Ten Tentative Conclusions from the Past Three Years

1. It is very difficult to predict an economic crisis. The most sophisticated statistical work has success rates of about 25-30 percent in predicting crises. But the fundamental problem with predicting crises is you have to worry about both mistakes: predicting crises that might not otherwise happen but for the prediction itself and not predicting crises that do happen. We often do well in saying there are elements of vulnerability, but the timing and likelihood of crisis are different matters. It is not always impossible to predict crises—in the Thai case it was clear that it was going to happen; similarly, the probabilities in the Russian case were very high.

2. It is hard to get a country to take action even when you do see a crisis coming. Crises occur in large part because the government hasn’t done a variety of things it should have done earlier. They are probably just as well informed as you about the underlying situation, but possibly more optimistic, possibly less willing to look the facts in the eye, possibly unable to do anything about it. There is a crisis today in Ecuador which we have seen coming for at least a year. The President had no support in the Congress; he simply couldn’t do anything. There was nothing that anybody on the outside could do to prevent that crisis.

The issues of whether the IMF should go public in these cases are very difficult. We have tended to speak out a bit more loudly lately, but we have always to be cautious, and to remember that forecasts are inexact and that our own forecasts can affect the outcome.

3. It is hard to have a program implemented when the government you are dealing with is divided and weak, and, accordingly, it is difficult to decide whether to help. You cannot know how fully the program will be carried out. The consequences of not helping are very serious, but the consequences of helping, providing aid and having the program fail, are also very serious. It is not an easy choice. In some cases I believe it could be very useful to have more bilateral assistance from those who have political reasons for wanting to help, rather than rely entirely on the international institutions.

At the same time, we all need to be a little modest. I reflect on the fact that the richest country in the world, the only superpower, has taken a very long time to get its poverty problems under control. It has a lot more money than the IMF and the World Bank combined, even relative to the scale of the economies with which we deal. It is reasonable to ask, “What is the rate of return that should be expected on dollars provided by the IMF to a country?” It could be presumably a little bit above the rate of return in the New York stock market. Well, even taking the last few years of, say 20 percent real stock market gains doesn’t change the universe because the amounts of money we are moving are not very large relative to the economies to which we lend.

4. I am struck by how little understanding there is of the IMF’s system of internal governance and accountability to our member governments. When I read the newspapers I get the impression Michel Camdessus makes the decisions about whom to help and how much money to give, and goes off and gets it all done. In fact, there is nothing that the management of the IMF does that does not have the support of its executive board, which sits in permanent session. The 24 Board members represent 182 countries; the 8 largest members have one-country constituencies; other members are grouped into self-chosen multi-country constituencies. The U.S. controls 17.5 percent of the votes. Important decisions typically require an 85 percent majority. Very little happens without the support of the G-7. We—the management of the IMF—hardly ever lose votes in the board of the IMF for a very simple reason: We don’t take things to the board until we have an understanding with our shareholders that they will support what is being proposed. The staff has a lot of expertise and therefore provides a lot of independent input, but, in the end, we are accountable to our member governments—and that accountability is exercised every single day.

5. Keynesianism is alive. In the mid 80s there was growing support for the view that fiscal contraction was expansionary. This conclusion was based on the fact that in Ireland and Denmark budget deficit cutting had very positive impacts on economic growth; and somehow this conclusion was generalized, the argument being that by tightening the budget you increase confidence in the economy and then there is more investment, etc. Well, we have seen in Asia that Keynesianism is alive in two regards. First, the budget cuts that took place in Thailand in the IMF program in July 1997, when we thought Asia was booming, were probably too large. Subsequently, all programs in Asia have switched to fiscal expansion on a very large scale. Second, the liquidity trap, which was thought of as a Keynesian curiosum, is indeed happening in Japan where interest rates are zero. The authorities have to figure out how to act in this environment and the questions are all Keynesian questions now.

6. The strategy of going fast on bank restructuring and corporate debt restructuring is much better than regulatory forbearance. I don’t doubt that within a couple of years the banking systems and the corporate debt structures in Korea, Thailand, and Indonesia will be in far better shape than they would have been if these countries had waited to restructure. Indeed, they probabaly wouldn’t have done these things had they not been forced to undertake them in the heat of the crisis. At least in those cases, without trying to propose a general rule, I believe that rapid action on these profound structural problems is the right way to go and that waiting until the macrosituation is right is the wrong way to go.

