Skip to main content

Interview with Joseph E. Stiglitz

September 1, 1997


Arthur J. Rolnick Senior Vice President and Director of Research, 1985-2010
Interview with Joseph E. Stiglitz

Photo: Joseph Stiglitz When Joseph Stiglitz was named to the president's Council of Economic Advisers in 1993, one pundit named the group "the dream team of economics," and another writer later dubbed him the council's "idea man."

Wherever Stiglitz has gone he has made important contributions, beginning with his work in academia, which included professorships at Oxford, Princeton and Stanford. His efforts in economic research earned him the John Bates Clark Medal of the American Economic Association in 1979, an award given to the most distinguished economist under 40.

After serving the administration through 1996, Stiglitz joined the World Bank as its senior vice president and chief economist. At the World Bank, Stiglitz has focused his efforts on how countries can grow, especially those that have been mired in poverty or that are trying to establish market economies after years of state-run systems. "We can make a difference," Stiglitz told officials from around the world at a recent World Bank conference. "The challenge is to understand what are the appropriate policies and how to target our assistance to promote growth and poverty alleviation most effectively."

In the following interview with Arthur Rolnick, Minneapolis Fed senior vice president and director of research, Stiglitz discusses the question of appropriate policies for economic growth, as well as questions relating to trade, banking, inflation and other issues.

ROLNICK: In our 1996 annual report essay we suggested that one of the keys to understanding the differences in economic growth in developing countries is openness to new technology. We argue that countries that are more open to new ideas tend to develop faster. What advice is the World Bank giving to developing countries? What do you see as a government's role in promoting growth?

STIGLITZ: Well, I looked at it in two parts. One is the perspective that you raised, that one of the major sources of economic growth is the transfer of knowledge. We used to think of the gap between developed and less-developed countries to be an object gap—the lack of capital. And we now realize that is one of the factors, but there's another factor that's very, very important, which is the gap in knowledge. And closing that gap is one of the most important strategies for development. It is being reflected even in the way the World Bank conceives of itself. It talks about itself as being a knowledge bank, as one of its main roles in promoting development.

The next question is: How do you close that knowledge gap? What are the strategies that can facilitate the closing of the knowledge gap? One of the ingredients of that is human capital, something that the bank and the developing countries have been discussing for a long time. In order to close the knowledge gap you have to have the skills to absorb that knowledge. That actually requires a two-pronged approach. It's important to make sure that all the populations reach the level of primary education in order to achieve the basic level of literacy that is essential for a modern society. You also need to have technical skills. You have to have engineers. Korea is successful, in part, because they promote those skills, they have enough engineers to facilitate the transfer of technology. So clearly that's one aspect of the strategy—human capital and very broad, essential skills.

The second part of the approach is openness, being part of the world community. And again, I think one of the lessons is that if you look at the enormous success of East Asia, it has really given a lot of inspiration that says that development is possible. A country like Korea closing the gap in the space of four decades shows that with the right policies you can do it.

ROLNICK: How would you explain the success of the South Korean economy relative to India's? Was openness a key factor?

STIGLITZ: Historically, one of the interesting things is that 40 years ago per capita income in these two countries was comparable, and now the gap is enormous. Openness is one of the factors, yes. India has a well-educated population at the high end, so that ingredient they had in some sense. And, clearly, an essential aspect of Korea was not only openness but actual aggressive attempts to acquire knowledge. So it was more than a passive openness. It was a government saying that we need to do everything we can to transfer that technology. And so at various periods they made big efforts to make sure that a lot of their students went abroad. They had an active policy of encouraging exports. And the interesting thing about encouraging exports was it did two things. One, it ensured that their firms were competitive and met international standards. And it made them focus on competition, which I view as one of the real ingredients here. Even though they were large firms, they were entering into an environment where they were competing with other firms all over the world. So competition was one of the ingredients. And the second one is that by entering into the international market and meeting international standards they came into contact with knowledge that inevitably, given their abilities, facilitated importing knowledge. I view that as a key ingredient.

