Let me begin by providing a brief overview of the proceedings so far in order to identify some of the issues our panelists may want to address. The seminar to date has covered the main linkages between macroeconomic stabilization and reform and has included case studies from both developing countries and transition economies. Two well-known principles have been reaffirmed, namely, that macroeconomic stabilization is a necessary, but not a sufficient, condition for sustained growth; and that structural reforms are necessary in order for macroeconomic stabilization to be effective.

Patrick de Fontenay

Let me begin by providing a brief overview of the proceedings so far in order to identify some of the issues our panelists may want to address. The seminar to date has covered the main linkages between macroeconomic stabilization and reform and has included case studies from both developing countries and transition economies. Two well-known principles have been reaffirmed, namely, that macroeconomic stabilization is a necessary, but not a sufficient, condition for sustained growth; and that structural reforms are necessary in order for macroeconomic stabilization to be effective.

There are two ways of looking at this last principle. The first, which is quite simple and is often mentioned by the IMF Fiscal Affairs Department, holds that if a country is striving for macroeconomic stabilization and has a budget deficit, the government must put in place a tax administration and the necessary instruments for collecting revenue. The second is that reform allows the economy to stabilize at a higher level of output than it would otherwise, making the stabilization doubly effective and allowing a consensus to emerge around the stabilization effort.

Possibly because of the persuasiveness of some of the speakers, and particularly of Anne Krueger, there has been agreement among participants on certain issues covered during the seminar. One such issue is trade reform. I think Ms. Krueger managed to convince her audience that trade reform is a necessary ingredient of any reform process in a developing or transition economy. The second point of agreement is the link between monetary policy and financial reform. And the third, which was analyzed by Alan Tait, is the necessity to undertake fiscal reform as part of the macroeconomic stabilization package and the structural reform itself.

At the same time, there were issues on which views differed, and I propose that the panel discussion focus mainly on these topics. The most obvious and perhaps most contentious was the exchange rate, on which very different views (sometimes from within the same IMF department) were expressed. For example, Jacques Artus from the European Department appeared skeptical about using the exchange rate as an anchor for inflation, while his colleague Michael Deppler emphasized how effective Poland’s anchor has been. Manuel Guitián also emphasized the usefulness of the exchange rate as an anchor, focusing on the macroeconomic aspects of exchange rate policy rather than on microeconomic factors such as competitiveness.

The second issue on which participants could not reach consensus was the design and implementation of social safety nets. The third was the much-discussed subject of sequencing. Several important points were raised on this topic, including the question of whether enterprise reform should precede banking reform—that is, whether hard budget constraints must be introduced for enterprises before banking reform is undertaken. A similar question was brought up in terms of price liberalization. Is it meaningful to talk about price liberalization when a country has no market and no competition and when enterprises have not been restructured?

The last area I will mention is fiscal policy. The most important question here is how to determine the appropriate stance of fiscal policy, given that such policy may be affected by, for instance, banking reform, which often translates into a transfer of claims from the banks to the budget. Similarly, price liberalization may affect the budget significantly, because some price controls may be replaced by direct subsidies from the budget itself, as Marcelo Selowsky explained.

These areas should be sufficient for the panel to discuss. The panel comprises three people who play a major role in IMF economic assessment and policy: Michael Mussa, Economic Counselor and Director of the Research Department; Sterie Beza, Economic Counselor and Director of the Western Hemisphere Department; and John Odling-Smee, Director of the European II Department.

Michael Mussa

Let me mention two specific points that I regard as essential to understanding the process of successful economic reform. First, the experience around the world demonstrates quite dramatically that successful economic reform in terms of generating economic growth, new jobs, and rising living standards depends to a great extent on the development and expansion of new businesses in the private sector. When the economic environment encourages new businesses and does not burden new entrepreneurs unduly, then the process of economic growth tends to take hold.

In China, for example, the state farms continue to operate, but much more like private sector enterprises. And what really has accounted for China’s phenomenal growth in recent years under the economic reform program has been the development of new private businesses. Small private firms are also generating much of the growth in East European countries—for example, in the Czech Republic, Hungary, and Poland—as well as in many Asian economies.

