Interview with Martin Feldstein

Interview conducted July 10, 2006

THE ART OF MONETARY POLICY

Region: In recent articles reviewing the tenures of Alan Greenspan at the Fed and Otmar Issing [former chief economist of the European Central Bank], you observed that monetary policy is as much an art as a science, that judgment must supplement forecasting models and policy rules. Given that, what is your judgment on the current course of monetary policy? I know you're not an advocate of strict inflation targeting, but then what should be done to anchor inflation expectations, either through explicit policy measures or improved communications?

Feldstein: The rhetoric that has worked for the last 20-plus years since Paul Volcker took over was an emphasis on "price stability." That didn't come with a precise number, but I think he defined price stability—and Greenspan used similar language—as meaning that price changes should be so low that people ignore them. And "price stability" actually has, in many ways, a better ring than "2 percent inflation." Why 2 percent inflation? Why not 3 percent inflation? Why not zero? I think what the public wants is price stability. They don't want to have prices rising. In practice, because of imperfections in measurement and because of concerns about deflation, you end up with a number globally now around 2 percent, and that strikes me as quite plausible.

The financial markets may like something reasonably precise, like 1, 2, 3 percent as measured. I say "as measured" because of the statistical bias in the numbers so that true inflation is lower than the measured number. The man on the street doesn't want to be concerned about small numbers and/or about "core" inflation versus "regular" inflation. He wants price stability; he wants the purchasing power of the money that he has to stay the same. And so that's what the Fed's message to him ought to be.

Region: How can that best be communicated?

Feldstein: I think it was successfully communicated during the Volcker and Greenspan eras. And Volcker started with much higher inflation rates, so it was the more difficult sell. By the time he left, the inflation rate was down to about 4 percent as headline inflation is conventionally measured, and Greenspan took it down to roughly 2 percent and then it overshot a little bit on the low side.

So I think it got communicated that the problems of inflation we lived with in the '70s were history and that the Fed was committed to not letting that happen again. It was done without putting a precise number on it, but by reacting to increases in inflation. The public looks at how the Fed responds, not just at what they say. It's easy to say, "I believe in price stability." But if you don't do something, then no matter what number you put out there, they're not going to believe it.

TIME CONSISTENCY IN FISCAL POLICY

Region: Economists have devoted a lot of attention to time consistency in monetary policy. Some suggest that time consistency in fiscal policy would also be a good idea. Given your research on public finance, do you think there's wisdom in such a stance? And given your experience in Washington, do you think it's pragmatic?

Feldstein: When I look at the current fiscal situation, in contrast to what we experienced in the '80s when the fiscal deficits were larger and rising, and debt-to-GDP ratios were rising, we're currently at a relatively comfortable level. The federal deficit-to-GDP ratio this year will be under 3 percent, probably low enough that the debt-to-GDP ratio will actually come down.

The problems are not that very far into the future, though, with increases in Social Security and Medicare costs relative to the tax revenue that comes in. The markets seem to be ignoring that, which is a puzzle, but there's nothing about long-term interest rates that suggests that the markets are afraid that Social Security and Medicare are really going to create large fiscal deficits. Now maybe they're right. And maybe the political process will raise taxes or cut benefits. What has to be done is to reform those programs. I wouldn't set my goal in terms of the fiscal deficit. I'd set it in terms of limiting the tax levels that are going to be needed to support them.

SOCIAL SECURITY REFORM

Region: For nearly 40 years, you've been a powerful advocate for reform of the Social Security system, especially for personal retirement accounts.

Feldstein: Right.

Region: But a case could be made that the system is largely unchanged from when it was first created. Are you discouraged by the lack of progress that's been made, or heartened by the incremental changes?

Feldstein: There've been no good incremental changes. There have been incremental changes, but they've gone the wrong way until about 10 years ago. Since then there have been changes, particularly getting rid of the distortions in retirement incentives by raising the benefits if you work longer, reducing the benefits if you retire earlier. So the distortions in retirement behavior that affect the European economies are no longer present in the United States.

But the reforms of Social Security and the movement to personal retirement accounts are happening around the world now, and the United States is really a laggard. And it's not just some of the developing countries, like Chile and Mexico. It's Australia and even Sweden that have moved to mixed systems with an investment-based component. So I'm encouraged that at some point the United States is going to move in that direction.

