excellent. But if someone said to me, “Well, there should only be five Feds with a staff, but, say, you should have complete access to all of them. If macro was mainly about data rather than new theories” — and I agree with that — “five staffs should be enough.” It’s not like the old days where you needed the Minnesota School, monetarism in St. Louis.
GOOLSBEE: Then why any? Why not just one? Why not say, “Hey, here are the data sets. We’ll give you all the data sets. We only need one opinion.” You, of all people, Tyler, you know how dangerous that is. Look, we need to let some flowers bloom because monocultures are prone to groupthink.
COWEN: That’s why I say five. You could talk me down to three, perhaps.
[laughter]
GOOLSBEE: Look, as I say, the thing was invented. There’re two of them in Missouri at the time in 1913 —
COWEN: It’s like baseball teams, right?
GOOLSBEE: There’s nothing in this. Yes, it’s like baseball teams. In an era where almost all the teams are in New York City, baseball’s the worse for it.
COWEN: What should the regional Feds be doing managerially to prepare for the arrival of AGI? Or just very strong AI, whatever you want to call it.
GOOLSBEE: I don’t know. I’m still curious, and I hope we’ll . . . Let’s have a little extended discussion about AI, and what it means for productivity, and how real is it? The thing that banks do — I always say it’s a handful of things. It’s five basic functions, one of which is monetary policy. I consider that the opposable thumb. That’s what separates us from the animals.
You’ve got monetary policy, and then there’s directing the payment system of the United States. People probably don’t understand, unless they get into it, that the plumbing of the financial system — wired transfers, ACH direct deposit, FedNow — there’re a whole bunch of payment system things that are operated by the Fed system that are critically important.
The biggest thing in terms of employees is bank supervision and regulation and oversight to make sure that our official banks are safe and sound. We are a bank to banks, so the fourth is financial services, like we send cash for their ATM machines. We do a whole bunch of financial services for the member banks.
Then the fifth is to participate, be upstanding members of the community. We have a community development function. We have a regional economics focus. We are absolute experts in Chicago, for example, on the auto industry. We have by far the most auto production and the highest manufacturing intensity of all the districts, so we have expertise in that.
I think all five of those are quite important that we maintain them. Where AI fits in — I could see it fitting in supervision and alerting you, at least red flagging, here’s a place to look more. As of now, though, there is a rule forbidding the use of AI tools for general Fed usage, and I understand it. If you take secret information, confidential information, and start plugging it into these and it starts getting out, I think that would be a problem.
COWEN: You could put this on your hard drive, right? We’re very close to being able to do that. The Fed has incredible data. You could take everything, all the Feds know about banking failures, build an AI, keep it on the Fed hard drive, and they would do much better than any human in predicting bank failures.
GOOLSBEE: Maybe. That’s a hypothesis. I had this theory before we ever started talking about LLM models. I had this theory with academics on the job market that would come in with a structural IO paper. I believe that down deep in the human psyche, there are what I always call the fundamental human. Some people believe in magic, and some people fundamentally don’t believe in magic.
How it would play out is, if a person had a paper that was so complicated that you could not understand it, there would be some people who would say, “Wow, was that job candidate impressive. They had such a machinery. I couldn’t even tell what was going on.” Then there would be people like Gene Fama who would say, “That guy was a complete bullshitter. I couldn’t understand anything that they were saying.”
That idea, do you believe in magic or not believe in magic, now transpires a little bit into the AI world. You might believe in magic, Tyler, if you think that that will automatically be, in near term, better than any human. The only caution I would have is, back up to 2023 and ask the question, “Could an LLM replace the FOMC and do it better?” I have some severe qualms that it could because you’re only as good as the training sample.
The training sample would’ve said, as you went into 2023, inflation’s way too high. What should happen? What is going to happen? The LLM would’ve said, “It’s a lead-pipe cinch guarantee. You need to jack up the interest rate and have a massive recession. Otherwise, inflation will not come down.” It’s only the thinking, hey, wait a minute, might there be a supply shock here, that we’re getting a positive supply shock, a massive labor force participation increase, more flexible labor markets, so disabled workers can come into the workforce, women are coming back into the workforce. The LLM would not catch that.