A related point is that it is important to recognize the ways in which people finance their new businesses. The single most important source of financing is what I like to call the “mattress.” That is, many businesses are financed by the entrepreneurs themselves or by parents, brothers, sisters, and other family members who provide the initial seed money. If the business proves successful, the proceeds and profits are reinvested. The official financial sector—banks, securities markets, and so forth—is of course also important, and it becomes increasingly so as enterprises expand. But it is essential for policymakers to recognize the key role the informal financial sector plays in supporting new businesses.

Only one policy element is really necessary to mobilize that support. The government must ensure that people who invest their own and their families’ savings in new businesses will be able to enjoy a substantial share of the profits if the business succeeds. When the economic system does not provide this assurance, people look for other ways to safeguard their wealth—holding gold or foreign currencies, for example, or hiding it in some way so that it will not be taxed or stolen. If the economic and political systems do not provide entrepreneurs with the degree of security they need in order to invest freely, the process of economic growth will stagnate.

This principle is not new. Writing on the Peloponnesian Wars, Thucydides comments that there was no real economic progress in the Peloponnese until Sparta was able to establish a reasonable degree of military and political security. This essential element in the stimulation of private business activity should not be disregarded.

The second point I want to emphasize at the outset has to do with the existing enterprises, which I refer to as “dinosaurs.” These firms are a difficult problem in every economy. They developed under an outmoded economic system and were often heavily subsidized by the government, but they are no longer economically efficient. What can be done with them? Unfortunately, when they are big and politically important, they must be “fed” (at least for a while), because otherwise they tend to threaten the reform process. However, it is difficult to manage the “feeding” of these dinosaurs in a way that does not consume the entire reform process. I think these dinosaurs are probably the number one challenge for economic policy in reforming countries, because they put enormous pressure on the budget, on monetary policy, and on the exchange rate, among other things.

I would like to tell you that there is an easy solution to the dinosaur problem. There is not. Anyone who does not believe me need only look at the problems the United States is having with military base closures. The U.S. Congress and the administration face major difficulties in deciding which bases to close, so a special commission has been appointed to relieve the government officials of the political responsibility for making these unpopular decisions. In this case, the issue is reducing U.S. military spending as a share of GNP by about 2 percentage points in real terms over five years, and base closings are a relatively small fraction of this consolidation effort. It is not hard to imagine the difficulties of restructuring or closing public enterprises in the transition economies, especially when these enterprises have long been and sometimes remain a significant part of economic activity and employment. But however difficult the task, it is nonetheless absolutely essential to the economic process.

Let me turn briefly to the issues Patrick de Fontenay raised. First, which exchange rate and exchange rate policies are appropriate for reforming economies? As in most issues in economics, the answer is “it depends.” Some countries are able to maintain a fixed exchange rate that provides an important anchor for economic stability. However, some governments rely on the inflation tax to finance subsidies for large-scale state enterprises, as the economic and political systems are not yet able to enforce hard budget constraints. Thus, these governments resort to nominal credit creation, however destructive and self-defeating this approach may be. If the problems cannot be solved in the short term, then a fixed exchange rate is really not going to be viable, and trying to adopt an exchange rate policy that is not viable is almost always a mistake.

I should add that there may be circumstances in which adopting a fixed exchange rate does provide an essential element of monetary and fiscal discipline despite other economic problems. In Estonia, for example, the policy of adopting what is essentially a currency board system seems to be working. The Estonians have been able to refuse to grant subsidies, claiming that the government does not have the money. If a government can take this stance, then perhaps such a system can be made to work. However, it would probably be a mistake to say that this feat can be achieved in all countries.

Briefly, in terms of social safety nets, the key fiscal problem is that in an economic system where the budget is already enormously strained by other demands, finding the necessary financing is always difficult. Equally unfortunate is the fact that the social safety net in a number of the reforming countries operates not so much for people as for enterprises. A great deal of money and resources are being wasted maintaining enterprises that produce negative value added at world prices. However, as I said before, this problem is an extraordinarily difficult one for the political system to resolve.