If you'd asked me two years ago, I would have said we have a president who's committed to this and sees this as his major domestic legacy, and therefore I think he's going to get it. I can't say that now. I can say the first half of that, but not the second half. The Democrats have been completely unwilling to come and discuss the subject. It doesn't look like anything the administration can do will get them to negotiate about it. There aren't enough Republicans to do it. And it shouldn't be done on a partisan basis. The administration tried, by setting up a bipartisan commission with Pat Moynihan as one of its leaders, to come up with a way of getting to that mixed system. But they've not been able, and since Moynihan died there's been no leadership on that side of the aisle.

EUROPEAN SOCIAL INSURANCE

Region: You've referred to several European countries. I would think that the demographics of Europe would be even less amenable to a favorable future for government pension programs. What is your sense of Europe's future relative to social insurance for the elderly?

Feldstein: You're right. Their demographics are worse. We would see the tax rate required to support the U.S. Social Security pension system rise to about 20 percent from the current 12 percent if we wanted to maintain the same benefit rules on a pure pay-as-you-go basis. Europeans are already up there. They have much higher replacement rates. They have earlier retirement. And for them it's going to get even worse.

But as I said, we've seen Sweden—which I think of as sort of the leading edge of welfare states—backing off the traditional pure pay-as-you-go system and moving to this kind of mixed system. Britain has very much a mixed system. It's not exactly the same structure, but it's very much a modest pay-as-you-go part plus an investment-based part.

So, I think at some point it will happen here. I don't know exactly what the formula will have to be to cause that to occur, but I think it will happen. Israel, another country with a tradition of very strong social welfare programs, has made this transition so that new people entering the labor force contribute to a mixed system.

EUROPEAN UNION

Region: Another question about Europe, if I may. In 1997, you wrote that "on balance, a European monetary union would be an economic liability." What is your judgment of where the EMU stands today, particularly since the recent reforms of the Stability and Growth Pact?

Feldstein: Well, I think it is a liability. I think that the one-size-fits-all monetary policy is seen as not working either for the countries that have weak demand and ought to have a more stimulative policy or, on the other hand, for those that are discovering they're becoming less competitive because their domestic prices are rising and they can't adjust monetary policy. I wrote a piece on the Stability Pact issue before the recent reforms, but even before the reforms, they weren't paying any attention to the Pact. The basic problem is there's no market feedback to discipline a country that doesn't have its own exchange rate or its own interest rate to warn them against fiscal deficits.

Normally, a European country that started running large fiscal deficits would see that the increased risk of their bonds would push up the interest rate they had to pay, there would be a flight from the currency and they would see that that translates into inflation. So they would have a lot of market-driven warning signals. All that is gone now. There's a 30-basis-point difference between Italian and German long-term interest rates because the market doesn't believe that it can discipline the Italians. If the Italians run a large fiscal deficit, it's effectively a European fiscal deficit. It's not a specifically Italian one.

THE RETURN OF SAVING

Region: In a recent Foreign Affairs article, you argue that the downward trend in savings by U.S. households will likely soon reverse, and that that could cause some near-term disruption. Could you explain that prediction, and tell us how that ties in with your recent op-ed in the Wall Street Journal calling for, I think your term was, a "competitive dollar abroad."

Feldstein: Right, strong dollar at home, a competitive dollar globally. Well, the brief history of our savings rate is that from a relatively low level it has been falling ever since the early '90s. This is household saving, not national saving. Household saving was coming down from around 7 percent of disposable income, and by 2003 it had gotten to 2 1/2 percent of disposable income—a remarkably low number.

That was not surprising since people's wealth had increased quite substantially, both because the stock market—despite the fall in 2000—was up substantially and because home values were up. So people felt they didn't have to save by reducing consumption. They just looked at their 401(k)s and IRAs, and at a time when asset prices were going up, the wealth effect dominated.

The fall in the savings rate is a reflection of the fact that we used to have defined benefit pension plans and we now have defined contribution plans, plus the IRAs. All of that put the increase in stock market value into the individuals' hands rather than into the companies' hands as it would have under defined benefit plans.