There is some research that suggests that entering into exports is more effective in transferring technology than openness to imports. But I think both of them are important. One of the reasons why openness on imports is important is that a lot of modern technology requires specialized intermediate goods. And if you're going to have a technology that's similar enough to other countries' technologies, you have to have the input. To give you an example, modern technology is based on computers, fairly sophisticated computers. A country that tried to keep out computers just couldn't play in the game of modern advanced technology.

ROLNICK: So while there may be different impacts between export policies and import policies, you're essentially arguing for open trade and it's not just in physical capital, it's in ideas as well?

STIGLITZ: That's right. But, it's a very different argument from the traditional arguments for trade. There is a little bit of evidence that the gains from trade do not come from inter-sectoral reallocation of resources—the traditional economist's view. The idea being that the gains from trade really come from these technology spillovers.

ROLNICK: This is the new literature that suggests there are gains to trade that we in the profession haven't quantified yet.

STIGLITZ: And this enters very much into thinking about regional trade arrangements, what the gains and the costs and benefits that go beyond the traditional debate of trade diversion vs. trade creation. They address issues like: What will be the impact of this particular regional trade agreement on technology transfer?

ROLNICK: What about NAFTA [North American Free Trade Agreement]? NAFTA is one of the more recent regional trade agreements. There are still many critics of NAFTA. Where do you stand?

STIGLITZ: My general stand on regional trade agreements is that they need to be viewed as a step toward a more open international set of arrangements. They are intermediate steps. And as intermediate steps, they need to be evaluated not where they are but where they lead. So if an intermediate step blocks off the link to the final step, then it's a step backwards. But if it's really intermediate . . .

ROLNICK: Some have argued NAFTA was just that, an intermediate step that was going to lead to a full-blown North American free trade agreement and then eventually open up to the rest of the world. Is that a fair view?

STIGLITZ: That, I think, is a fair interpretation of what those of us who are optimistic about NAFTA believe. There have been some setbacks over the last couple of years. A lot of us hoped for fast track authority in Congress that would have facilitated Chile being brought into the free trade area. And, in a way, I wish the debate were on the issue of whether the best way of reaching the free trade area, embracing all of the Americas, was through fast track accession or a big bang agreement that would embrace everybody at the same time. That kind of debate would have made clear what the goal was and the best way of obtaining it. With some of the protectionist sentiment in Congress, one worries about whether we've lost sight of what the goal is. And the fact that there's still much resistance to even a fast track for Chile suggests that the goal may not be in sight.

ROLNICK: One reason for this resistance is that when governments make new trade agreements there are going to be winners and losers. NAFTA attempted to take care of those that would lose. Has this attempt been successful?

STIGLITZ: Any change has winners and losers. New innovations have winners and losers. Our attitude toward opening up trade, which is like an innovation, should be like our attitude toward somebody discovering a super laser that makes the economy more productive.

ROLNICK: Well, that's easy for an economist to say, but U.S. politicians have to worry about firms that have to compete with Mexican firms.

STIGLITZ: I agree. However, once you begin by bringing the issue of lost opportunities in our society, you can phrase a question like: How are we going to react to the automobile that threw the blacksmith out of work? Are we going to say we're not going to allow the automobile? And from a historical perspective, do we think it was a mistake to allow the automobile? I think most of us probably think, aside from maybe being worried about pollution, that our society is much better off. Now the same argument goes for opening up trade. We're going to be better off. But, just like the blacksmiths were hurt, opening up trade leads to some people being hurt. It really is basically the same principle. Innovations that make our society more productive have mostly winners but some losers. And as a society it is incumbent upon us to ease the transition of those who are hurt. And that principle was encompassed in NAFTA, and it's one of the things that we need to continue to focus on.

ROLNICK: Some argue that we should only negotiate free trade agreements when we're at full employment. Do you agree?

STIGLITZ: No. My view is that good macro-economic policy overall is the most important part of a jobs program. You've got good macro-economic policy. You've got the unemployment rate under 5 percent without inflationary pressure. Whatever the cause of unemployment, whether it's new technology or new trade, people have to move from one job to another. And so, some of this issue I think is a red herring. I think we ought to have job transition programs, period. And the cost of the programs, if we have our economy running at full employment, is not going to be that great. People don't want to spend a lot of time in a training program when they could be out in the labor force. So, I'm not worried about the overall cost. When the economy has ups and downs, the cost will go up a little bit, but that's again not bad. People are training for where the economy goes.

