Unraveling the Washington Consensus

An Interview with Joseph Stiglitz

Joseph Stiglitz served as chief economist and senior vice president for development economics at the World Bank until the end of 1999. At the World Bank, he gained notoriety by criticizing the development economics enshrined in "the Washington Consensus" and by leveling harsh attacks on the International Monetary Fund (IMF). Prior to serving at the World Bank, Stiglitz was a member and then chair of the U.S. Council of Economic advisers. He is on leave from Stanford University, where he is a professor of economics.

Multinational Monitor: What was the "Washington Consensus?"

Joseph Stiglitz: It was a consensus formulated between 15th Street and 19th Street in Washington among members of the International Monetary Fund (IMF), the U.S. Treasury Department, and the World Bank. It argued that the keys to success in developing countries were three things: macro-stability, liberalization (lowering tariff barriers and market deregulation) and privatization. It was largely formulated out of experience with Latin America.

MM: Does it still exist as a consensus?

Stiglitz: No. There is a consensus that those precepts, while important, are neither necessary nor sufficient for successful development.

If you look around the world, the most successful economic performer in recent decades has been China. Roughly 75 percent of the increase in aggregate GDP among all the low-income countries has occurred within China in the last 20 years. China followed some elements of the Washington Consensus -- such as macro-stability -- but clearly did not follow other elements, such as privatization and liberalization.

It did liberalize in certain dimensions. For instance, it did encourage competition among the township and village enterprises (TVEs). It provided a fertile ground for a new entrepreneurship at the village level. But it didn't follow the standard precepts.

Other countries in the most successful region in the world -- East Asia -- by and large also did not follow all the doctrines.

At the same time, there were several countries that followed most of the precepts almost perfectly and have not grown.

There is also an emerging consensus that the Washington Consensus was not only faulty in its narrow economic strategies, but also excessively narrow in its objectives. It focused mainly on increasing GDP, not on broader concepts of increasing living standards or democratic, equitable, sustainable development.

MM: Among whom is the new consensus emerging? Does it extend to the IMF and the Treasury Department?

Stiglitz: Nowadays there is at least lip-service being paid to issues such as democratic and sustainable development. Even the IMF managing director says that they are going to be concerned with poverty and the poverty impacts of their programs.

It is also the case that even at the IMF there is a recognition that capital market liberalization can expose countries to enormous amounts of volatility risk, and that countries therefore ought to proceed down the path of capital market liberalization with far more caution than the IMF would have previously suggested.

MM: Should developing countries aspire to increasing levels of capital market liberalization?

Stiglitz: As the economies of the more advanced nations become more open, it's very hard to control capital flows. So it isn't a question of aspiring, but rather feasibility. As countries become more integrated into the global economy, it becomes more difficult to intervene in some of these capital flows.

However, the underlying facts are that, for developing countries, capital market liberalization is not associated with faster economic growth and is associated with greater volatility. So the lesson is that one has to proceed very cautiously down that path.

In many countries, the form that interventions in capital markets takes can give rise to distortions and to corruption. So you're playing a very fine game where you're trying to maintain the stability of the economy but you're also trying to avoid the adverse effects that some of the regulatory regimes have had in the past.

MM: Is there an overall approach or specific mechanisms that you think are generally advisable?

Stiglitz: The Chilean regime is based on a philosophy that I think is quite sound. It recognizes that capital inflows can have adverse affects on innocent bystanders -- people besides those engaging in the process, and that it is worthwhile as a result to try to stabilize those flows.

It's like a dam. In the absence of the dams you can have a flood of water from the top of the mountain down to the sea, which in the process can cause a lot of death and destruction. You put a dam in, the water still goes from the top of the mountain to the sea, but meanwhile you are able to channel that water and make it productive.

The kinds of interventions that Chile is trying to do are intended to stabilize capital flows. That is to say, to reduce the inflows in peak periods of excess supply, but not to do that in other periods.

There are several countries that are trying to use the banking system as a vehicle through which to stabilize capital flows. That's a good vehicle, because the banking system is often the first victim of excessive capital flow and volatility. If you have a banking system that's designed to limit those flows, you are protecting the banking system itself at the same time.

MM: For what purposes should developing countries borrow foreign money?

Stiglitz: The basic principle in general is that you should borrow for productive purposes, and the returns on those investments should be sufficiently high that they more than pay the interest that's charged. Another consideration relevant in some public borrowing cases is that a government should be able to appropriate through its tax power enough income to pay back those loans.

