Transcript: The History of Financial Crises – Lecture 1 w/ Dr. Charles Calomiris

So that’s really the first, I think, most important question about financial crises, addressing this possibility that we shouldn’t be having them still.

I’m going to begin with a discussion of a financial crisis that happened 2000 years ago, and then we’ll be talking about crises that happened more than 300 years ago. And so you get the picture pretty quickly that the history of financial crises is something that goes back a long way in time. How come we don’t learn?

Theories of Financial Crises: Minsky, Kindleberger, and the Historians

Now, of course, 1 possibility is we’re incapable of learning. And 1 of the most influential theories of financial crises by Hyman Minsky and Charles Kindleberger is that view. It’s the view that human beings behaviorally in some mass psychotic waves oscillate between massive greed and massive fear. And when massive greed is happening, assets get bid up because people have a low perception of risk. They’re very optimistic and they’re excessively optimistic.

And then, almost naturally, through a behavioral cycle of mass hysteria, that then changes from excessive optimism to excessive pessimism. So if you take that view, crises are all quite similar. They come from a common behavior oscillating between fear and greed, and there’s no way to fix them because it’s part of human nature. You may say it’s irrational, but maybe that doesn’t matter because humans are the way they are.

Now, historians don’t necessarily agree with that view. Some historians would say, all crises are different. So there’s really no way to learn how to alter your behavior. And the Minsky-Kindleberger view is wrong because it just lumps together all these very different events. It’s really more that there’s particularity to each of these crises, and so there’s little that really can be learned. We shouldn’t jump to the conclusion that people are behaving irrationally through these cycles of greed followed by fear.

However, other historical opinions aren’t of that same view. That is, it’s not necessarily the case that all these crises are unique. In fact, there’s a literature that’s identified some commonalities.

For example, when crises involve the collapse of banks, we tend to see that banks lend a lot in advance of the collapse. That is, the growth rate of loans tends to be unusually high. We also tend to find that banking systems that have been given a lot of observable government protection — for example, very generous deposit insurance of banks by the government — also tend to be much more likely to have a banking crisis.

So there are observable commonalities statistically that tend to forecast banking crises. Similarly, with exchange rate collapses, where exchange rates depreciate massively and suddenly, we tend to find that unsustainable fiscal and monetary policies tend to be an important part of that story. Again, statistical evidence of commonality. And in stock market collapses, statistical evidence tends to point to what’s sometimes called momentum effects or returns extrapolation, which just means that when returns are high, there is a tendency for them to be high in the adjacent next period.

That’s sometimes viewed as a behavioral response that might be, in some sense, related to this Minsky-Kindleberger view of greed. That is, you see stock prices go up and you extrapolate out of optimism that if you get into the market, they will go up again, and that forces them up again.

Common Features Across Crises: Narrative and Statistical Evidence

So, we have some disagreement theoretically, and we have different views about crises. In addition to the statistical evidence about some common features among crises, there’s also narrative evidence that shows that they do have some common features.

For example, we’re going to talk about the Mexican crisis of 1994-1995 in this class. And those of us who were there at the time noticed a remarkable similarity between the Mexican experience of 1994 and the Chilean experience of 1983. One could also say that the Greek crisis of 2010 had a lot in common with the East Asian crisis, which we’re going to study in some detail in the case of Korea in 1997.

Now, how do we sort between these views that, I guess, we’d say they’re kind of 2 extreme views. All crises are the same — that’s the Minsky-Kindleberger view, all driven by a common factor of human nature. Or the historical particularist view — all crises are different, you can’t learn from 1 about another.

And, of course, in this course, we’re going to find a middle ground there. That there are discernible similarities across crises, but that doesn’t mean that all crises are the same. And because we’re going to point in that direction — that there is the possibility that you can learn from the experience of 1 crisis to another — then that raises the puzzle. Why is it that we don’t learn to avoid them?

Of course, in the Minsky theory, it’s just because we’re incapable of learning as a society. We’re driven by irrational greed and fear, we can’t learn. That seems strange to me as a historian and as an economist because, of course, we have collective action through government and we could learn through government policies and regulations to try to improve our situation.

A Surprising Argument: Crisis Risk as Adaptive

And so we’re going to take a different approach to this question of why learning is so limited. And it’s a very — you might even say — a surprising approach. What I’m going to argue is that crisis risk, in other words, allowing as a society for us to have the risk of a crisis despite the cost of that crisis, might actually be adaptive. And I mean that in this sort of evolutionary sense.

So why would it possibly be that having these explosions occasionally occur in financial crises might actually be adaptive? And I want to point to 5 examples to give you a sense of what I mean by adaptive.

So first of all, I’m not saying that the losses in the crisis, that the destruction is a good thing. I’m not saying that. What I’m saying is that the bad thing of the crisis might be stapled to other things that we want. And when I say we want, I don’t mean that I can tell you are desirable. What I mean is we may want some things as a society despite the fact that they’re not desirable.

Democracy, Coalitions, and Rent Extraction

And let me give you an example. Political systems, our democracy here in the United States, are designed to achieve outcomes. Those outcomes aren’t always perfect. And in fact, you could think about the way our system works as forming coalitions that win elections. And those coalitions may not be best for everyone. They may not choose policies that are the best for everyone. But that’s not the way a democracy works.

A democracy chooses things that are best for the winning coalition, maybe at the expense of the people who aren’t part of the winning coalition. That’s a successful democracy. Democracy isn’t designed to produce the sort of perfect situation for everyone on average. What it’s designed to do is to allow majorities to win. And as part of their winning, they then get to make choices that are going to benefit those who are part of the winning coalition, maybe at the expense of those who aren’t.

And so, it could be that that coalition wants to do what we in economics call rent extraction. That is, to create systems, to create regulations, to create ways of organizing ourselves that systematically take money from 1 group and give it to another. You could say, well, that sounds horrible. That’s a failure. It’s not a failure of democracy. It actually is often designed to do that. In fact, almost always does that.

Subsidizing Risk: The Veterans Administration Example

So, why would a crisis be related to that process? Because creating a subsidy for risk, which might lead to excessive risk taking, might be a very effective way to create rents, to take from some people and give to others.

Seeing from that perspective, when I, for example, subsidize mortgage risk by telling people, “Oh, you can take out a mortgage with no down payment.” By the way, we do that for veterans in the Veterans Administration. So veterans are able to borrow and have a house with 100 percent mortgage at an interest rate that’s artificially low because of the insurance coming from Veterans Administration mortgage programs. Now, does that make veterans more likely — all things equal — does that fact make them more likely to default on their mortgages?