Then saving fell very sharply in the next two years, 2003 to 2005, about as much as it had fallen over the last decade and a half. It went from 2 1/2 percent to minus 1 1/2 percent. It's not entirely clear what caused that. There was a spurt in home prices, so there was a sharp wealth effect. But the main thing that drove it, I believe, was mortgage refinancing. Mortgage refinancing gave people a chance not only to cut their monthly payments but also to take out cash and use it to buy things. Not all of it went into purchases. Some of it went into paying off other debt. Some of it went into financial assets. But I suspect that enough of it went into buying things to drive the savings rate from plus 2 1/2 to minus 1 1/2 percent.

Well, that process of mortgage refinancing is reversing now, and with mortgage rates significantly higher, a full percentage point higher than they were a year ago, there isn't the incentive for people to refinance. So I think we will see a return to higher savings rates. Whether it'll be 2 1/2 percent or it'll be higher than that, I don't know, but I think there'll be a natural turnaround in the savings rate.

If that happens relatively quickly, it will cause a significant slowdown in aggregate demand in this country. Of course, many people are saying that we need to improve our trade imbalance, and in order to improve our trade imbalance, we have to save more. I think this is where the saving will come from. It will not come from reductions in fiscal deficits, which are already relatively low. Nor will it come from increases in business savings, which are quite strong.

But to convert the rise of savings to an improvement in our trade balance, the dollar has to be more competitive. If we simply have an increase in our savings rate and nothing changes in the exchange rate, then we will have, depending on how much the savings rate goes up, a slowdown or a downturn in aggregate activity. We need to translate those extra resources, that extra saving, into an increase in exports and a reduction in imports so that the total demand for U.S. goods and services remains on track, and the economy continues to expand. And that's the way the market ought to work: When the savings rate goes up, the exchange rate becomes more competitive, and Americans consume more American-made products and services. So that's the case for a more competitive dollar.

In the [Wall Street Journal] article, I said the government during the Clinton and Bush years has been sending out a confusing message by saying that the United States believes in a "strong dollar." Now what exactly does that mean? It seems to me what it ought to mean is we believe in a dollar that is strong at home, meaning that we believe in low inflation. We believe in protecting the value of the dollar when the consumer goes to shop.

Region: Which is why you distinguish between "strong" and "competitive"?

Feldstein: Right. At home versus globally. If we want the dollar to stay strong relative to the euro and Chinese renminbi and Japanese yen, then when the savings rate goes up, we're going to have a serious problem. So I think somehow the message has to get out that the government isn't just concerned with the renminbi, which is the only currency they seem to talk about, but is concerned with making the dollar competitive against all currencies.

Region: If I'm not mistaken, you moved markets a bit when people might have presumed that a competitive dollar meant a weak dollar. You were quoted by the Financial Times as estimating that a 30 percent to 40 percent devaluation might be needed to help narrow the trade deficit. Does it surprise you that you can move markets in that way, that you're that influential?

Feldstein: I don't know that I moved markets, and I never made a personal forecast of what it would do. I have said that in the '80s, when the dollar moved and the current account deficit was relatively smaller—it was 4 1/2 percent of GDP then as opposed to 6 1/2 percent of GDP now—it moved almost 40 percent in two years.

Some of my colleagues who have done detailed analytic calculations would say that's the kind of number that it might take now. That would not get us back to balance, but just get the current account deficit down to something like 2 or 3 percent of GDP, which would be consistent with our external debt-to-GDP ratio not rising. So I've quoted those numbers, but I've never said those were my forecasts. But these numbers are from smart people who've put a lot of effort into trying to estimate that.

ECONOMICS OF HEALTH AND HEALTH CARE

Region: A change of subject. Some of your first academic papers were on the economics of health care, in the United Kingdom in particular. More recently, you wrote a paper about health care economics that was subtitled "What Have We Learned? What Have I Learned?" How do you answer those questions?

Feldstein: The first paper I ever published was about health care in Britain. I was a graduate student in Britain at the time, and I said what they needed to do was to introduce some economic thinking. Remember, it's a state-run system. You need some cost-benefit analysis. Look at the costs of doing things, look at the benefits.

