An Econ Nobel for research that saved the world
Bernanke, Diamond and Dybvig explained why banks collapse -- and take the rest of the economy with them
The 2022 Econ Nobel went to Ben Bernanke, Douglas Diamond, and Philip Dybvig for their work on banking and financial crises. This award is, in my opinion, richly deserved — very few economists have done research that saved the economy in real time.
But the award also came as a bit of a surprise to me, because the Econ Nobel — unlike the natural science prizes — is usually not given for a specific discovery about the economy. Usually, it’s a prize for developing a new research method. In other words, the Nobel committee usually seems to judge the importance of a piece of economic research based on how much other economics research it leads to. This is especially true in macroeconomics. This is not a knock on the Econ Nobel — econ, and really all of social science, is still a very young endeavor, and discovering new ways to discover things is often the highest-value work.
The 2022 prize is different. Yes, the famous Diamond-Dybvig model, and Bernanke’s work on financial macroeconomics, have generated substantial follow-on literatures. But their primary importance is in the here and now. Diamond and Dybvig developed a model of banking crises that — pretty much everyone now agrees — explains the basic reason why banks tend to collapse. And Bernanke showed how those collapses could bring down the real economy. Those insights were then successfully applied to save the economy from a second Great Depression in 2008-10 — in part by Bernanke himself.
This is a stunning achievement for a field that usually finds itself struggling to predict what inflation will be three months from now. Most of macroeconomics is still a mystery to us, despite all our fancy math. But one crucial piece of the mystery — the ever-present danger posed by our fragile financial system — is now something we think we sort of understand. And that understanding gives humanity enormous power to make our economies safer and more prosperous, to safeguard the livelihoods of millions, and to make the world a more stable place.
Why banks fail
If you’ve ever seen the movie It’s a Wonderful Life (pictured above), or the movie Mary Poppins, you’re familiar with the concept of a bank run. The scenes are iconic — a bunch of people hear false rumors that a bank is about to collapse, and they rush to the bank in terror, afraid that they won’t be able to get their money back. The bank wasn’t about to collapse, but now, because everyone panicked, it does collapse. Some version of this scene has repeated itself in real life, in country after country, since the dawn of banking.
But why does this happen? Why can’t banks just give everyone their money if they all want it back at the same time? The answer is that they don’t have it.
The reason has to do with why banks exist in the first place. Banks aren’t just safe deposit boxes where people keep their cash. They make profit by lending out money. Banks lend to homebuyers, small businesses, big businesses, the government — basically to anyone.
Most of these loans are highly illiquid — you can’t just go ask for cash back from homeowners and businesses any time you want. Imagine if someone from Wells Fargo came to your door and said “Uhh, hey…um…we need you to pay off your whole mortgage, like, right now.” You couldn’t do it. Mortgage borrowers, businesses, and the government pay off their loans in small increments over time, not in one big lump of cash whenever you feel like it. These loans pay higher interest rates than the loan on a savings account, precisely because they’re less liquid. The spread between the high interest rates on a bank’s illiquid loans and the low interest rates on a bank’s deposits is called the net interest rate spread, and it’s how a bank makes profit.
So imagine if banks had to take $1 in deposits for every $1 they lent out (which isn’t true, but just suppose). Even in this world, they would still not be able to pay out all their depositors at once, if they all came running. Because most of the money is locked up in illiquid assets. Therefore, banks are inherently fragile creatures — if a bunch of people suddenly get scared that they won’t be able to get their money back, all at the same time, that fear becomes a self-fulfilling prophecy and the bank collapses.
This insight is the basis of the famous Diamond-Dybvig model of bank runs, published in 1983. Diamond and Dybvig use some simple game theory concepts to show how runs work. But the model has important implications beyond simply “runs happen” — it also gives us some insight into how to stop them from happening.
In the old days, when runs happened, banks would simply stop paying depositors’ money out, and tell them they just had to wait. Diamond and Dybvig show that the threat of halting payouts (“suspending convertibility”) is enough to decrease the frequency of bank runs, but not enough to end it entirely — sometimes runs still happen. This is especially bad when one bank failure scares a bunch of people and leads to more bank runs, causing chaos in the whole financial system. This is why one of Franklin D. Roosevelt’s first acts as President was a banking holiday that shut down the whole system for a while. This was successful, but it’s not the kind of thing you want to have to do a lot.