COWEN: I want you to talk me into the correct view on a central bank–issued digital currency. Let me explain to you my dilemma.
GOOLSBEE: [laughs] Okay.
COWEN: It seems, on one hand, the Fed or other central banks — if they do nothing, stablecoins simply proliferate. You lose control. There are prudential issues, maybe, eventually money supply issues. If you do something and, say, create a CBDC through the Fed, there’s the risk of disintermediating community banks, regional banks, smaller banks, and Americans seem to hate the idea for reasons that to me are partly irrational, maybe somewhat justified. What exactly is the right view to have on this? I find myself torn. You must think about this.
GOOLSBEE: I’ve seen you ask. I’ve seen you ask Mark Carney and others about this very topic. The first thing I’ll say, I’m at Chicago Fed. We don’t set policy on stuff like that. That’s Washington has to set that.
COWEN: Yes, but they ask you for advice. You’re a renowned economist.
GOOLSBEE: Congress has to set that policy. I think the thought that Grandma is going to have a CBDC and forget her password, and the Fed is going to have a customer service line that Grandma’s going to call and say, “How do I get my password back?” I don’t know. I don’t see that. That seems pretty far from the world that the Fed’s going to focus in. CBDCs that are a wholesale thing, I at least have more understanding, and some of the central banks do that. But there, I wonder, how different is that from wire transfer? What does it mean? What does a wholesale —
COWEN: I mean retail, like we have retail stablecoins.
GOOLSBEE: You’re talking about retail.
COWEN: A true CBDC, yes, like a $10 bill.
GOOLSBEE: Do you advocate postal savings accounts or government-run bank alternatives? There are countries where they have that. I feel like it’s approaching that space.
COWEN: It’s worked okay in the past. I suppose I’m willing to live with the proliferation of stablecoins, but all the paths make me nervous.
GOOLSBEE: All the paths make you nervous. Then all the paths make me nervous.
COWEN: Grandma calling on the Fed line you might prefer to managing all these stablecoins, and Grandma emailing her rep or Grandma’s AI emailing the rep saying, “Hey, I couldn’t get my stablecoins back because the private supplier screwed me. What’s the Fed going to do about this?” That’s not a fun phone call, either.
GOOLSBEE: That’s not a fun phone call, either. I agree with that. To the extent that there are non- . . . In a way, it’s back to our thing of what’s a bank and what’s not a bank. Your argument is, in some sense, stablecoins are shadow banks, and the disintermediation of community banks or the like will create a lot of upheaval, and there will be a lot of jockeying in a world like that, and I agree. The thing is, the risk so far associated with cryptocurrency makes me nervous.
Again, I want to reiterate, I’m just at a Reserve Bank. We don’t set policy on this. Congress has to set this policy, and in conjunction with the Board of Governors, Washington sets the policy here. I find the prospect that the stablecoins — if we’re going to go back to the free banking era in the United States, they end the Second Bank of the United States, and we go through decades where anybody can start a bank.
If you go look at the free banking era, it is an era of massive numbers of bank failures and recurrent financial crises. The Panic of 1847, the Panic of 1857, and close to panic in 1860, massive numbers of recessions in the flavor of a person shows up, starts a bank. They’re supposed to be backed by the equivalent of treasuries, but they start the bank, print their own money, and get out of town. That seems problematic to me, to be in the space.
Anything where you’re going to have cyberattacks and fraud, associating the Fed’s name with it, we just need to be careful. In a way, the Fed will always be the fuddy-duddy of the financial system, and that’s how it should be.
You might have seen my colleague at the University of Chicago, Eric Budish, had this paper about, let’s call it the blockchain in centrally controlled versus not centrally controlled — not controlled but central oversight crypto versus non-central oversight crypto.
It’s just a theory paper, but it goes off of the idea that everybody collectively agrees to history, what happened, and who made a transaction. That’s the essence of what’s happening on the blockchain, and the potential attack, that if you could get 50 percent plus 1 of the miners or of the blockchain to say, “No, no, there was a transaction, and all of that coin is actually, they sent it to us for nothing,” you could get 51 percent of the instantaneous flow, and that would allow you to snipe-attack a giant stock of crypto. That’s a very interesting paper, and it’s —
COWEN: Oh, I know that piece. It’s great.