Let me turn next to sequencing. A typical sequencing problem involves deciding whether to initiate enterprise reform first (and thus impose hard budget constraints early on) or to begin by reforming the banking system. This dilemma is actually easy to resolve, because most of the time there really is no choice. The two issues need to be approached simultaneously.

There has been a great deal of confusion about the issue of sequencing in general and, in particular, about the question of whether liberalizing prices all at once or taking a more gradual approach to economic stabilization and reform is the preferred approach. Some analysts have suggested that the former Soviet Union (FSU) would have been better off adopting a more gradual approach, as China appears to have done quite successfully. This issue may have been relevant in 1986 or even in 1988. But with the complete political collapse of the old regime in the FSU, the mechanism of centralized economic control simply evaporated, and there was no logical way that it could be gradually dismantled.

Similar observations can be made about many East European countries, which, like the FSU, experienced an enormous buildup of monetary and inflationary pressures just as the old economic system was collapsing. These countries could not have gradually decontrolled that system: either they had to liberalize prices more or less officially or their economic systems would disintegrate. Again, there was relatively little choice about sequencing in the process of price liberalization, although there have been some unfortunate, significant delays in freeing certain prices, particularly in the energy sector and for some public services, rents, and the like. In many important respects, the degree of relative price distortion in these economies has probably increased rather than decreased during the initial course of economic reform and liberalization.

Finally, what about fiscal policy? How is it linked to banking reform and enterprise debt, among other things? Fiscal policy is proving to be an extraordinarily difficult area to reform. But macroeconomic stabilization is not possible unless a government has some degree of reasonable control over both its fiscal position and the rate of credit creation by the central bank and the banking system. As Mr. de Fontenay said, this control is not the only thing required for successful economic reform, but it is an absolutely essential ingredient. Unfortunately, there is no simple answer to the question of how to impose fiscal discipline and hard budget constraints in economic systems that have not been accustomed to functioning with them.

Sterie Beza

When I consider the topic for this session, “the lessons from reform,” the first point that comes to mind is that one should be rather modest about lessons that one draws. Thus, I am not sure that all of the points I am going to make are lessons.

The revolution in reform that has taken place in Latin America and the Caribbean has been astonishing. Virtually every country has undertaken a substantial reform program and, although each program is unique, the trend toward structural adjustment is widespread. It is not easy to say what sparked these reforms. The World Bank, the Inter-American Development Bank, bilateral donors and creditors, and the IMF have all pushed for a variety of reforms over time, but what is happening now goes beyond these initial efforts. Much of the impetus has come from countries themselves as they examine and exchange views on each other’s experiences. The IMF area departments also do this type of “networking.” IMF staff try to apply what they have learned from the experience of one country to other countries, knowing that the authorities will be more readily convinced by working policies than by theoretical arguments. In Latin America, not only do policymakers study the experiences of each other’s countries, but also a number of the ministers of economics, finance, and planning know each other well and often exchange views on economic policies. This closeness is frequently very helpful.

What has made Latin America’s reform programs work? My first response would be that countries where reform has succeeded often have firm direction from the top. It helps to have technical people in high-level positions in the ministries, as Latin American governments often do. These people are much more than technicians, but they do have training in the areas they are managing. For instance, many of the ministers of finance in Latin America are economists familiar with the problems of reform. However, even skilled ministers ready to push for reform have little chance of succeeding unless the president is also very interested in reform.

My second point would be that divisions in the top economic team or the cabinet can have negative consequences. A unified leadership with a long-term commitment to reform is needed, because many reforms take considerable time. In the interim, the efficiency of production and the flow of foreign direct investment should not only indicate whether the reforms are having the desired effect but also serve as a signal if policies need adjusting. Success requires a persistent application of the policies necessary for reform. And finally, I would say that in most cases, structural adjustment in Latin America has been associated with the return of democratic governments.

Now let me turn to some other “lessons” from reform. Patrick de Fontenay noted that a supportive macroeconomic environment is necessary in order to carry out reforms. A high and volatile rate of inflation and a precarious balance of payments do not make for such an environment; a country with a high inflation rate looks as if it is not being governed, despite any efforts the authorities might be making. Furthermore, in such circumstances the general public also does not expect the reform effort to be sustained, so the situation is a poor one both for building domestic confidence and for attracting foreign capital.