A better solution, Diamond and Dybvig show, is federal deposit insurance. If the FDIC will step in and give you your money back if your bank fails, you don’t really need to worry much about your bank failing. So since people worry less, they don’t tend to cause runs in the first place. In fact, the FDIC has been quite successful — it hasn’t eliminated bank runs completely, but it has cut down on them quite a lot. Diamond and Dybvig also suggest that the Fed act as a lender of last resort, lending money to banks in order to give people cash.
Now, both of these policies were already in place long before Diamond and Dybvig published their paper. But the model they created is more than just a vindication of things we were already doing. There are plenty of other bank-like institutions and situations in the world, such as:
savings and loans (which had a crisis in the 80s)
shadow banks (which had a crisis in 2008)
interbank loans (which also had a crisis in 2008).
In fact, George Bailey’s bank in It’s a Wonderful Life is actually an S&L, which is why it doesn’t have FDIC protection and is thus vulnerable to a run.
For these institutions — where deposits and withdrawals sometimes don’t work the same way they do for a bank — it can be very helpful to do the game theory of a run, and design ways to protect them without killing their business models. Extensions of the Diamond-Dybvig model can be used to do that. Diamond-Dybvig also creates a framework for people to study other questions about financial crises, like why people need liquidity in the first place, or how panics get started.
But the most important effect of Diamond-Dybvig is probably just to explain why bank runs are a fundamental feature of the economic universe — why no matter what we do, we’ll always have to worry about the possibility of financial institutions that borrow short and lend long suddenly collapsing. And that means we must think about how those inevitable collapses will affect the real economy.
This is where Ben Bernanke’s research comes in.
Why financial crises cause economic crashes
Why should it matter for the whole country if a bunch of banks fail? If half the mini-golf businesses or bowling alleys in America suddenly went under, it wouldn’t tank the economy. It’s true that big waves of bank failures tend to be followed by a recession, but this doesn’t mean they cause the recession, any more than roosters cause the sun to rise. In fact, some economists who study business cycles have argued that finance is just a “veil” for the real economy, and that we shouldn’t be particularly worried that the banking system itself will be the cause of real economic downturns.
Ben Bernanke was one of the first macroeconomists who put forward some very clear and concrete reasons why bank failures could be a major cause of recessions. He wasn’t the first to think about these topics, but he was one of the first to put them into the language of mathematical models based on clear assumptions about individual economic behavior. These aren’t the kind of models you can use to make quantitative predictions; instead, they’re just very detailed explanations of forces you think could be at work in the economy.
Bernanke’s theories about how finance could wreck the economy were put forward in a series of papers — one in 1989 with Mark Gertler, one in 1996 with Gertler and Simon Gilchrist, and a third in 1999 with Gertler and Gilchrist, as well as a few others. The basic idea throughout all of these papers is that there’s a feedback loop between asset prices and corporate borrowing. When asset prices are high, companies can easily borrow a lot of money, because they can use the expensive assets as collateral, so banks don’t have to think too hard about whether the borrower will default on the loan.
But when asset prices drop, companies suddenly find it much harder to borrow from banks, because now banks suddenly have to worry that they won’t pay the money back. Companies could just go to the bond markets instead of banks, but those will charge them a higher interest rate (because there’s no collateral). So companies just borrow less and invest less, which reduces economic activity. This reduces asset prices even more (because asset prices depend on the real economy), which hurts lending even more, in a vicious cycle. That cycle is called a “financial accelerator”.
This obviously isn’t the only way that financial crises can hurt the real economy. You could also have a bunch of Diamond-Dybvig style bank runs that collapse a bunch of banks. This could A) damage the valuable long-term lending relationships between banks and companies, and B) deprive the economy of bankers’ ability to tell good borrowers from bad ones. Both of those things would make it hard for companies to borrow. In fact, Bernanke, who is also an economic historian of the Great Depression, wrote a 1983 paper arguing that this was one reason the Depression got so bad. But the financial accelerator papers were especially influential in convincing the economists of the 90s and 00s that financial crashes were a big danger.
Which turned out to be very, very important.