The problem in many cases in the past has been that loans have been made to governments where governments were corrupt and funds were taken out of the country or were used in a way that did not increase the productivity of the economy, and so the economy did not generate the income that would allow the country to pay back the loan.

MM: Is there any distinction between borrowing domestically and internationally? For example, does it make sense to borrow foreign currency for an educational system that's not going to directly generate foreign currency to pay the loan back?

Stiglitz: As countries go to more open, integrated economies, the basic function of borrowing is to give you more resources. Domestic borrowing by the government doesn't change the amount of resources the country has, it just changes who has those resources to spend.

So if the government borrows from the private sector, it may generate more savings, allowing the country to invest more. Or the government investment in education, for instance, may crowd out private investment in, for instance, factories. In which case you get more human capital but less physical capital. By borrowing from abroad, you are able to have both more human capital and more physical capital.

MM: Do you see reasons not to immediately cancel the debts of the poorest countries?

Stiglitz: The fundamental argument in favor of doing that is that these countries are very poor and that without some debt forgiveness it's very hard for them to get on with their development program. All their export revenues are being used to pay off the money they have borrowed.

The two arguments against immediate and blanket debt forgiveness is that it costs resources to forgive debt. Take, for instance, money that the countries owe the World Bank through International Development Association (IDA) loans. That money, when it gets repaid, gets recirculated and then lent out to other poor countries. If that money doesn't get repaid, the World Bank has less money to lend to other countries. So debt forgiveness is basically a redistribution from some poor countries to other poor countries.

The question then becomes which countries are more able or deserving to use the funds. A lot of people say it is unfair that those countries that struggled hard to repay their loans will now get less money, and the countries that were profligate and didn't repay their loans are going to get off scot free.

The second argument is the incentive issue. If people thought they could always get loans written off, they would have no incentive to repay loans.

There is a legitimate question that one wants to ask: Should the developed countries be lending money to the developing countries or should they just be giving them grants?

My own view is that there is an argument for making loans, because you want it to cost the country something to undertake a particular project. It is a way of getting engagement and commitment -- they have to make choices.

MM: Do you think the IMF's Enhanced Structural Adjustment Facility (ESAF) process Ð now known as the Poverty Reduction and Growth Facility Ð is proceeding in the right way, or should it be revised?

Stiglitz: There are several separate and related issues here. One is if it's a good thing to focus more on the impact on the poor of the programs that are being put forward. I think the answer is clearly yes. The real test will not be the rhetoric, but the reality.

There was an ESAF review committee put together about a year and a half ago, which looked at a whole host of issues, of which the lack of focus on poverty was only one problem. They also identified the excessive focus on stabilization relative to growth, and a lot of technical mistakes with respect to sequencing, such as financial liberalization before the country had the regulatory structure to make sure the financial markets worked. Many of those issues have not yet been fully resolved. Let me give you one issue in particular that is of enormous importance and has not yet been fully resolved.

On the budgetary front, the IMF has always held that good policy requires that you have a reasonably balanced budget during normal times. The question is: Do you include foreign aid as part of revenue?

There is a tendency on the part of the IMF to only include tax revenues. Why? The best argument they've made so far is that you can't count on foreign aid. It's here today, gone tomorrow. We did a statistical analysis where we compared the volatility of foreign aid to tax revenue. Tax revenue was more volatile, so based on that argument you should not include tax revenue as revenue either.

The right argument is that if you have volatile foreign aid, you need to have a flexible expenditure program. If you receive money for building schools, when that money stops you stop building schools. That's not that difficult and most countries do that automatically. So that whole argument is basically fallacious.

For a country like Ethiopia, which has been using foreign aid to build schools, it looks in IMF terms like their budget is out of balance. If tax revenue plus foreign aid equals expenditure, then they have a balance. If you can't include a large fraction of that revenue as revenue, it looks like expenditures exceed revenue. Then they're told they're out of balance and are told to cut back on their construction of schools. That leads to poverty or inhibits their ability to address poverty. It could be health clinics or any development program.

To those of us who are committed to reducing poverty and enhancing growth, this is an extremely pernicious policy. Yet it is still part of the IMF's basic economic dogma. It makes absolutely no economic sense.

There is a more fundamental issue that the Meltzer Commission [the U.S. Congressionally appointed commission to review international financial institutions] raised and that Treasury Secretary Summers also raised in his talk: Should the IMF focus on crisis situations? I think there is a very persuasive case that can be made for that.