These points are what I consider the most important “lessons” of the Latin American reforms. The list may not be substantial, but as I said earlier, one should be modest about drawing such conclusions. Latin America is still, after all, in the middle of the reform process. And the experience of reform varies widely from country to country, despite the similarities in what is being attempted.

Let me turn next to the questions Mr. de Fontenay raised, beginning with issues concerning the appropriate exchange rate policy. I think the main point Michael Mussa made is correct: the appropriate exchange rate for a given country will depend on economic circumstances. I would expand on that idea by adding that the country authorities must feel comfortable with the policy they adopt. Economists can give them advice and suggestions and point out obvious flaws, but in the final analysis, the authorities themselves must feel at home with their policy. Telling a country with a fixed exchange rate that the policy does not make sense in light of domestic economic circumstances—suggesting, for example, a float—is not enough, because the authorities may not be ready to handle such a change in their regime. Laying out all the alternatives may involve a lengthy process of dialogue and persuasion, but in the end the authorities need to feel they can manage the regime they have selected. If they do not, a dangerously unstable economic situation can develop.

My experience in the IMF has provided me with other lessons in the area of exchange rates. I worked on certain countries involved in the creation of the crawling peg exchange rate system, which has been used primarily to keep countries competitive in relatively high-inflation situations. Many years ago, several countries with high inflation (up to 50 percent annually) were in the habit of devaluing their currency only once a year. For six months after the devaluation, capital would pour in because investors thought the exchange rate could be sustained for some time. But if inflation stayed high, capital would flow out again in expectation of yet another devaluation. If the devaluation occurred in December, for instance, by July capital would begin leaving. By November, it would be pouring out, and the IMF staff would hold rather tense meetings with the authorities. If the country had an economic program with the IMF, the authorities would be urged to adjust the rate, even if they felt it was not time yet. Countries that found themselves in this position realized at about the same time that there were serious drawbacks to their approach, and they adopted crawling peg systems under which the exchange rate was adjusted more frequently.

Mr. de Fontenay also mentioned social safety nets. The problem I have in discussing social safety nets with countries needing them is that it is impossible to borrow money at market interest rates to pay for these nets. Perhaps a country can obtain grants for a while, but it will not have access to grants forever. Creating social safety nets requires raising taxes or otherwise finding the resources internally, and convincing the authorities to raise taxes is difficult, despite the high returns a reasonably well-functioning social safety net that minimizes political tensions can generate. Even when political tensions are not overwhelming, a government’s opponents can exploit economic difficulties in order to discredit a reform program. Thus, social safety nets have a dual effect that is quite important: first, they meet a genuine economic need; and second, they show that the authorities are conscious of the problems of certain segments of the society. This said, however, one still has to find the necessary resources, a task that is not easy.

On the issue of sequencing, Latin America does not have problems comparable to those in some of the transition economies, where a number of public enterprises are, as Mr. Mussa pointed out, producing negative value added. Sequencing in Latin America—and probably in other countries as well—has often been determined by what the authorities have been prepared to do. The authorities may have reforms under way in several areas and thus will be better prepared in some of these areas than in others. Effective reforms should not be held up because other areas also need attention or should be investigated more thoroughly.

John Odling-Smee

I would like to say a few things about the lessons from reform programs in the former Soviet Union (FSU), with special reference to the coordination of structural reform and macroeconomic stabilization. Many of my comments will have to do with the issue of sequencing. The first and rather obvious remark is that these reform programs are very recent, and thus many of the lessons they offer must remain tentative. However, that said, one can obviously come to some interim conclusions.

The FSU countries all faced the same initial situation, although now they are beginning to follow rather divergent paths. They all faced huge structural problems, but not just in the conventional economic sense. Not only were there problems with the structure of industry, which was too heavily oriented toward military and investment goods, but there were also geographical distortions. Under the old Soviet system, industries had been spread around the various regions of the FSU with little regard for the principle of comparative advantage.