The Adaptive Nature of Financial Risk-Taking

And the answer is, oh, of course. If I create a policy that makes the mortgage riskier and that subsidizes taking risk, you will take more risk. And if you take more risk, you’ll be more likely to default, all things equal. But that choice of subsidizing that risk might be part of an overall social bargain that politically works for us. And if it politically works, that might make it adaptive, even though it leads to some financial fragility.

So, the first ingredient of understanding what I mean by adaptive is that there are some politically adaptive ways to create risk subsidies that might serve the purposes of democracy by making the formation of political coalitions sustainable, despite the fact that things blow up and are economically costly. They may not be ideal in some abstract economic sense, but they might be adaptive politically.

Adaptive Geopolitics

A second example is adaptive geopolitics. Suppose that you’re in a country — we’re going to talk about 2 examples today, from the early eighteenth century in early modern Europe, but we could also later will point to examples from emerging markets today — where countries are in an extreme competitive global situation, where countries are competing with each other. Competing for trade, competing for market share.

And as part of that competition, they may decide to take risks because failing to take the risk might lead to an even worse outcome. It’s not that I want the risk per se, but I’m involved in a competitive situation where if I don’t take the risk, something else happens that I don’t like. Think about it this way. Suppose that I’m lagging behind as a competitor in the early eighteenth century, like France or Britain is lagging behind Spain, Portugal, the Netherlands.

If they don’t take some risks, maybe they lose even more territory in the global competition. By the way, that competition has both a military aspect and a mercantilist aspect. It’s about competing for establishing markets. The way you do that is you invade different parts of the world, you take them over, then you establish trade markets. So, if you’re behind in that game, you’re trying to figure out, how do I catch up quickly?

Because I don’t have forever to wait. If I don’t catch up quickly, I’m gone. And then, not only do I lose the ability to have an empire, but I might get invaded eventually myself. So if you’re France or England at the turn of the eighteenth century, you’re not just thinking, what risks do I want to take from some sort of isolated economic policy standpoint. You’re thinking, what risks can I take? What things must I do to be able to survive global competition?

So that can be, therefore, adaptive to you from a survival standpoint to take risks that might lead to financial crises because they produce something else.

Learning Through Risk: The Florida Example

A third example of an adaptive risk-taking strategy comes from the advantages of learning. So there are risks that are brand new. We’re going to talk about Florida, where we’re filming this class. We’re going to talk about Florida in the 1920s. New technologies allowed people to come to Florida and to think about developing it. There were dreams that it would be a paradise. By the way, those dreams were correct. But they thought it would happen a little faster than it did.

But the reason that it was so interesting, as we’ll describe, is that nobody really knew how things were going to play out. And so if you didn’t take the risk, you wouldn’t learn. You wouldn’t learn what is the market for Florida land. What is this land going to turn out to be? How much is it going to be worth? Where are things going to be developed? What are we going to do about all these swamps? By the way, the coastline was dramatically changed by technology.

So Florida was a complete puzzle, right? We didn’t know in the 1920s what Florida was going to turn out to be. How did we learn and make it happen? By taking huge risks, which in the short term didn’t turn out so well.

Similarly, when we look at the stock market in the US in the 1920s, because we were such an innovative society, we had a boom in the stock market, and then we saw that that boom collapsed. Why exactly it collapsed, we’ll see, is not so easy to explain. But what is clear is that it was something about learning about new technologies that involved that new risk, both in the 1920s Florida land market and in the stock market in the 1920s. And you could say we should have avoided those risks, but to avoid those risks, you would have had to avoid learning. You would have had to avoid new technologies.

The Role of Fraud

What about fraud? That’s going to be my fourth example. So you might say, well, there are a lot of episodes where fraud on a massive scale — Florida turns out to be one of them in the 1920s — substantially magnified the losses that were going on. And you might say, isn’t there a way for us to reorganize ourselves so that we make fraud less likely?

The problem there is, as we’ll see when we talk about some of the examples, including the land speculation in Florida in the 1920s, and the frauds that are happening in the government, actually, as well as in the banking system at that time, you might say, well, there should be a way to prevent these. But on the other hand, notice that hidden actions — the ability of people to keep their actions somewhat secret and to themselves, which is necessary for fraud to happen — has a lot of favorable qualities. It allows private information. It rewards entrepreneurship.

We don’t want to have Big Brother monitoring us so closely and restricting us so closely and sharing all of our information about everything so that there’s no possibility of investing in and creating private information. A market-based society actually depends on decentralized information and private information. Fraud is an undesirable consequence of something that is in fact highly desirable: privacy.

Fiat Money and Central Bank Policy

Finally, another example of something that creates a lot of risk but has a silver lining is fiat money. We’re all used to living in a society where central banks create paper currency, and it’s through the central bank’s production of the paper currency over which they have a monopoly. The Federal Reserve determines the amount of dollars.

We might say, well, there are going to be lots of examples in this class — in 1929, in the period 2002 to 2007 — where we’re going to see that monetary policy made big mistakes. Monetary policy was too tight or too loose. And there’s a connection, which we’ll explore almost right away today, between those monetary policy errors and that sudden repricing of risk.

So, when monetary policy is too loose, it tends to, for some temporary period of time, make risk too low. In other words, people are underestimating risk when monetary policy is weak. And when monetary policy is tight, it has the opposite effect. It tends to cause risk perceptions to rise very quickly. So another phenomenon, another source of this oscillation of perceptions of low risk changing to perceptions of high risk, economic research has shown, comes from central bank policies.

So you might say, let’s get rid of those guys so that we don’t have such oscillations. But then, the reason we have across the whole world agreed to have fiat money regimes as opposed to, let’s say, gold or other kinds of monetary systems that one could imagine, is because we think it’s the best of available technologies. That doesn’t mean that there isn’t room for reform of central banks. Many economists, myself included, believe that central banks, including the US Central Bank, could substantially improve its monetary policy to reduce these oscillations of risk.

By the way, Milton Friedman, a famous economist in the 1970s, in the wake of the inflationary episode that happened in the US, made a lot of progress convincing people that there was big room for improvement. And some of that improvement happened, and some remains to be done.