ROLNICK: Let me switch to another controversial issue, excessive risk-taking by the banks and the role of government in protecting bank deposits and being a lender of last resort. Of course, the Federal Reserve was established to be the lender of last resort in the United States, to prevent systemic problems. One thing we have found in country after country, whether developed or underdeveloped, when the government starts to get involved in the banking sector it tends to overinsure and create moral hazard in the extreme. The S&L debate in this country is often cited as a classic example of moral hazard. Last year The Economist did an article listing five or six major countries that were having serious banking problems and suggested that over-insurance was part of the problem. My question here is: In advising governments, how do you suggest they balance safety for small depositors against the need for market discipline?

STIGLITZ: Let me challenge your underlying hypothesis. If you look around the world, there have been banking crises in countries in which there's no insurance.

ROLNICK: I'm not just talking about explicit insurance, I'm also talking about the implicit kind, where depositors assume the government's going to rescue a bank if it gets into trouble.

STIGLITZ: I remember conversations in one country, I won't mention the name, where they were very proud that they didn't have deposit insurance. And they said that's what provides the discipline. And I said, "Well, you know, it's always been a theory that depositors are good monitors of the behavior of the bank." In general, the evidence is that the individuals don't have the requisite information. Furthermore, the financial state of the bank is a public good, monitoring of that kind is a public good, and there is therefore a strong rationale for governments to have a role.

The general perspective that I would have is that the problem comes from a disjunction between providing insurance and providing the monitoring by the insurance companies. And the government in most of these cases is an implicit insurer, whether it provides a formal insurance or not, of the insured. Now, any company that provides fire insurance to a commercial building puts a stipulation that there are sprinklers and inspects to make sure the sprinklers are there. An essential part of all insurance is monitoring. Now, what has happened in many of these countries is a disjunction between the implicit or explicit insurance and the monitoring.

And my interpretation of the problems of the '80s in the United States was exactly that kind of disjunction. It was an ideologically based regulatory lapse, without an understanding of the fundamental incentive issues that underlie bank behavior. Of course, there's a political problem. But ex-post in most of these economies, there are relatively few banks and the principle of too-big-to-fail is an inherent part of the problem. So you're not going to walk away from the issue.

One of the things that we are getting to understand is how best to undertake regulation, that is, we have a better understanding of banks. Some of us understood this before, but I think it's getting out—what motivates banks and what kinds of information regulators can use to help them undertake more effective regulation. For instance, we know the incentives of a bank to undertake excessive risk depends, in part, on the net worth of the bank. And if the net worth is very low you get to these issues like gambling on resurrection. That was part of the problem with the S&Ls.

ROLNICK: The S&Ls with very low net worth are what some economists called "Zombie thrifts."

STIGLITZ: Zombie thrifts, exactly. And so you recognize that if banks are in that kind of financial position, their incentive—whether or not there is insurance—is to gamble.

ROLNICK: Suppose an insured financial institution has substantial net worth, but the owners are risk neutral or risk lovers. You still have a moral hazard problem.

STIGLITZ: That's right. So, going back to the theme I said before of the government being in a better position to monitor, it can monitor net worth—which is a different kind of monitoring than monitoring individual transactions—and as monitor of net worth, it can increase the likelihood that banks undertake reasonable action.

ROLNICK: Can't we rely on the large, sophisticated depositors to monitor banks' behavior toward risk?

STIGLITZ: One of the things that good banking regulation may do is to establish a layer of uninsured capital, which would presumably be provided by large depositors, which would, in my view, complement government monitoring. I don't view these as either/ors. I think good monitoring requires an array of instruments—large, private monitoring, government monitoring of net worth and some government monitoring of transactions brings this monitoring of self-dealing. A rule we've established, the intent of which is to reduce the likelihood of bad behavior, and in which we complement what you might call the ordinary civil incentive (the criminal incentive is when you violate this) is again a rule for government. Government has the role and tries to discourage fraud and makes fraud a criminal offense. And here is a similar set of actions that you don't want any bank to undertake because when they undertake it, it creates a public bad. People won't have confidence in the banking system.