Other structural features of the economy were also far removed from what is required in a market economy. In 14 out of the 15 FSU countries, there was very little in the way of a central government before the Soviet Union broke up, and an administrative structure with the appropriate skills and knowledge had to be created. Government policies were formulated in a way that encouraged government intervention in microeconomic decisions but did not explicitly create a macroeconomic framework to support private economic agents. The governments had to learn how to create and manage effective macroeconomic policy. The legal framework was inadequate, as were the accounting systems for government finances and banks. In addition, the attitudes of economic agents had been formed under a centrally planned system, and many of these agents had no experience with market economies, except perhaps in the Baltic states.

Quite apart from these huge structural problems, the macroeconomic disequilibria were quite serious by the time the Soviet Union broke up at the end of 1991. During the last few years of the Soviet Union, budget deficits rose as the government of Mikhail Gorbachev tried to placate various interest groups by spending more money. The growing strength of the republics allowed local authorities to obtain more and more central bank credit, undermining central budget discipline. As a result, there were severe shortages and suppressed inflation. On the external side, the Gorbachev regime reacted to the increasing political and economic difficulties by borrowing abroad on a large scale in the mid- and late 1980s. By the early 1990s, no more commercial credit was available, and in 1991 commercial banks started withdrawing credit quite rapidly. The external situation deteriorated, and a very serious balance of payments crisis occurred toward the end of 1991.

Thus, the 15 countries that emerged from the Soviet Union faced enormous structural and macroeconomic problems. Under the circumstances, all the necessary reforms had to be undertaken at once. Given the difficulties of the situation, policymakers developed a practical rule of thumb: take as much action as possible in the right direction in all areas at once. And by and large, that is what they did.

In the debate about sequencing, it is necessary to identify the speed with which the policies could be implemented. Among structural reforms, there were certain components that could be implemented quickly: for example, price and trade liberalization and the freeing of the exchange rate. Perhaps the most important thing in the FSU countries was simply that the governments needed to stop controlling everything. Such an approach has something to recommend it, and some countries moved in this direction early on, beginning with Russia and the Baltic states. Others then followed, in some cases because they had no choice once Russia had done it. The freeing of prices in Russia, for example, forced other countries to free prices as well.

Some structural reforms, however, take quite a long time. Privatization is not something that can be done quickly, although some countries have found ways of speeding up the process. Countries are now quite rightly moving away from West European-style privatization, which is a very slow and careful process involving stock markets. New legislation has had to be devised in all the countries that are struggling to introduce new democratic reforms, requiring extensive parliamentary debate and new accounting and legal systems, among other things. Creating new laws and administrative systems takes time. Thus, structural reforms can be divided into two categories: those that can be implemented quickly and those that cannot. On the macroeconomic side, stabilization policies fall into the category of reforms that can be implemented quickly.

Much of the debate in the FSU about sequencing has involved these issues and, especially, whether countries should go ahead with reforms that are easily implemented before undertaking long-term measures. Some have argued that it is necessary to wait until all enterprises are privatized and market-oriented legislation is in place before liberalizing prices. Others have argued against this approach. By and large, the argument that governments should move quickly on what they can—namely price liberalization and macroeconomic stabilization—has won the day in Armenia, the Baltic states, and the Kyrgyz Republic, and among the reformers in Russia. However, the argument that governments should move as slowly as possible until the long-term structural reforms are in place is still quite strongly held by policymakers in Azerbaijan, Ukraine, and some other FSU countries, and also by some groups in Russia.

The parliaments in all the FSU countries were elected under the Soviet system in 1990 and have tended to be conservative on issues such as price liberalization and macroeconomic stabilization. They have used the argument—either from conviction or to support their political agendas—that the sequence of reforms is wrong if prices are liberalized before market-oriented institutions are created.