So I’m not saying that we live in an ideal world. I’m just saying that we have decided as a society that when people like me say, let’s make monetary policy more predictable and systematic and remove these oscillations from it, that’s not a popular view. You voters are not supporting people who are running for office who are advocating that. Whether you know it or not, you’re supporting people who advocate what’s called discretionary monetary policy, where central banks get to decide without much accountability while they’re making the decision to change things in one direction or another.

We all just in the US lived through a major inflation. It’s still happening, where the Fed has deviated from its target massively. No one doubts that the Federal Reserve made a huge mistake from 2021 to 2023. No one doubts that. That’s not controversial. The Fed doesn’t say we didn’t make a mistake that caused that high inflation. The primary objective of the Fed is to not allow those kinds of inflations to happen.

And the question is, well, you could say, well, A, maybe we should just get rid of the Federal Reserve and go back to a gold standard. That’s no one’s advocating that. B, you might say, well, let’s change monetary policy to make it more systematic so there isn’t so much discretion. That’s my view, personally. That’s not a popular view. The most popular view is, despite the fact that the Fed made a big mistake, we like the status quo of the way this is done. And politically, this is currently the winning operation.

Financial Crises as Adaptive Political Choices

So notice that all 5 of the things I talked about — domestic subsidies as a way to reward constituents in a political equilibrium, geopolitical competition, innovation and learning about new risks related to innovation, the fact that risk is hard to observe because we keep information private, and the fact that we have central banks that manage monetary policy through discretionary fiat money regimes — those are all things that may not be economically optimal, but they seem to be part of what I would call an adaptive political equilibrium. That is, countries and politicians that support these things seem to succeed.

So we might say, well, things aren’t perfect. We’re having financial crises. But I like to think about it a little differently. I like to think that the reason we’re having financial crises is because we want them. Not for themselves, but because they’re stapled to things that we’re very attached to. We’re very attached to rent seeking in our democracies as a way to allocate benefits. We’re very attached to surviving global competition with other countries. We’re very attached to innovation and learning about new risks, and to privacy, and to monetary policy through fiat money and discretion.

I’m not saying we live in the best of all possible worlds. What I’m saying is that we live in a world in which all of those things are politically chosen, and they seem to be successful. That is, politicians who would advocate differently from those wouldn’t win. Our system is designed to actually promote those kinds of choices. And it’s not just our system. I think if you look around the world, you’ll see these are common choices.

So if you were just a biologist looking at this situation from the standpoint of what looks like it’s being selected as part of a biological equilibrium, if you apply that same logic to social choice, you would say, these are choices that seem to be common around the world, and they seem to be part of an adaptive social political system. Again, I’m not saying that means they’re ideal economically. I’m just saying that we can understand them as choices that many countries are making, not just some failed political regime, but this seems to be a very successful political regime.

Know Thyself: The Motto of This Course

So that’s why the motto of this course, I think, is “know thyself.” Instead of approaching financial crises as some sort of a foreign invasion by an alien being that we could somehow defeat or learn we don’t want that, instead of tutting about, “oh, look how we keep repeating those same mistakes,” or blaming people on Wall Street — instead of doing any of that, what we should do is say, let’s look in the mirror. Let’s learn to know who we are and why we’re making these choices repeatedly.

It’s not — I’ll just come out right now and tell you — it’s not an irrational cycle of greed and fear that’s driving all this. It’s not that at all. It is a conscious set of choices that seem to be adaptive, that seem to be part of who we want to be.

Deposit Insurance and Banking Risk

To give you a few windows into how we as economists analyze these states of the world where crises happen more or less across time and space, I’ll just point to some graphs. Simple graphs can lead to a lot of self-knowledge, can pose questions, make you ask questions about why your society that you’re living in is different from other societies, and why choices have been made.

So this is a graph of the number of countries in the world that have some kind of deposit insurance system for their banks. One theme of this course will be deposit insurance. This is maybe the least controversial empirical statement in financial economics: deposit insurance makes banking systems much riskier.

Now, you didn’t expect that probably. You expected to hear that deposit insurance, which removes the incentive of depositors to withdraw their funds because the funds are guaranteed by the government, should make banks less risky because the banks don’t have to worry about somebody getting a massive number of people coming in and trying to withdraw their money all at once. Well, that’s an interesting theory. That theory is wrong.

In fact, as you’ll see, the most robust and agreed upon empirical fact about deposit insurance is that it makes banking systems more likely to collapse. Very interesting. And of course, that’s related to the fact that when you remove the discipline of people worrying about whether their deposits are safe, then bankers end up taking a lot more risk. Deposit insurance is a risk subsidy for the banking system. And when you subsidize risk, you get more of it.

So notice in this graph, what you see is that as of about 1970, there really were only a handful of countries.

The Origins of Deposit Insurance and Its Global Spread

The US is the first in 1933. So prior to 1933, no country has deposit insurance. So we’ve been having sort of depository banking in the world since about the seventh century BC in Athens. And when did we first get deposit insurance? Not till the US in 1933.

And then for another 40 years, only a very small number of countries followed the US example. But then all of a sudden, from the 1980s and the 1990s, we see deposit insurance spreads like wildfire throughout the world. That’s a really interesting political choice. Why does that happen? When and where it happens?

Believe it or not, we will be talking about that toward the end of this class. But the point is, if I said to most people in the world who now live in a system with deposit insurance, you know, deposit insurance is a bad idea. We can accomplish better the same objectives that you think we’re achieving with deposit insurance with something else and without promoting a huge amount of risk in the banking system. So if I said to most Americans, I want to get rid of the deposit insurance you have, most of them wouldn’t like that. That wouldn’t be popular.

And so as a person, you can either decide, well, that means that people aren’t being smart. They should listen to professors. Right? When I tell them something, they should obey. Do what I say.

Or you could say, professors aren’t smart because professors aren’t understanding that there’s something adaptive about this political equilibrium. It’s happening all over the world. We may think it’s a bad idea economically. By the way, I do. But now we have to somehow respect it too.

Even if we don’t think it’s making the world a better place economically, we have to try to understand, why is this happening? It doesn’t seem like it ever gets reversed. So if I think about evolution, biological evolution. Suppose I told you there’s a time when all over the world, organisms created adopted a new trait, and that trait was never reversed. You’d say it was adaptive.

Right? So we have to learn to think that way in social science too. Even if we, as economists, tend to think a little differently about what’s optimal, we have to think about not just what’s optimal economically, but what’s happening. So, this is an interesting part of our thinking, is to be able to come to grips with things that seem to be happening all over the world at the same time, not getting reversed, and being fairly popular as policies, despite the fact that they might produce financial crises. This is the prime example.