GOOLSBEE: — a bit challenging. You know that piece.
COWEN: I get that problem, but just concretely, Tether is chartered abroad, right?
GOOLSBEE: Yes.
COWEN: What are you all going to do about it?
GOOLSBEE: [laughs] Like I said —
COWEN: There’s no great firewall in the United States.
GOOLSBEE: — I don’t set the policy.
COWEN: But they come to you for advice. What do you tell them they should do?
GOOLSBEE: Who? Tether? I would tell them —
COWEN: No, the US government.
GOOLSBEE: Well, the US government — do everything you can to make sure that they have the assets that they claim to have.
COWEN: Can we do anything? They’re chartered abroad. Is it Cayman Islands?
GOOLSBEE: I don’t know. I don’t know. The question of what’s inside the safety net and what’s outside the safety net is a super important problem. You probably know, I was very close friends with Paul Volcker, and he was a mentor to me and really one of my personal heroes. We went through the analysis of the financial crisis of ’07 and into ’08 and Bear Stearns and the financial crisis and the response.
He, early, early on, identified this question that it’s all going to come out. At some point soon, they’re going to be like, “Who gets rescued and who doesn’t?” Anything that’s too risky has got to stay outside the shadow of rescue, if you want to think of it that way. If it’s risky, it can’t get it. Where I start to get nervous — and as I say, I’m not anti-crypto. I like innovation and financial innovations.
If we start attaching risky things to the payment system, now I’m getting nervous because we know that in financial crises, we must make sure that people can still make their payments, that they’re not getting evicted, and they make their house payment, or they can buy groceries, or they can buy whatever. If stablecoins started turning into actually the money that we’re using to pay —
COWEN: And they are turning into that.
GOOLSBEE: They want to, but I don’t know if they are. But to the extent that they start to do that, I think we’ve got to take seriously the risk that that poses to financial stability.
COWEN: That’s the argument, though, for CBDC because, otherwise, they can proliferate abroad.
GOOLSBEE: Is it, though? Yes, I guess maybe it is.
COWEN: If there were a Fed stablecoin, you would crush the market, right? That would dominate.
GOOLSBEE: Maybe.
COWEN: I don’t think it’s a maybe. People would love to use a well-run Fed stablecoin.
GOOLSBEE: Well, you added the “well-run.” Then we’re back to, well, would Grandma be calling the Fed for a retail stablecoin, asking, “What’s the login again?” I don’t know.
COWEN: If you can manage a payment system and prudential supervision of US banks, you can run a stablecoin.
GOOLSBEE: Maybe, but in our operation of payments, we’re fundamentally a behind-the-scenes plumbing operation where you have a bank, and that bank sends wire transfers a lot of times on the Fed rails. But you don’t know that that’s going over the Fed rails, and I think that’s fine. This is critical national security financial infrastructure in my view, and I understand why the Fed operates that.
Look, it’s Edmund Burke. There is daylight. Though twilight be long, there is a difference between day and night. There’s somewhere in there where a central bank should be doing it, and there’s somewhere that the central bank clearly should not be doing it. I don’t know exactly where that line is, but I like safe.
COWEN: How much do you think strong AI will boost rates of economic growth or productivity growth? Any number you care to give? High uncertainty, but if you had to make a stab at a number.
GOOLSBEE: Over what time period?
COWEN: The next 10 to 20 years. Not next year. It’s zero for next year, right?
GOOLSBEE: I think it could be a decent . . . I gave a speech at Stanford that was kind of about the topic. We have seen, over the last two years, a pretty remarkable surge in productivity growth, faster than the pre-COVID trend, and on the first glance, a lot of people said, “Yes, of course, it’s faster than trend because it’s going to bring us back up to the trend of where we would’ve been, and then it will slow down.” But it didn’t slow down yet. It may still, but we’re actually up above the pre-COVID trend extended to the future.
