In an Uncertain World

by Robert Rubin

  Did the response make the crisis worse? Many people who shared the view that First World investors bore a significant share of the responsibility for what had happened in Asia were antagonistic toward the IMF for forcing “austerity” programs on developing countries. I think this criticism misses the point. The IMF—and we at Treasury in working with them—sometimes did make mistakes in the technical design and implementation of its programs. In the face of rapidly changing and in many ways unprecedented circumstances, it would be amazing if we hadn’t. But by and large, the IMF got it right. What triggers an economic crisis is a loss of confidence. This results from underlying macroeconomic and structural problems as well as an often sudden reevaluation by domestic and foreign investors of the attractiveness of a country as a place to put savings. The cause of the hardship in crisis countries was not the IMF-backed programs but the crisis itself—the fact that capital was fleeing.

  A common complaint was that the IMF’s programs forced countries to tighten their fiscal and monetary policies, which sent their economies into a tailspin, when they should in fact have followed a Keynesian policy and tried to stimulate growth. A developed country might have sufficient credibility with credit markets to take stimulative measures to fight off recession. Unfortunately, a developing country gripped by financial panic and fleeing capital does not have the luxury of borrowing more or of cutting interest rates without worsening the panic.

  In general, a restrained fiscal regime is required because investors dumping a country’s bonds are likely to be further spooked by additional debt. The IMF believed that fiscal policy should be tighter to make room for more exports, without inflation, and to counteract the huge costs that would be incurred in cleaning up banking failures. But fiscal adjustment can be overdone, and in a subsequent reexamination the IMF has concluded that, in the Asian crisis, the initial recommended fiscal or budgetary policies were often more stringent than necessary. Officials at the Fund, and many other people, underestimated how much the crisis itself would cause these economies to contract. However, the Fund moved promptly to correct the excessive requirements and evidence is that this did not contribute significantly to the hardship. The economic impact of the fiscal tightening was dwarfed by the effect of capital outflows.

  As to high interest rates, I don’t think there was any ambiguity, although others disagree. Savers and investors would regain confidence and be willing to leave their money only in a country that would provide a reward for the risks involved and that was committed to fighting inflation. In the short term, higher interest rates did hurt indebted companies and banks. But they were essential for restoring financial stability, which in turn was required for recovery. Moreover, rates could be—and were—quickly brought down once confidence was reestablished. On the other hand, failing to tighten monetary policy during a crisis risked provoking more of a run. Deeper currency declines could, in turn, lead—in a worsening spiral—to higher inflation, more capital flight, and still further exchange rate decline. Either solution—higher interest rates or a weaker currency—can put a strain on indebted companies and banks. But with higher interest rates, the strain would be temporary if a government succeeded in reestablishing confidence. Treasury kept track throughout the crisis of how interest rates compared with pre-crisis levels. Surprisingly soon, the governments in South Korea and Thailand, which addressed their problems convincingly, were able to bring rates back down, even to below pre-crisis levels, and begin the process of recovery.

  Critics who argued that “structural” reforms were intrusive and unwarranted said that passing financial market fashions were being treated as imperatives. But attending to the question of what factors were affecting confidence was not just a matter of following the whims of fickle investors. Markets may worry too much, but their worries generally tend to focus on serious issues. It was always difficult to determine which measures were truly needed to restore confidence and put the economy on a stronger path. If we sometimes erred on the side of excessive pressure for reform, that may have been wiser than erring on the other side and being ineffective. As Mexican President Ernesto Zedillo said, “Markets overshoot, so sometimes policies have to overshoot as well.”

  Finally, some people who argued against the IMF programs believed that a market panic fed on itself and the problem in many cases was primarily a lack of liquidity. I felt then—and feel even more firmly today—that this was an oversimplification. Experience showed that market confidence, and capital flows from foreign and domestic investors, tended to turn around only when governments implemented reform. Without that, continued and deepening outward flows could swamp whatever money the IMF and others might provide. On the other hand, the combination of official backing and strong reform embedded in the IMF-led approach was able to stem the crisis and eventually spur recovery.

  Did the IMF-led “rescues” make future crises more likely? In Mexico, and again in the Asia-crisis countries, the IMF lent far greater amounts than was its usual practice. In a number of cases, the United States, other governments, and the World Bank lent additional funds on top of that. In capital account crises, where market psychology was often of critical importance, the firepower had to be sufficient to restore confidence. To some of our critics, these large loans only encouraged private creditors and investors to take more risks in emerging-market economies. While I had a lot of sympathy for moral-hazard concerns, I thought they were overplayed. I had never heard anyone say that they had been more inclined to invest in emerging-market economies because of the Mexican support program. Moreover, the surge of money into these markets was matched by a surge of money into a broad range of higher-yielding, riskier assets. In addition, those who argued for smaller IMF packages and greater “private-sector burden sharing”—which I agreed with to the extent feasible—were ignoring the enormous practical difficulties in managing a debt restructuring and the costs of allowing crises to widen.