The Two Gorillas of Financial Crisis: Banking Risk and Mortgage Risk

Here’s another example, and I will talk about these later in the course as the 2 gorillas of financial crisis in the world and especially in the US, but all over the world. These 2 risks that are getting subsidized, 1 is banking risk, default risk coming through deposit insurance, and the other is mortgage risk coming from programs all over the world that subsidize risk taking by allowing people to borrow to fund the acquisition of a house at an artificially low rate by taking on a lot of risk that they don’t have to pay for. I gave you the Veterans Administration as an example earlier. So here’s a graph that shows that at the end of World War II, around most of currently developed economies, so this is a list of about 17, I think, 18 countries, all of which are developed economies, so their total ratio of mortgage loans to total loans in their banking system, as of 1950, mortgages were about 30 percent. Today, they’re more than double that.

So how did that happen? What’s going on there? And the answer, of course, is mortgage risk is being subsidized. That’s why you got more of it. Now, this housing collapse that we’re living through, and we had 1 in the US in 2008, you might say, well, it seems like that’s a bad idea.

That was a pretty disastrous experience. Subsidizing housing risk leading to this sort of excessive risk taking was very costly. But I will tell you that I don’t see any evidence that this trend that we’re seeing is reversing. Subsidizing mortgage risk is now happening again in the US. We are once again undertaking policies to allow people to make what you might call financially unsound investments in a house by borrowing because it’s the tool that the government can most use most politically effectively and easily to help promote borrowing for houses and because that’s politically popular.

You might say, that’s stupid. We shouldn’t be doing that. By the way, it’s also the case that subsidizing mortgages doesn’t seem to actually increase the house ownership rate. I’ll show you evidence about that. Because when you subsidize mortgage risk, it tends to bid up the price of houses so that you’re chasing your own tail in some sense.

Right? But that doesn’t mean that we don’t want to do it politically. Just because it doesn’t make sense doesn’t mean we don’t want to do it.

Monetary Policy and the Predictability of Risk

Okay. A third example, this has to do with monetary policy. So what you’re looking at here is a measure of risk. There are 2 measures of risk. The orange line is a measure of risk in terms of spreads — I’m sorry, is the VIX. That is a measure of the volatility of the stock market. And the blue line is a measure of bond risk spread.

So a particular physical risk in the bond market, how much extra yield do you get for that risk? So that spread, you can see, varies. And in fact, notice that the VIX and the bond spreads tend to vary quite closely together, don’t they? And you can see that from 2002 to 2007, those risks declined dramatically. Remember when I said financial crises are often a time when there’s a dramatic reappraisal of risk.

So look at 2008. You see the dramatic reappraisal of risk. But that reappraisal of risk was preceded by 5 years of dramatically declining perceptions of risk. The pricing of risk was very low and declining for 5 years, and then it suddenly went up. When we look at that, we know that we’re not just talking about mortgage risk.

We’re not just talking about mortgage risk subsidization. So what’s driving that bus for the stock market and the bond market? Well, it’s monetary policy. And we will get back to that. We’ll talk about that specific example later in the class.

But the reason I’m pointing to it right now is that’s an example of fiat money regimes under discretion, the Federal Reserve, choosing to enact monetary policies that had predictable consequences for risk. This is a predictable regular phenomenon of monetary policy. When monetary policy is excessively loose, and it was observably excessively loose from 2002 to 2006, as we’ll see. It was unusual. It wasn’t just looking back at that and saying, oh, it was loose.

It was during the time we were living through it. It was obvious that they were loosening monetary policy excessively, and that had big consequences for these markets. And yet, we haven’t decided to stop the Fed from doing that, and I don’t think that we’re going to do that anytime soon.

Know Thyself: Crises Are Chosen

So these are examples. When I say know thyself, these are the United States that we’re talking about or other developed economies or in the first graph, it was worldwide. These are phenomena that are happening all over the world. They’re quite similar. It’s not like there’s somebody out there with some better idea that’s very popular that’s getting elected. And so why? Why is this adaptive?

Well, that’s a question which we’ll get into in part. But what I want to get you to think about, which is very different from the normal tutting about those unwise people who take risks on in advance, I want you to think, no. There are a lot of these risks that are predictable. They’re going to often end badly. They’re done on purpose. And now the even more controversial statement. They’re adaptive. The next politician will make the same choice because we choose it. And he might say, we should get smart and not do that.

I would agree with you. But that’s not the question. The question I’m here as a positivist, as an economist, and historian trying to understand — why do these things happen? And the answer is they’re chosen. And there doesn’t seem to be much evidence that that’s going to change.

It seems like the political equilibrium likes them. Why? Well, because it promotes learning. That’s good. Because it promotes private information. That’s good. Because it promotes fiat money. I would say that’s good. But that doesn’t mean that it’s the best we can do. So I’m not saying that. It just means we’re doing it on purpose.

Toward a Crisis Taxonomy

And so, as we get through talking about these various crises, I want to suggest a middle ground between the historical particularist view. All crises are different. You can’t learn anything from 1 from looking at another. That’s not true. But I also want to suggest that the opposite extreme is not true either. All crises are the same. That’s not true.

When we talk about these influences on crises — international competition, domestic rent seeking political subsidization, learning, fiat money policies — some of these influences are more important in some crises than in others. And so what that suggests to me is that just like I like making biological analogies, I think we want to think about constructing a crisis taxonomy. That is where some crises are more similar to other crises than they are to others. They’re more similar to some crises than they are to others. So there are these multiple dimensions that we’ve talked about that tend to influence crises, make them happen, make them more likely, but they’re not always present. It’s more like some are present in some events, some are present in others.

And so, I’m writing a book right now with Matt Cheremski, and we’ve already published a preliminary article looking at 10 crises this way, which are the ones we’re going to talk about. And we have in our book 35. And what we find is it’s extremely useful to ask a common set of questions that help us organize our thinking about these crises along the lines I’ve just developed with you. But then notice how some crises are really driven by 1 of these or 2 of these phenomena and others by very different ones.

So, the middle ground is crises aren’t all unique, but neither are they all the same. Just like we think about organisms that we can usefully divide the world using some taxonomy that’s reflected in zoology, we’re going to think we’re going to be able to do the same thing for financial crises.