When macroeconomics saved the world
One very important fact about the Great Recession that began with the fall of Lehman Brothers and the financial crisis of 2008 is that it ended up being much, much less bad than the Great Depression. In 2009, it wasn’t clear that this would be the case. The initial shock looked just as bad, along a number of dimensions, as the crash of 1929. But by 2010, it was clear that the world would do much better this time around. Brad DeLong summed it up in 2018:
To be sure, nothing happened in the years after 2008 that compares to the four-year slide after 1929, when U.S. nonfarm unemployment rose to 28 percent and German joblessness topped out at 33 percent. The numbers on output reveal much the same story. Four years after the business cycle peak of 1929, national income per capita was down 28 percent, and it did not return to 1929 levels for a full decade. By contrast, after the financial crash in 2008, per capita income fell by only 5 percent and was back to its pre-crash level in six years.
Why did things turn out so much less awful this time around, despite a similar initial shock? There were a number of factors, but the most likely answer is that this time, the banking system did not fail. In the Great Depression, half of American banks failed by 1933; in the U.S., only 0.6% of banks failed by 2009.
And why did our banks not fail this time? Because we bailed them out. The most famous bailout by far was the $431 billion Troubled Asset Relief Program (TARP), in which Congress bought crappy housing-backed bonds from U.S. banks. But another quasi-bailout was probably quantitative easing itself. In the name of stimulating the economy with easy money, the Fed bought over $3 trillion of assets from banks. Many of these were government bonds, but much was mortgage-backed bonds; transferring these from banks to the Fed helped restore confidence in the big banks. And on top of all that, the Fed stepped in to do some of the work of banks, lending money to consumers under the Term Asset-Backed Securities Loan Facility (TALF).
To say that these bailouts made people angry would be something of an understatement; they seemed to reward the bad actors who had brought the crisis upon us in the first place. But it was absolutely the right thing to do, because if the U.S. had had 50% of its banks fail instead of 0.6%, a second Great Depression could easily have been a reality. And the reason we did it, even in the face of all that popular rage, has a lot to do with the fact that the most important economic policymaker in the country was Fed Chair Ben Bernanke.
Most people who are forced to deal with momentous historical events do not have the luxury of preparing for the particular challenges they face. Franklin Roosevelt, for example, did not come into office expecting to fight World War II. Ben Bernanke is an exception to this rule. More than almost any other economist of his time, he had spent his career thinking about the Great Depression—the closest analogue for the crisis he would eventually face…
[I]t was eerily providential that when the Great Recession hit, America's most powerful economic policymaker (Bernanke was appointed Fed chair in 2006) was the economist who had spent more time than almost anyone thinking about the main historical precedent for this sort of crisis. At a time when the financial sector threatened to collapse, the Fed was headed by one of the only macroeconomists who realized how dangerous a financial collapse could be. Nearly anyone else—for example, Martin Feldstein or Glenn Hubbard, who were widely mooted for the top Fed job—would have been more blasé about letting the big banks collapse under the weight of their own bad decisions. Bernanke, on the other hand, bailed out big banks quickly and decisively.
In the months after Lehman fell and lending dropped off a cliff, Americans were furiously debating whether banks were facing a liquidity crisis or a solvency crisis. Bernanke realized that these were one and the same. Thanks to Diamond-Dybvig style bank runs and financial accelerator effects, fear of a solvency crisis could generate a liquidity crisis, which could then quickly become an actual solvency crisis. The key was to provide liquidity and ensure solvency at the same time, thus short-circuiting the vicious cycle.
The results of the bank bailouts, and of QE, appear to validate the theories of this year’s Nobel winners. Freed from both imminent collapse and from the fear that they were holding bad collateral, banks began lending again as early as 2011:
Asset prices recovered, which bolstered collateral and enabled more bank loans. The U.S. government even made a profit on TARP, and eventually made a profit on QE as well — which doesn’t really matter, but which shows the strength and swiftness of the recovery. And saving the U.S. financial system meant saving the world financial system — especially when Europe was able to (belatedly) apply the lessons from Bernanke’s financial rescues to its own crisis in the 2010s.
In other words, in the Great Recession, we saw, in real time, the direct application of macroeconomic theories to the saving of the actual macroeconomy, by one of the researchers who created the theories in the first place. And if you believe that the severity of the Great Depression contributed substantially to the rise of fascism in the 1930s, those economic theories might just have saved not just tens or even hundreds of millions of jobs, but the world itself. Thanks to advances in our understanding of macroeconomics, we did better this time around.
That’s all we can hope for, really. To understand a little more of the world, and to do a little better.