There are several explanations that the economists have come up with for that, most of which are one-time explanations, like people started working hybrid or working from home, and that raised productivity by X percent. If you’re a believer in that — I know you talked to Nick Bloom. Nick Bloom — in his head, he absolutely believes that, but that’s a one-time.
The labor reallocation — suddenly we had the Great Resignation, and people could reallocate to jobs that they’re better suited to. Again, one-time. Increase in the level of business dynamism. New firm creation went from whatever — 100,000 to 150,000. Mostly one-time.
Only a new technology has the potential, I think, to slowly work its way through the economy, going from sector to sector. If AI is a general-purpose technology like computers, like electricity, like telephones — that sort of thing — those played out over decades.
There’s work done here at Chicago Fed, looking at the industry concentration of this surge of productivity. Is it broad-based or is it concentrated in certain industries? The answer is, it’s concentrated in certain industries. Despite the skepticism of economists that AI is not big enough yet to explain, it looks to be concentrated in a bunch of tech-related, AI-intensive areas.
If true, it’s entirely possible that this goes like computers, where first you saw the productivity growth in the computer production sector. Then five, ten years later, you saw a surge of productivity in the computer-using sectors. Then, after a decade or more, Walmart incorporates IT into the inventory management system. The rental car companies — people are equipped with handheld computers. They’re coming out and checking in, and checking in your car, and you get productivity that’s spread through there.
Could be a high number, but so far, I think the adoption rate is not high enough to explain very big.
There is a risk of over-anticipation of productivity growth that, if equity values surge and business investment surges in AI infrastructure-type capital investment on the anticipation of the bounty that is to come, I think there’s a high danger that you encounter the over-capacity, the capacity constraint problem that we always face in the short run when you get a surge in demand. Or if wealth effect leads to a big increase in consumption here and now based on these high equity valuations, again, you can get overheating if you’re not careful.
On 01 Pro
COWEN: If I gave, say, a 50-question quiz to new assistant professors at top-20 departments, and I gave the same quiz to o1 Pro, how many of the professors do you think could beat o1 Pro?
GOOLSBEE: At what?
COWEN: Getting the quiz right.
GOOLSBEE: What’s your question? No, what’s the quiz? What’s the quiz?
COWEN: Some’s micro, some’s macro, some econometrics. You and I draw up the quiz.
GOOLSBEE: If this is a BuzzFeed listicle, then whoever’s looking on BuzzFeed is going to succeed the best on the quiz.
COWEN: Analysis.
GOOLSBEE: If it’s judgment, fundamentally, I’m skeptical of magic. I think, in my experience, where the greatest opportunities arise in the short run are on things where, in a way, there’s not a correct answer. The hallucination problem, I think, is actually an extremely damaging problem for a lot of applications of AI. I’ve seen you describe the joy that you felt in being able to ask about historical — what was the cause of the Boer War? — and that it would answer questions. But how do you balance that off of, what if it’s giving you bogus answers?
COWEN: I think o1 Pro hallucinates less, say, than what’s in JSTOR.
GOOLSBEE: I hope that’s true. However, though, the hallucinating in JSTOR — I will abide no insults to JSTOR. We ride at dawn. That’s the hill I’ll die on.
The thing is, you are old enough to remember, as am I: Before there were laser printers, there were those dot-matrix printers.
COWEN: Sure.
GOOLSBEE: The thing is, the essence was adding more and more dots, and the advocates of dot-matrix printers would say, “There’re so many dots. You can’t even tell that this is not a typewriter.” Only, the thing is, you could. You would look and say, “No, look, this is better than it used to be, but I can tell this isn’t a typewriter.” “No, you can’t.” Then he says, “No, you can’t. It’s indistinguishable.”
If I say to AI’s biggest advocates, “Here have been some exciting use cases, but here have been ones that I’ve found troubling.” If your first reaction is to say, “Well, what version did you use?” That’s, to me, kind of the dot-matrix printer problem. If you’re going to steer me to o1 Pro, all we’re going to identify is that I’m too cheap to shell out the money for the higher and higher levels.
COWEN: The Fed won’t pay it for you?