This table, which I know you can’t read, the point of is not for you to read it, but just to get the ideas I’m telling you about it, which is so what we do is we look at each crisis and we ask a set of questions. On the left hand extreme column, we have a list of questions we’re asking about each crisis. Like, was the collapse predictable? And when I say was it predictable, I mean predictable using some kind of ex ante tool.

We’re going to talk about the Korean crisis, which happens toward the end of 1997. But in March and April of 1997, both the Financial Times and The Economist predicted a crisis in Korea, 6, 7 months before it happened. And so did I, and so did many economists, because we saw common features. Right? So that’s what I mean. Predictable based on some kind of ex ante criteria that we can also see reflected in news reports long before the crisis happened. The Mexican crisis of 1994 was predicted by lots of economists 2 years ahead of the crisis happening. So, is the crisis predictable? But there are some for which it wasn’t.

Secondly, can we point to a kind of price boom just before the bust? In some cases, we can. I’ve mentioned the housing crisis of 2008, and as you could see from that graph where we looked at monetary policy and its effect on bond markets and stock markets, all asset prices were elevated by the very low perceptions of risk from 2002 to 2007, and then asset prices declined precipitously when those risk perceptions changed. There was an excess price boom in the case of 2008. When we talk about Korea, or some other examples, but Korea in 1997, there was no excess price boom.

So it isn’t just 1 phenomenon. In most cases, we’re going to see examples of political risk subsidies of 1 kind or another. But there are examples where there isn’t a political risk subsidy. The Florida land boom, for example, or the stock market crash boom and crash of 1929. So, what’s interesting is by asking these narrative questions that each crisis has to answer, we can then create a kind of data set of what makes crises different or similar.

And what we see is there isn’t just 1 phenomenon, but there are some clusters. There are some things that kind of go together. For example, political risk subsidies tend to be highly positively covary with the predictability of the excessive pricing. That’s not surprising. If you’re subsidizing something politically, it’s not surprising that the price goes up a lot.

So that’s our approach. And our conclusions are that this taxonomic approach really leads us to some useful insights.

The Plan for This Class: Deep Dives into Financial Crises

What’s our plan for this class? So, we are going to do deep dives because so far, I’ve just told you a lot of things in a very general way. I’ve just given you the introduction to how I’m thinking about presenting this material to you and what I think we can learn from thinking about financial crises. But to really show you what I’m talking about, and for you to actually learn it, you have to do what I call a deep dive. A deep dive means that you’re going to have to get into the sordid details of each of these crises. And some of them are quite arcane and complicated. But without that, it’s just vague talk. So we’re going to talk about the banking and land market collapse in Rome in the year 33 AD.

The Mississippi Bubble, South Sea Bubble, and Banking System Comparisons

We’re going to talk about the Mississippi bubble in France in 1720, the South Sea bubble happening immediately thereafter in England. We’re going to talk more generally about the situation of bank chartering and crises with 2 different comparisons. One between England and Scotland, where England is the crisis prone, relatively highly crisis prone banking system in the eighteenth and nineteenth century, early nineteenth century, and Scotland much less so. And then we’ll do another similar comparison between the crisis prone banking system of the US in the nineteenth and twentieth century, and the Canadian system, which we’re still waiting to have a banking crisis. Canada, since it starts chartering banks in the 1810s, has not had a banking crisis yet.

What’s going on? These 2 very similar countries in so many respects have completely different banking crisis histories. And so by making those comparisons between England and Scotland and between the US and Canada, we’re going to learn some things. We’re going to learn that crises are caused by political choices, and that’s why they can be different across countries. Even countries that are very similar and neighbors, like England versus Scotland or the US versus Canada.

We’re also going to learn how to think about why learning doesn’t happen. So it’s not that the US people weren’t looking at Canada. It’s not that the English people weren’t looking at Scotland and noticing that their banking system wasn’t exploding. Why can’t we have a banking system like the Scots? Say many people living in England.

Let’s say circa 1820 or even before then, 1800. And similarly, many people in the US are kind of wondering, why can’t we have a banking system that’s as stable as Canada’s? And yet, they’re asking those questions for many decades, and there is no imitation. So, it’s a great example of this phenomenon of the puzzling persistence that we started with. When you see adjacent countries with drastically different experience where there isn’t learning going on.

I can’t learn from your history because guess what? We’re in a different political equilibrium from your political equilibrium.

US Banking Crises: The National Banking Era to the Great Depression

And then we’re going to talk in some more depth about the US crisis — very crisis prone — but we’re going to do some nuanced discussion there about how the shapes of those crises are different over time. So we’re going to compare the era from 1873 to 1907, the so called national banking era crises, of which there are 6. We’re going to compare that to the Great Depression, which was a much more severe event for US bank failures.

Then we’re going to move on to discuss the Florida land boom and bust of the 1920s and the stock market boom and bust that happens also in the 1920s in the US.

The Rise of Deposit Insurance and Its Consequences

We are going to then talk about deposit insurance because it turns out to be such an important topic. So we’re going to do an analysis of the growth of deposit insurance in the 1920s in the US. Here’s another political or economic puzzle. All of the states that adopted — not the country as a whole, but the states that adopted it — ended up seeing complete collapse.

So that by 1930, it was absolutely clear that every state that had chosen deposit insurance, their banking system collapsed. It was a complete disaster. And then, in 1933, the federal government adopted federal deposit insurance. That seems kind of strange.

Just to add to the puzzle, and I don’t want to tell you everything at the introduction, in the 1932 election, FDR, who ended up winning the election, said there was definitely one thing he would not do when he was president, and that is to support deposit insurance. Because it was so obvious what a catastrophe it had been for the states. And he went on to — Roosevelt was actually fairly knowledgeable about banking. And he went on to talk about how deposit insurance by removing market discipline means that depositors aren’t concerned enough about their money, which makes bankers act excessively risky.

So he not only was opposed to it, but explained his opposition, showing that he had been paying attention to all the collapses that had happened in the 8 states in the 1920s. Nevertheless, he’s the president who signs federal deposit insurance into being.

Is that because his mind changed about the economics of deposit insurance? No, it is not. So that’s really, I think, a great lesson for us as we’re thinking about how learning takes place, what it means for something to be adaptive versus economically optimal.

Risk Subsidies, the Farm Credit System, and Residential Mortgage Subsidies

Then we’re going to move on to talk about other risk subsidies that are consciously chosen. The farm credit system, it starts in the first decade of the twentieth century, and residential mortgage subsidies that begin in the New Deal.