  Have we fixed the system for the future? At the same time as the IMF, we at Treasury, the Fed, and other governments were dealing with the crisis on a day-to-day basis, we became very focused on a longer-run effort to improve the workings of the global economy—what came to be called the “international financial architecture.” We had begun the reform effort after the Mexico crisis at the 1995 G-7 summit in Halifax, Nova Scotia. But now this took on a new sense of urgency, and in 1998 it developed into a broad effort involving many countries, both industrialized countries and emerging markets. My view of what had caused the Asian crisis carried implications for both response and prevention. If borrowers and lenders are both at fault, why should the borrowing countries, and their much poorer people, bear the entire cost of repairing the damage? My view was that the burden should be shared to the greatest extent possible. This is an issue of fairness but, even leaving that aside, goes to the essential logic of capitalism. To allocate resources efficiently, a free-market system must allow people who take risks and lose to suffer accordingly, so that future decisions will have the best likelihood of being made with due concern for risk. Unfortunately, this theoretical framework did not provide an effective answer to many of the problems we faced in the muddled reality of the actual crisis.

  Too often, well-meaning but simple and sweeping proposals for dramatic change in the system weren’t congruent with the complex realities we faced. They often ignored market behavior and unintended consequences and tended to focus on only one aspect of crisis. Looked at most broadly, the system needed change that would work practically to strengthen emerging-market economies and make them less vulnerable, to promote better risk management among banks and financial institutions, whose poor judgment had first masked and then exaggerated the underlying weaknesses in many countries, and to improve the international community response to financial crises.

  The first key to prevention is to reduce countries’ vulnerability to crisis. There is no simple way to promote this, although a number of ideas were put forward. Some focused on tying future IMF and
World Bank money to the adoption of good policies in advance. Others looked at new systems that would send an official signal about countries’ policies—for example, some kind of red flag that the IMF should wave when it was concerned about a country. Neither type of scheme was practical. In a time of crisis, the IMF and its shareholder governments will always need flexibility, if only because of the risk of contagion. I thought that an IMF warning system simply wasn’t sensible. Predicting what will happen to a country’s financial circumstances is highly uncertain at best, and IMF warnings could precipitate crises that might otherwise not occur.

  One decision we did make was to focus attention on the dangers of a fixed-exchange-rate system. Attempts to hold a currency fixed when economic fundamentals were not in line had led to trouble time and again. These situations were inherently unstable and provided a kind of one-way bet for market speculators, as a central bank would first defend a particular rate but then had to let the currency move sharply when its foreign exchange reserves were exhausted. Exchange rates that are allowed to fluctuate with market movements are less subject to such bets. More important, they allow countries’ currencies and economies to adjust to outside shocks more gradually. In this case, we did call for the Fund to adopt a policy that prohibited lending to countries that maintain fixed exchange rates except under very unusual circumstances. Though that policy was not officially adopted at the IMF, the practical effect of the crisis, and of the general change in thinking that was taking place, was to eliminate almost all fixed-exchange-rate systems. By now, there are only a few economies left, such as those of China and Hong Kong, with the reserves and the policy determination to hold the pegs. And in the case of China, at least, this peg has become quite controversial, though for trade reasons, not financial stability reasons. My own view is that in time China will move to a more flexible exchange rate regime in its own interest. China has a large trade imbalance with the United States, but that’s because so much of non-Japan Asia’s exports to the United States pass through China. China’s global trade surplus is relatively small and has been declining.

  Another idea for improving country policies was to develop a catalogue of best practices in such areas as debt management, bankruptcy, public statistics including disclosure of levels of international reserves, deposit insurance, and bank supervision. U.K. chancellor of the exchequer Gordon Brown pressed particularly hard for codes and standards in these areas and some progress has been made. The question going forward is how to encourage countries to adopt these policies, with due allowance for differences in cultural, economic, and historical circumstances.

  The crisis also showed how much countries need adequate social safety nets, such as unemployment insurance and programs for the poor, to mitigate the hardship of financial turmoil and to protect the most vulnerable from economic downturns. The World Bank has greatly stepped up its work in this area, both to develop understanding of the policies that are needed and to provide financing for countries to implement better social programs.

  Since poor lending and investment decisions helped cause the crisis, it is important to search for ways to improve risk assessment and management by industrial-country creditors and lenders. Again, this is very hard to effect with official policy. But greater transparency—involving both information on the IMF’s views and analysis of a country and data about that country’s economic situation—could be helpful, and progress has been made. For example, today any country that hopes to attract foreign investment feels obligated to disclose data about its currency reserves and forward obligations. It would be close to impossible for a country to do what Thailand or South Korea or, earlier, Mexico did and hide its true reserve position. The IMF also now publishes information that countries provide about their adherence to the various codes and standards of best practice that have been developed, reinforcing the pressure to put such standards into place. Greater disclosure should reduce capital flows into riskier situations. This in turn should encourage countries to reform before a crisis occurs. At the same time, transparency on its own will not prevent Asian-type crises. The United States is generally regarded as having the most advanced transparency regime in the world, and it still experiences speculative excesses.

