Q & A With Martin Feldstein

Few other academics have had as much impact in recent years as economists, in turn attracting the best and the brightest to the field. The American reports that the number of students graduating with degrees in economics increased by 40 percent in five years in 2004.

And no institution is as responsible for this renaissance as the National Bureau of Economic Research, the private, nonprofit, and nonpartisan group based in Cambridge, Massachusetts. Founded in 1920, the bureau currently provides research support for over 500 top economists nationwide.

Researchers at the NBER have pursued influential research in fields as varied as health care, education, aging, and finance -- to name just a few.

And the man largely credited with creating the environment that allowed this research to flourish is Martin Feldstein, the NBER's president and a professor of economics at Harvard University. Feldstein, 67, announced last week that he was stepping down from the post at the NBER after 30 years next June. He says he wanted to concentrate less on day-to-day management and pursue more writing and research.

As for accomplishments, he's had quite a few. Feldstein received the prestigious John Bates Clark Medal in 1977, formalized the NBER's role as arbiter of business cycles, was the chairman of President Ronald Reagan's Council of Economic Advisers from 1982 to 1984, and is currently a member of President Bush's Foreign Intelligence Advisory Board. And if you care about such things, Feldstein is the 8th ranked economist in the world, on this experimental ranking system.

I talked with him recently about how he helped transform the NBER into a research powerhouse. Here are some choice excerpts:

How has the NBER changed since you came on board?
It was basically a New York-based small research organization with a couple of dozen people, a small budget, and a fairly narrow range of activities.

My role was to decide what changes needed to be made and to recruit the people to do it. At the beginning, people were not so sure they wanted to be associated, even though the bureau had been around since 1920. Several people said "Marty why are you doing this? Why aren't you continuing your research?" But they quickly said, "If you're foolish enough to do it, sign me up."

We started a series of research programs (on labor studies, taxation, economic fluctuations, asset markets, etc.). We started with a half of dozen of these and there are now more than 18.

How did you recruit researchers?
We said to people that we're not going to tell you what to do, we're going to offer you the opportunity to get together with other researchers in your field a few times a year to talk about research, and then from time to time we'll offer you a chance to participate in research programs.

That wasn't a bad offer, so some very good people accepted it.

It's important to mention that academic economists working in their own departments usually are one of at most two or three people in a particular field. It's a very rare department that would have two people work on international finance, so there's nobody to talk to, and that affects the character of your research. It becomes difficult to work on things where you need to understand the data because you could get it wrong pretty easily and there's nobody down the corridor who can help you -- and not only that, they're not particularly interested. So instead, if you work on more abstract theoretical things, well, then you have something to talk about with your colleagues.

One of the things we did was by bringing together what may have started as 10 or 12, but soon became 20 to 30 people in a program was to give them a chance to interact with others in the same specialty from around the country who could be mutually helpful. "Have you considered such and such a data set?" or "Do you understand that those data don't mean exactly what they appear to mean?" and "What about the recent legal change that caused such and such?" I think we encouraged more applied empirical research than would otherwise have occurred.

Was it a hard decision to give up some of your time doing research?
It turned out that it was much less of a sacrifice then some people made me think it would be. If you look at my CV, I don't think you find a point of inflection where suddenly I became president of the bureau and the research stopped. If anything, it had the opposite effect, because it kept me very much in touch with what was going on.

It's been documented by NBER research that top economists are publishing less than they used to in top journals. How has this played out at the NBER where most of these top economists do their work? (Note: This may already be obvious, but NBER working papers are not peer reviewed.)
We had more than two million downloads last year from all around the world, so it is the way that people read research. In fact if you look at academic papers you see that many of them cite the bureau working paper even if it's subsequently got published because they didn't bother to follow up and see it.

Now, what peer review does is to say is that this paper, especially if it's in a good journal, is a quality piece of research. Being in the bureau working paper series is another kind of certification. It says you may not know professor Smith or assistant professor Smith, but he or she must be pretty good or they wouldn't be part of the NBER family, and so the quality control comes in that way.

What's your take on the popularization of economics that's been brought about thanks to the work of people like Steven Levitt?
What Steve really did was not to talk about subjects that others hadn't thought about, although occasionally he's done that. He found ways of being persuasive about what the data are trying to say.

The technical issue is whether you have correlation or causation, and a good example is the question of do more police on the streets reduce crime?

When there's a crime wave they put more police on the streets so you can't tell. Steve had the clever idea that before elections, regardless of what's happening to crime, the mayor says to the police department "Put more cops on the street" in the hope that that will reduce crime. The question is: Does it? And lo and behold, Steve found that it does, so there you've got a nice, clean experiment where you can tell the direction of causation.
(Note: You can read Levitt's 1995 paper here.)

What is the most under-researched area of economics?
What I say to graduate students and people at the NBER is that the economics of national security is the most neglected important field in economics. Just a generation ago we didn't have people working on health issues or education issues and now there are so many economists who work on it. Surely national security in its various dimensions deserves the same kind of attention.

You were in the running to take over for former Fed chairman Greenspan. What's your assessment of his tenure?
Hindsight is really very helpful in doing economic policy. At the time, when they cut rates down to one percent in 2003 they were very concerned that we could move into a deflationary spiral as Japan had. So, [the FOMC] made a decision to push rates down to try to stimulate the economy.

Did they keep them down too long?

By promising to keep them low I think that may have been a mistake. That led to the teaser rate promises of mortgage companies because they knew they would have a few years of low rates. That's not a brand new thing but it has exacerbated the problems that we've seen.

Photograph by Steven E. Frischling/Polaris