So 1934 is an amazing year for federal government decisions to subsidize risk in the residential mortgage area. And that was a landmark change, and it culminated quite predictably in the thrift crisis in the 1980s, just as the farm credit system culminated in the agricultural bank and land crisis of the 1980s.

So what’s really interesting is, when we create these political subsidies for agricultural borrowing against land or for residential borrowing against land, they were quite conscious. They weren’t done because people were making some economic choice, rather because of a political choice. They resulted ultimately in greater crisis risk and the crises happened.

And by the way, we haven’t gotten rid of those. We didn’t learn to get rid of those. And then we end up finally talking about that graph that I showed you where deposit insurance is rising throughout the world in the 1980s and 1990s. And so what I’ll do is I’ll explain to you why that happened, and then also show you that the consequences of that are well known to be making banking crises much greater and more severe around the world.

So this period from 1980 to 2010 was the most severe period all around the world for the banking crisis. That is, we had more of them, and they were of greater severity than any other time in human history. That was the result of a conscious choice that we have not reversed. We have done some reforms to try to mitigate those risks, especially in emerging market countries. But what’s interesting is those choices were made despite the fact that the economic research was telling us that these risks were going to happen. There was nothing surprising, in a sense, about this period.

The Ex Ante Approach: Avoiding Hindsight Bias

So, as I said, our approach is going to be a deep dive. The other thing about our approach is we’re going to avoid ex post thinking. That is, we’re not going to be looking at this — if you read the books by John Kenneth Galbraith’s terrible book about the stock market crash of 1929, it’s a sort of exercise in superiority thinking. He’s making fun of everyone. Oh, how silly. How stupid.

By the way, he wasn’t much of a historian about the 1920s, and most of the views that he expressed in that book, I would say, have been shown to be false. So a lot of times, what people who tell you they’re analyzing financial crisis are really doing is looking back and making fun of the people who took risks that didn’t work out well, pretending that they’re much smarter than those people were. That’s not the approach we’re going to take.

We’re going to take the ex ante approach. When I say that the risk was predictably too high, it’s because based on evidence that you could have prior to the collapse, you would have known so. And surprisingly, there are some events where that’s true. So we’re going to keep ourselves with ex ante, not ex post thinking. We’re going to look at how people actually were gauging risk. We’re going to try to ask what they were thinking, what information they had, and how reasonable their actions were on an ex ante basis, not an ex post basis.

The Roman Banking Panic of AD 33

Okay. Let’s begin with the Roman banking panic of AD 33. This is the oldest crisis that we’re going to talk about. But the Roman world, despite the fact that it’s happening 2000 years ago, seems remarkably familiar.

So there are 2 kinds of banks or lenders, I would say, in Rome at this time. One are deposit banks, maybe more familiar to us. The other were money lenders who were really members of the political elite who just ran their own lending operations as individuals. Lending of both groups was regulated, as it is today, for political purposes.

I would say that most of the literature by scholars of Rome would say that the goal of this regulation was primarily to sustain the empire and to sustain the elite who controlled the empire. So the people who were running the empire, the Roman senators, were wealthy landowners. And the regulations — we’ll see very clearly in the example we’re going to talk about — the regulations were geared, of course, to help the members of the elite because they’re in control, but also to promote the interests of the empire, a way to kind of coordinate a good future for all of their investments because they all cared about promoting the future of the empire.

And the empire, at the time this is happening, has been expanded recently in the recent decades, as you all know if you know Roman history. But with that expansion came new concerns about the stability of the empire, about the ability of the empire to hold itself together as a political entity spanning the whole Mediterranean area and beyond. And so the banking regulations reflected that.

There was also perhaps unwitting macroeconomic variation. So at the time of Julius Caesar, generally speaking, the Roman policies resulted in making, through lots of government expenditures, interest rates very low, and credit and money were quite abundant. But under Tiberius, just prior to this crisis, that had been reversed and interest rates were rising, which made, as we’ll see, one of the regulations — the usury ceiling — become binding.

Roman Banking Regulations: The Usury Ceiling and Land Requirements

So we’ve got some background. One is the empire run by the elite and for the elite, but also what that means is wanting to preserve the stability of the newly expanded empire. There are 2 main bank regulations. One of them is limits on the interest rate that you can charge. This is called the usury ceiling.

Now, that had not been enforced, partly because interest rates were generally pretty low. The purpose of a usury ceiling in most of the finance literature is seen as favoring the elite. Because if you are a member of the elite, your risk level tends to be lower than some new entrant. So by keeping interest rates low, by saying you can’t charge high interest rates, you implicitly are forcing lenders to lend to members of the elite.

Now, that might be justified. Usury laws might be justified by politicians for a lot of reasons. But as economists and as economic historians, we tend to find that constituent interests tend to explain regulations pretty well. And in this case, the constituent interests — this is true also of the US in the nineteenth century studies of usury laws there — what we find is that the best explanation is that usury laws favored low risk borrowers, and in this case, members of the elite.

The other regulation was requiring that when you lent money, you had to make sure that a large amount of your own wealth as a lender, as a bank, was actually in the form of Italian land. So, it limited the amount of loans you could make because you had to make loans in proportion to your holdings of Italian land.

So why were they so interested in making sure that lenders, the wealthy people in the economy, held Italian land? I think one interpretation is fairly obvious and the historians have made it: that Italian land is at the center of the Roman Empire. And so as the empire is scattering, they want to make sure that people who are part of the key governance are, as much as they may be drawn to distant places, still investing locally, still part of a critical mass of their wealth being near to the center of the empire. So those were the 2 regulations.

The Crisis Unfolds: Credit Contraction and Falling Land Prices

Alright. Now here’s more of the background. So what happens is, as under Tiberius, as I said, interest rates start rising. And some of the people who are borrowers who are also in the Senate start complaining that the interest rates are rising. Oh, and by the way, there’s this dormant law out there that says interest rates can’t rise above a certain point. And so they complain, and then they enforce the law.

When they enforce the law, the amount of credit declines in the economy. I’ll show all this to you on a graph. We’ll review it together. The amount of credit declines. When the credit declines, another action is taken. Because when the credit declines, land also declines.

So, what’s happening is, as credit is declining, that also means that credit available for buying land is also getting more scarce. And so, what happens to land prices? They fall. So when the usury ceiling is initially enforced, it leads to a scarcity of credit, which tends to also be reflected in a decline in land values.