And there are two things that we've learned from our general research. One is that good financial institutions promote economic growth, that financial mediation is a valuable asset and aspect of the organization of society. That is how economies grow. And the other is banking crises inhibit economic growth and have impacts that last long after the financial crisis.

ROLNICK: The Congress is now considering expanding banking powers and allowing nonbanks to own banks. In a world in which the government is going to be the chief monitor, some are worried that we have enough trouble in monitoring traditional banking activities. What's your position on how far Congress should go in deregulating banking?

STIGLITZ: I think that one should proceed very cautiously in this area. The argument for breaking down some of the restrictions I find fairly unpersuasive, if only in terms of the way economists put it. The argument for putting people together are synergy, economies of scope, I believe is the technical term. I'm not convinced that somebody who knows how to manage an automobile company well has learned something that makes him a better manager of a financial institution. And so I don't see a lot of economies of scope. I do see lots of potential for danger. That is to say, an owner of an automobile company might want to use access to the bank to get loans that it couldn't otherwise get, or to provide loans to customers it that may not otherwise be able to take. So, though I can see some dangers, I don't see a lot of advantages from it. Why should I, as a shareholder, buy a bank through buying shares in an automobile company that then buys a bank? If I want to own a bank I probably would do it directly. Why this indirect route, unless you can show some big managerial economies or you can take advantage of getting a better rate for the bank.

ROLNICK: Of course a shareholder of an automobile company would want its company to own a bank.

STIGLITZ: Those are the very reasons why, from a societal point of view, it's not efficiency enhancing. It's probably using your differential assets with a government guarantee as a source of problems.

ROLNICK: The nonbanks/financial institutions relationships are a problem in the United States. Is this a problem in the developing countries?

STIGLITZ: The issue of nonbanks/financial institutions relationships is an issue of debate in all countries. One of the driving forces in this debate, though, is that the traditional lines between industries are breaking down. And what is going on in our economy forces you to reexamine these issues constantly. Take General Electric, a major industrial company; G. E. Credit is one of the largest financial institution employers. There are lots of the traditional lines, even though we had a division between ownership of banks. In fact, the division between finance and industry has already broken down. I think you put your finger on the right issue, though. Banks are in a special position partly because of their systemic effect on the economy, and partly because of that systemic effect we have an explicit or implicit role of government insurance. And that raises certain sets of dangers. My own view is, therefore, let's keep that narrow and only expand, if at all, gradually and in response to necessary changes for efficiency.

ROLNICK: There is one other economic development issue that I would like you to discuss. We're seeing sub-governments (cities, counties and states) competing for private firms. The most visible competitive battles have come over professional sports teams, but the competition extends to virtually every type of business. Is this the appropriate role of government?

STIGLITZ: That's a really interesting question. There is a well-developed theory in economics about how competition among communities can lead to an efficient delivery of public services. That theory, when carried to its extreme, said competition should be very fierce and when it's very fierce it turns out that effectively you can't tax capital at all. Because what happens is if you tax capital that's mobile, the community next to you says, "Come to me. I will tax you a little bit less." And then the first community says, "I'll tax you a little bit less." And you wind up with no tax on mobile capital. In the extreme then, that says the only taxes that can be imposed by local communities are on immobile factors, which are typically land and labor that is grossly immobile, which tends in many cases to be the least skilled labor because skilled labor is a national market. What I think worries some of us is that the prediction of that kind of intense competition among communities is being realized. Communities are discovering the principles of competition without thinking through where that's going to lead. It is in the interest of each of them to do it.

ROLNICK: It's tough to unilaterally withdraw.