Over the last several decades in this country, where of course it's much more of a decentralized and market-driven health care system than in the United Kingdom, the combination of insurance companies and employers has driven home the message that a doctor, in deciding what to do when a patient presents, is no longer to ask, "What are all of the things I could possibly do to help my patient?" but rather, "What is the cost-effective thing to do? Where do I draw the line?"

So what do I think I've learned, and we've learned? I think I've learned that preferences were left out of all this, that it was treated like an engineering problem. Maybe preferences ought to be in engineering problems as well, but in any case, health care should reflect individual differences in preferences.

In that article, I emphasized that people have different preferences when it comes to health. Of course, everybody wants to be healthy. We all know that smoking is bad for your health. We all know that being overweight is bad for your health. We all know that exercise is good for your health. But many people enjoy smoking and they enjoy eating and they'd rather not exercise. And they're aware of the consequences. So it's a trade-off of preferences. And it may well be that when I go to see the doctor, in answering questions of how much I want to spend, I may differ from other people. Not because I have a higher income, but because I have different preferences about this. And somehow the system ought to reflect this. It ought to reflect the fact that individuals have different tastes for health versus other things and that a one-size-fits-all kind of health care is a mistake.

EXECUTIVE COMPENSATION

Region: You sit on the boards of three major corporations: AIG (American International Group), Eli Lilly and HCA (Hospital Corporation of America). I guess my segue here is health and insurance. What is your view of current debates over executive compensation and corporate governance? Are additional restrictions needed, or should there be some loosening of government oversight of corporations?

Feldstein: I think the recent SEC, NYSE and accounting rules have strengthened the role of the board in a good way. I think boards are working harder and treating themselves as more independent of management. I don't think there was a big problem before. To me, the most basic aspect of corporate governance is whether the outside directors meet alone and do that on a regular basis. Not much happens during those meetings, but you have the meeting so you have a chance to say, without management present, "Well, how do you think management is doing? And what message should we give management that would make things better?"

A board doesn't run a company, but it can provide useful feedback to management. That aspect of board independence I think is a useful thing. And when management is failing, a good independent board will force a change of management or even sell the company.

Region: Does recent controversy about stock options reflect a problem with the economics of executive compensation and incentive structures?

Feldstein: I guess in the boards that I've served on I've never felt that excess compensation was a serious problem. There was a recent NBER paper explaining why there's been this n-fold increase in executive compensation for the top companies. Well, lo and behold, the top companies are n-fold larger than they were 20 years ago, and so if you think about the compensation in proportion to the size of the business, in part because of mergers, in part because of just growth and the big ones growing more than others, that gives you an explanation of what's going on.

SUPPLY-SIDE ECONOMICS

Region: In the mid-1980s, you pointed out that "supply-side economics" was really just a return to basic ideas about creating capacity and removing government impediments. But as used in current parlance, the term seems to have a lot to do with the elasticity of taxable income. What's your rough estimate of that elasticity, and what does that imply about current tax policy?

Feldstein: Let me back up first to the '80s and then talk about taxes. I wrote a piece back then called "The Retreat from Keynesian Economics." In it I said that the broad outlines of the Keynesian economics that had come to dominate policy were an attitude that output depended on demand, that high savings were a bad thing, that a big government was necessary for stabilization, for maintaining aggregate demand, and could do more than that, could manage the economy in all kinds of ways.

By the time I wrote that piece [Summer 1981], there was beginning to be a retreat from all of those ideas. We came to understand that what really drove output in the long run, even in the medium run, was not demand, but was capacity and that a large part of that was saving and investment. So contrary to this earlier view, we were going back to earlier ideas.

At that time, people—particularly in the press—were looking for the new vision. Who was going to be the new Keynes? Where was this New Great Idea going to come from? And I said, "No new Keynes; no New Great Idea. Keynesianism was a passing intellectual phase associated with the Depression. Let's go back to the basics that economists have believed in more or less since Adam Smith—with some modifications, certainly; we've learned some things along the line." That's what supply-side economics was about: It was about creating capacity.

Now much of my own work over the years has been about taxes and about the response of households and businesses to taxes in various ways. And in particular if you look at the household response to marginal tax rates, the typical professional economist's view and also that of most tax policy officials is that people don't seem to respond very much. If you look at the relationship between labor force participation and tax rates, or working hours and tax rates, there's not much there. There is for married women, who have more discretion, but for single women, or men between 25 and 60, there's virtually no response of labor force participation.