7. Exchange rate systems. The major crises we have had in the last couple of years have all involved a fixed or a crawling peg exchange rate. Although Thailand, Indonesia, and Korea were not really fixed, they were assumed to be fixed. Brazil and Russia were formerly crawling pegs. From these experiences, the conclusion has been drawn by many that pegged exchange rates cannot be maintained and that we need to move to almost universal floating exchange rates. There is no question that pegged exchange rate systems are crisis prone and that crisis is more likely to be avoided if the rate is free to float. Quite a few countries would undoubtedly have had a crisis if they had had a pegged rate system—in particular, South Africa, Turkey, and Mexico. But let me register three reservations:

• First, if a country, for historical reasons, has a pegged rate that it is able and willing to defend and has enshrined, for instance, in the law, in a way that makes it even more credible, as have Argentina and Hong Kong, that may well be a good way of maintaining a peg. These are the two extreme versions of exchange rate systems: either float or peg them very, very hard.
• Second, I am not sure that this particular period of enthusiasm for pure unvarnished floating rate systems is going to last. Views in the economics profession fluctuate. This is one of those problems that doesn’t have a good solution. And we have seen, for example, in the case of the two largest economies, the United States and Japan, the exchange rate move from 80 yen to the dollar in 1995 to 147 yen to the dollar two years later. It’s not clear that small developing countries can put up with movements like that. I doubt we have the final answer. Whether we go to a system with broad bands, which seemed to work in Europe, or something else, I think we will probably move on from the pure floating regime, but not right now.
• Third, I do believe that if the Euro succeeds—and there is no reason to think it won’t succeed—we will see fewer and fewer national currencies. The Argentine dollarization move and the quiet Mexican dollarization campaign now under way are likely to make progress in the years ahead. One’s views on exchange rate systems are affected a little by whether you’re a young optimist or a more mature skeptic. When you’re young and optimistic you’re very impressed by all the good things that could happen if you’re free to let your exchange rate move. But, when you’ve watched what countries do with their floating exchange rates over long periods, you become much more concerned about the damage they can do by having their own exchange rate to manipulate.

8. Globalization is here to stay. The most extraordinary result of this crisis is that after all the noise, after all the statements about how the universe could not put up with this, how countries could not withstand this type of system, how we would have something else, that essentially only one country has tried to move away from integration into the global system—namely, Malaysia—and now even it is coming back from that position and seeking to reintegrate.

That is certainly not to say that the current system cannot be improved. What happened to several emerging market countries in the last few years is extremely undesirable, and we must be able to do better. But even in these circumstances, countries have simply not withdrawn from the international system in the way that the pessimists predicted. There have been some changes in national policies with regard to capital flows, and there should be such changes. There are more countries now that do put some controls on international capital flows. Only a week ago, Thailand, which has remained open, imposed some controls on what could be done with offshore contracts. But those changes do not represent a major withdrawal. Rather they’re something which will make the system work better.

The fundamental fact is that globalization will strengthen and will deepen. No country in Latin America and only one country in Asia has sought to withdraw from it. It is an extraordinary outcome of the debate which most people thought would go the other way.

9. Of all the difficult issues in dealing with the international financial architecture, the need to involve the private sector in the solution of international crises is probably the most difficult. We are struggling with this now in the IMF in a variety of programs. We are currently engaged with the Brazilian government in an effort to get the international banks to commit to keeping their credit and trade and interbank credit lines at the levels they were at at the end of February. The effort is going well, but it is a very delicate exercise. I was told by a banker the other day as the discussion heated up, “We are not willing to let the IMF make our portfolio decisions.” That’s an appropriate view. The banker’s obligation is to maximize shareholder returns, not to satisfy the international institutions, and that is a balance that we have to solve by getting them to see that it is in their individual interest to stay in if the others stay in.

10. The IMF is here to stay. When I was in the World Bank a decade ago, I thought it would not be a good idea to merge the Fund and the Bank, for two fundamental reasons. One is that these agencies are immensely powerful relative to the countries they deal with, and putting them together would make them even more powerful. We and the Bank exert control on each other by occasionally disagreeing and that’s a control mechanism that we’d lose if we unified. And the second reason, and I am far more aware of this now that I am in the management of the IMF, is that it is very hard to control an agency of 2,700 people, which is the size of our institution. I am not sure it is possible to control an agency of 10,000 people, which is Jim Wolfensohn’s agency. However, I am quite sure that 12,700 would be beyond the control of mere mortals, even with a permanent board in residence representing 182 countries riding herd on us. So merger is not a good idea and we shouldn’t go that way. What we should do is define our separate responsibilities, minimize the overlap, and cooperate when we overlap. We are doing that pretty well, as we go our mutually supportive ways. So, the IMF is here to stay. And we will make it work better.