STIGLITZ: Yes, it is tough to unilaterally withdraw. But when they all do it, the consequence will be that the benefit doesn't go to one city or the other, the equilibrium will be over taxes on businesses and on, in general, low taxes in the pure theory. Now there are two ways of responding to this. One of them is an agreement at a national level that we will not allow this kind of competition. As a practical matter, that's very difficult because even if you restrict competition in actual tax rebates, there is a whole set of amenities that communities can provide that basically reduce costs. And some of them a city ought to provide. There's drawing a line between what the community ought to provide and what is a teaser or gift. It's a very fine line.

I'm going to give you an example. You want good transportation facilities. But where do you draw the line between transportation facilities that the town should provide and transportation facilities that the company should provide? How about making good school facilities that you then get to use for your training programs. Good school facilities, we all agree on that, but using them for training? These programs are pretty good things, you know, night school and so on. So, the lines are going to be very difficult to draw and coming up with a meaningful compact would be difficult. But maybe that's one direction we should go. The other one—and these are not exclusive strategies—is to recognize that in fact both the communities that have relatively little scope for taxing capital and moving at the local level more to broad-based use of income taxes, or broad sales taxes, and recognizing that in the modern life we can't use some of the kinds of taxes that we have. Globalization, and the world has its counterpart in mobility within the country, and some of the tax structures that we have used in the past are really not going to be viable.

ROLNICK: How is this problem being addressed internationally? Are there parts of trade agreements that prohibit subsidies or other financial inducements to lure companies away from other countries?

STIGLITZ: There are—it's part of the World Trade Organization. In the Uruguay Round there was a whole set of provisions dealing with this issue, where if one country subsidizes an industry you're allowed to offset that subsidy. So that is already part of the trade structure. There is a related issue, however, that I am concerned about. It is a danger in the way the United States is proposing implementing the subsidy policy that goes too far and will impede the transition of some of the nonmarket economies, like the former Soviet Union, into market economies. The subsidy policies can be used as protectionist measures. A key issue on this is in a socialist system you couldn't tell what a subsidy was because they didn't have prices on everything. Now you have in these economies nationalized enterprises, socialized enterprises, and you privatize them by auctioning them off. When you auction them off, do previous subsidies get carried forward or do they get wiped out at the private auction? Almost all economists would say that if you have a competitive auction they get wiped out. There are some groups in the US government that are trying to argue that this is not the case. And if they prevail in this debate, we will have established a real barrier to auctioning off these enterprises and to these companies entering into market economies. So, implementing this in a fair and effective way is going to be very difficult.

ROLNICK: Another current trade-related issue is the proposed European monetary union. European countries are talking about a single currency. It looks like they're going to get there, despite all the economic and political hurdles in their way. Do you think a single currency is a good idea for the European communities, or should they stick with floating rates and multiple currencies?

STIGLITZ: There's a large literature that talks about the issue of what is the appropriate basis of a common currency. The United States can be viewed as an area in which we use a common currency. We've had it for so long we take it for granted. But clearly, each of the 12 districts could have its own currency.

ROLNICK: We do. We have our numbers on them, but we have a fixed exchange rate.

STIGLITZ: So, it's a common currency. And the question is: What are the concerns? The real issue is that you have regional shocks that leave one part of the country in a recession. And how do you equilibrate? The problem is that right now we can't use monetary policy. Say Texas has a problem. We can't use monetary policy very effectively to help Texas recover from a downturn because it doesn't have a separate interest rate. It's linked to the national interest rate, which is linked to concerns about national levels of inflation, not the Texas problem. Texas also may have a problem because it may have a constitutional amendment—many states have a constitutional provision that says they can't have deficit spending. So you can't use fiscal policies to stimulate Texas. But in the United States there are two other forces that help Texas recover. One of them is we have a very nationally integrated product market. Texas can export to the rest of the country very easily. And so the rest of the economy is doing well, and that's the reason why the Fed has a lower interest rate. Markets in the rest of the economy are strong, so Texas starts doing exports. This is one thing in their favor that helps their recovery.

Secondly, there's migration. People start leaving and entering, and they all speak the same language. So we have a variety of ways that we adjust in response to a shock in one part of the economy even though we don't have monetary policy and we have what I call restrictive fiscal policy. Now, if you look at Europe, they're giving up monetary policy in the same way. Fiscal policy is restrained, particularly with conversion criteria. However, they don't have the constitutional provisions that we have and they don't have an integrated product market yet.