I've argued that that's really looking in the wrong place. The measure of labor supply that matters is not just hours. The relevant labor supply includes human capital formation, choice of occupation, willingness to take risk, entrepreneurship and so on. All of these affect income and tax revenue.

What's more, taxes cause a further distortion that causes a "deadweight loss," that is, an economic inefficiency. Taxes change the way people choose to be compensated. I get compensated in fringe benefits rather than taxable cash because I have the choice between 65 cents of spendable cash or a dollar of fringe benefits. That choice of fringe benefits that are worth less than a dollar for every dollar that they cost to produce implies economic waste. It shows up as lower taxable income. A reduction in taxable income, whether it occurs because I work less or because I take my compensation in this other form, creates the same kind of inefficiency.

Economic analysis shows that if you want a single measure of the inefficiencies created by the tax on labor income, you can just look at taxable labor income. You don't have to distinguish whether a higher tax rate reduces taxable income because I work fewer hours or I bring less human capital to the table or I get compensated in the form of fringe benefits and nice working conditions.

Therefore, we should look at the data on how taxable income relates to marginal tax rates. I looked at the experience before and after the 1986 tax cut, because that was a very big, bold one. The Treasury provided data that allowed one to track individual taxpayers over time. So you could look at an individual a few years before the 1986 Tax Reform Act and at that same individual a few years later. And that comparison suggested quite a large response: Taxable income responded with an elasticity of about 1, meaning that a 10 percent increase in the after-tax share that an individual got to keep, say, going from 60 percent to 66 percent, would increase their taxable income by 10 percent. So those are big numbers.

Think about an across-the-board tax cut. Let's say you cut all tax rates by 10 percent, so that the 25 percent rate goes to 22 1/2 percent, 15 percent rate goes to 13 1/2 percent, and so on. That raises taxable incomes. The revenue cost of that tax cut is only about two-thirds of the so-called static result that you'd get if you didn't take behavior into account. So both in terms of thinking about the economic efficiency, which is very hard to explain to the lay public—I've been bending my sword trying—and also in terms of tax revenue, these are very large effects.

Of course, cutting the 15 percent rate to 13 1/2 percent has a much smaller proportional effect on the net-of-tax share than doing it at higher rates. So if this were not a change across the board but a change in the top rate of the sort that we had in '86, that would be an even bigger behavioral impact.

TAX REFORM PANEL

Region: Last year the president created a tax reform advisory panel that held public meetings around the country, with economists and others testifying before it. In November the panel came out with a report and a series of recommendations. As someone who has studied tax reform for years, how do you view the panel's recommendations, and why do you think they've gained so little traction?

Feldstein: Their proposals were pretty sensible. Of course, they didn't have a single set of recommendations. They emphasized ways of reducing the taxes on income on savings, which I think is a good thing to do. We didn't talk about that, but that's a very significant thing. They did not opt for any sort of radical flat tax reforms, and I think that's probably the right thing also. Flat taxes are a wonderful dream, but not a practical policy.

So why did it not have more traction? I don't know the answer to that question. To say that the White House is concerned with a wide variety of other things would be an understatement. Why didn't the Treasury push it more independently? I suppose that wasn't their job. Their job was to receive it and pass it on to the White House, and the White House chose for a variety of reasons not to do more.

THE NBER

Region: You've been the director of the NBER for 28 years, nearly one-third of its history. What have been the Bureau's most significant accomplishments in that period? And what is your vision for the NBER over the next quarter century?

Feldstein: A major accomplishment has been to encourage empirical research. Empirical research is a risky strategy. A researcher starting to do an empirical project has to know the data and the institutions. There's a good chance he gets it wrong. So after he's done all this work and he presents it, people will say, "Well, you didn't know about such and such." Or, "those data aren't really measuring what you thought they were measuring." So economists were reluctant to [engage in] serious empirical research.