GOOLSBEE: Yes, it’s true. I’m going off of these older ones that are free or cheap, but I don’t accept that it’s like, “Oh, well, if you just upgrade to the higher-priced one, then those hallucination problems would stop.” So far, they’ve got to work on that side of the equation. Otherwise, it still could have amazing impacts, but there are some things for which making a mistake is an unforgivable error. You see what I mean? You’ve got to change the penalty function in that thing.
COWEN: Put aside opinions on AI. Just think of your knowledge as an economist, econometrician. You can just run a few AI models and aggregate and get rid of almost all the hallucinations that way. You can do that now. It costs a little more, but it’s obvious you can get hallucinations down as much as you want.
GOOLSBEE: You’re getting dangerously close, I think, to the position that if we have enough observations, then we have solved the causality problem. “All we need, if we get enough observations, we’ll just be able to predict anything with 100 percent accuracy.” I don’t know if it still flies too close to the sun.
I’m not sure what just happened there. Did that motion create hearts?
COWEN: Created hearts. It created hearts.
GOOLSBEE: Oh.
COWEN: I think that’s from your system.
GOOLSBEE: That is deep in my heart I believe —
COWEN: The Chicago Fed loves us.
GOOLSBEE: That you better be careful. Here’s my second question about AI and its impact on productivity. How much of the improvement in AI . . . I’ll tell you the thought experiment I like to do is to ask, what jobs would be replaced with the AI that exist right now, not extrapolating it forward? Because a lot of the grandest dreams for AI and rapid adoption, I think, are premised on extrapolating a growth rate.
To the extent that some of the improvement in AI is not coming from improvements of AI theory or new algorithms, but instead from bigger and bigger data sets and more and more computing power, data sets and computing power have diminishing returns that will kick in pretty quickly. So, it would behoove us to remember that there was a time 15 years ago when self-driving cars were improving so rapidly that people were predicting that within five years, there would not be a single professional driver in the United States.
Then, basically, what happened is, the rate of improvement didn’t go negative; it didn’t go to zero. It just slowed way down. Now, we’re having some self-driving taxis in different cities, but we’re nowhere near what, 15 years ago, there was a group of real advocates who said, by 15 years from now, people are going to be like, “Dad, what do you mean people used to drive their own car? How dumb were people?” We’re still a long, long way from that.
COWEN: How should we reform the MBA business graduate education? You’re an expert on that.
GOOLSBEE: I am only an expert in that I taught the MBAs, and I taught a class on platform competition. I had hundreds and hundreds of students over the years. I found the MBAs to be very data-minded and very intellectually curious. Now, it might be the University of Chicago MBAs are a high-powered group. I think the market for MBAs, as you know, all the pressure is on the opportunity cost side, trying to shorten the programs. There are new master’s programs in finance and management and things that are one year in nature or a year and a half in nature.
I think the MBA, whether you use it as a market test — what does the market demand — or you just go look at the content of what people know, I think it’s a pretty well-functioning market. It’s a valuable degree, and it teaches a lot of skills.
COWEN: Final question. What’s the story of how you beat Ted Cruz as a debater, and he was a national champion, as were you, right? How was it you won?
GOOLSBEE: Well, I was big at speech and debate. I was in high school speaking and then went to college. I was a year older than Ted. My partner and I — we were the National Team of the Year, and Ted and his partner were second. The thing is, I kind of had kryptonite on poor Ted, which was, when we would debate . . . You don’t debate face to face; there’s a judge.
Ted was an excellent debater, and his greatest strength was plotting. “We’ll say this, and that’ll lay a trap and get them to say this, and then we’ll respond in this way.” His weakness was thinking on his feet, especially if he got mad, so I would, to the judge, start teasing. I would tell jokes at Ted’s expense, and many of our rounds would end with Ted red in the face. “How dare you say that?” I was, “Ah, we got him again.” That was the secret.
COWEN: Austan Goolsbee, thank you very much.
GOOLSBEE: Tyler, you come to Chicago. I’m going to have to take you on the food tour. I have, for many decades, followed your advice, and I’ve got some great places out here that are going to be right up your alley.
COWEN: That would be great. Thanks so much.
GOOLSBEE: Thank you.