  Perhaps the greatest challenge in improving financial architecture is finding more effective ways to induce investors and creditors to exercise discipline and focus appropriately on risk during good times. One key here is for creditors and investors to bear the consequences of their decisions to the greatest extent practical. Some pointed to the big losses that many lenders and investors had suffered on their emerging-market positions in the Asian crisis and argued that the lesson about risk had been learned. But I felt strongly—based on all of my own market experience—that the moral-hazard argument had not been adequately addressed, and I fully supported private-sector burden sharing conceptually. But it was no easy matter to allocate more of the burden to private-sector foreign lenders and less to the poor of the affected countries, without making the situation worse for the poor by scaring creditors off from the country in crisis or from other emerging-market countries. Also, there were often legal problems in doing this.

  In South Korea, where most of the foreign debt was held by banks, private-sector involvement was feasible because banks could be involved in an organized fashion and came to see that their interest was to restructure the debt. But in Mexico, and later Russia and Brazil, where foreign debt was in the form of securities widely dispersed among individual investors and institutions, there was no practical mechanism to create a “bail-in.” And the risk of provoking a wider pullback was much greater. This leads to another important rule of mine that Tim and his colleagues neglected to put on their list: number 12: Reality is always more complex than concepts and models.

  In domestic markets, bankruptcy procedures serve the purpose of sorting out the allocation of losses in an orderly fashion and giving those who fail financially a chance to start over. The Federal Reserve and Treasury had explored this idea after the Mexico crisis in concert with the G-10 group of industrialized countries but had eventually concluded that it wasn’t practical for a host of reasons, and that idea seems to have fallen by the wayside at least for now.

  We also looked at the possibility of a “collective action” clause in bonds that would authorize some party to negotiate for bondholders in the event of default. This kind of binding arbitration would promote more cohesive action among creditors and could provide for private-sector burden sharing when there are thousands of bondholders. I always believed such clauses would be introduced when there was sufficient market pressure for them, and a number of emerging-market countries have issued new bonds with these clauses. Most important to me was that IMF programs should have the maximum practicable focus on private-sector burden sharing, and we pursued this where feasible, as in South Korea. Finally, I supported a new framework for IMF decisions that stressed burden sharing and laid out considerations to guide the Fund on a case-by-case basis, given the practical difficulties that made a rigid mandate infeasible.

  The greater focus on private-sector burden sharing is just one of the changes that has taken place since the Asia crisis. In addition, as we had hoped, most countries have shifted to floating exchange rates and have become much more transparent, as have the IMF and the World Bank. The enormous leverage in the system that contributed to the Asia crisis spreading so rapidly has also been reduced, at least for now. This does not, however, warrant any degree of complacency. Financial crises have continued to rock emerging markets and are likely to remain a factor in the decades ahead. The international community must continue to try to improve its means of prevention and response.

  FROM EARLY 1998, we faced the intensification of a long-simmering issue in Congress over support for the IMF. Even as the criticism of IMF actions in Asia increased, we had to go to Congress to ask for more funding for it. I had signed on to a routine replenishment of the IMF’s resources in September 1997, at the annual
meeting in Hong Kong, but fulfilling that agreement required congressional action. And now this money was urgently needed, in case still more countries fell victim to the spreading financial market contagion and had to borrow from an IMF that was beginning to face the possibility of running short of funds. The President’s January budget asked Congress for an $18 billion contribution, our share of a $90 billion increase in the IMF’s resources. Although the amount seemed large, it did not come at the expense of other programs. Our contribution to the Fund is more like a deposit at a credit union. We get an asset in exchange—a claim against the IMF—and we have never suffered a loss. But niceties of this kind were hard to explain in the heated atmosphere of the budget debate that year.

  The debate over IMF funding was overwhelmed by another story that broke that January, the one concerning Monica Lewinsky. The Lewinsky revelations became the subject of a media firestorm that lasted through 1998 and resulted in far less attention being paid to events overseas. This disproportionate focus on the Lewinsky matter didn’t make any difference to what we did in responding to the crisis. In another sense, though, it did real damage by crowding out media coverage of what was happening in Asia. As a consequence, the American people didn’t learn what they might have learned about the critically important issues raised by the crisis. The distraction kept the public from recognizing the threat from crisis abroad to the stability of the U.S. economy. Thus, when we most needed support from a well-informed public, that public was nowhere to be found.