Economics is different from some of the natural science departments or medical schools where there may be a dozen people who work on very similar things. In an economics department, it's unusual to have more than one or two labor economists or public finance economists. And even then, they probably do slightly different things. So there's nobody to talk to in your own department about the details of empirical research. In contrast, of course, you can talk about economic theory and get useful feedback from colleagues.

I'm not belittling that, but I'm saying that the profession had moved very far in that direction and was doing much less empirically because of this natural tendency for people to do things that were safe and reliable.

In addition, the National Bureau brings people together. Of course, it's normally smaller groups than this Summer Institute. The NBER program meetings bring together researchers from a dozen or more universities who are expert in the area you're working on. A researcher can present his research and get feedback. You get to know people better in this subdiscipline. You can e-mail them and say, "I'm working on this, and do you think these data are the best?"

So bringing people together and encouraging empirical research, I think, have been two of the really important things we have done. Also, as an organization, we launch projects on important economic issues like exchange rate adjustment or monetary policy in an open economy. So the Bureau directs attention on what we think are important problems. We did a lot of work on debt crises, for instance, and we were able to get people who would say, "Well, I'm really a trade specialist, I don't know very much about that, but if there's really a good group of people working on it and you'd like me to participate, yes, I'll participate." So I think we're able to focus resources.

Again, that's something you can't do in a single department. If I as a professor at Harvard—forgetting the Bureau—were to call up economists in six other departments and say, "How would you like to work on my study?" They would say, "I don't understand; I'm a professor at Yale (or Princeton or Chicago)." It would be like the Harvard football team calling some guy from Yale and saying, "How would you like to play for us?" Nobody thinks that about the Bureau. It's perfectly natural to participate in an NBER study. It's a neutral ground on which people can come together and do research.

So I think we're able to add value in terms of bringing resources to bear on important questions. We did a lot on Social Security, for example. We continue to do various things in the tax area. We have a Washington conference called "Tax Policy and the Economy." We do not advocate policies, so it's quite different from all of the other [think tanks]. We work on policy-related issues, and we can say, "If you do this policy, that's likely to be the outcome," but we stop short of saying, "Therefore, you ought to adopt a particular policy." Individually, I and other researchers can advocate policies, but as an organization we don't. But we can bring attention to it, and we keep bringing attention to issues like Social Security and health care costs and tax policy.

Region: And your vision of the Bureau 25 years from now?

Feldstein: Somebody else will be running it, so I'm not sure how it will develop. The economic problems will change. The technology may change. We may do things like more video conferencing or offering opportunities for small groups to get together.

Still, there is something about holding [in-person] meetings that brings people from all over the world here to Cambridge. And it's because of the coffee breaks and the lunches and everything else as well as just hearing the papers. It's not just to know what's in the paper or even to hear the discussion. It's to talk to others about your own work and what they're doing and get a sense of where the profession is moving.

Paul Samuelson was here today, which is really quite unusual, and I was sitting next to him in the meeting room next door. After everybody had gone around the room and introduced themselves, he said, "This must be what Copenhagen was like in the 1920s when Niels Bohr was bringing physicists together from around the world." I really thought that was a nice sentiment.

So I don't know where we will go and how it will change. You would not have predicted that it would look like this 25 years ago. In part it's the technology. I mean, the Internet is an amazing tool for us in terms of shared data and collaborative work, and distribution of the work. We had 21Ú2 million downloads of NBER working papers last year.

Region: It's a little daunting how frequently those long lists of new papers come out. But I have to guess that leadership, as much as technology, will guide the Bureau's future.

Feldstein: Well, yes. But you know we're a very decentralized organization, so there are these separate programs and program directors. So the meeting next door on monetary economics was run by David and Christine Romer from [the University of California at] Berkeley. The meeting in [this room] was the international finance and macro program, which has had a series of program directors over the years, and they've all done a very good job.

And people have been prepared to do it. The prestige and the chance to influence their part of the profession is what I think draws them to do it. I mean, I can call somebody up and say, "I think it would be good to do a project on X. Would you be willing to do it? We'll take care of all the logistics. So you can concentrate on the intellectual part of the task, picking authors and so on." I enjoy that aspect of it very much because I work with the project directors on both what questions are going to be asked and who the researchers are. There are now so many NBER project meetings that I can't go to them all. But I go to a lot.

Region: Thank you very much.