ROLNICK: It's tough to unilaterally withdraw.

ROLNICK: But both could develop.

STIGLITZ: They could develop. But it will take a long period of time; there's been a lot of mobility but it's not the magnitude that we have between Texas and California or the rest of the country. All of that says that the challenges in integrating are substantial. Then the question is: What are the benefits that accrue? Interestingly, some of the innovations that have occurred in modern capital markets have reduced the magnitude of the advantages of a common currency. Because one of the disadvantages is the risk of fluctuations in value; one of the disadvantages of having different exchange rates is that you don't know the value. That makes trade difficult, right? And because we have improvements in foreign exchange markets, futures markets, that risk can be spread out more widely.

ROLNICK: But the risk is still there.

STIGLITZ: Yes, but it gets spread among a large number of people and so the real question is: If it's spread among enough people, then it doesn't act as a very effective impediment to trade. What you're doing is choosing one form of risk for another. In one case you have the shock to a particular country or region and you allow for prices to adjust when you have flexible exchange rates. The other one is if prices are fixed you're more likely to get quantity adjustments. Price fluctuations aren't real consequences, aren't real risks. And so you're not getting rid of risk. It's really how risk gets spread through the economy. Modern financial measures have allowed financial markets to do a better job of distributing the risk.

ROLNICK: I agree that we've become more efficient in dealing with that risk. The literature, though, suggests that these exchange rate fluctuations are deadweight loss. The empirical work shows that short-run exchange rate fluctuations look like a random walk; they do not reflect changes in economic fundamentals. Going to one currency gets rid of that deadweight loss. That's a benefit.

STIGLITZ: That's right. And it's really a question of the role of exchange rates. I view it as an open question, and an interesting experiment. If you look at the data and how difficult it was even for the United States to make some adjustments, how long it took California to move from 10 percent unemployment down to 7 percent unemployment with a major role adjustment from '92 to '96, a major role adjustment being migration patterns, it makes you worry.

ROLNICK: The transition may be a problem. But, if I think of doing business in Europe, say I have a major company and I am supplying my product to 10 different countries. From that point of view, even if the markets are more efficient in handling this risk, there's still a big cost to me in dealing with the exchange rate risk. I have to pay for these hedges.

STIGLITZ: And this goes back to what I said earlier—optimum currency here—20 may be too much for Europe. We have, by the way, fiscal methods within the United States that we use. The federal government is sensitive to unemployment and tries to adjust. If Europe starts thinking more about these adjustment problems, then they have some fundamental reasons why they're at a disadvantage to the United States.

ROLNICK: A question about research. We have limited resources for research, whether it's the World Bank or the Federal Reserve. There's been a debate as to how to allocate that research recently. Some argue that we should continue trying to put our resources into better understanding the business cycle. Others say that the return on making any improvement in the business cycle is relatively low, that the real benefits are going to come from studying why some economies grow faster than others. That is, research on economic growth is where we should be putting our resources. Your thoughts?

STIGLITZ: I think that business cycles are still very important. In fact, there's actually some relationship between cycles and growth. We've done work that shows that when the economy goes into a downturn, one of the first things that gets cut out is research; when a firm experiences a shock in its cash flow, it cuts back research. This is consistent with certain models. The consequence is that recessions have an adverse effect on long-run growth. So I actually see the study of both business cycles and growth as necessary and complementary.
Now, I think it would be shortsighted for us to say that we've slain the dragon of the business cycle. Europe still has unemployment over 10 percent. We have a better functioning economy—and I could tell you all the wonderful things the administration has done over the last four years to facilitate that—but I want to know the whole set of things so that when the next administration comes in, I can tell them what they ought to do so they don't kill this wonderful economy that we've helped create.

ROLNICK: One of the things the Fed argues that helped to create this robust economy is low inflation. Has the Fed been following the right policy?