But then the government’s reaction to this is, well, we don’t want land values in Italy to decline. So what we’re going to do is increase the regulation, increase the proportion that money lenders have to hold in land. But then what happens when they do that? Well, it means that the money lenders have to lend out even less. And what that does is it makes land prices fall even more. So I’ll show this to you in a graph.

The Roman Credit Crisis and Its Modern Parallels

So the point is, these were bad policies in the sense that they led to an undesirable contraction of credit. And then when they tried to fix the effect of that on land prices by using the land holding requirement, by boosting the land holding requirement, they made the supply of credit fall even more, causing land prices to collapse. At this point, the banks and everyone in the economy is in chaos, and the emperor resolves the problem by becoming the first known lender of last resort, making a 3 year interest free loan to all the lenders from the Roman treasury to bring an end to the crisis.

So it’s a pretty interesting example of a crisis that’s clearly related to government policies that have clear political constituencies. Here’s a graph that illustrates this.

So the initial equilibrium, pre crisis, is the intersection of the 2 blue lines. So pre crisis, you can see the interest rate is above the legal usury ceiling, which is the flat orange line. And then you can see the first thing that happens when they impose the usury ceiling, the new equilibrium — so we can understand this evolution through this graph.

If we go over here, you can see before the crisis, these 2 blue lines, this is the supply of loans, this is the demand for loans, pre crisis are intersecting at this interest rate, which is above the usury ceiling interest rate that’s not being enforced at that time. And then when the usury ceiling is enforced, well, loan demand stays the same, but now the intersection of loan demand and loan supply is here.

So you can see a huge decline in credit equal to this distance. That means that land prices are declining initially because credit is more scarce, because we see why credit is more scarce, because interest rates are being kept too low. But then, to fix it, the Senate says, well, we’re going to force lenders to hold more land, which we hope will drive up the price of land. But that means if you’re forcing them to hold more land, they have to hold less loans. So that shifts the loan supply curve inward, restricting the amount of credit, which means that the new equilibrium over here is even worse.

So the first policy of imposing the usury law shifts credit negatively by this much. The next policy by boosting the land ownership requirement reduces credit by this additional amount. So what this analysis shows is that 2 regulations that we can completely understand from a political perspective — usury laws and limitations on wealth holding that try to make sure that wealthy people are holding a portion of their wealth in politically important assets — we can make perfect sense of those.

By the way, those regulations weren’t repealed. And we’re still living with them in most of the world. Regulations that are doing one or the other of those things. Despite the fact that those regulations were very costly.

The Political Logic Behind Economic Regulations

So that leads us then to ask, well, should we have learned differently? Of course, there was no repeat of the specific circumstances. But the same political choices, the same regulatory choices are reflected in capital control regulations and usury laws that are still used as political tools today.

By the way, economists tend to complain about them, myself included, saying these are generally not good for economic efficiency. There are better ways — even if you want to do redistribution, there are economically better ways to achieve those. And yet, what we see is, just as these tools were used in ancient Rome, they’re used by economies today.

And the reason, I would say, is that there’s some kind of political purpose that they tend to serve, and so they continue to be useful. It wasn’t that the Roman Empire had these regulations in place because they were trying to optimize some economic achievement of growth. It was more that they were trying to maintain a political equilibrium that rewarded members of the political elite and that made sure that the members acted cohesively to have critical mass of their wealth invested in the center of the Roman Empire. So you can make sense of that politically. It isn’t about economic growth of the empire. It’s about trying to create a stable political coalition.

When we — if we were going to talk about, for example, Mexico around the year 1900, we would be talking about the dictator Porfirio Diaz, and what was called the Porfiriato. And you could say, well, that was a very inhospitable kind of regime. It wasn’t pro growth, but it was very much pro members of the elite that preserved the stability of the country until 1912 when it didn’t preserve the stability of the country anymore. That choice though actually made economic success in Mexico much greater than it could have been probably with any other system.

So when you’re looking at the question of does this promote economic growth? Don’t compare it to a vacuum. Compare it to what are all reasonable alternatives. Because you can’t have economic growth or a banking system occurring in a vacuum.

For example, after 1912, there was no banking system in Mexico for decades. Because if you have a complete chaotic situation of bandits running throughout the country and no centralized government in control of the peace, you can’t have a banking system. It’s like an invitation. Put up a sign: “Come steal my money.” Banks disappear.

So, when you’re asking the question, did this promote the growth of the Roman Empire? Which is a very good question. It’s not so obvious how to answer that. Compared to what? Because you can’t have an empire if it’s not stable. So, if the purpose of these regulations was to create a stable coalition to run the empire, well, that’s a precondition for even being able to talk about economic growth.

So, economic policies can’t be understood in isolation without thinking about how they affect the ongoing political equilibrium.

The Rise of the Modern World

Now we want to pivot to talk about the rise of the modern world. So we’re going to skip a lot of history. The whole Middle Ages just got done away with here. But we’re going to talk now about the rise of the modern world, and the modern world is the world that we still live in that starts, let’s say, around 1600.

And what defines this modern world are changes in technology for weapons, shipping, navigation, which centralized power. So, instead of a medieval world where power is very fragmented, think about all those beautiful castles that you’ve seen throughout Europe. That’s what decentralized power looks like. Somebody can sit in a castle out in the middle of nowhere and be pretty much in charge of that local area.

But if you have a cannon, a castle is kind of useless. So, that’s a nice way to think about the transition from decentralized power in Europe to centralized power — the creation of a cannon, which makes the king not just the head of a coalition of local fiefdoms. But now the king, if he can raise enough money to fund an army and the weapons and the navigation and the shipping, can now exert power in a much more centralized way. Not just within the country, but now across the whole world.

That’s what defines modernity as we use it in economic history and political history. It’s a change of power that makes the center of power more centralized and it makes the breadth of the power more sweeping. And it’s fundamentally a technological change.

And sometimes this era, the early modern era, is referred to as the era of mercantilism imperialism because there’s a coalition between these new powerful centralized leaders in Spain, in Portugal, in the Netherlands, later in Britain and France. A coalition between them and the merchants who are going to execute — not just the conquest, that’s for soldiers — but they’re going to set up trade routes so that the consequence of conquest can be the increase in trade. This is also physically associated with a complete shifting of the center of gravity of the world from the Mediterranean to the Atlantic Ocean. I’m not saying other oceans weren’t important too, but the Atlantic Ocean was really the key.