My own view, and the one that guides much of the staff’s work in the IMF, is that governments should move on price liberalization and macroeconomic stabilization as quickly as they can. The outcomes to date in the transition economies support this position. In general, the Baltic states are perceived as the most successful economies in the FSU today. In Latvia and Estonia, for example, inflation has been reduced with remarkable success. In Latvia, the inflation level in mid-1992 when the new currency was introduced was around 10 percent a month. It was falling, it is true, but it was still around 10 percent. By April 1993, however, the inflation rate was only 0.3 percent a month, and in May it was actually negative as prices fell. Inflation was also around 10 percent a month in Russia and most of the other ruble area countries in mid-1992, but since then it has risen and is now in the range of 20–25 percent per month in most of these countries. Estonia and Latvia have been very successful on the inflation side in absolute terms and also relative to other reforming economies. They have succeeded by freeing all prices and at the same time introducing restrictive macroeconomic policies.

What about exchange rate policies and the IMF’s experience with exchange rate regimes in the FSU? Again, Latvia and Estonia stand out. As Michael Mussa said, Estonia chose a currency board system; fixed its currency, the kroon, at eight to the deutsche mark in June 1992; and did not deviate from that level. Latvia, on the other hand, chose a floating currency. But despite this difference, both currencies have remained strong. There was some initial depreciation of the Latvian ruble, but it was followed by an appreciation. I conclude from this experience that no one exchange rate regime is necessarily better than another when a country with high rates of inflation embarks on reform. Both a fixed and a floating regime can work if the willingness to make them work is there, and especially the willingness to put in place the necessary supporting monetary and fiscal policies.

This point really follows from what Sterie Beza said earlier: that the government must feel comfortable with the exchange regime it chooses. In Estonia and Latvia, when the authorities told the IMF what type of regime they had chosen, much discussion ensued. In both cases, the authorities understood the potential risks of the systems they had chosen and still felt comfortable with their position. The IMF supported these choices, and so far the outcomes have been good. This example is particularly interesting because of the similarities between the initial positions and economic structures of these two economies.

My second conclusion is that a fixed rate is more risky than a floating rate. A fixed rate may help to the extent that it gives the market more confidence that a policy will be sustained, and it will also impose more fiscal discipline on policymakers. However, if the rate cannot be sustained, or policymakers are unable to maintain the necessary discipline, the loss of credibility can be very damaging.

More generally, the experiences of Latvia and Estonia suggest that macroeconomic stabilization can succeed in the difficult situations faced by the FSU countries even before the long-term structural reforms are in place. Of course, the Latvians and the Estonians are working hard on these long-term reforms, but the process will take some time.

One important question that is always asked in the debate about the timing of macroeconomic stabilization concerns the impact that the reforms will have on output and employment. It is much too early to reach anything like a final judgment on this matter for the FSU. However, there has been a huge drop in output in all the FSU countries, regardless of whether there has been macroeconomic stabilization. This decline is a consequence of the systemic changes that these countries are undergoing. It is also true that the fall in output has actually been greater in Latvia and Estonia than in most other countries, prompting the authorities of some of these other countries to conclude that while introducing a national currency may help control inflation, as Latvia and Estonia have shown, it leads to large losses in output and employment.

However, this is the wrong conclusion. The truth is that a very large portion of the decline in output, and perhaps the entire amount that exceeded the FSU average, can be attributed to the much larger terms of trade shock the two Baltic countries have experienced. As a result of the freeing of prices—and especially of the prices Russia, the major energy producer, charges for energy exports to other FSU countries—Latvia and Estonia have suffered a terms of trade deterioration of about 40 percent. This deterioration has had a large impact on real incomes, with major adverse consequences for output in the short run. In addition, there have been difficulties with the supply of energy from Russia to the Baltics that many other FSU countries have not experienced, or at least not to the same degree. These problems have adversely affected output independently of macroeconomic policy. So my judgment is that up to now, macroeconomic stabilization in these countries, which has been successful, has been much less significant as a factor in the evolution of output and employment than might first appear.

In the longer term, of course, macroeconomic stabilization will make it easier for output to recover and growth to begin again. I will not go through all the explanations that show why inflation, if unchecked, becomes hyperinflation and damages output, growth, and investment, because these are well-known points. But I will say that at the empirical level, there is some very early evidence of a turnaround in Estonia, where output has begun to grow again. Estonia is the first FSU country where such a turnaround has occurred, although in 1991 similar successes began to occur in some East European countries that had stabilized inflation or at least avoided the very high levels of inflation that persist in the ruble area.