MM: Would you endorse that recommendation?

Stiglitz: I think there's an increasing consensus behind that, as Secretary Summers himself seemed to indicate. It enhances focus and it means that you have an institution that has a clearly defined objective and you can judge its performance based on that objective rather than a multiplicity of objectives. It's also based on the fact that they've made some very striking mistakes in terms of development policies, as the ESAF review pointed out.

MM: If the IMF were limited to addressing crisis situations, would a penalty rate for IMF loans be a sound policy?

Stiglitz: The Meltzer Commission was greatly influenced by the observation that the bailout packages that were put forward in East Asia, Russia and Brazil didn't work. In the case of East Asia, they were intended to stop the decline of the exchange rate -- but if you look at the data, you can't see any perceptible effect. In the case of Russia, we know what happened: the money went out of the country to Switzerland and Cyprus. A few weeks later the currency fell. In the case of Brazil, it may have held off the decline of the exchange rate, but it only did that for a few months. You have to ask if that was a good investment. Are the people in Brazil better off after $50 billion was invested? I have not been able to find an economist who thinks that that investment paid off from Brazil's point of view.

So you approach the subject by saying, first, these bailouts don't seem to work very well. Secondly, the bailouts may in fact be part of the cause of the problem. The bailouts lead to a moral hazard problem -- a lack of due diligence in lending. There's a further argument that the bailouts may lead to speculative attacks. Much of the bailout money is a transfer from the governments to speculators. So you are feeding the sharks. The more you feed them, the more there are. The more there are, the more unstable the world is.

The reason why a lot of people find the recommendations troubling is that the IMF was originally established for the purpose of liquidity lending, to help when the markets have irrationally lost faith in a country.

Take a country, for example, which is an exporter of oil. Oil prices are temporarily depressed for a year, well below what anybody expects to be the case in the long run. Futures markets are telling you that the prices are going to go up, but capital markets don't fully integrate that information, so a country which is quite wealthy because it owns oil right now is in very bad shape because the price of oil is low and its revenues have gone down.

To many people, it seems cruel to take a country that's in that kind of situation and say that you'll give them money but you're going to charge very high interest rates.

MM: Where do you come down?

Stiglitz: I come down wishing that it were possible to differentiate between the situation that I just described and one where you would issue a clear warning to governments to think twice before borrowing. The track record, the inability to differentiate those cases, makes me lean more towards the Meltzer commission perspective in favor of penalty rates. Also bear in mind that when they say penalty rates, the rates are typically not that high -- they're well below commercial rates.

MM: How would you deal with the problem of allocating responsibility and costs between the lenders and borrowers in moments of crisis?

Stiglitz: In the case of private sector flows which dominated the flows in East Asia -- where it was private sector lending to private sector -- we have within our countries a well-defined way that you do it: bankruptcy. You have a bankruptcy law that provides the backdrop for negotiations that go on between the creditor and the lender, and they renegotiate. If they can't renegotiate, it goes to a court for determination.

I believe that there should be a similar framework internationally with the same basic kind of framework that we have in the United States.

There ought to be what I call a Super Chapter 11. Under Chapter 11, the U.S. bankruptcy code, when a firm goes bankrupt there is a reorganization. The creditors take more equity in the firm. Management proposes a restructuring but the firm continues to operate. It isn't stripped apart or dissolved. A Super Chapter 11 would, like Chapter 11, be designed to keep the firm in business. But it would be designed for these macro-economic shocks where even good, well-managed firms would have a hard time surviving. In these cases, the presumption that you want to keep these firms operating is very strong.

Basically at that point you would make the creditors take a haircut.

MM: When the World Bank does structural adjustment in the context of heavily-indebted poor countries, how do you differentiate it's performance from the IMF?

Stiglitz: There is an allocation of responsibility in which the IMF is primarily responsible for the macroeconomic assessments whereas the Bank is responsible for the structural and poverty assessments.

If, for instance, you were to look at a country like Ethiopia where inflation has been effectively zero and where the government, inclusive of foreign aid, has been able to make it on a balanced budget, most outsiders would say that that country has good macroeconomic policies. The IMF has at times said it does not have good macroeconomic policies because it has this perverted view of how you count revenues.

So when it's said that a heavily indebted poor country (HIPC) fails because the IMF says it doesn't meet its macroeconomic policies, then the appropriate place to focus your attention is the IMF and their macroeconomic assessment process.

