Bernanke's 1980s Computer Model Predicts Crisis A set of equations drawn up in the 1980s by Federal Reserve Chairman Ben Bernanke considered the effects of shocks to the financial system. A co-creator says that the model predicts disaster if there's not a quick intervention, and it guided Bernanke to take action now.
NPR logo

Bernanke's 1980s Computer Model Predicts Crisis

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript
Bernanke's 1980s Computer Model Predicts Crisis

Bernanke's 1980s Computer Model Predicts Crisis

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript


It's Morning Edition from NPR News. I'm Steve Inskeep.


And I'm Linda Wertheimer. Wouldn't it be great if there were a crystal ball you could consult about our economic woes? You could ask it, will there be a recession, how bad, should we launch a $700 billion rescue mission? Well, the closest thing we have in the real world are computer models. We have them for the weather and, it turns out, for the economy. In fact, one was written by a guy named Ben Bernanke. You know him now as chairman of the Federal Reserve. NPR's David Kestenbaum reports.

DAVID KESTENBAUM: That computer model is called the Financial Accelerator. I went to see it at the office of Mark Gertler, a professor of economics at New York University. Gertler worked with Ben Bernanke on it in the 1980s. I was hoping for a sort of video game like that "SimCity" thing. But when I asked Gertler to show me the model, he started to print something out.

(Soundbite of printer)

KESTENBAUM: It turns out he and Bernanke were equations guys, not programmers.

Dr. MARK GERTLER (Economics, New York University): For that we have highly intelligent and ambitious graduate students.

KESTENBAUM: The research paper he prints out is filled with what look like physics equations. Are there people in your computer model?

Dr. GERTLER: Yeah. There are people and their firms.

KESTENBAUM: Is there a stock market in there?

Dr. GERTLER: Absolutely. The firms issue equity and debt.

KESTENBAUM: Now, there are lots of models of various aspects of the economy out there. This one is sort of unique because it was designed to incorporate the rippling effects of shocks to the financial system. Exactly what is going on now?

Dr. GERTLER: Ultimately, we would like to be able to make precise predictions. Yes, we know the economy will get into trouble or can get into trouble through these forces. But we'd like to be able to say how much.

KESTENBAUM: So, does the model say, you know, we're heading for doomsday unless something happens?

Dr. GERTLER: Roughly speaking, yes.

KESTENBAUM: The model calculates how a shock to the system like the credit crisis happening now might snowball dangerously. For example, in a credit crisis, it's hard to borrow money. To take out a loan, you have to agree to pay a lot of interest. That hurts. The result is that people and companies have less money. So, they spend less. The economy slows. The problem is the system feeds back on itself. The slowdown can make houses worth less, companies worth less, which makes it even harder for everyone to borrow money, because their collateral is worth less. The whole cycle begins again.

Dr. GERTLER: We called it a financial accelerator.

KESTENBAUM: Sort of a financial decelerator in this case, though.

(Soundbite of laughter)

Dr. GERTLER: That's what my wife said, but we called it accelerator because it's amplifying the downturn.

KESTENBAUM: Amplifying the downturn?


KESTENBAUM: Now, I know what you're thinking. How can a computer model an economy? Millions of people all making their own, let's face it, sometimes crazy decisions. Other economists take the model's predictions with a big grain of salt. But they also can't resist asking what it has to say. Gertler says there is reason to heed its warnings. It did a good job explaining what happened during the Southeast Asia financial crisis of the late 1990s. South Korea's gross domestic product plunged 12 percent. More than half of that, the model says, was due to the financial-accelerator problem. Gertler says he thinks the logic behind the model is definitely affecting Ben Bernanke's thinking. They've talked about it.

Dr. GERTLER: There's no question about it. I mean, the one thing that I can let on is they said, well, the only good news is the financial-accelerator theory seems to be working. So...

KESTENBAUM: Darn it, we were right.

(Soundbite of laughter)

Dr. GERTLER: Yeah, absolutely. Yes. No, about - no, in all seriousness, this has guided his thinking and really motivated him to act very quickly.

KESTENBAUM: Because otherwise things get very bad very fast.

Dr. GERTLER: Yeah.

KESTENBAUM: Gertler hasn't asked the model to predict what comes next. But his colleague Simon Gilchrist at Boston University, who's worked on the project from the beginning, has been training the model on data from the past 20 years.

Dr. SIMON GILCHRIST (Economics, Boston University): It hasn't really predicted out in the future with this model. That's probably my exercise for Monday. But what the model is saying currently is that we're experiencing about a two percent decline in GDP growth because of these credit-tightening effects.

KESTENBAUM: In other words, growth would be two percent higher if we didn't have this problem. One other question he wants to ask the model about that $700 billion the administration says it needs to fix the economy, is that enough? David Kestenbaum, NPR News.

Copyright © 2008 NPR. All rights reserved. Visit our website terms of use and permissions pages at for further information.

NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.