New Institutional Tools of Mercantilism and Imperialism

Now, in addition to the new technologies, there were new institutional tools of mercantilism imperialism, which were fundamental and extremely novel. What were these? The granting by sovereigns to particular groups of people monopoly rights over some kind of activity. So, it could be a monopoly right to form a bank, like the Bank of England. Or it could be a monopoly right to organize trade with a certain region, like the East Indies Company. Or it could be other kinds of monopoly rights, to run a lottery, for example.

So, granting monopoly rights — that was something new. As part of the strategy, new partnerships between particular individuals or groups or merchants and the sovereign.

Secondly, the chartering of companies, initially called joint stock companies, basically what we would call corporations, which allow the pooling of funding by a wide variety of investors. Again, it’s all part of this need for large sums to help to fund centralized power.

The issuance of new types of debt. So, we’ll see that a lot of these sovereigns are issuing lots of different debt, like IOUs here, short term this, a little bit of that. And all of a sudden, we see big advantages to creating much more homogeneous types of debt, long term bonds, what eventually became in Britain known as consuls. And we’ll see that there are big advantages to the sovereign of creating a single class of homogeneous debt, mainly because it raises the price by making the debt more liquid.

And some of those corporations that are now being created are banks. Because a major idea that’s emerging in the modern era is that there’s a big advantage to the sovereign of chartering banks that have a mission — from the sovereign’s perspective — of making the empire great. And from the banker’s perspective, having the sovereign as your ally is a very helpful thing.

So, when you’re thinking about these new tools, don’t just think technological tools, but new innovations of institutions. Again, what’s the list? Monopoly privileges. Joint stock companies or corporations with wide ranges of investors. New types of sovereign debt. What will emerge in Britain, the consul. And the chartering of privileged banks.

Now, of course, Europe already had banks, but they were just privately created things. The Medici. This is not that. This is where the state says, “We’re creating a bank, and it has special privileges.” Now, we could decide to charter lots of them. Some countries do that. Or we could decide to charter only 1. And so we’re going to talk about that distinction too.

So the point is the modern era is all about initially technologies and later institutional innovations that are promoting this new centralized government with wide ranging imperial ambitions.

Sovereign Risk and Financial Crises in the Early Modern Era

When we’re thinking about financial crises in the early modern era, it’s all about sovereign risk. It’s all about the competition among these countries. Because that’s what’s defining the risk of the era.

There are individual companies. They’re issuing stock on the London Exchange. Sometimes it goes up and goes down. But the thing that really defines the crises we’re going to talk about, the risks that make systemic problems, are the sovereign risks.

And the first 2 examples we’re going to focus on, the Mississippi Bubble in France and the South Sea Bubble, are wonderful examples of how backward sovereigns — sovereigns that are late at the mercantilism imperialism game — Britain and France are desperately trying to use these innovations, these new institutional innovations, to catch up with the incumbents who were running the show at the time.

John Law and the Mississippi Bubble

The thing that makes the Mississippi Bubble so much fun to study is the character who’s at the center of it, who was perhaps still to this day the most colorful character in the history of finance, a Scot named John Law.

And John Law had been a schemer for a long time prior to this. But I think what might surprise you, if you’ve only read superficially about the Mississippi Bubble and about John Law, is that financial historians that are studying what John Law did, and economists, do not characterize John Law as some insane person who managed to, inexplicably drag the whole French Empire down through his cockamamie schemes. No. That is not what I’m going to teach you. And that is not the received wisdom.

The received wisdom is John Law actually was an extremely creative person who understood all of the possibilities of these new institutional innovations of modernity. The corporations, the banks, the funding schemes — and who saw all of the mechanisms that others were also using at the time, understood them, and wanted to invent them in a new way. His particular ideas, qualitatively, all made sense. And I think now the consensus view is he just took it a little too far.

And in particular, the thing that made it a little too far was when he thought that he could also fix the market prices of things. So, he had his scheme for his company, and he also was running a bank. And so then he said, “Well, I don’t like the current price that the shares of the company are getting, so why don’t I just print a lot of notes in the bank and set the price higher than what the market wants to set it at.”

And that was really the idea — or doing other kinds of behind the scenes, you might even say quasi fraudulent, in the sense that, if you print up a lot of money, it’s obvious that you’re probably going to fuel inflation. He knew that. He was smart enough to know that. But people sometimes get stuck wanting to achieve a short term objective and convincing themselves that, “Oh, all I have to do is just help the market a little bit, see the wisdom of that price, and I won’t have to flood the market with notes to achieve that.” And it turned out that he was wrong about that.

So, the lack of market discipline, the fact that 1 person’s in charge of the whole country financially and can make, through centralized decision making, whatever decision he wants and can decide to try some scheme that’s going to say, “That price, I want to keep that price high. And I have all the tools I need to do that?”

The Limits of Monetary Power and the Collapse of Law’s System

I can print money like crazy. But, of course, when you overdo it, then it doesn’t work because ultimately, you can’t fix the stock price by just issuing money, at least not in real terms. And so as he learns the hard lesson — there’s also a market out there that decides prices. Just because you have all the powers of France doesn’t mean that you can set the price of something. That’s a very important lesson.

But the fact that he didn’t get that lesson means that the whole thing came tumbling down. He is driven out of France. And for — I think it’s more than a century — France doesn’t even use the word bank in any chartered company, because it’s such a disreputable term. So it was a very major financial crisis and disaster.

But what I want to — when we come back to talk about it — what I want to emphasize is that it was not a crazy idea by some lunatic who randomly got control. We understand why he got control. We understand that his ideas were quite sound, were related to other things that people were doing throughout Europe, that they were original too. They were probably quite smart. And if you hadn’t taken them a little too far, they might have been very successful.

In fact, they might have put France on a much better historical trajectory in terms of its global competition. So, let’s take a break, and we’ll come back to pick up where we left off.

Related Posts

• You’ll Stop Using ChatGPT After Listening To This: Jonathan Pageau (Transcript) 
• Transcript: The 100 Year History Of ‘Woke’ (In 15 Minutes) – Paul Marshall
• Transcript of The Why Files: The Asteroid Behind the Bible’s Most Famous Disaster
• Transcript: “This Is A Civilisational Failure” – Katharine Birbalsingh at ARC26
• ARC 2026: Anthropic’s Chloe Lubinski on AI (Transcript)