STIGLITZ: I actually think that the success of our economy, that low inflation, is a symptom of some more fundamental successes and not something that I attribute really to Fed policy, as much as you would like to take credit for it. I think a more competitive economy is likely to be more disciplined in prices. We've got a more competitive economy through a number of features. I think the complete openness of the economy, which is partly a change in mentality—we began talking exports, we passed NAFTA, the Uruguay Round, all the things we've talked about—some of it has actually changed trade barriers; some of it has just changed the way we look at the world, the way we enter the world and engage the world. And classification, in that sense, creates a more competitive environment in which firms realize that this process of globalization means that you don't have three automobile companies, you have a world of automobile companies and that means you can't raise your prices. So, it's something that hopefully we did, but it's something really that is the evolution of the economy here. Classification has provided the tool that makes competition more effective. It's not only that we have made classification more effective, but classification means that I can find out what the prices are all over the world.

The work that I did earlier shows the relationship between information and competition and shows that imperfections in information lead to imperfections in competition; whereas, more perfect information leads to more perfect competition. That has been what has led to a big fall in the NAIRU [nonaccelerating inflation rate of unemployment]. There's a lot of evidence of this shift in NAIRU, very little evidence that there is a relationship through the rate of growth of the economy and the levels of inflation so long as inflation remains low. So that's why . . .

ROLNICK: So you're not concerned whether we're aiming at 4 percent or 1 percent inflation?

STIGLITZ: For inflation in that range, the evidence shows very little effect that is discernible to even a highly trained eye. Some people argue that if you let it get too low there may be some adverse effect. I'll leave that as an open question. But within the range we're talking about, inflation is not a major source of concern. A lot of the evidence that we produced at the Council suggests that if you happen to miss the inflation target you were aiming at, in a low-inflation environment, the cost of undoing that is very, very low; the cost of undoing it is less than the benefit you got when you did it. In other words, there's a steep slope of the Phillip's curve such that wringing out inflation is relatively low cost. As long as you do these things moderately. So I've been advocating what I call a policy of cautious activism.

ROLNICK: Related to the pursuit of good monetary policy is the need for an independent central bank. How important is this independence?

STIGLITZ: I think the econometric evidence on that is somewhat ambiguous. There's not a lot of evidence that it improves the trade-offs in a significant way. I think there are some deep philosophical issues on whether, or the extent to which, macro-economic policies which are a fundamental responsibility of central banks should be delegated to nonelected officials. Now, there are arguments against putting macro-economic policy in the heat of the fray. And so the question is arriving at this balance between independence and accountability; it seems to be a question a democratic society has to address. People should be aware of the balance in principle. We should not, for instance, allow authority to be delegated to a group of people whose only focus are bankers. Bankers' concern is inflation. That may have a very disastrous consequence for working people. That, I think, would be a mistake.

ROLNICK: Actually, I think if you go back to the debate on the Federal Reserve Act, that's exactly what the issue was. That's why the district banks are required to have nonbanker directors.

STIGLITZ: We're within the ambit of what I would call a balance. But, if we started debating, I would argue for more accountability than we currently have.

ROLNICK: It's healthy to have the discussion about the importance of independence. The discussion itself promotes accountability.

STIGLITZ: Discussion is part of accountability.

ROLNICK: Thank you, Mr. Stiglitz.

More about Joseph E. Stiglitz

Currently senior vice president and chief economist at the World Bank, Washington, D.C.

Served on the president's Council of Economic Advisers, including stint as chairman, from 1993-1996.

After receiving his doctorate from the Massachusetts Institute of Technology in 1966, he began his teaching career there, and later served as a professor at Yale, Oxford, Stanford and Princeton.

He received the John Bates Clark Medal of the American Economic Association in 1979, and has also received international prizes in economics, including the Union des Assurances de Paris prize in 1989.

He has been an editor and associate editor of many professional journals, including the Journal of Public Economics, Review of Economic Studies and American Economic Review.

Stiglitz has written many journal articles, with his most path-breaking papers summed up in "Information and Competitive Price Systems," American Economic Review, May 1976.

His books include textbooks, such as Lectures in Public Economics (1980), and others, including The Theory of Commodity Price Stabilization (1981).

Nobel Prize Winner in Economics, 2001