Moving away from sequencing, I will refer briefly to developments in the enterprise sector in the FSU. The enterprise sector and the problem of enterprise reform itself are at the heart of the transformation in the FSU. They are also at the heart of the interaction between macroeconomic stabilization and structural reform. In some FSU economies, governments have been reluctant to push enterprises to reform. These enterprises have not changed their behavior, their output mix, or inputs; improved productivity or reduced overstaffing; or found new markets, even though orders from the defense sector are disappearing. The governments that have neither pushed enterprises into making these changes nor created a situation in which the enterprises are forced to change their behavior themselves have had to provide large subsidies or credits that are undermining the process of macroeconomic stabilization.

The IMF’s view is that, consistent with introducing macroeconomic stabilization as early as possible, it is also necessary to impose a hard budget constraint on enterprises. In fact, the hard budget constraint is the working out of macroeconomic stabilization policies at the enterprise level. This approach requires the finance ministers and governors of central banks to refuse subsidies and credits to enterprise managers and to members of parliament acting for these managers. Apart from anything else, these refusals involve a huge cultural shift. Under the old Soviet culture, any enterprise manager who needed help simply contacted the sponsoring ministry to request credit or other assistance. In a sense, it was the ministry’s job to do something for that enterprise. Now, however, it is the job of the government and the ministry to provide the financial framework and then tell the enterprises to operate within it. It is proving difficult for the FSU countries to make this cultural shift. Even the Baltic states, where the cultural changes have been most profound, have experienced difficulties. However, the reality of the hard budget constraint is that ministers must say “no” on the telephone 20 times a day, despite threats of massive unemployment and other unattractive consequences.

At the same time, the hard budget constraint will be an easier discipline to live with if enterprise reform in the microeconomic sense is pushed forward as quickly as possible. Governments should therefore be doing all they can to restructure enterprises. Initiatives are needed, ranging from privatization and ownership changes to new incentive structures for the managers of enterprises that must remain in the state sector.

Let me end my remarks about enterprise reform on an optimistic note. In the first part of 1992, many observers were discouraged by the seeming inability of enterprises to change their behavior in any way. Price liberalization had suddenly presented these firms with a new set of relative prices, and still the old behavior did not change. The enterprises went on producing the old output—which, in many cases, nobody wanted, in part because much of it was destined for the defense industry, which was experiencing huge budget cuts. However, the enterprises could not afford to continue producing, fell behind in paying their bills, and often tried to obtain more credits from the central bank or the government. It looked as though a huge financial crisis was being generated. However, gradually it became clear, largely through anecdotal evidence at first, that enterprise managers were beginning to understand that businesses would have to change their behavior to survive in a market economy. Enterprises started changing their output mix, producing refrigerators instead of rockets. They sought out new markets in the West and within the FSU area. These enterprises have the potential to do quite well in the West because their current labor costs are extremely low (the average monthly wage in Russia is approximately $20–$30).

Finally, I shall turn very briefly to the question of social safety nets. The problem in the FSU is, in a sense, the opposite of the problem elsewhere. The issue is not putting in place a social safety net that did not exist before; on the contrary, the old Soviet regime was really quite generous in terms of social welfare. In addition, even today the politicians are inclined to make sure that the people are protected, and the debates in the Russian Parliament often refer to this imperative. The problem the government now faces is not how to create a social safety net, but how to make the present one more affordable. It could become extremely expensive if unemployment rises to the levels that some analysts have projected—and unemployment certainly will rise during the transition period as the old industries, the dinosaurs, are shut down. Making the net more affordable means targeting it more effectively. One of the first goals should be to remove the generalized price subsidies, especially on food and, to a lesser extent, on housing and public transport, and to target benefits in these areas to the truly needy groups within the population. In terms of social safety nets, this challenge is the most important one these countries face. Most FSU countries recognize this fact, but there are a few where the old paternalistic approach still prevails.