MM: Do you think Bank has been doing its poverty-alleviation job well?

Stiglitz: The problems that we so often face are two-fold. One, we don't have the requisite data to measure what we would like to measure. Secondly, we also don't have the requisite theories.

We would like to be able to make some clear links between policies and the impacts on poverty. It many cases that turns out to be very difficult. Quite often, from an analytic point of view, you undertake a policy but you don't know what would have happened had you not taken undertaken that policy.

There are other issues that have come up in places like East Asia that have not come up very often before. For example, we know a lot about how to handle individual bankruptcies. But in Indonesia, somewhere between 65 and 70 percent of the firms are bankrupt. We don't have any good experience in trying to get an economy out of a situation where more than half of the firms are bankrupt. So it is not surprising that different people would have different views on the appropriate way to proceed. What was clear to me was that you needed to proceed differently from the way you would proceed if it were an isolated bankruptcy. One of my complaints about some of the strategies that have been pursued is that they did not sufficiently make that differentiation.

MM: What about the sectoral adjustment loans? One criticism would be that there has been too much marketizing of things like health care and water supply that should not be marketized.

Stiglitz: Those issues are also difficult and complex. There could have also been a better effort at nuancing the policy.

In the case of water, one of the problems in the past is that public water systems often only supply the wealthier districts and the poor have been left without water. And the government water agency says it doesn't have any resources to extend the water supply. The poor are left to buy water privately. So the current situation is one where the poor are under a private system and the better-off are under a public system.

If it's done right, privatization often has the advantage of providing more revenue that enables the extension of services into poor areas.

Another example is the electrical sector. One of the problems of public electricity companies in many developing countries is that they can't collect revenues. Because they can't collect revenues, supply is very bad. It's a vicious circle. Because supply is very bad, there are inadequate hook-ups. People then steal the electricity because that's the only way they can get it. Of course that creates a culture in which people say, "Why should I pay for electricity, when I can steal it?" The private guys have a strong incentive to make sure people pay for the electricity that they get. It then gives them revenue to extend the services.

Private telecommunications -- cell phones -- have similarly extended telecommunications to areas that the public sector never went. It would have been years, maybe decades, before they would have had access. That's a big impact.

In each of these cases you might have said that there's an alternative strategy -- reforming the public sector in ways that actually get the service down to the poor people more effectively than the private sector can do. But it seems very difficult in many of these countries to get them to undertake these reforms. They have not occurred.

There are some other cases that give you a different perspective. Many people used to say that charging small fees for primary education would generate revenues and not discourage education very much.

A lot of that thinking has been really challenged by Museveni and Uganda's initiative to make universally free public education. Uganda increased enrollment enormously by doing two things: making it completely free, and making a big push for primary education. I think if there had been charges it would not have been at all successful.

So I think charges can have adverse effects under certain circumstances.

MM: With regards to labor-related issues and the Washington Consensus view, what are some of the things that have been done wrong in the past and what should be done in the future?

Stiglitz: There are many dimensions to this issue. A lot of theory treated labor like any other commodity without recognizing some of the critical ways in which labor is different.

As part of the doctrine of liberalization, the Washington Consensus said, "make labor markets more flexible." That greater flexibility was supposed to lead to lower unemployment. A side effect that people didn't want to talk about was that it would lead to lower wages. But the lower wages would generate more investment, more demand for labor. So there would be two beneficial effects: the unemployment rate would go down and job creation would go up because wages were lower.

The evidence in Latin America is not supportive of those conclusions. Wage flexibility has not been associated with lower unemployment. Nor has there been more job creation in general.

The mind-set of many people within the Washington Consensus has been very much influenced by a kind of flexible, full-employment dynamic in an advanced country like the United States. In the United States, almost anyone who wants a job can get a job. We have a good capital market. If you have a new idea, it's remarkable how easy it is to find finance. New enterprises are being created all the time. That's not the case elsewhere. Capital markets in many countries don't function. There's very little enterprise creation, very little job creation. When you lose a job in one part of the economy it isn't the case that you will automatically create a job elsewhere.

Labor market flexibility was designed to move people from low productivity jobs to high productivity jobs. But too often it moved people from low productivity jobs to unemployment, which is even lower productivity.

In a world in which there are large mobility costs, the issue of workers' rights becomes very important. Those issues -- workers' rights, safety legislation and so forth -- are appropriately within the purview of government. The right to collectively bargain and organize should be viewed as a basic right that is needed to redress the balance of power.